Influence of different rol gene products on the chain length of Shigella dysenteriae type 1 lipopolysaccharide O antigen expressed by Shigella flexneri carrier strains.

Introduction of the rol genes of Shigella dysenteriae 1 and Escherichia coli K-12 into Shigella flexneri carrier strains expressing the heterologous S. dysenteriae type 1 lipopolysaccharide resulted in the formation of longer chains of S. dysenteriae 1 O antigen. In bacteria producing both homologous and heterologous O antigen, this resulted in a reduction of the masking of heterologous O antigen by homologous lipopolysaccharide and an increased immune response induced by intraperitoneal immunization of mice by recombinant bacteria. The rol genes of S. dysenteriae 1 and E. coli K-12 were sequenced, and their gene products were compared with the S. flexneri Rol protein. The primary sequence of S. flexneri Rol differs from both E. coli K-12 and S. dysenteriae 1 Rol proteins only at positions 267 and 270, which suggests that this region may be responsible for the difference in biological activities.     

I-CeuI fragment analysis of the Shigella species: evidence for large-scale chromosome rearrangement in S. dysenteriae and S. flexneri

I-CeuI fragments of four Shigella species were analyzed to investigate their taxonomic distance from Escherichia coli and to collect substantiated evidence of their genetic relatedness because their ribosomal RNA sequences and similarity values of their chromosomal DNA/DNA hybridization had proved their taxonomic identity. I-CeuI digestion of genomic DNAs yielded seven fragments in every species, indicating that all the Shigella species contained seven sets of ribosome RNA operons. To determine the fragment identities, seven genes were selected from each I-CeuI fragment of E. coli strain K-12 and used as hybridization probes. Among the four Shigella species, S. boydii and S. sonnei showed hybridization patterns similar to those observed for E. coli strains; each gene probe hybridized to the I-CeuI fragments with sizes similar to that of the corresponding E. coli fragment. In contrast, S. dysenteriae and S. flexneri showed distinct patterns; rcsF and rbsR genes that located on different I-CeuI fragments in E. coli, fragments D and E, were found to co-locate on a fragment. Further analysis using an additional three genes that located on fragment D in K-12 revealed that some chromosome rearrangements involving the fragments corresponding to fragments D and E of K-12 took place in S. dysenteriae and S. flexneri.     

Possible mechanisms underlying the slow lactose fermentation phenotype in Shigella spp.

A Southern hybridization analysis revealed that the region homologous to Escherichia coli lacZ was present on the chromosomal DNAs of beta-galactosidase-positive Shigella strains, such as Shigella dysenteriae serovar 1 and Shigella sonnei strains, whereas this region was absent from chromosomal DNAs of beta-galactosidase-negative strains of Shigella flexneri and Shigella boydii. We found that the lacY-A region was deficient in S. dysenteriae serovar 1 and believe that this is the reason for the slow fermentation of lactose by this strain. S. sonnei strains possessed the region which hybridized with E. coli lacY-A despite their slow hydrolysis of lactose. The whole lactose-fermenting region was cloned from S. sonnei and compared with the cloned lac operon of E. coli K-12. Both clones directed the synthesis of beta-galactosidase in an E. coli K-12 strain lacking indigenous beta-galactosidase activity (strain JM109-1), and we observed no difference in the expression of beta-galactosidase activity in S. sonnei and E. coli. However, E. coli JM109-1 harboring the lactose-fermenting genes of S. sonnei exhibited the slow lactose fermentation phenotype like the parental strain. S. sonnei strains had no detectable lactose permease activities. E. coli JM109-1 harboring the lactose-fermenting genes of S. sonnei had a detectable permease activity, possibly because of the multicopy nature of the cloned genes, but this permease activity was much lower than that of strain JM109-1 harboring the lac operon of E. coli K-12. From these results we concluded that slow lactose fermentation by S. sonnei is due to weak lactose permease activity.     

Possible mechanisms underlying the slow lactose fermentation phenotype in Shigella spp.

