boydii. Colicin U, a novel colicin produced by ... - Semantic Scholar
Colicin U, a novel colicin produced by Shigella boydii. D Smajs, H Pilsl and V Braun J. Bacteriol. 1997, 179(15):4919.
These include: CONTENT ALERTS
Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
Updated information and services can be found at: http://jb.asm.org/content/179/15/4919
JOURNAL OF BACTERIOLOGY, Aug. 1997, p. 4919–4928 0021-9193/97/$04.0010 Copyright © 1997, American Society for Microbiology
Vol. 179, No. 15
Colicin U, a Novel Colicin Produced by Shigella boydii ˇMAJS,1 HOLGER PILSL,2 DAVID S
AND
VOLKMAR BRAUN2*
Department of Biology, Faculty of Medicine, Masaryk University, Brno, Czech Republic,1 and Mikrobiologie/Membranphysiologie, Universita ¨t Tu ¨bingen, Tu ¨bingen, Germany2 Received 17 March 1997/Accepted 22 May 1997
colicins suggested that this was a new type of colicin, for which, in accordance with the original Fredericq classification system (24), the name colicin U was proposed (31). This paper presents the molecular characterization of this novel colicin and its plasmid.
Colicins are antibacterial proteins whose genes are usually located on plasmids. The bacteriocin 28b gene of Serratia marcescens is the only colicin known to be localized on the chromosome (59). Colicins are produced by certain bacterial strains of the family Enterobacteriaceae, particularly by Escherichia coli. The toxic effects of colicins are limited to sensitive strains of this family and are most active within the species of the producer strain (43, 44). Colicins consist of single polypeptide chains with molecular masses of 29 to 75 kDa (11). Their interaction with susceptible bacterial cells occurs in three steps: attachment to a specific receptor of the outer membrane, translocation through the cell envelope, and lethal action on the cell target (2, 11, 15, 46). The three-step mechanism of colicin action is reflected by the three-domain structure of the polypeptide. The central domain mediates binding to cell surface receptors, the N-terminal sequence is responsible for the uptake of colicins across the cell envelope, and the C-terminal part exerts the lethal effects. The domains function largely independently of each other. Nature assembled colicins by exchanging DNA segments that encode domains (39, 40, 47). Colicins are classified according to receptor specificity, immunity, and type of translocation through the cell envelope of susceptible cells (17, 18). Group A colicins utilize the Tol system, which consists of the proteins TolA, TolB, TolQ, and TolR (55), and group B colicins use the Ton system, which consists of the proteins TonB, ExbB, and ExbD (9). In 1992, V. Hora´k (31) demonstrated the production of a colicin in two cross-immune strains of Shigella boydii (serovars 1 and 8). The absence of cross-resistance with any of the known
MATERIALS AND METHODS Bacterial strains. E. coli strains, plasmids, and the bacteriophage used in the experiments are listed in Table 1. S. boydii M592 (serovar 8) and S. boydii 215/92 (serovar 1) were from V. Hora´k, Department of Microbiology, Faculty of Medicine, Charles University, Hradec Kra´love´, Czech Republic. S. boydii Shp132/56 (serovar 8) and S. boydii Shp166/58 (serovar 1) were from the Czech National Collection of Type Cultures, Prague, Czech Republic. Colicins A, B, K, N, and L were produced by E. coli BZB 2101(pColA-CA31), BZB 2102(pColB-K260), BZB 2116(pColK-K235), BZB 2123(pColN-284), and Serratia marcescens JF246 (J. Foulds), respectively. The BZB strains were obtained from A. P. Pugsley, Institut Pasteur, Paris, France. Growth media. Bacteria were grown at 37°C in TY medium containing 8 g of Bacto Tryptone (Difco Laboratories), 5 g of yeast extract, and 5 g of NaCl per liter (pH 7). For maintenance of plasmids, 50 mg of chloramphenicol per ml was added to liquid medium and to 1.5% (wt/vol) TY agar. Microbiological methods. Serial dilutions of a stock solution of bacteriophage C21 were mixed with 0.1 ml of bacteria (109 per ml) and plated to determine the number of plaques. Crude extracts of colicins were prepared from colicinogenic strains grown in TY medium supplemented with mitomycin (0.5 mg per ml). Cells were harvested, suspended in distilled water, washed, and sonicated. Colicin activity assays were performed as described previously (39). The colicin binding activities of both parental and mutant strains of E. coli were assayed by mixing 1 ml of cell suspension in distilled water (2 3 109 per ml) with 1 ml of colicin solution (103 dilution titer). The mixtures were incubated for 20 min at 37°C and then centrifuged. The supernatant was decanted, and unbound colicin was assayed by the punch-hole method (36, 58). Recombinant DNA methods. Standard techniques were employed for restriction endonuclease analysis, ligation, isolation of plasmids, and transformation of plasmid DNA (48). pColU genes were cloned into the vector pBCSK1 (Stratagene). DNA was sequenced by using the dideoxy chain termination method (49), fluorescence-labeled nucleotides (AutoRead Sequencing Kit), and the A.L.F. sequencer. Site-specific mutagenesis was performed by PCR (33). The mutagenized fragments were examined by DNA sequencing.
* Corresponding author. Mailing address: Mikrobiologie/Membranphysiologie, Auf der Morgenstelle 28, 72076 Tu ¨bingen, Germany. Phone: (49) 7071 2972096. Fax: (49) 7071 294634. E-mail: volkmar [email protected]. 4919
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
A novel colicin, designated colicin U, was found in two Shigella boydii strains of serovars 1 and 8. Colicin U was active against bacterial strains of the genera Escherichia and Shigella. Plasmid pColU (7.3 kb) of the colicinogenic strain S. boydii M592 (serovar 8) was sequenced, and three colicin genes were identified. The colicin U activity gene, cua, encodes a protein of 619 amino acids (Mr, 66,289); the immunity gene, cui, encodes a protein of 174 amino acids (Mr, 20,688); and the lytic protein gene, cul, encodes a polypeptide of 45 amino acids (Mr, 4,672). Colicin U displays sequence similarities to various colicins. The N-terminal sequence of 130 amino acids has 54% identity to the N-terminal sequence of bacteriocin 28b produced by Serratia marcescens. Furthermore, the N-terminal 36 amino acids have striking sequence identity (83%) to colicin A. Although the C-terminal pore-forming sequence of colicin U shows the highest degree of identity (73%) to the pore-forming C-terminal sequence of colicin B, the immunity protein, which interacts with the same region, displays a higher degree of sequence similarity to the immunity protein of colicin A (45%) than to the immunity protein of colicin B (30.5%). Immunity specificity is probably conferred by a short sequence from residues 571 to residue 599 of colicin U; this sequence is not similar to that of colicin B. We showed that binding of colicin U to sensitive cells is mediated by the OmpA protein, the OmpF porin, and core lipopolysaccharide. Uptake of colicin U was dependent on the TolA, -B, -Q, and -R proteins. pColU is homologous to plasmid pSB41 (4.1 kb) except for the colicin genes on pColU. pSB41 and pColU coexist in S. boydii strains and can be cotransformed into Escherichia coli, and both plasmids are homologous to pColE1.
4920
ˇMAJS ET AL. S
J. BACTERIOL. TABLE 1. E. coli strains, plasmids, and bacteriophage used
Strain, plasmid, or bacteriophage
Genotype and/or phenotype
Source or reference
hsdR lacZ rpsL ser thi thr 58-161 metB1 rpsL l1 Fdef 5K(pColU pSB41) tolA fhuA21 lacY1 thi leuB6 supE44 thr-1 HfrC K10C4 tolC ara D(lac pro) thi F9 lac pro GM1 tolQ GM1 tolR tolB fhuA21 lacY1 thi leuB6 supE44 thr-1 tonB aroB malT tsx thi tolQ exbB tolR exbBD ompA araD139 D(argF-lac)U169 rpsL150 relA fibB5301 deoC1 pts F25 rbsR MC4100 ompFC tsx thr ala leu proA lacY galK argE rpsL xyl thi supE mtl ompABC ompFC lamB KS26 ompA ompFC lamB Row ompF 5K ompA 5K ompA rfa KB426 ompF 5K rfa 5K rfa thi pyrD glt galK trp recA rpsL UH99 ompA F2 thi argE proA thr leu mtl xyl galK lacY rpsL supE non P400 non1 his ompB101 ompA2000 P400 ompA2005 ompA of Shigella dysenteriae ompA of Enterobacter aerogenes ompA of Serratia marcescens
G. Schrempf P. Fredericq This work 55 60 60 55 55 55 55 K. Hantke 8 10 H. J. Krieger-Brauer 13 E. Bremer 30 R. Benz R. Koebnik R. Benz This work This work This work This work This work This work 16 12 51 28 37 5 6 7
E. coli B BL21
F2 hsdS gal
54
E. coli O8:K27 F464 F539 F588
met his pro mtl Strr F2 rfa met his pro mtl Strr F2 rfa met his pro mtl Strr F2
50 50 50
Plasmids pBCSK1 pDS1 pDS2 pDS3 pDS4 pDS5 pHP90 pJBS636 pHP91
Phage T7 gene 10 promoter pBCSK1 carrying pColU in ClaI restriction site pBCSK1 carrying cua, cui, and cul in 4.5-kb HindII-ClaI fragment of pColU pBCSK1 carrying cua in 2.3-kb HindII-EcoRV fragment of pColU pBCSK1 carrying cui and cul in 2.7-kb SpeI-ClaI fragment of pColU pBCSK1 carrying cul in 1.5-kb EcoRV-EcoRV fragment of pColU pDS2; ColUD382–425 b-Lactamase fusion vector ColE1 ori, Neor pT7-7 polylinker T7 promoter pJBS636 carrying cui in 1.3-kb SspI-PstI fragment of pColU
Stratagene This work This work This work This work This work This work T. Focareta This work
Bacteriophage C21
Core LPS receptor
45
Protein nalytical procedures. Proteins were labeled radioactively in E. coli BL21 (54) with [35S]methionine (56) and were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described previously (40). Proteins were labeled in vitro with [35S]methionine in a bacterial cell-free transcription-translation system (Promega, Madison, Wis.). SDS-PAGE of LPS. Lipopolysaccharide (LPS) samples were prepared by using the SDS–proteinase K–whole-cell lysate method (29). Samples were run on 17% polyacrylamide gels, and then the gels were silver stained (57). Computer-assisted sequence analysis. Computer-assisted sequence analysis was performed as described previously (39). Nucleotide sequence accession number. The nucleotide sequences reported in this study were deposited in the EMBL data bank under accession no. Y11823.
RESULTS Genes on the pColU plasmid. E. coli 5K was transformed with the colicin U-determining plasmid of S. boydii M592 (serovar 8). Selection for immunity to added colicin U resulted in the isolation of transformants that carried two plasmids: a 7.3-kb plasmid designated pColU and a 4.1-kb plasmid designated pSB41. No transformants with a single plasmid were obtained. One such transformant was chosen for further studies and was designated strain U8T19. Sequencing of one DNA
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
E. coli K-12 5K Row U8T19 A592 K10C4 C65 GM1 TPS13 TPS300 A593 BR158 HE2 HE10 KB426 MC4100 HF24 HO830 KS26 KO16 KS26-2 DS2 DS1 DS3 KB426H1 HP77 HP78 UH99 UH100 P400 P530-1cII P400-2.2. UH100(l ompA-Sh) UH100(l ompA-En) UH100(l ompA-Se)
VOL. 179, 1997
COLICIN U OF S. BOYDII
4921
strand of each plasmid revealed partial sequence identity and, in addition, similarity to the pColE1 sequence. Sequence similarity between pSB41 and pColE1 was high in the rep and mob regions of pColE1 (bp 596 to 3733) (Fig. 1) (14). The sequence of the remaining 1-kb region of pSB41 was not similar to sequences of pColE1, with the exception of a 0.1-kb sequence, which was similar to that of exc from pColE1. pColU and pColE1 had high sequence similarity from bp 496 to 5140, extending from the second third of the gene for the lytic protein (see below) to the start codon of the colicin activity gene (Fig. 1). The remaining 2.6-kb pColU region did not show any sequence similarity to pColE1. The pSB41 and pColU sequences differed in the origin of replication, which may allow their coexistence in cells. Transformation of pColU alone was achieved after the plasmid was cleaved by ClaI and ligated into pBCSK1, which resulted in pDS1. The cloned 2.3-kb HindII-EcoRV fragment on pDS3 (Table 1) encoded the gene necessary for colicin U synthesis. However, the transformant carrying this fragment was not stable because, as sequencing revealed, it lacked the immunity gene. The SpeI-ClaI (2.7 kb) fragment in pDS4 conferred immunity to colicin U, which indicated the presence of a functional immunity gene on the cloned fragment. Bacteria carrying the cloned EcoRV-EcoRV fragment (1.5 kb) in pDS5
either grew or lysed, depending on the polarity of the colicin U lytic gene relative to the lacZ promoter of pBCSK1. pColU was sequenced between HindII and AluI (2.9 kb) on both strands. Three open reading frames, one each for colicin U activity (cua), immunity (cui), and lysis (cul), were identified (Fig. 2). The cui transcription polarity is opposite to that of cua and cul, which is characteristic for colicins whose target is the cytoplasmic membrane. The cua gene codes for a protein consisting of 619 amino acid residues (Mr, 66,289), cui codes for a protein of 174 residues (Mr, 20,688), and cul codes for a polypeptide of 45 residues (Mr, 4,673). The promoter region upstream of cua contains two overlapping SOS boxes, as is found on pColE1, and a highly conserved 235 region and a less conserved 210 region, as is observed in other colicin determinants. The promoter region from pDS1 was cloned into pDS2 by using HindII, which replaced the optimal 235 consensus region TTGACA with TCGACA and resulted in a considerable decrease of colicin U synthesis upon SOS induction (data not shown). SDS-PAGE of radiolabeled proteins from cells synthesizing the colicin activity and immunity proteins showed bands that corresponded to the expected size of the proteins (Fig. 3, lane 1 [65 kDa], and lane 3 [21 kDa]). pColU of another S. boydii strain was selected for compar-
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 1. Restriction map of plasmid pColU. The arrows indicate the transcriptional polarities of the cua, cui, and cul genes. Numbers correspond to the nucleotide sequence of pColE1. The positions of the genes rep, mob, cer, and exc regions are those of pColE1 (14) and are at the same sites on pColU, as revealed by the sequencing of pColU performed in this study.
4922
ˇMAJS ET AL. S
J. BACTERIOL.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 2. Nucleotide sequence of the HindII-AluI fragment (2.9 kb) of pColU. The arrows indicate the transcriptional polarities of the cua, cui, and cul genes. Putative promoter regions (210 and 235), SOS boxes, ribosome binding site sequences (S.D.), transcription termination sites (T1, T2, T3), and a predicted TolA box are indicated.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 2—Continued.
4923
4924
ˇMAJS ET AL. S
J. BACTERIOL.
FIG. 3. SDS-PAGE of radiolabeled proteins from colicinogenic (colicin U) and immune strains. The arrows indicate the positions of colicin U (lane 1), ColUD382-425 (lane 2), and the Cui immunity protein (lane 3). The genes for colicin U (on pDS3) and ColUD382-425 (on pHP90) were expressed in vitro; cui (on pHP91) was expressed in vivo. The 25-kDa bands (marked with dots) in lanes 1 and 2 correspond to the chloramphenicol transacetylase encoded on pBCSK1, and the 30-kDa band in lane 3 corresponds to the neomycin phosphotransferase encoded by pJSB636. The molecular masses (in kilodaltons) of marker proteins are indicated.
ison. The restriction maps of the pColU plasmids and the nucleotide sequences of the cua, cui, and cul genes were identical. Colicin U protein. A comparison of the colicin U amino acid sequence with those of known colicins showed the highest degree of similarity to colicin A (36.8% identity, 10.1% similarity). However, the degree of sequence similarity varied along the polypeptide chain. A dendrogram of sequence similarities of pore-forming colicins is shown in Fig. 4. Since colicins are composed of domains, the degree of sequence similarity was determined for each domain. The 200 C-terminal amino acids of colicin U exhibited the greatest sequence similarity to the corresponding region of colicin B (73.0% identity,
FIG. 4. Dendrogram of the sequence alignment of the pore-forming colicins.
12.5% similarity) (Fig. 5); they had 65.5% identity to the poreforming domain of colicin A and 49.0% identity to the bacteriocidal domain of colicin N. Therefore, colicin U belongs to the group of pore-forming colicins (2, 15), which are characterized by a long hydrophobic sequence near the C terminus (38). This hydrophobic sequence, extending from residue 571 to 599, displays the lowest sequence similarity to the corresponding region of colicin B (Fig. 5) and therefore may determine the specificity of inactivation of colicin U by its cognate immunity protein. Colicins U and B displayed no cross-immunity. The N-terminal sequence of colicin U shows the highest sequence similarity to bacteriocin 28b (Fig. 6), with 54.2% identity and an additional 16.0% similarity in the 130 N-terminal residues (27). A high degree of identity of the N-terminal sequences of colicins U and A (83.3%) was confined to the first 36 amino acid residues. The putative receptor binding domain in the central region of colicin U revealed only low sequence similarity to that of other colicins. Nevertheless, significant sequence similarity between colicin U and colicin K (Fig. 7), which was restricted to a region ranging from residue 280 to 386 (34.6% identity), was observed. A higher sequence similarity (restricted to 44 residues) between colicin U and the TolA protein (Fig. 8), extending from residue 382 to 425 of the colicin U molecule (68.2% identity), was found. Short KAAAG stretches homologous to the KAAAD/E sequence, repeated in TolA, is contained once and only in colicin U. KA-rich regions occur also in other colicins. Immunity protein of colicin U. The immunity protein of colicin U is homologous to the immunity proteins of poreforming colicins, with the highest sequence similarity to colicin A (identity, 45.4%; similarity, 14.9%) and lower sequence sim-
FIG. 6. Sequence comparison of the N-terminal 130 amino acids of colicin U with the corresponding region of bacteriocin 28b and with the 36 amino acids of colicin A. Asterisks denote identical residues; dots indicate similar amino acids.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 5. Sequence comparison of the C-terminal 200 amino acids of colicins U and B. Asterisks denote identical residues; dots indicate similar amino acids. a1 to a10 indicate a-helices of colicin B postulated by sequence comparison with colicin A (25), for which an X-ray structure of the pore-forming domain in aqueous solution has been determined (38).
VOL. 179, 1997
COLICIN U OF S. BOYDII
4925
TABLE 2. Sensitivities of ompA, ompF, and rfa mutants to colicins U, A, L, K, and N E. coli straina
FIG. 7. Sequence similarity of the central 107 amino acids of colicins U and K. Asterisks denote identical residues; dots indicate similar amino acids.
FIG. 8. Sequence similarity of the central 44 amino acids of colicin U and the TolA protein. Asterisks denote identical residues; dots indicate similar amino acids.
Wild type ompA ompF rfa ompA rfa Wild type rfa rfa
U
A
L
K
N
6 3 5 5 r 5 4 4
5 5 1 5 5 — — —
4 — — 4 — 3 2 1
3 2 2 3 2 — — —
4 4 2 5 4 — — —
a Strains DS1, DS2, DS3, and HP77 were isolated from the colicin U inhibition zone on plates seeded with E. coli K-12 5K. E. coli F464 and its derivatives F539 and F588 arose from E. coli O8:K27 (50). b The numbers indicate the highest colicin dilutions active on bacteria (e.g., 5 5 105). r, resistant; —, not determined.
sitivity, and an ompA rfa mutant (DS3) was resistant to colicin U (Table 2). The killing by colicins U, K, and L of bacterial strains from various genera of the family Enterobacteriaceae that synthesize OmpA and of E. coli strains that synthesize OmpA proteins with defined point mutations is shown in Table 3. Colicin U displayed the same pattern of activity as colicin K, which indicated that the two colicins recognize a similar, if not identical, region of OmpA. In contrast, the sensitivity of these strains to colicin L differed from the sensitivity to colicin U and K (Table 2). Binding of colicin U to ompA, ompF, and rfa mutants. Mutations in ompA, ompF, and rfa could affect binding of colicin U to cells, uptake of colicin U, or both. Therefore, we tested the adsorption of colicin U to several of the mutants that displayed a reduced sensitivity to colicin U (Table 4). Adsorption decreased proportionally to the sensitivity of the various cell types. Presumably, colicin U can bind to OmpA, OmpF, and the outer core of the LPS. However, even the colicin U-resistant ompA rfa mutant adsorbed 26% of the amount of colicin U that adsorbed to the wild type, which probably reflects unspecific binding. Uptake of colicin U. To examine whether colicin U is taken up by the Ton or the Tol system, we tested the colicin U sensitivities of various mutants (Table 5). A tolA mutant was insensitive to colicin U, while a tonB mutant was unaffected. The Tol proteins TolA, TolQ, TolR, and TolB were required for colicin U uptake, whereas TolC was not. The residual susceptibility of the tolQ and tolR mutants came from the partial replacement of their activities by the ExbB and ExbD
FIG. 9. SDS-PAGE of outer membrane LPSs from rfa mutant strains isolated from the inhibition zone of colicin U on plates seeded with E. coli 5K. Lanes: 1, E. coli 5K; 2, E. coli HP77 (rfa); 3, E. coli HP78 (rfa).