A Southern hybridization analysis revealed that the region homologous to Escherichia coli lacZ was present on the chromosomal DNAs of beta-galactosidase-positive Shigella strains, such as Shigella dysenteriae serovar 1 and Shigella sonnei strains, whereas this region was absent from chromosomal DNAs of beta-galactosidase-negative strains of Shigella flexneri and Shigella boydii. We found that the lacY-A region was deficient in S. dysenteriae serovar 1 and believe that this is the reason for the slow fermentation of lactose by this strain. S. sonnei strains possessed the region which hybridized with E. coli lacY-A despite their slow hydrolysis of lactose. The whole lactose-fermenting region was cloned from S. sonnei and compared with the cloned lac operon of E. coli K-12. Both clones directed the synthesis of beta-galactosidase in an E. coli K-12 strain lacking indigenous beta-galactosidase activity (strain JM109-1), and we observed no difference in the expression of beta-galactosidase activity in S. sonnei and E. coli. However, E. coli JM109-1 harboring the lactose-fermenting genes of S. sonnei exhibited the slow lactose fermentation phenotype like the parental strain. S. sonnei strains had no detectable lactose permease activities. E. coli JM109-1 harboring the lactose-fermenting genes of S. sonnei had a detectable permease activity, possibly because of the multicopy nature of the cloned genes, but this permease activity was much lower than that of strain JM109-1 harboring the lac operon of E. coli K-12. From these results we concluded that slow lactose fermentation by S. sonnei is due to weak lactose permease activity.     

The expression of lipopolysaccharide by strains of Shigella dysenteriae, Shigella flexneri and Shigella boydii and their cross-reacting strains of Escherichia coli

Strains of Shigella dysenteriae, Shigella flexneri and Shigella boydii express lipopolysaccharides, that enable the serotyping of strains based on their antigenic structures. Certain strains of S. dysenteriae, S. flexneri and S. boydii are known to share epitopes with strains of Escherichia coli ; however, the lipopolysaccharide profiles of the cross-reacting organisms have not been compared by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) lipopolysaccharides profiling. In the present study, type strains of these bacteria were examined using SDS-PAGE/silver staining to compare their respective lipopolysaccharide profiles. Strains of S. dysenteriae, S. boydii and S. flexneri all expressed long-chain lipopolysaccharide, with distinct profile patterns. The majority of strains of Shigella spp., known to cross-react with strains of E. coli , had lipopolysaccharide profiles quite distinct from the respective strain of E. coli . It was concluded that while cross-reacting strains of Shigella spp. and E. coli may express shared lipopolysaccharide epitopes, their lipopolysaccharide structures are not identical.     

Production of Shiga toxin and a cytolethal distending toxin (CLDT) by serogroups of Shigella spp.

Abstract 56 strains of Shigella including 12 Shigella dysenteriae (serotypes 1, 2, 9, 11 and 12), 23 Shigella flexneri (serotypes 1, 2, 3, 4, 6, var. X and var. Y), 19 Shigella boydii (serotypes 1, 2, 4, 5, 7, 11, 13, 14, 15 and 18), and 2 Shigella sonnei were screened for their ability to produce both classic Shiga toxin and a new heat-labile cytolethal distending toxin (CLDT). Whereas extracellular Shiga toxin was only detectable in filtrates of five S. dysenteriae type 1 strains, CLDT was produced by four strains of S. dysenteriae type 2 and an isolate of S. boydii type 7. No cytotonic enterotoxins similar to Escherichia coli LT were observed in this study. None of the S. flexneri or S. sonnei isolates tested were found to produce extracellular cytotoxic factors. The Shiga toxin produced by the S. dysenteriae type 1 was neutralizable by anti-toxin to verotoxin 1 of E. coli O157 : H7. The Shigella CLDT was neutralizable by antisera prepared to a CLDT-producing E. coli O55 : H4.     

Gene transfer in enteric bacteria through the formation of R-prime plasmids by an RP4: :mini-Mu element.

Gene transfer in seven pathogenic enteric bacteria was studied using an RP4: :mini-Mu element, the plasmid pULB113. From the E. coli K-12 host strain the plasmid could be efficiently transferred to these enteric bacteria, but its transfer back to E. coli K-12 was not as efficient, being detected only in Shigella dysenteriae 1, S. flexneri and the 'smooth' variant of S. sonnei. In these three species, transposition of chromosomal fragments into the plasmid to produce R-prime plasmid was also detected at a frequency of approximately 10(-5). Transposition was random as suggested by the recovery at approximately the same frequency (10(-5) to 10(-6)) of R-primes involving 20 different auxotrophic markers from widely separated chromosomal locations. Formation of R-prime plasmids expressing toxicity in the E. coli K-12 recipient strain was also efficient in S. dysenteriae 1 but the toxin-activity was rapidly lost from these R-primes. In our experiments, the plasmid pULB113 incorporated relatively small amounts of chromosomal DNA as determined by restriction endonuclease digestion. For a Thy+ R-prime that we analyzed, the amount of cloned DNA was approximately 15 kb.     