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
ilarity to colicin B (identity, 30.5%). The colicin U producer strain was partially immune to colicins A and N (10- to 100-fold and 10-fold reduction of sensitivity, respectively), but the colicin A and N producer strains were fully sensitive to colicin U (data not shown). Lytic protein of colicin U. The colicin U lytic protein (4.7 kDa) showed the highest sequence similarity to the lytic proteins of colicins 5 and K (76.7% identity), of colicin 10 (72.1% identity), and of cloacin DF13 and the E-type colicins (62.2 to 71.1% identity). The mature proteins display the highest sequence similarity after residue 16, which is the cleavage site for lipoprotein signal sequences. Complex requirement of outer membrane determinants for colicin U entry into susceptible cells. Mutants of E. coli K-12 insensitive to colicin U were isolated from growth inhibition zones on plates onto which colicin U at a dilution titer of 106 was applied. Insensitive mutants isolated from the center of the inhibition zones, which contained the highest colicin concentration, were usually mutated in tol genes, as identified by their insensitivity to colicin E1, K, or L at the highest colicin concentrations available. Colonies isolated from the edge of the inhibition zone displayed a lower colicin U insensitivity (Table 2). These colonies carried mutations in ompA or ompF, as revealed by SDS-PAGE, which demonstrated the lack of OmpA or OmpF (data not shown). The patterns of resistance to colicin U of the mutants showed some overlap with the patterns of resistance to colicins A, K, L, and N, but no two mutants had identical resistance patterns (Table 2). The colicin U sensitivity of multiple mutants from our strain collection was also tested. The sensitivity of strain HF24 ompF ompC was reduced 100-fold, that of strain KB426H1 ompA ompF was reduced 103-fold, and that of strain HO830 ompA ompF ompC was reduced 105-fold. Strain KS26-2 ompC ompF lamB and strain KO16 ompA ompF ompC lamB were fully resistant. The sensitivity of the porin mutants to colicin B remained unchanged, which indicated that resistance to colicin U did not result from a general disturbance of outer membrane integrity. In addition, rfa mutants, which do not synthesize complete LPS but synthesize core LPS, were isolated from the colicin U inhibition zone, as shown by their sensitivity to bacteriophage C21 (45), to which the parent strain was resistant, and by silver staining after SDS-PAGE of isolated LPS (Fig. 9). An rfa mutant (HP77) had a 10-fold-reduced sensitivity to colicin U, an ompA mutant (DS1) displayed a 103-fold reduction in sen-
5K DS1 DS2 HP77 DS3 F464 F539 F588
Sensitivityb to colicin: Plasmid genotype
4926
ˇMAJS ET AL. S
J. BACTERIOL.
TABLE 3. Sensitivities to colicins U, K, and L of E. coli K-12 strains that synthesize different OmpA proteins Strain
Relevant phenotype
E. coli K-12 5K UH99 P400-2.2 P530-1cII UH100(l ompA-Sh) UH100(l ompA-En) UH100(l ompA-Se)
Wild type Wild type OmpA (E68K) OmpA (V110D) OmpA from Shigella dysenteriae OmpA from Enterobacter aerogenes OmpA from Serratia marcescens
TABLE 5. Sensitivity of tol and tonB mutants to colicin U
Sensitivitya to colicin: U
K
L
3 3 3 2 3 1 0
3 3 3 2 3 1 0
3 3 1 0 0 3 r
a
proteins (8, 10), as shown by the complete insensitivity of tolQ exbB and tolR exbB exbD mutants. Deletion of 44 amino acid residues in the colicin U protein. To estimate the biological function of the region homologous to the TolA protein, the 44 amino acids of this region in colicin U were deleted, resulting in the production of ColUD382-425, which had the size expected (Fig. 3, lane 2). ColUD382-425 was as active as wild-type colicin U on E. coli 5K, and its toxicity to omp, rfa, and tol mutant strains was decreased to the same extent as for wild-type colicin U. A colicin U-producing strain was as immune to ColUD382-425 as to wild-type colicin U (data not shown). DISCUSSION Colicin U is an addition to the list of bacteriocins (S1, S4, and Js) found in Shigella strains (23, 52). The U colicinogenic serovars 1 and 8 of S. boydii display cross-immunity to each other and to the colicinogenic strains S. boydii Shp 132/56 (serovar 8) and S. boydii Shp 166/58 (serovar 1). Shigella is very similar to E. coli, and plasmids can easily be exchanged between the two genera, as has been found for Shigella sonnei, in which 70 colicin types have been identified, most of which are also found in E. coli (32). The ColU plasmids of the two S. boydii serovars investigated are identical with regard to the restriction map and the nucleotide sequence of the region that encodes the cua, cui, and cul genes. pColU (7.3 kb) and pSB41 (4.1 kb) coexist in all four S. boydii strains. In addition, pColU could not be transformed into E. coli 5K without pSB41. These findings suggest that pSB41 may be required for replication of pColU or for another function important for pColU maintenance, such as partitioning among progeny cells. The high sequence similarity of the
TABLE 4. Binding of colicin U to E. coli 5K ompA, ompF, and rfa mutants Strain
Relevant genotype
Adsorption (%)a
Sensitivity to colicin Ub
5K Row HP77 DS2 DS1 DS3
Wild type Wild type rfa ompF ompA ompA rfa
100 6 11 100 6 13 74 6 7 72 6 14 52 6 4 26 6 12
6 6 5 5 3 r
a
95% confidence interval for the mean. The numbers indicate the highest colicin dilutions active on bacteria (e.g., 6 5 106). r, resistant. b
5K BR158 C65 A592 TPS13 TPS300 A593 HE2 HE10
Relevant genotype
Wild type tonB tolC tolA tolQ tolR tolB tolQ exbB tolR exbBD
Sensitivity to colicin Ua
6 6 6 t 1 2 3 t t
a The numbers indicate the highest colicin dilutions active on bacteria (e.g., 6 5 106). t, tolerant.
plasmids in the region outside the colicin genes of pColU suggests that a pSB41-like plasmid acquired the colicin determinant, thereby forming pColU. pColU and pColE1 are homologous in the region extending from the second third of the cul gene to the start codon of the cua gene, which covers a total length of 4.7 kb. The plasmids differ entirely in the region that determines the colicin genes. In pCol10, homology with pColE1 is found in the region coding for the lytic protein and in the sequence upstream of the start codon of the colicin activity gene. The homologous regions presumably serve as sites of recombination between the colicin plasmids. In the colicin plasmids pCol5, pCol10, pColK, and pColE1 (39–41) and the pesticin plasmid pPCP1 (42), homologous DNA fragments precisely determine functional domains, which suggests that these colicins evolved by exchange of DNA fragments. Sequence comparison of colicin U with other colicins also revealed sequence similarities of functional domains, but these were less pronounced than in the examples described above. The high sequence similarity of the C-terminal domain of colicin U to the corresponding regions of colicin B (73% identity), colicin A (65.5% identity) (26), and colicin N (49% identity) assigns colicin U to the pore-forming colicins. The high sequence similarity is interrupted between residues 571 and 599; this region may be the specific recognition site of the colicin U immunity protein, which does not confer immunity to colicin B. In analogy to the X-ray structure of the pore-forming domain of colicin A, the region from residue 571 to 599 is located where the hydrophobic helix 8 extends into helix 9 (Fig. 5) (38, 53). Uptake of colicin U and bacteriocin 28b is dependent on the tolA, -B, -Q, and -R genes. The N termini of the colicins have high sequence similarity (Fig. 6). Colicin U and colicin A contain the sequence DGTGWS, and bacteriocin 28 contains the sequence DGTNWS. These sequences are highly similar to the previously proposed TolA box DGSGS (39), which may be responsible for interaction of the translocation domain of group A colicins with TolA proteins. Colicin U exhibits the complex dependence on several integral outer membrane proteins that is typical for group A colicins. It requires OmpA, OmpF, and core LPS and in this regard is similar to the sequenced bacteriocin 28b (20), which is similar if not identical to colicin L used in this study for comparison (27). A rather long stretch of amino acids close to the N terminus of colicin U, extending beyond the proposed TolA box, is homologous to that region of bacteriocin 28b and may confer the requirement for OmpF and core LPS (Fig. 10). An OmpF interaction sequence has been localized to the N terminus of colicin A (3), but this sequence is not similar to the
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
The numbers indicate the highest colicin dilutions active on bacteria (e.g., 5 5 105). r, resistant.
Strain
VOL. 179, 1997
COLICIN U OF S. BOYDII
4927
assumed OmpF interaction sequence of colicin U and bacteriocin 28. Either the colicins interact with OmpF differently or the homologous sequence in colicin U and bacteriocin 28 specifies only interaction with the core LPS. The homologous sequence of 107 amino acids in the central part of colicins U and K may determine the identical requirement for OmpA, as revealed by the use of ompA mutants and OmpA proteins from different species. None of the single deletions in ompA and ompF conferred complete insensitivity to colicins U, A, L, K, and N; this supports previous findings that several outer membrane proteins are involved in the uptake of group A colicins, in contrast to group B colicins, for which deletion of single outer membrane proteins confers complete resistance. The pore-forming group A colicins remain bound to the outer membrane receptors while they insert into the cytoplasmic membrane (4). This also seems to apply to group B poreforming colicins, since the three-dimensional X-ray structure of colicin Ia reveals two very long a-helices, one of which connects the receptor binding domain with the pore-forming domain (61) and thus serves to bridge the periplasmic space. A very long a-helix in the TolA protein, required for the uptake of group A colicins, was proposed (35); this helix easily spans the periplasm from the cytoplasmic membrane to the outer membrane. Excision of a fragment extending from residue 68 to 209 of TolA does not inactivate the protein (19). Forty-four amino acids of colicin U showed a high sequence similarity to the central domain of the TolA protein; these amino acids could be excised (forming ColUD382-425) without inactivating colicin U. We did not find any changes in ColUD382-425 interaction with sensitive, tolerant, resistant, or immune bacterial cells, as compared with that of wild-type colicin U. Deletion of 35 amino acids in colicin A between residues 337 and 371 also caused no changes in the biological activity of colicin A (1). A repetitive Pro-X motif in the TonB protein is suggested to be involved in bridging the periplasmic space. Excision of the Pro-X motif (residues 66 to 100) in TonB slightly reduces TonB activity only under hyperosmolar conditions in which the periplasmic space is expanded (34). Although rigid structures designed to span the periplasmic space may be important for the action of certain colicins, TolA, and TonB, they are apparently not essential. It seems that there are sites of shorter distances between the outer membrane and cytoplas-
mic membrane, presumably fusion sites, through which TolAand TonB-dependent colicins can enter cells, and TonB-dependent infection by certain phages and uptake of ferric siderophores and vitamin B12 might occur. Colicin U-insensitive derivatives of E. coli K-12 that were not mutated in the tol genes were mutated in ompA, ompF, and rfa. Colicin U-resistant ompC and lamB mutants were not isolated in our experiments, even though existing mutants with these genotypes displayed a reduced sensitivity to colicin U. The porins can functionally substitute for each other, as shown by E. coli KS26-2 ompC ompF lamB, which is resistant to colicins U, L, A, K, and N. Binding of colicin N to the porins OmpF, OmpC, and PhoE has been demonstrated by microcalorimetry (22). The very low sensitivity of ompA deletion mutants to colicin U indicates a major role of OmpA in colicin U entry. The immunity protein of colicin U (Cui) is 45% identical to that of colicin A (21), which explains the partial immunity of U colicinogenic strains to colicin A. Although the sequence similarity between Cui and the colicin N immunity protein is much lower, Cui confers some immunity to colicin N. The immunity proteins of colicins U, B, A, and N exhibit a much higher sequence diversity than the pore-forming domains of the colicins with which they interact. The immunity proteins have to recognize the small differences between the pore-forming regions, implying selection for difference rather than similarity. ACKNOWLEDGMENTS D.S. thanks the Deutsche Akademische Austauschdienst (DAAD) for a short-term fellowship and the Deutsche Forschungsgemeinschaft (SFB 323) for financial support. ˇmarda for encouragement, advice, and reading of the D.S. thanks J. S manuscript. We are grateful to V. Hora´k for providing the S. boydii strains, and we thank K. Hantke, A. Angerer, and H. Killman for helpful discussions and K. A. Brune for critical reading of the manuscript. REFERENCES 1. Baty, D., R. Frenette, R. Lloubes, V. Geli, S. P. Howard, F. Pattus, and C. Lazdunski. 1988. Functional domains of colicin A. Mol. Microbiol. 2:807–811. 2. Benedetti, H., and V. Geli. 1996. Colicin transport, channel formation and inhibition, p. 665–691. In W. N. Konings, H. R. Kaback, and J. S. Lolkema (ed.), Handbook of biological physics, vol. 2. Elsevier Sciences, Amsterdam, The Netherlands. 3. Benedetti, H., N. Frenette, M. Baty, M. Knibiehler, F. Pattus, and C. Lazdunski. 1991. Individual domains of colicins confer specificity in colicin
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 10. Tentative assignment of structural regions to functional domains of the colicins indicated. Bacteriocin 28b is very similar if not identical to colicin L. In the colicin U mutant D382-425, residues 382 to 425 are deleted. aa, amino acids.
4928
4. 5. 6. 7. 8.
10. 11. 12.
13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
uptake, in pore-properties and in immunity requirement. J. Mol. Biol. 217: 429–439. Benedetti, H., R. Lloubes, C. Lazdunski, and L. Letellier. 1992. Colicin A unfolds during its translocation in Escherichia coli cells and spans the whole cell envelope when its pore has formed. EMBO J. 11:441–447. Braun, G., and S. T. Cole. 1982. The nucleotide sequence coding for major outer membrane protein OmpA of Shigella dysenteriae. Nucleic Acids Res. 10:2367–2378. Braun, G., and S. T. Cole. 1983. Molecular characterization of the gene coding for major outer membrane protein OmpA from Enterobacter aerogenes. Eur. J. Biochem. 137:495–500. Braun, G., and S. T. Cole. 1984. DNA sequence analysis of the Serratia marcescens ompA gene: implication for the organisation of an enterobacterial outer membrane protein. Mol. Gen. Genet. 195:321–328. Braun, V. 1989. The structurally related exbB and tolQ genes are interchangeable in conferring tonB-dependent colicin, bacteriophage, and albomycin sensitivity. J. Bacteriol. 171:6387–6390. Braun, V. 1995. Energy-coupled transport and signal transduction through the gram-negative outer membrane via the TonB-ExbB-ExbD dependent receptor proteins. FEMS Microbiol. Rev. 16:295–307. Braun, V., and C. Herrmann. 1993. Evolutionary relationship of uptake systems for biopolymers in Escherichia coli: cross-complementation between the TonBExbB-ExbD and the TolA-TolQ-TolR proteins. Mol. Microbiol. 8:261–268. Braun, V., H. Pilsl, and P. Gross. 1994. Colicins: structures, modes of action, transfer through membranes, and evolution. Arch. Microbiol. 161:199–206. Bremer, E., E. Beck, I. Hindennach, I. Sonntag, and U. Henning. 1980. Cloned structural gene (ompA) for an integral outer membrane protein of Escherichia coli K-12. Localization on hybrid plasmid pTU100. Mol. Gen. Genet. 179:13–20. Casabadan, M. J., and S. N. Cohen. 1979. Lactose genes fused to exogenous promoters in one step using a Mu lac bacteriophage: in vivo probe for transcriptional control sequence. Proc. Natl. Acad. Sci. USA 176:4530–4533. Chan, P. T., H. Ohmori, J. Tomizawa, and J. Lebowitz. 1984. Nucleotide sequence and gene organization of ColE1 DNA. J. Biol. Chem. 260:8925–8935. Cramer, W. A., J. B. Heymann, S. L. Schendel, B. N. Deriy, F. S. Cohen, P. A. Elkins, and C. V. Stauffacher. 1995. Structure-function of the channel-forming colicins. Annu. Rev. Biophys. Biomol. Struct. 24:611–641. Datta, D. B., B. Arden, and U. Henning. 1977. Major proteins of the Escherichia coli outer cell envelope membrane as bacteriophage receptors. J. Bacteriol. 131:821–829. Davies, J. K., and P. Reeves. 1975. Genetics of resistance to colicins in Escherichia coli K-12: cross-resistance among colicins of group B. J. Bacteriol. 123:96–101. Davies, J. K., and P. Reeves. 1975. Genetics of resistance to colicins in Escherichia coli K-12: cross-resistance among colicins of group A. J. Bacteriol. 123:102–117. Derouiche, R., M. Gavioli, H. Benedetti, A. Prilipov, C. Lazdunski, and R. Lloubes. 1996. TolA central domain interacts with Escherichia coli porins. EMBO J. 15:6408–6415. Enfedaque, J., S. Ferre, J. F. Guasch, J. Tomas, and M. Regue´. 1996. Bacteriocin 28b from Serratia marcescens N28b: identification of Escherichia coli surface involved in bacteriocin binding and translocation. Can. J. Microbiol. 42:19–26. Espesset, D., P. Piet, C. Lazdunski, V. Geli. 1994. Immunity proteins to pore-forming colicins: structure-function relationships. Mol. Microbiol. 13: 1111–1120. Evans, L. J. A., A. Cooper, and J. H. Lakey. 1996. Direct measurement of the association of a protein with a family of membrane receptors. J. Mol. Biol. 255:559–563. Fredericq, P. 1948. Action antibiotiques reciproques chez les Enterobacteriaceae. Rev. Belge Pathol. Exp. Med. 19 Suppl. 4:1–107. Fredericq, P. 1965. A note on the classifications of colicins. Zentralbl. Bakteriol. Hyg. A 196:140–142. Ge´li, V., and C. Lazdunski. 1992. An a-helical hydrophobic hairpin as a specific determinant in protein-protein interaction occurring in Escherichia coli colicin A and B immunity systems. J. Bacteriol. 174:6432–6437. Ge´li, V., D. Baty, F. Pattus, and C. Lazdunski. 1989. Topology and function of the integral membrane protein conferring immunity to colicin A. Mol. Microbiol. 3:679–687. Guasch, J. F., E. Enfedaque, S. Ferrer, D. Gargallo, and M. Regue´. 1995. Bacteriocin 28b, a chromosomally encoded bacteriocin produced by most Serratia marcescens biotypes. Res. Microbiol. 146:447–483. Henning, U., I. Hindennach, and I. Haller. 1976. The major proteins of the Escherichia coli outer cell envelope membrane: evidence for the structural gene of protein II*. FEBS Lett. 61:46–48. Hichcock, P. J., and T. M. Brown. 1983. Morphological heterogeneity among Salmonella lipopolysaccharide chemotypes in silver-stained polyacrylamide gels. J. Bacteriol. 154:269–277. Hoffmann, H., E. Fischer, H. Kraut, and V. Braun. 1986. Preparation of the FhuA (TonA) receptor protein from cell envelopes of an overproducing
J. BACTERIOL. strain of Escherichia coli K-12. J. Bacteriol. 166:404–411. 31. Hora ´k, V. 1992. Personal communication. 32. Hora ´k, V. 1994. Seventy colicin types of Shigella sonnei and an indicator system for their determination. Zentralbl. Bakteriol. Hyg. A 281:24–29. 33. Killmann, H., R. Benz, and V. Braun. 1993. Conversion of the FhuA transport protein into a diffusion channel through the outer membrane of Escherichia coli. EMBO J. 12:3007–3016. 34. Larsen, R. A., G. E. Wood, and K. Postle. 1993. The conserved proline-rich motif is not essential for energy transduction by Escherichia coli TonB protein. Mol. Microbiol. 10:943–953. 35. Levingood, S. K., W. B. Beyer, and R. E. Webster. 1991. TolA: a membrane protein involved in colicin uptake contains an extended helical region. Proc. Natl. Acad. Sci. USA 88:5939–5943. 36. Mayr-Harting, A., A. J. Hedges, and R. C. W. Berkeley. 1972. Methods for studying bacteriocins. Methods Microbiol. 7A:315–442. 37. Morona, R., M. Klose, and U. Henning. 1984. Escherichia coli K-12 outer membrane protein (OmpA) as a bacteriophage receptor: analysis of mutant genes expressing altered proteins. J. Bacteriol. 159:570–578. 38. Parker, M. W., J. P. Postma, F. Pattus, A. D. Tucker, and D. Tsernoglou. 1992. Refined structure of the pore-forming domain of colicin A at 2.4 Å resolution. J. Mol. Biol. 224:639–657. 39. Pilsl, H., and V. Braun. 1995. Novel colicin 10: assignment of four domains to TonB- and TolC-dependent uptake via the Tsx receptor and to pore formation. Mol. Microbiol. 16:57–67. 40. Pilsl, H., and V. Braun. 1995. Evidence that the immunity protein inactivates colicin 5 immediately prior formation of the transmembrane channel. J. Bacteriol. 177:6966–6972. 41. Pilsl, H., and V. Braun. 1995. Strong function-related homology between the pore-forming colicins K and 5. J. Bacteriol. 177:6973–6977. 42. Pilsl, H., H. Killmann, K. Hantke, and V. Braun. 1996. Periplasmic location of the pesticin immunity protein suggests inactivation of pesticin in the periplasm. J. Bacteriol. 178:2431–2435. 43. Pugsley, A. P. 1984. The ins and outs of colicins. I. Production, and translocation across membranes. Microbiol. Sci. 1:168–175. 44. Pugsley, A. P. 1984. The ins and outs of colicins. II. Lethal action, immunity and ecological implications. Microbiol. Sci. 1:203–205. 45. Rapin, A. M. C., and H. M. Kalckar. 1971. The relation of bacteriophage attachment to lipopolysaccharide structure, p. 267–307. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. 4. Academic Press, New York, N.Y. 46. Reeves, P. 1965. The bacteriocins. Bacteriol. Rev. 29:24–45. 47. Roos, U., R. E. Harkness, and V. Braun. 1989. Assembly of colicin genes from a few DNA fragments. Nucleotide sequence of colicin D. Mol. Microbiol. 3:891–902. 48. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 49. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. 50. Schmidt, G., B. Jann, and K. Jann. 1970. Immunochemistry of R lipopolysaccharides of Escherichia coli. Eur. J. Biochem. 16:382–392. 51. Skurray, R. E., R. E. W. Hancock, and P. Reeves. 1974. Con2 mutants: class of mutants in Escherichia coli K-12 lacking a major cell wall protein and defective in conjugation and adsorption of a bacteriophage. J. Bacteriol. 119:726–735. ˇ marda, J., J. Petrzelova 52. S ´, and B. Vyskot. 1987. Colicin Js of Shigella sonnei: classification of type colicin “7”. Zentralbl. Bakteriol. Hyg. A 263:530–540. 53. Song, H. Y., and W. A. Cramer. 1991. Membrane topography of ColE1 gene products: the hydrophobic anchor of the colicin E1 channel is a helical hairpin. J. Bacteriol. 173:2927–2934. 54. Studier, F. W., and B. A. Moffat. 1986. Use of bacteriophage T7-RNApolymerase to direct selective high level expression of cloned genes. J. Mol. Biol. 189:113–130. 55. Sun, T. P., and R. E. Webster. 1987. Nucleotide sequence of a gene cluster involved in entry of E colicins and single-stranded DNA of infecting filamentous bacteriophages into Escherichia coli. J. Bacteriol. 169:2667–2674. 56. Tabor, S., and C. C. Richardson. 1985. A bacteriophage RNA polymerase/ promoter system for controlled exclusive expression of specific genes. Proc. Natl. Acad. Sci. USA 82:1074–1078. 57. Tsai, G. M., and C. E. Frasch. 1982. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119:115–119. 58. Van Horn, E. A. 1961. The assay of colicin as a special case of a slowly diffusing substance and the application of the assay methods to the study of the colicin production. Thesis. University of Bristol, Bristol, United Kingdom. 59. Viejo, M. B., D. Gargallo, S. Ferrer, J. Enfedaque, and M. Regue´. 1992. Cloning and DNA sequence analysis of a bacteriocin gene from Serratia marcescens. J. Gen. Microbiol. 138:1737–1743. 60. Whitney, E. N. 1971. The tolC locus in Escherichia coli K12. Genetics 67:39–53. 61. Wiener, M., D. Freymann, P. Ghosh, and R. M. Stroud. 1997. Crystal structure of colicin Ia. Nature 385:461–464.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
9.