Molecular cloning and genetic analysis of the rfb region from Shigella flexneri type 6 in Escherichia coli K-12

The rfb gene cluster which determines the biosynthesis of the Shigella flexneri serotype 6 O-antigen specificity has been cloned in pHC79, generating plasmids pPM3115 and pPM3116. These plasmids mediate expression, in Escherichia coli K-12, of lipopolysaccharides (LPS) immunologically similar to the S. flexneri type 6 LPS as judged by SDS-PAGE and Western-immunoblot analysis using S. flexneri type 6 specific antisera. Thus, unlike other S. flexneri serotypes, no additional loci are required for serotype specificity. This expression is independent of E. coli K-12 rfb genes. Southern-hybridization analysis using the 16.2-kb BglII probe from S. flexneri type 6 rfb region detected very little sequence homology in S. flexneri serotypes 1-5, however, some homology was detected with E. coli O2 and O18, but not in E. coli 0101 strains, Salmonella and Vibrio cholerae.     

The stability of O-antigen plasmid is determined by a chromosomal region of Shigella dysenteriae 1

It is well established that plasmids are involved in the expression of lipopolysaccharide in certain species of Shigella. In Shigella sonnei, both the biosynthesis of oligosaccharide side chains (O antigen), and cell invasiveness are controlled exclusively by a 120 megadalton (MDa) plasmid. In Shigella dysenteriae 1, a 10 kilobase (kb) plasmid is required for O-antigen production. Shigella dysenteriae 1 strains devoid of this plasmid lose the ability to synthesize O antigen. Interestingly, this 10-kb plasmid is not stably maintained in Escherichia coli K-12 strains, where it is lost spontaneously at a high frequency. Our genetic analyses of Shigella dysenteriae 1 strain IDBM11 and its derivatives indicate that the stability of this plasmid is associated with the histidine region of the chromosome which is unique to Shigella dysenteriae. Furthermore, the 10-kb plasmid is stably maintained in wild-type IDBM11 with an intact histidine locus. However, this plasmid is not stable in IDBM11 derivatives (e.g., IDBM11-1 and IDBM11-2), in which the his locus has been substituted with the histidine region of an E. coli K-12 chromosome. The S. dysenteriae IDBM11 strain, and its derivatives (lacking a 10-kb plasmid), displayed an invasive property as demonstrated by their internalization by HeLa cells in an in vitro assay. Thus the 10-kb plasmid of Shigella dysenteriae 1 is required for O-antigen synthesis but not for cell invasion.     

Structure of the Shigella dysenteriae haem transport locus and its phylogenetic distribution in enteric bacteria

The ability to transport and use haemin as an iron source is frequently observed in clinical isolates of Shigella spp. and pathogenic Escherichia coli . We found that many of these haem-utilizing E. coli strains contain a gene that hybridizes at high stringency to the S. dysenteriae type 1 haem receptor gene, shuA . These shuA -positive strains belong to multiple phylogenetic groups and include clinical isolates from enteric, urinary tract and systemic infections. The distribution of shuA in these strains suggests horizontal transfer of the haem transport locus. Some haem-utilizing pathogenic E. coli strains did not hybridize with shuA , so at least one other haem transport system is present in this group. We also characterized the chromosomal region containing shuA in S. dysenteriae . The shuA gene is present in a discrete locus, designated the haem transport locus, containing eight open reading frames. Several of the proteins encoded in this locus participate with ShuA in haem transport, as a Salmonella typhimurium strain containing the entire haem transport locus used haem much more efficiently than the same strain containing only shuA . The haem transport locus is not present in E. coli K-12 strains, but the sequences flanking the haem transport locus in S. dysenteriae matched those at the 78.7 minute region of E. coli K-12. The junctions and flanking sequences in the shuA -positive pathogenic E. coli strains tested were nearly identical to those in S. dysenteriae , indicating that, in these strains, the haem transport locus has an organization similar to that in S. dysenteriae , and it is located in the same relative position on the chromosome.     

Cloning and expression in Escherichia coli K-12 of the genes for major outer membrane protein OmpA from Shigella dysenteriae, Enterobacter aerogenes, and Serratia marcescens.