ˇMAJS ET AL. S
These include: CONTENT ALERTS
Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
Updated information and services can be found at: http://jb.asm.org/content/179/15/4919
JOURNAL OF BACTERIOLOGY, Aug. 1997, p. 4919–4928 0021-9193/97/$04.0010 Copyright © 1997, American Society for Microbiology
Vol. 179, No. 15
Colicin U, a Novel Colicin Produced by Shigella boydii ˇMAJS,1 HOLGER PILSL,2 DAVID S
AND
VOLKMAR BRAUN2*
Department of Biology, Faculty of Medicine, Masaryk University, Brno, Czech Republic,1 and Mikrobiologie/Membranphysiologie, Universita ¨t Tu ¨bingen, Tu ¨bingen, Germany2 Received 17 March 1997/Accepted 22 May 1997
colicins suggested that this was a new type of colicin, for which, in accordance with the original Fredericq classification system (24), the name colicin U was proposed (31). This paper presents the molecular characterization of this novel colicin and its plasmid.
Colicins are antibacterial proteins whose genes are usually located on plasmids. The bacteriocin 28b gene of Serratia marcescens is the only colicin known to be localized on the chromosome (59). Colicins are produced by certain bacterial strains of the family Enterobacteriaceae, particularly by Escherichia coli. The toxic effects of colicins are limited to sensitive strains of this family and are most active within the species of the producer strain (43, 44). Colicins consist of single polypeptide chains with molecular masses of 29 to 75 kDa (11). Their interaction with susceptible bacterial cells occurs in three steps: attachment to a specific receptor of the outer membrane, translocation through the cell envelope, and lethal action on the cell target (2, 11, 15, 46). The three-step mechanism of colicin action is reflected by the three-domain structure of the polypeptide. The central domain mediates binding to cell surface receptors, the N-terminal sequence is responsible for the uptake of colicins across the cell envelope, and the C-terminal part exerts the lethal effects. The domains function largely independently of each other. Nature assembled colicins by exchanging DNA segments that encode domains (39, 40, 47). Colicins are classified according to receptor specificity, immunity, and type of translocation through the cell envelope of susceptible cells (17, 18). Group A colicins utilize the Tol system, which consists of the proteins TolA, TolB, TolQ, and TolR (55), and group B colicins use the Ton system, which consists of the proteins TonB, ExbB, and ExbD (9). In 1992, V. Hora´k (31) demonstrated the production of a colicin in two cross-immune strains of Shigella boydii (serovars 1 and 8). The absence of cross-resistance with any of the known
MATERIALS AND METHODS Bacterial strains. E. coli strains, plasmids, and the bacteriophage used in the experiments are listed in Table 1. S. boydii M592 (serovar 8) and S. boydii 215/92 (serovar 1) were from V. Hora´k, Department of Microbiology, Faculty of Medicine, Charles University, Hradec Kra´love´, Czech Republic. S. boydii Shp132/56 (serovar 8) and S. boydii Shp166/58 (serovar 1) were from the Czech National Collection of Type Cultures, Prague, Czech Republic. Colicins A, B, K, N, and L were produced by E. coli BZB 2101(pColA-CA31), BZB 2102(pColB-K260), BZB 2116(pColK-K235), BZB 2123(pColN-284), and Serratia marcescens JF246 (J. Foulds), respectively. The BZB strains were obtained from A. P. Pugsley, Institut Pasteur, Paris, France. Growth media. Bacteria were grown at 37°C in TY medium containing 8 g of Bacto Tryptone (Difco Laboratories), 5 g of yeast extract, and 5 g of NaCl per liter (pH 7). For maintenance of plasmids, 50 mg of chloramphenicol per ml was added to liquid medium and to 1.5% (wt/vol) TY agar. Microbiological methods. Serial dilutions of a stock solution of bacteriophage C21 were mixed with 0.1 ml of bacteria (109 per ml) and plated to determine the number of plaques. Crude extracts of colicins were prepared from colicinogenic strains grown in TY medium supplemented with mitomycin (0.5 mg per ml). Cells were harvested, suspended in distilled water, washed, and sonicated. Colicin activity assays were performed as described previously (39). The colicin binding activities of both parental and mutant strains of E. coli were assayed by mixing 1 ml of cell suspension in distilled water (2 3 109 per ml) with 1 ml of colicin solution (103 dilution titer). The mixtures were incubated for 20 min at 37°C and then centrifuged. The supernatant was decanted, and unbound colicin was assayed by the punch-hole method (36, 58). Recombinant DNA methods. Standard techniques were employed for restriction endonuclease analysis, ligation, isolation of plasmids, and transformation of plasmid DNA (48). pColU genes were cloned into the vector pBCSK1 (Stratagene). DNA was sequenced by using the dideoxy chain termination method (49), fluorescence-labeled nucleotides (AutoRead Sequencing Kit), and the A.L.F. sequencer. Site-specific mutagenesis was performed by PCR (33). The mutagenized fragments were examined by DNA sequencing.
* Corresponding author. Mailing address: Mikrobiologie/Membranphysiologie, Auf der Morgenstelle 28, 72076 Tu ¨bingen, Germany. Phone: (49) 7071 2972096. Fax: (49) 7071 294634. E-mail: volkmar [email protected]. 4919
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
A novel colicin, designated colicin U, was found in two Shigella boydii strains of serovars 1 and 8. Colicin U was active against bacterial strains of the genera Escherichia and Shigella. Plasmid pColU (7.3 kb) of the colicinogenic strain S. boydii M592 (serovar 8) was sequenced, and three colicin genes were identified. The colicin U activity gene, cua, encodes a protein of 619 amino acids (Mr, 66,289); the immunity gene, cui, encodes a protein of 174 amino acids (Mr, 20,688); and the lytic protein gene, cul, encodes a polypeptide of 45 amino acids (Mr, 4,672). Colicin U displays sequence similarities to various colicins. The N-terminal sequence of 130 amino acids has 54% identity to the N-terminal sequence of bacteriocin 28b produced by Serratia marcescens. Furthermore, the N-terminal 36 amino acids have striking sequence identity (83%) to colicin A. Although the C-terminal pore-forming sequence of colicin U shows the highest degree of identity (73%) to the pore-forming C-terminal sequence of colicin B, the immunity protein, which interacts with the same region, displays a higher degree of sequence similarity to the immunity protein of colicin A (45%) than to the immunity protein of colicin B (30.5%). Immunity specificity is probably conferred by a short sequence from residues 571 to residue 599 of colicin U; this sequence is not similar to that of colicin B. We showed that binding of colicin U to sensitive cells is mediated by the OmpA protein, the OmpF porin, and core lipopolysaccharide. Uptake of colicin U was dependent on the TolA, -B, -Q, and -R proteins. pColU is homologous to plasmid pSB41 (4.1 kb) except for the colicin genes on pColU. pSB41 and pColU coexist in S. boydii strains and can be cotransformed into Escherichia coli, and both plasmids are homologous to pColE1.
4920
ˇMAJS ET AL. S
J. BACTERIOL. TABLE 1. E. coli strains, plasmids, and bacteriophage used
Strain, plasmid, or bacteriophage
Genotype and/or phenotype
Source or reference
hsdR lacZ rpsL ser thi thr 58-161 metB1 rpsL l1 Fdef 5K(pColU pSB41) tolA fhuA21 lacY1 thi leuB6 supE44 thr-1 HfrC K10C4 tolC ara D(lac pro) thi F9 lac pro GM1 tolQ GM1 tolR tolB fhuA21 lacY1 thi leuB6 supE44 thr-1 tonB aroB malT tsx thi tolQ exbB tolR exbBD ompA araD139 D(argF-lac)U169 rpsL150 relA fibB5301 deoC1 pts F25 rbsR MC4100 ompFC tsx thr ala leu proA lacY galK argE rpsL xyl thi supE mtl ompABC ompFC lamB KS26 ompA ompFC lamB Row ompF 5K ompA 5K ompA rfa KB426 ompF 5K rfa 5K rfa thi pyrD glt galK trp recA rpsL UH99 ompA F2 thi argE proA thr leu mtl xyl galK lacY rpsL supE non P400 non1 his ompB101 ompA2000 P400 ompA2005 ompA of Shigella dysenteriae ompA of Enterobacter aerogenes ompA of Serratia marcescens
G. Schrempf P. Fredericq This work 55 60 60 55 55 55 55 K. Hantke 8 10 H. J. Krieger-Brauer 13 E. Bremer 30 R. Benz R. Koebnik R. Benz This work This work This work This work This work This work 16 12 51 28 37 5 6 7
E. coli B BL21
F2 hsdS gal
54
E. coli O8:K27 F464 F539 F588
met his pro mtl Strr F2 rfa met his pro mtl Strr F2 rfa met his pro mtl Strr F2
50 50 50
Plasmids pBCSK1 pDS1 pDS2 pDS3 pDS4 pDS5 pHP90 pJBS636 pHP91
Phage T7 gene 10 promoter pBCSK1 carrying pColU in ClaI restriction site pBCSK1 carrying cua, cui, and cul in 4.5-kb HindII-ClaI fragment of pColU pBCSK1 carrying cua in 2.3-kb HindII-EcoRV fragment of pColU pBCSK1 carrying cui and cul in 2.7-kb SpeI-ClaI fragment of pColU pBCSK1 carrying cul in 1.5-kb EcoRV-EcoRV fragment of pColU pDS2; ColUD382–425 b-Lactamase fusion vector ColE1 ori, Neor pT7-7 polylinker T7 promoter pJBS636 carrying cui in 1.3-kb SspI-PstI fragment of pColU
Stratagene This work This work This work This work This work This work T. Focareta This work
Bacteriophage C21
Core LPS receptor
45
Protein nalytical procedures. Proteins were labeled radioactively in E. coli BL21 (54) with [35S]methionine (56) and were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) as described previously (40). Proteins were labeled in vitro with [35S]methionine in a bacterial cell-free transcription-translation system (Promega, Madison, Wis.). SDS-PAGE of LPS. Lipopolysaccharide (LPS) samples were prepared by using the SDS–proteinase K–whole-cell lysate method (29). Samples were run on 17% polyacrylamide gels, and then the gels were silver stained (57). Computer-assisted sequence analysis. Computer-assisted sequence analysis was performed as described previously (39). Nucleotide sequence accession number. The nucleotide sequences reported in this study were deposited in the EMBL data bank under accession no. Y11823.
RESULTS Genes on the pColU plasmid. E. coli 5K was transformed with the colicin U-determining plasmid of S. boydii M592 (serovar 8). Selection for immunity to added colicin U resulted in the isolation of transformants that carried two plasmids: a 7.3-kb plasmid designated pColU and a 4.1-kb plasmid designated pSB41. No transformants with a single plasmid were obtained. One such transformant was chosen for further studies and was designated strain U8T19. Sequencing of one DNA
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
E. coli K-12 5K Row U8T19 A592 K10C4 C65 GM1 TPS13 TPS300 A593 BR158 HE2 HE10 KB426 MC4100 HF24 HO830 KS26 KO16 KS26-2 DS2 DS1 DS3 KB426H1 HP77 HP78 UH99 UH100 P400 P530-1cII P400-2.2. UH100(l ompA-Sh) UH100(l ompA-En) UH100(l ompA-Se)
VOL. 179, 1997
COLICIN U OF S. BOYDII
4921
strand of each plasmid revealed partial sequence identity and, in addition, similarity to the pColE1 sequence. Sequence similarity between pSB41 and pColE1 was high in the rep and mob regions of pColE1 (bp 596 to 3733) (Fig. 1) (14). The sequence of the remaining 1-kb region of pSB41 was not similar to sequences of pColE1, with the exception of a 0.1-kb sequence, which was similar to that of exc from pColE1. pColU and pColE1 had high sequence similarity from bp 496 to 5140, extending from the second third of the gene for the lytic protein (see below) to the start codon of the colicin activity gene (Fig. 1). The remaining 2.6-kb pColU region did not show any sequence similarity to pColE1. The pSB41 and pColU sequences differed in the origin of replication, which may allow their coexistence in cells. Transformation of pColU alone was achieved after the plasmid was cleaved by ClaI and ligated into pBCSK1, which resulted in pDS1. The cloned 2.3-kb HindII-EcoRV fragment on pDS3 (Table 1) encoded the gene necessary for colicin U synthesis. However, the transformant carrying this fragment was not stable because, as sequencing revealed, it lacked the immunity gene. The SpeI-ClaI (2.7 kb) fragment in pDS4 conferred immunity to colicin U, which indicated the presence of a functional immunity gene on the cloned fragment. Bacteria carrying the cloned EcoRV-EcoRV fragment (1.5 kb) in pDS5
either grew or lysed, depending on the polarity of the colicin U lytic gene relative to the lacZ promoter of pBCSK1. pColU was sequenced between HindII and AluI (2.9 kb) on both strands. Three open reading frames, one each for colicin U activity (cua), immunity (cui), and lysis (cul), were identified (Fig. 2). The cui transcription polarity is opposite to that of cua and cul, which is characteristic for colicins whose target is the cytoplasmic membrane. The cua gene codes for a protein consisting of 619 amino acid residues (Mr, 66,289), cui codes for a protein of 174 residues (Mr, 20,688), and cul codes for a polypeptide of 45 residues (Mr, 4,673). The promoter region upstream of cua contains two overlapping SOS boxes, as is found on pColE1, and a highly conserved 235 region and a less conserved 210 region, as is observed in other colicin determinants. The promoter region from pDS1 was cloned into pDS2 by using HindII, which replaced the optimal 235 consensus region TTGACA with TCGACA and resulted in a considerable decrease of colicin U synthesis upon SOS induction (data not shown). SDS-PAGE of radiolabeled proteins from cells synthesizing the colicin activity and immunity proteins showed bands that corresponded to the expected size of the proteins (Fig. 3, lane 1 [65 kDa], and lane 3 [21 kDa]). pColU of another S. boydii strain was selected for compar-
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 1. Restriction map of plasmid pColU. The arrows indicate the transcriptional polarities of the cua, cui, and cul genes. Numbers correspond to the nucleotide sequence of pColE1. The positions of the genes rep, mob, cer, and exc regions are those of pColE1 (14) and are at the same sites on pColU, as revealed by the sequencing of pColU performed in this study.
4922
ˇMAJS ET AL. S
J. BACTERIOL.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 2. Nucleotide sequence of the HindII-AluI fragment (2.9 kb) of pColU. The arrows indicate the transcriptional polarities of the cua, cui, and cul genes. Putative promoter regions (210 and 235), SOS boxes, ribosome binding site sequences (S.D.), transcription termination sites (T1, T2, T3), and a predicted TolA box are indicated.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 2—Continued.
4923
4924
ˇMAJS ET AL. S
J. BACTERIOL.
FIG. 3. SDS-PAGE of radiolabeled proteins from colicinogenic (colicin U) and immune strains. The arrows indicate the positions of colicin U (lane 1), ColUD382-425 (lane 2), and the Cui immunity protein (lane 3). The genes for colicin U (on pDS3) and ColUD382-425 (on pHP90) were expressed in vitro; cui (on pHP91) was expressed in vivo. The 25-kDa bands (marked with dots) in lanes 1 and 2 correspond to the chloramphenicol transacetylase encoded on pBCSK1, and the 30-kDa band in lane 3 corresponds to the neomycin phosphotransferase encoded by pJSB636. The molecular masses (in kilodaltons) of marker proteins are indicated.
ison. The restriction maps of the pColU plasmids and the nucleotide sequences of the cua, cui, and cul genes were identical. Colicin U protein. A comparison of the colicin U amino acid sequence with those of known colicins showed the highest degree of similarity to colicin A (36.8% identity, 10.1% similarity). However, the degree of sequence similarity varied along the polypeptide chain. A dendrogram of sequence similarities of pore-forming colicins is shown in Fig. 4. Since colicins are composed of domains, the degree of sequence similarity was determined for each domain. The 200 C-terminal amino acids of colicin U exhibited the greatest sequence similarity to the corresponding region of colicin B (73.0% identity,
FIG. 4. Dendrogram of the sequence alignment of the pore-forming colicins.
12.5% similarity) (Fig. 5); they had 65.5% identity to the poreforming domain of colicin A and 49.0% identity to the bacteriocidal domain of colicin N. Therefore, colicin U belongs to the group of pore-forming colicins (2, 15), which are characterized by a long hydrophobic sequence near the C terminus (38). This hydrophobic sequence, extending from residue 571 to 599, displays the lowest sequence similarity to the corresponding region of colicin B (Fig. 5) and therefore may determine the specificity of inactivation of colicin U by its cognate immunity protein. Colicins U and B displayed no cross-immunity. The N-terminal sequence of colicin U shows the highest sequence similarity to bacteriocin 28b (Fig. 6), with 54.2% identity and an additional 16.0% similarity in the 130 N-terminal residues (27). A high degree of identity of the N-terminal sequences of colicins U and A (83.3%) was confined to the first 36 amino acid residues. The putative receptor binding domain in the central region of colicin U revealed only low sequence similarity to that of other colicins. Nevertheless, significant sequence similarity between colicin U and colicin K (Fig. 7), which was restricted to a region ranging from residue 280 to 386 (34.6% identity), was observed. A higher sequence similarity (restricted to 44 residues) between colicin U and the TolA protein (Fig. 8), extending from residue 382 to 425 of the colicin U molecule (68.2% identity), was found. Short KAAAG stretches homologous to the KAAAD/E sequence, repeated in TolA, is contained once and only in colicin U. KA-rich regions occur also in other colicins. Immunity protein of colicin U. The immunity protein of colicin U is homologous to the immunity proteins of poreforming colicins, with the highest sequence similarity to colicin A (identity, 45.4%; similarity, 14.9%) and lower sequence sim-
FIG. 6. Sequence comparison of the N-terminal 130 amino acids of colicin U with the corresponding region of bacteriocin 28b and with the 36 amino acids of colicin A. Asterisks denote identical residues; dots indicate similar amino acids.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 5. Sequence comparison of the C-terminal 200 amino acids of colicins U and B. Asterisks denote identical residues; dots indicate similar amino acids. a1 to a10 indicate a-helices of colicin B postulated by sequence comparison with colicin A (25), for which an X-ray structure of the pore-forming domain in aqueous solution has been determined (38).
VOL. 179, 1997
COLICIN U OF S. BOYDII
4925
TABLE 2. Sensitivities of ompA, ompF, and rfa mutants to colicins U, A, L, K, and N E. coli straina
FIG. 7. Sequence similarity of the central 107 amino acids of colicins U and K. Asterisks denote identical residues; dots indicate similar amino acids.
FIG. 8. Sequence similarity of the central 44 amino acids of colicin U and the TolA protein. Asterisks denote identical residues; dots indicate similar amino acids.
Wild type ompA ompF rfa ompA rfa Wild type rfa rfa
U
A
L
K
N
6 3 5 5 r 5 4 4
5 5 1 5 5 — — —
4 — — 4 — 3 2 1
3 2 2 3 2 — — —
4 4 2 5 4 — — —
a Strains DS1, DS2, DS3, and HP77 were isolated from the colicin U inhibition zone on plates seeded with E. coli K-12 5K. E. coli F464 and its derivatives F539 and F588 arose from E. coli O8:K27 (50). b The numbers indicate the highest colicin dilutions active on bacteria (e.g., 5 5 105). r, resistant; —, not determined.
sitivity, and an ompA rfa mutant (DS3) was resistant to colicin U (Table 2). The killing by colicins U, K, and L of bacterial strains from various genera of the family Enterobacteriaceae that synthesize OmpA and of E. coli strains that synthesize OmpA proteins with defined point mutations is shown in Table 3. Colicin U displayed the same pattern of activity as colicin K, which indicated that the two colicins recognize a similar, if not identical, region of OmpA. In contrast, the sensitivity of these strains to colicin L differed from the sensitivity to colicin U and K (Table 2). Binding of colicin U to ompA, ompF, and rfa mutants. Mutations in ompA, ompF, and rfa could affect binding of colicin U to cells, uptake of colicin U, or both. Therefore, we tested the adsorption of colicin U to several of the mutants that displayed a reduced sensitivity to colicin U (Table 4). Adsorption decreased proportionally to the sensitivity of the various cell types. Presumably, colicin U can bind to OmpA, OmpF, and the outer core of the LPS. However, even the colicin U-resistant ompA rfa mutant adsorbed 26% of the amount of colicin U that adsorbed to the wild type, which probably reflects unspecific binding. Uptake of colicin U. To examine whether colicin U is taken up by the Ton or the Tol system, we tested the colicin U sensitivities of various mutants (Table 5). A tolA mutant was insensitive to colicin U, while a tonB mutant was unaffected. The Tol proteins TolA, TolQ, TolR, and TolB were required for colicin U uptake, whereas TolC was not. The residual susceptibility of the tolQ and tolR mutants came from the partial replacement of their activities by the ExbB and ExbD
FIG. 9. SDS-PAGE of outer membrane LPSs from rfa mutant strains isolated from the inhibition zone of colicin U on plates seeded with E. coli 5K. Lanes: 1, E. coli 5K; 2, E. coli HP77 (rfa); 3, E. coli HP78 (rfa).
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
ilarity to colicin B (identity, 30.5%). The colicin U producer strain was partially immune to colicins A and N (10- to 100-fold and 10-fold reduction of sensitivity, respectively), but the colicin A and N producer strains were fully sensitive to colicin U (data not shown). Lytic protein of colicin U. The colicin U lytic protein (4.7 kDa) showed the highest sequence similarity to the lytic proteins of colicins 5 and K (76.7% identity), of colicin 10 (72.1% identity), and of cloacin DF13 and the E-type colicins (62.2 to 71.1% identity). The mature proteins display the highest sequence similarity after residue 16, which is the cleavage site for lipoprotein signal sequences. Complex requirement of outer membrane determinants for colicin U entry into susceptible cells. Mutants of E. coli K-12 insensitive to colicin U were isolated from growth inhibition zones on plates onto which colicin U at a dilution titer of 106 was applied. Insensitive mutants isolated from the center of the inhibition zones, which contained the highest colicin concentration, were usually mutated in tol genes, as identified by their insensitivity to colicin E1, K, or L at the highest colicin concentrations available. Colonies isolated from the edge of the inhibition zone displayed a lower colicin U insensitivity (Table 2). These colonies carried mutations in ompA or ompF, as revealed by SDS-PAGE, which demonstrated the lack of OmpA or OmpF (data not shown). The patterns of resistance to colicin U of the mutants showed some overlap with the patterns of resistance to colicins A, K, L, and N, but no two mutants had identical resistance patterns (Table 2). The colicin U sensitivity of multiple mutants from our strain collection was also tested. The sensitivity of strain HF24 ompF ompC was reduced 100-fold, that of strain KB426H1 ompA ompF was reduced 103-fold, and that of strain HO830 ompA ompF ompC was reduced 105-fold. Strain KS26-2 ompC ompF lamB and strain KO16 ompA ompF ompC lamB were fully resistant. The sensitivity of the porin mutants to colicin B remained unchanged, which indicated that resistance to colicin U did not result from a general disturbance of outer membrane integrity. In addition, rfa mutants, which do not synthesize complete LPS but synthesize core LPS, were isolated from the colicin U inhibition zone, as shown by their sensitivity to bacteriophage C21 (45), to which the parent strain was resistant, and by silver staining after SDS-PAGE of isolated LPS (Fig. 9). An rfa mutant (HP77) had a 10-fold-reduced sensitivity to colicin U, an ompA mutant (DS1) displayed a 103-fold reduction in sen-