The outer membranes of many gram-negative bacteria contain a major heat-modifiable protein which shows serological cross-reactivity with the OmpA protein of Escherichia coli K-12. Using the cloned gene for the E. coli K12 protein as a DNA-DNA hybridization probe, we were able to identify the corresponding genes from Shigella dysenteriae. Enterobacter aerogenes, and Serratia marcescens. These were cloned in a phage lambda vector, and their expression in E. coli K-12 was studied. All three OmpA proteins were fully produced and correctly exported to the outer membrane. In several cases, complete or partial restoration of known function of the E. coli K-12 protein was observed.     

Acetylation of O-specific lipopolysaccharides from Shigella flexneri 3a and 2a occurs in Escherichia coli K-12 carrying cloned S. flexneri 3a and 2a rfb genes.

Most of the Shigella flexneri O-specific serotypes result from O-acetyl and/or glucosyl groups added to a common O-repeating unit of the lipopolysaccharide (LPS) molecule. The genes involved in acetylation and/or glucosylation of S. flexneri LPS are physically located on lysogenic bacteriophages, whereas the rfb cluster contains the biosynthesis genes for the common O-repeating unit (D.A.R. Simmons and E. Romanowska, J. Med. Microbiol. 23:289-302, 1987). Using a cosmid cloning strategy, we have cloned the rfb regions from S. flexneri 3a and 2a. Escherichia coli K-12 containing plasmids pYS1-5 (derived from S. flexneri 3a) and pEY5 (derived from S. flexneri 2a) expressed O-specific LPS which reacted immunologically with S. flexneri polyvalent O antiserum. However, O-specific LPS expressed in E. coli K-12 also reacted with group 6 antiserum, indicating the presence of O-acetyl groups attached to one of the rhamnose components of the O-repeating unit. This was confirmed by measuring the amounts of acetate released from purified LPS samples and also by the chemical removal of O-acetyl groups, which abolished group 6 reactivity. The O-acetylation phenotype was absent in an E. coli strain with an sbcB-his-rfb chromosomal deletion and could be restored upon conjugation of F' 129, which carries sequences corresponding to a portion of the deleted region. Our data demonstrate that E. coli K-12 strains possess a novel locus which directs the O acetylation of LPS and is located in the sbcB-rfb region of the chromosomal map.     

Lipopolysaccharide core structures and their correlation with genetic groupings of Shigella strains. A novel core variant in Shigella boydii type 16

Bacteria Shigella, the cause of shigellosis, evolved from the intestinal bacteria Escherichia coli. Based on structurally diverse O-specific polysaccharide chains of the lipopolysaccharides (LPSs; O-antigens), three from four Shigella species are subdivided into multiple serotypes. The central oligosaccharide of the LPS called core is usually conserved within genus but five core types called R1-R4 and K-12 have been recognized in E. coli. Structural data on the Shigella core are limited to S. sonnei, S. flexneri and one S. dysenteriae strain, which all share E. coli core types. In this work, we elucidated the core structure in 14 reference strains of S. dysenteriae and S. boydii. Core oligosaccharides were obtained by mild acid hydrolysis of the LPSs and studied using sugar analysis, high-resolution mass spectrometry and two-dimensional NMR spectroscopy. The R1, R3 and R4 E. coli core types were identified in 8, 3 and 2 Shigella strains, respectively. A novel core variant found in S. boydii type 16 differs from the R3 core in the lack of GlcNAc and the presence of a D-glycero-D-manno-heptose disaccharide extension. In addition, the structure of an oligosaccharide consisting of the core and one O-antigen repeat was determined in S. dysenteriae type 8. A clear correlation of the core type was observed with genetic grouping of Shigella strains but not with their traditional division to four species. This finding supports a notion on the existing Shigella species as invalid taxa and a suggestion of multiple independent origins of Shigella from E. coli clones.     

The occurrence of duplicate lysyl-tRNA synthetase gene homologs in Escherichia coli and other procaryotes.