5K DS1 DS2 HP77 DS3 F464 F539 F588
Sensitivityb to colicin: Plasmid genotype
4926
ˇMAJS ET AL. S
J. BACTERIOL.
TABLE 3. Sensitivities to colicins U, K, and L of E. coli K-12 strains that synthesize different OmpA proteins Strain
Relevant phenotype
E. coli K-12 5K UH99 P400-2.2 P530-1cII UH100(l ompA-Sh) UH100(l ompA-En) UH100(l ompA-Se)
Wild type Wild type OmpA (E68K) OmpA (V110D) OmpA from Shigella dysenteriae OmpA from Enterobacter aerogenes OmpA from Serratia marcescens
TABLE 5. Sensitivity of tol and tonB mutants to colicin U
Sensitivitya to colicin: U
K
L
3 3 3 2 3 1 0
3 3 3 2 3 1 0
3 3 1 0 0 3 r
a
proteins (8, 10), as shown by the complete insensitivity of tolQ exbB and tolR exbB exbD mutants. Deletion of 44 amino acid residues in the colicin U protein. To estimate the biological function of the region homologous to the TolA protein, the 44 amino acids of this region in colicin U were deleted, resulting in the production of ColUD382-425, which had the size expected (Fig. 3, lane 2). ColUD382-425 was as active as wild-type colicin U on E. coli 5K, and its toxicity to omp, rfa, and tol mutant strains was decreased to the same extent as for wild-type colicin U. A colicin U-producing strain was as immune to ColUD382-425 as to wild-type colicin U (data not shown). DISCUSSION Colicin U is an addition to the list of bacteriocins (S1, S4, and Js) found in Shigella strains (23, 52). The U colicinogenic serovars 1 and 8 of S. boydii display cross-immunity to each other and to the colicinogenic strains S. boydii Shp 132/56 (serovar 8) and S. boydii Shp 166/58 (serovar 1). Shigella is very similar to E. coli, and plasmids can easily be exchanged between the two genera, as has been found for Shigella sonnei, in which 70 colicin types have been identified, most of which are also found in E. coli (32). The ColU plasmids of the two S. boydii serovars investigated are identical with regard to the restriction map and the nucleotide sequence of the region that encodes the cua, cui, and cul genes. pColU (7.3 kb) and pSB41 (4.1 kb) coexist in all four S. boydii strains. In addition, pColU could not be transformed into E. coli 5K without pSB41. These findings suggest that pSB41 may be required for replication of pColU or for another function important for pColU maintenance, such as partitioning among progeny cells. The high sequence similarity of the
TABLE 4. Binding of colicin U to E. coli 5K ompA, ompF, and rfa mutants Strain
Relevant genotype
Adsorption (%)a
Sensitivity to colicin Ub
5K Row HP77 DS2 DS1 DS3
Wild type Wild type rfa ompF ompA ompA rfa
100 6 11 100 6 13 74 6 7 72 6 14 52 6 4 26 6 12
6 6 5 5 3 r
a
95% confidence interval for the mean. The numbers indicate the highest colicin dilutions active on bacteria (e.g., 6 5 106). r, resistant. b
5K BR158 C65 A592 TPS13 TPS300 A593 HE2 HE10
Relevant genotype
Wild type tonB tolC tolA tolQ tolR tolB tolQ exbB tolR exbBD
Sensitivity to colicin Ua
6 6 6 t 1 2 3 t t
a The numbers indicate the highest colicin dilutions active on bacteria (e.g., 6 5 106). t, tolerant.
plasmids in the region outside the colicin genes of pColU suggests that a pSB41-like plasmid acquired the colicin determinant, thereby forming pColU. pColU and pColE1 are homologous in the region extending from the second third of the cul gene to the start codon of the cua gene, which covers a total length of 4.7 kb. The plasmids differ entirely in the region that determines the colicin genes. In pCol10, homology with pColE1 is found in the region coding for the lytic protein and in the sequence upstream of the start codon of the colicin activity gene. The homologous regions presumably serve as sites of recombination between the colicin plasmids. In the colicin plasmids pCol5, pCol10, pColK, and pColE1 (39–41) and the pesticin plasmid pPCP1 (42), homologous DNA fragments precisely determine functional domains, which suggests that these colicins evolved by exchange of DNA fragments. Sequence comparison of colicin U with other colicins also revealed sequence similarities of functional domains, but these were less pronounced than in the examples described above. The high sequence similarity of the C-terminal domain of colicin U to the corresponding regions of colicin B (73% identity), colicin A (65.5% identity) (26), and colicin N (49% identity) assigns colicin U to the pore-forming colicins. The high sequence similarity is interrupted between residues 571 and 599; this region may be the specific recognition site of the colicin U immunity protein, which does not confer immunity to colicin B. In analogy to the X-ray structure of the pore-forming domain of colicin A, the region from residue 571 to 599 is located where the hydrophobic helix 8 extends into helix 9 (Fig. 5) (38, 53). Uptake of colicin U and bacteriocin 28b is dependent on the tolA, -B, -Q, and -R genes. The N termini of the colicins have high sequence similarity (Fig. 6). Colicin U and colicin A contain the sequence DGTGWS, and bacteriocin 28 contains the sequence DGTNWS. These sequences are highly similar to the previously proposed TolA box DGSGS (39), which may be responsible for interaction of the translocation domain of group A colicins with TolA proteins. Colicin U exhibits the complex dependence on several integral outer membrane proteins that is typical for group A colicins. It requires OmpA, OmpF, and core LPS and in this regard is similar to the sequenced bacteriocin 28b (20), which is similar if not identical to colicin L used in this study for comparison (27). A rather long stretch of amino acids close to the N terminus of colicin U, extending beyond the proposed TolA box, is homologous to that region of bacteriocin 28b and may confer the requirement for OmpF and core LPS (Fig. 10). An OmpF interaction sequence has been localized to the N terminus of colicin A (3), but this sequence is not similar to the
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
The numbers indicate the highest colicin dilutions active on bacteria (e.g., 5 5 105). r, resistant.
Strain
VOL. 179, 1997
COLICIN U OF S. BOYDII
4927
assumed OmpF interaction sequence of colicin U and bacteriocin 28. Either the colicins interact with OmpF differently or the homologous sequence in colicin U and bacteriocin 28 specifies only interaction with the core LPS. The homologous sequence of 107 amino acids in the central part of colicins U and K may determine the identical requirement for OmpA, as revealed by the use of ompA mutants and OmpA proteins from different species. None of the single deletions in ompA and ompF conferred complete insensitivity to colicins U, A, L, K, and N; this supports previous findings that several outer membrane proteins are involved in the uptake of group A colicins, in contrast to group B colicins, for which deletion of single outer membrane proteins confers complete resistance. The pore-forming group A colicins remain bound to the outer membrane receptors while they insert into the cytoplasmic membrane (4). This also seems to apply to group B poreforming colicins, since the three-dimensional X-ray structure of colicin Ia reveals two very long a-helices, one of which connects the receptor binding domain with the pore-forming domain (61) and thus serves to bridge the periplasmic space. A very long a-helix in the TolA protein, required for the uptake of group A colicins, was proposed (35); this helix easily spans the periplasm from the cytoplasmic membrane to the outer membrane. Excision of a fragment extending from residue 68 to 209 of TolA does not inactivate the protein (19). Forty-four amino acids of colicin U showed a high sequence similarity to the central domain of the TolA protein; these amino acids could be excised (forming ColUD382-425) without inactivating colicin U. We did not find any changes in ColUD382-425 interaction with sensitive, tolerant, resistant, or immune bacterial cells, as compared with that of wild-type colicin U. Deletion of 35 amino acids in colicin A between residues 337 and 371 also caused no changes in the biological activity of colicin A (1). A repetitive Pro-X motif in the TonB protein is suggested to be involved in bridging the periplasmic space. Excision of the Pro-X motif (residues 66 to 100) in TonB slightly reduces TonB activity only under hyperosmolar conditions in which the periplasmic space is expanded (34). Although rigid structures designed to span the periplasmic space may be important for the action of certain colicins, TolA, and TonB, they are apparently not essential. It seems that there are sites of shorter distances between the outer membrane and cytoplas-
mic membrane, presumably fusion sites, through which TolAand TonB-dependent colicins can enter cells, and TonB-dependent infection by certain phages and uptake of ferric siderophores and vitamin B12 might occur. Colicin U-insensitive derivatives of E. coli K-12 that were not mutated in the tol genes were mutated in ompA, ompF, and rfa. Colicin U-resistant ompC and lamB mutants were not isolated in our experiments, even though existing mutants with these genotypes displayed a reduced sensitivity to colicin U. The porins can functionally substitute for each other, as shown by E. coli KS26-2 ompC ompF lamB, which is resistant to colicins U, L, A, K, and N. Binding of colicin N to the porins OmpF, OmpC, and PhoE has been demonstrated by microcalorimetry (22). The very low sensitivity of ompA deletion mutants to colicin U indicates a major role of OmpA in colicin U entry. The immunity protein of colicin U (Cui) is 45% identical to that of colicin A (21), which explains the partial immunity of U colicinogenic strains to colicin A. Although the sequence similarity between Cui and the colicin N immunity protein is much lower, Cui confers some immunity to colicin N. The immunity proteins of colicins U, B, A, and N exhibit a much higher sequence diversity than the pore-forming domains of the colicins with which they interact. The immunity proteins have to recognize the small differences between the pore-forming regions, implying selection for difference rather than similarity. ACKNOWLEDGMENTS D.S. thanks the Deutsche Akademische Austauschdienst (DAAD) for a short-term fellowship and the Deutsche Forschungsgemeinschaft (SFB 323) for financial support. ˇmarda for encouragement, advice, and reading of the D.S. thanks J. S manuscript. We are grateful to V. Hora´k for providing the S. boydii strains, and we thank K. Hantke, A. Angerer, and H. Killman for helpful discussions and K. A. Brune for critical reading of the manuscript. REFERENCES 1. Baty, D., R. Frenette, R. Lloubes, V. Geli, S. P. Howard, F. Pattus, and C. Lazdunski. 1988. Functional domains of colicin A. Mol. Microbiol. 2:807–811. 2. Benedetti, H., and V. Geli. 1996. Colicin transport, channel formation and inhibition, p. 665–691. In W. N. Konings, H. R. Kaback, and J. S. Lolkema (ed.), Handbook of biological physics, vol. 2. Elsevier Sciences, Amsterdam, The Netherlands. 3. Benedetti, H., N. Frenette, M. Baty, M. Knibiehler, F. Pattus, and C. Lazdunski. 1991. Individual domains of colicins confer specificity in colicin
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
FIG. 10. Tentative assignment of structural regions to functional domains of the colicins indicated. Bacteriocin 28b is very similar if not identical to colicin L. In the colicin U mutant D382-425, residues 382 to 425 are deleted. aa, amino acids.
4928
4. 5. 6. 7. 8.
10. 11. 12.
13. 14. 15. 16. 17. 18. 19. 20.
21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
uptake, in pore-properties and in immunity requirement. J. Mol. Biol. 217: 429–439. Benedetti, H., R. Lloubes, C. Lazdunski, and L. Letellier. 1992. Colicin A unfolds during its translocation in Escherichia coli cells and spans the whole cell envelope when its pore has formed. EMBO J. 11:441–447. Braun, G., and S. T. Cole. 1982. The nucleotide sequence coding for major outer membrane protein OmpA of Shigella dysenteriae. Nucleic Acids Res. 10:2367–2378. Braun, G., and S. T. Cole. 1983. Molecular characterization of the gene coding for major outer membrane protein OmpA from Enterobacter aerogenes. Eur. J. Biochem. 137:495–500. Braun, G., and S. T. Cole. 1984. DNA sequence analysis of the Serratia marcescens ompA gene: implication for the organisation of an enterobacterial outer membrane protein. Mol. Gen. Genet. 195:321–328. Braun, V. 1989. The structurally related exbB and tolQ genes are interchangeable in conferring tonB-dependent colicin, bacteriophage, and albomycin sensitivity. J. Bacteriol. 171:6387–6390. Braun, V. 1995. Energy-coupled transport and signal transduction through the gram-negative outer membrane via the TonB-ExbB-ExbD dependent receptor proteins. FEMS Microbiol. Rev. 16:295–307. Braun, V., and C. Herrmann. 1993. Evolutionary relationship of uptake systems for biopolymers in Escherichia coli: cross-complementation between the TonBExbB-ExbD and the TolA-TolQ-TolR proteins. Mol. Microbiol. 8:261–268. Braun, V., H. Pilsl, and P. Gross. 1994. Colicins: structures, modes of action, transfer through membranes, and evolution. Arch. Microbiol. 161:199–206. Bremer, E., E. Beck, I. Hindennach, I. Sonntag, and U. Henning. 1980. Cloned structural gene (ompA) for an integral outer membrane protein of Escherichia coli K-12. Localization on hybrid plasmid pTU100. Mol. Gen. Genet. 179:13–20. Casabadan, M. J., and S. N. Cohen. 1979. Lactose genes fused to exogenous promoters in one step using a Mu lac bacteriophage: in vivo probe for transcriptional control sequence. Proc. Natl. Acad. Sci. USA 176:4530–4533. Chan, P. T., H. Ohmori, J. Tomizawa, and J. Lebowitz. 1984. Nucleotide sequence and gene organization of ColE1 DNA. J. Biol. Chem. 260:8925–8935. Cramer, W. A., J. B. Heymann, S. L. Schendel, B. N. Deriy, F. S. Cohen, P. A. Elkins, and C. V. Stauffacher. 1995. Structure-function of the channel-forming colicins. Annu. Rev. Biophys. Biomol. Struct. 24:611–641. Datta, D. B., B. Arden, and U. Henning. 1977. Major proteins of the Escherichia coli outer cell envelope membrane as bacteriophage receptors. J. Bacteriol. 131:821–829. Davies, J. K., and P. Reeves. 1975. Genetics of resistance to colicins in Escherichia coli K-12: cross-resistance among colicins of group B. J. Bacteriol. 123:96–101. Davies, J. K., and P. Reeves. 1975. Genetics of resistance to colicins in Escherichia coli K-12: cross-resistance among colicins of group A. J. Bacteriol. 123:102–117. Derouiche, R., M. Gavioli, H. Benedetti, A. Prilipov, C. Lazdunski, and R. Lloubes. 1996. TolA central domain interacts with Escherichia coli porins. EMBO J. 15:6408–6415. Enfedaque, J., S. Ferre, J. F. Guasch, J. Tomas, and M. Regue´. 1996. Bacteriocin 28b from Serratia marcescens N28b: identification of Escherichia coli surface involved in bacteriocin binding and translocation. Can. J. Microbiol. 42:19–26. Espesset, D., P. Piet, C. Lazdunski, V. Geli. 1994. Immunity proteins to pore-forming colicins: structure-function relationships. Mol. Microbiol. 13: 1111–1120. Evans, L. J. A., A. Cooper, and J. H. Lakey. 1996. Direct measurement of the association of a protein with a family of membrane receptors. J. Mol. Biol. 255:559–563. Fredericq, P. 1948. Action antibiotiques reciproques chez les Enterobacteriaceae. Rev. Belge Pathol. Exp. Med. 19 Suppl. 4:1–107. Fredericq, P. 1965. A note on the classifications of colicins. Zentralbl. Bakteriol. Hyg. A 196:140–142. Ge´li, V., and C. Lazdunski. 1992. An a-helical hydrophobic hairpin as a specific determinant in protein-protein interaction occurring in Escherichia coli colicin A and B immunity systems. J. Bacteriol. 174:6432–6437. Ge´li, V., D. Baty, F. Pattus, and C. Lazdunski. 1989. Topology and function of the integral membrane protein conferring immunity to colicin A. Mol. Microbiol. 3:679–687. Guasch, J. F., E. Enfedaque, S. Ferrer, D. Gargallo, and M. Regue´. 1995. Bacteriocin 28b, a chromosomally encoded bacteriocin produced by most Serratia marcescens biotypes. Res. Microbiol. 146:447–483. Henning, U., I. Hindennach, and I. Haller. 1976. The major proteins of the Escherichia coli outer cell envelope membrane: evidence for the structural gene of protein II*. FEBS Lett. 61:46–48. Hichcock, P. J., and T. M. Brown. 1983. Morphological heterogeneity among Salmonella lipopolysaccharide chemotypes in silver-stained polyacrylamide gels. J. Bacteriol. 154:269–277. Hoffmann, H., E. Fischer, H. Kraut, and V. Braun. 1986. Preparation of the FhuA (TonA) receptor protein from cell envelopes of an overproducing
J. BACTERIOL. strain of Escherichia coli K-12. J. Bacteriol. 166:404–411. 31. Hora ´k, V. 1992. Personal communication. 32. Hora ´k, V. 1994. Seventy colicin types of Shigella sonnei and an indicator system for their determination. Zentralbl. Bakteriol. Hyg. A 281:24–29. 33. Killmann, H., R. Benz, and V. Braun. 1993. Conversion of the FhuA transport protein into a diffusion channel through the outer membrane of Escherichia coli. EMBO J. 12:3007–3016. 34. Larsen, R. A., G. E. Wood, and K. Postle. 1993. The conserved proline-rich motif is not essential for energy transduction by Escherichia coli TonB protein. Mol. Microbiol. 10:943–953. 35. Levingood, S. K., W. B. Beyer, and R. E. Webster. 1991. TolA: a membrane protein involved in colicin uptake contains an extended helical region. Proc. Natl. Acad. Sci. USA 88:5939–5943. 36. Mayr-Harting, A., A. J. Hedges, and R. C. W. Berkeley. 1972. Methods for studying bacteriocins. Methods Microbiol. 7A:315–442. 37. Morona, R., M. Klose, and U. Henning. 1984. Escherichia coli K-12 outer membrane protein (OmpA) as a bacteriophage receptor: analysis of mutant genes expressing altered proteins. J. Bacteriol. 159:570–578. 38. Parker, M. W., J. P. Postma, F. Pattus, A. D. Tucker, and D. Tsernoglou. 1992. Refined structure of the pore-forming domain of colicin A at 2.4 Å resolution. J. Mol. Biol. 224:639–657. 39. Pilsl, H., and V. Braun. 1995. Novel colicin 10: assignment of four domains to TonB- and TolC-dependent uptake via the Tsx receptor and to pore formation. Mol. Microbiol. 16:57–67. 40. Pilsl, H., and V. Braun. 1995. Evidence that the immunity protein inactivates colicin 5 immediately prior formation of the transmembrane channel. J. Bacteriol. 177:6966–6972. 41. Pilsl, H., and V. Braun. 1995. Strong function-related homology between the pore-forming colicins K and 5. J. Bacteriol. 177:6973–6977. 42. Pilsl, H., H. Killmann, K. Hantke, and V. Braun. 1996. Periplasmic location of the pesticin immunity protein suggests inactivation of pesticin in the periplasm. J. Bacteriol. 178:2431–2435. 43. Pugsley, A. P. 1984. The ins and outs of colicins. I. Production, and translocation across membranes. Microbiol. Sci. 1:168–175. 44. Pugsley, A. P. 1984. The ins and outs of colicins. II. Lethal action, immunity and ecological implications. Microbiol. Sci. 1:203–205. 45. Rapin, A. M. C., and H. M. Kalckar. 1971. The relation of bacteriophage attachment to lipopolysaccharide structure, p. 267–307. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial toxins, vol. 4. Academic Press, New York, N.Y. 46. Reeves, P. 1965. The bacteriocins. Bacteriol. Rev. 29:24–45. 47. Roos, U., R. E. Harkness, and V. Braun. 1989. Assembly of colicin genes from a few DNA fragments. Nucleotide sequence of colicin D. Mol. Microbiol. 3:891–902. 48. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. 49. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. 50. Schmidt, G., B. Jann, and K. Jann. 1970. Immunochemistry of R lipopolysaccharides of Escherichia coli. Eur. J. Biochem. 16:382–392. 51. Skurray, R. E., R. E. W. Hancock, and P. Reeves. 1974. Con2 mutants: class of mutants in Escherichia coli K-12 lacking a major cell wall protein and defective in conjugation and adsorption of a bacteriophage. J. Bacteriol. 119:726–735. ˇ marda, J., J. Petrzelova 52. S ´, and B. Vyskot. 1987. Colicin Js of Shigella sonnei: classification of type colicin “7”. Zentralbl. Bakteriol. Hyg. A 263:530–540. 53. Song, H. Y., and W. A. Cramer. 1991. Membrane topography of ColE1 gene products: the hydrophobic anchor of the colicin E1 channel is a helical hairpin. J. Bacteriol. 173:2927–2934. 54. Studier, F. W., and B. A. Moffat. 1986. Use of bacteriophage T7-RNApolymerase to direct selective high level expression of cloned genes. J. Mol. Biol. 189:113–130. 55. Sun, T. P., and R. E. Webster. 1987. Nucleotide sequence of a gene cluster involved in entry of E colicins and single-stranded DNA of infecting filamentous bacteriophages into Escherichia coli. J. Bacteriol. 169:2667–2674. 56. Tabor, S., and C. C. Richardson. 1985. A bacteriophage RNA polymerase/ promoter system for controlled exclusive expression of specific genes. Proc. Natl. Acad. Sci. USA 82:1074–1078. 57. Tsai, G. M., and C. E. Frasch. 1982. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119:115–119. 58. Van Horn, E. A. 1961. The assay of colicin as a special case of a slowly diffusing substance and the application of the assay methods to the study of the colicin production. Thesis. University of Bristol, Bristol, United Kingdom. 59. Viejo, M. B., D. Gargallo, S. Ferrer, J. Enfedaque, and M. Regue´. 1992. Cloning and DNA sequence analysis of a bacteriocin gene from Serratia marcescens. J. Gen. Microbiol. 138:1737–1743. 60. Whitney, E. N. 1971. The tolC locus in Escherichia coli K12. Genetics 67:39–53. 61. Wiener, M., D. Freymann, P. Ghosh, and R. M. Stroud. 1997. Crystal structure of colicin Ia. Nature 385:461–464.
Downloaded from http://jb.asm.org/ on March 1, 2013 by PENN STATE UNIV
9.
ˇMAJS ET AL. S