The lysyl-tRNA synthetase (LysRS) system of Escherichia coli K-12 consists of two genes, lysS, which is constitutive, and lysU, which is inducible. It is of importance to know how extensively the two-gene LysRS system is distributed in procaryotes, in particular, among members of the family Enterobacteriaceae. To this end, the enterics E. coli K-12 and B; E. coli reference collection (ECOR) isolates EC2, EC49, EC65, and EC68; Shigella flexneri; Salmonella typhimurium; Klebsiella pneumoniae; Enterobacter aerogenes; Serratia marcescens; and Proteus vulgaris and the nonenterics Pseudomonas aeruginosa and Bacillus megaterium were grown in AC broth to a pH of 5.5 or less or cultured in SABO medium at pH 5.0. These growth conditions are known to induce LysRS activity (LysU synthesis) in E. coli K-12. Significant induction of LysRS activity (twofold or better) was observed in the E. coli strains, the ECOR isolates, S. flexneri, K. pneumoniae, and E. aerogenes. To demonstrate an association between LysRS induction and two distinct LysRS genes, Southern blotting was performed with a probe representing an 871-bp fragment amplified from an internal portion of the coding region of the lysU gene. In initial experiments, chromosomal DNA from E. coli K-12 strain MC4100 (lysS+ lysU+) was double digested with either BamHI and HindIII or BamHI and SalI, producing hybridizable fragments of 12.4 and 4.2 kb and 6.6 and 5.2 kb, respectively. Subjecting the chromosomal DNA of E. coli K-12 strain GNB10181 (lysS+ delta lysU) to the same regimen established that the larger fragment from each digestion contained the lysU gene. The results of Southern blot analysis of the other bacterial strains revealed that two hybridizable fragments were obtained from all of the E. coli and ECOR collection strains examined and S. flexneri, K. pneumoniae, and E. aerogenes. Only one lysU homolog was found with S. typhimurium and S. marcescens, and none was obtained with P. vulgaris. A single hybridizable band was found with both P. aeruginose and B, megaterium. These results show that the dual-gene LysRS system is not confined to E. coli K-12 and indicate that it may have first appeared in the genus Enterobacter.     

Comparison of the tryptophan synthetase alpha-subunits of several species of Enterobacteriaceae

Creighton, T. E. (Stanford University, Stanford), D. R. Helinski, R. L. Somerville, and C. Yanofsky. Comparison of the tryptophan synthetase alpha subunits of several species of Enterobacteriaceae. J. Bacteriol. 91:1819-1826. 1966.-The tryptophan synthetase alpha subunits of Escherichia coli K-12, E. coli B, Shigella dysenteriae, Salmonella typhimurium, and Aerobacter aerogenes have been purified and their structures compared. Each of these alpha subunits exhibits a sedimentation coefficient of about 2.7S. Peptide patterns of trypsin plus chymotrypsin digests of the alpha subunits have indicated that all of the alpha subunits have peptide regions in common. The patterns of E. coli K-12, E. coli B, and S. dysenteriae alpha subunits appear to be nearly identical, whereas the alpha subunits from S. typhimurium and A. aerogenes differ from those of E. coli and from each other. It has also been shown that the E. coli structural gene for the alpha subunit is translated identically in E. coli and S. typhimurium.     

Haem iron-transport system in enterohaemorrhagic Escherichia coli O157:H7

In this study, we identified the iron-transport systems of Escherichia coli O157:H7 strain EDL933. This strain synthesized and transported enterobactin and had a ferric citrate transport system but lacked the ability to produce or use aerobactin. It used haem and haemoglobin, but not transferrin or lactoferrin, as iron sources. We cloned the gene encoding an iron-regulated haem-transport protein and showed that this E. coli haem-utilization gene ( chuA ) encoded a 69 kDa outer membrane protein that was synthesized in response to iron limitation. Expression of this protein in a laboratory strain of E. coli was sufficient for utilization of haem or haemoglobin as iron sources. Mutation of the chromosomal chuA and tonB genes in E. coli O157:H7 demonstrated that the utilization of haemin and haemoglobin was ChuA- and TonB-dependent. Nucleotide sequence analysis of chuA revealed features characteristic of TonB-dependentFur-regulated, outer membrane iron-transport proteins. It was highly homologous to the shuA gene of Shigella dysenteriae and less closely related to hemR of Yersinia enterocolitica and hmuR of Yersinia pestis . A conserved Fur box was identified upstream of the chuA gene, and regulation by Fur was confirmed.     

Genomic subtraction to identify and characterize sequences of Shiga toxin-producing Escherichia coli O91:H21

To identify Shiga toxin-producing Escherichia coli genes associated with severe human disease, a genomic subtraction technique was used with hemolytic-uremic syndrome-associated O91:H21 strain CH014 and O6:H10 bovine strains. The method was adapted to the Shiga toxin-producing E. coli genome: three rounds of subtraction were used to isolate DNA fragments specific to strain CH014. The fragments were characterized by genetic support analysis, sequencing, and hybridization to the genome of a collection of Shiga toxin-producing E. coli strains. A total of 42 fragments were found, 19 of which correspond to previously identified unique DNA sequences in the enterohemorrhagic E. coli EDL933 reference strain, including 7 fragments corresponding to prophage sequences and others encoding candidate virulence factors, such a SepA homolog protein and a fimbrial usher protein. In addition, the subtraction procedure yielded plasmid-related sequences from Shigella flexneri and enteropathogenic and Shiga toxin-producing E. coli virulence plasmids. We found that lateral gene transfer is extensive in strain CH014, and we discuss the role of genomic mobile elements, especially bacteriophages, in the evolution and possible transfer of virulence determinants.     

Identification of an Escherichia coli gene homologous to virR, a regulator of Shigella virulence.

Virulence in Shigella spp., as well as in strains of enteroinvasive Escherichia coli, is regulated by growth temperature. Previously, virR had been identified as the gene controlling the temperature-regulated expression of Shigella virulence. Since Shigella spp. and E. coli are also known to share greater than 90% DNA sequence homology, we sought to determine if nonpathogenic E. coli K-12 C600 contains a gene homologous to the Shigella flexneri 2a gene virR. Through the use of transduction and molecular cloning of strain C600 chromosomal DNA we have shown that E. coli K-12 does indeed contain a gene functionally homologous to the virR of S. flexneri.     

Genetics and regulation of heme iron transport in Shigella dysenteriae and detection of an analogous system in Escherichia coli O157:H7.

Shigella species can use heme as the sole source of iron. In this work, the heme utilization locus of Shigella dysenteriae was cloned and characterized. A cosmid bank of S. dysenteriae serotype 1 DNA was constructed in an Escherichia coli siderophore synthesis mutant incapable of heme transport. A recombinant clone, pSHU12, carrying the heme utilization system of S. dysenteriae was isolated by screening on iron-poor medium supplemented with hemin. Transposon insertional mutagenesis and subcloning identified the region of DNA in pSHU12 responsible for the phenotype of heme utilization. Minicell analysis indicated that a 70-kDa protein encoded by this region was sufficient to allow heme utilization in E. coli. Synthesis of this protein, designated Shu (Shigella heme uptake), was induced by iron limitation. The 70-kDa protein is located in the outer membrane and binds heme, suggesting it is the S. dysenteriae heme receptor. Heme iron uptake was found to be TonB dependent in E. coli. Transformation of an E. coli hemA mutant with the heme utilization subclone, pSHU262, showed that heme could serve as a source of porphyrin as well as iron, indicating that the entire heme molecule is transported into the bacterial cell. DNA sequences homologous to shu were detected in strains of S. dysenteriae serotype 1 and E. coli O157:H7.     

Role of Escherichia coli K-12 rfa genes and the rfp gene of Shigella dysenteriae 1 in generation of lipopolysaccharide core heterogeneity and attachment of O antigen.

The rfp gene of Shigella dysenteriae 1 and the rfa genes of Escherichia coli K-12 and Salmonella typhimurium LT2 have been studied to determine their relationship to lipopolysaccharide (LPS) core heterogeneity and their role in the attachment of O antigen to LPS. It has been inferred from the nucleotide sequence that the rfp gene encodes a protein of 41,864 Da which has a structure similar to that of RfaG protein. Expression of this gene in E. coli K-12 results in the loss of one of the three bands seen in gel analysis of the LPS and in the appearance of a new, more slowly migrating band. This is consistent with the hypothesis that Rfp is a sugar transferase which modifies a subset of core molecules so that they become substrates for attachment of S. dysenteriae O antigen. A shift in gel migration of the bands carrying S. dysenteriae O antigen and disappearance of the Rfp-modified band in strains producing O antigen suggest that the core may be trimmed or modified further before attachment of O antigen. Mutation of rfaL results in a loss of the rough LPS band which appears to be modified by Rfp and prevents the appearance of the Rfp-modified band. Thus, RfaL protein is involved in core modification and is more than just a component of the O-antigen ligase. The products of rfaK and rfaQ also appear to be involved in modification of the core prior to attachment of O antigen, and the sites of rfaK modification are different in E. coli K-12 and S. typhimurium. In contrast, mutations in rfaS and rfaZ result in changes in the LPS core but do not affect the attachment of O antigen. We propose that these genes are involved in an alternative pathway for the synthesis of rough LPS species which are similar to lipooligosaccharides of other species and which are not substrates for O-antigen attachment. All of these studies indicate that the apparent heterogeneity of E. coli K-12 LPS observed on gels is not an artifact but instead a reflection of functional differences among LPS species.