Recent work indicates that the growth and behavior of cancers are ultimately determined by a small subpopulation of malignant stem cells and that information about the properties of these cells is urgently needed to enable their targeting for therapeutic elimination. A key feature of normal stem cells is their asymmetrical division, the mechanism that allows stem cell self-renewal while producing hierarchies of amplifying and differentiating cells that form the bulk of the tissue. Most cancer deaths result from epithelial malignancies, but the extent to which the hierarchical proliferative stem and amplifying cell patterns of normal epithelia are actually retained in epithelial malignancies has been unclear. Here we show that even cell lines generated from carcinomas consistently produce in vitro colony patterns unexpectedly similar to those produced by the stem and amplifying cells of normal epithelia. From the differing types of colony morphologies formed, it is possible to predict both the growth potential of their constituent cells and their patterns of macromolecular expression. Maintenance of a subpopulation of stem cells during passage of cell lines indicates that the key stem cell property of asymmetrical division persists but is shifted towards enhanced stem cell self-renewal. The presence of malignant epithelial stem cells in vivo has been shown by serial transplantation of primary cancer cells and the present observations indicate that stem cell patterns are robust and persist even in cell lines. An understanding of this behavior should facilitate studies directed towards the molecular or pharmacologic manipulation of malignant stem cell survival.

The rag locus of Porphyromonas gingivalis W50 encodes RagA, a predicted tonB-dependent receptor protein, and RagB, a lipoprotein that constitutes an immunodominant outer membrane antigen. The low G؉C content of the locus, an association with mobility elements, and an apparent restricted distribution in the species suggested that the locus had arisen by horizontal gene transfer. In the present study, we have demonstrated that there are four divergent alleles of the rag locus. The original rag allele found in W50 was renamed rag-1, while three novel alleles, rag-2 to rag-4, were found in isolates lacking rag-1. The three novel alleles encoded variants of RagA with 63 to 71% amino acid identity to RagA1 and each other and variants of RagB with 43 to 56% amino acid identity. The RagA/B proteins have homology to numerous Bacteroides proteins, including SusC/D, implicated in polysaccharide uptake. Monoclonal and polyclonal antibodies raised against RagB1 of P. gingivalis W50 did not cross-react with proteins from isolates carrying different alleles. In a laboratory collection of 168 isolates, 26% carried rag-1, 36% carried rag-2, 25% carried rag-3, and 14% carried rag-4 (including the type strain, ATCC 33277). Restriction profiles of the locus in different isolates demonstrated polymorphism within each allele, some of which is accounted for by the presence or absence of insertion sequence elements. By reference to a previously published study on virulence in a mouse model (M. L. Laine and A. J. van Winkelhoff, Oral Microbiol. Immunol. 13:322-325, 1998), isolates that caused serious disease in mice were significantly more likely to carry rag-1 than other rag alleles.

The rag locus of Porphyromonas gingivalis encodes a putative TonB-dependent outer membrane receptor, RagA, and a 55-kDa immunodominant antigen, RagB. Inactivation of either ragA or ragB prevented expression of both RagA and RagB. Both the ragA and ragB mutants were significantly less virulent than wild-type strains in a murine model of infection.The rag locus of the periodontal pathogen Porphyromonas gingivalis encodes RagA, a 115-kDa outer membrane protein with features of a TonB-dependent receptor, and RagB, a 55-kDa antigen to which periodontal patients demonstrate an elevated immunoglobulin G response; together, these proteins are predicted to constitute a membrane transporter system (4,8). Four variants of the rag locus have been detected among clinical isolates of P. gingivalis (7), and a significant association was observed between carriage of the rag-1 allele and a highly virulent phenotype in a murine model of soft tissue destruction (7, 11). Indeed, in a variety of animal models of infection, P. gingivalis strains have been found to differ in their degrees of virulence, but rag-1 strains W50 and W83 (for which the complete genome sequence has been reported [15]) are consistently reported to be among the most virulent (1,6,11,14). P. gingivalis produces a number of well-characterized virulence factors, including proteases, fimbriae, and capsule (9). In order to determine whether the rag locus represents a further virulence factor for P. gingivalis, we have created insertion mutants with the rag-1 genes and tested their effect in a murine model of soft tissue destruction. Since interstrain polymorphism had also been detected in PG0183, the gene upstream of rag (7), we additionally investigated the effect of mutation in this locus.Genes were inactivated by insertion of the erm (ermFermAM) cassette from plasmid pVA2198, using an allele replacement strategy as described by Fletcher et al. (5). Briefly, the erm cassette (obtained by SacI plus PstI or SphI plus EcoRI digestion of pVA2198) was either inserted into the cloned genes (8) or ligated to PCR products to produce constructs in which the central region of each gene was replaced by the erm cassette. (PCR primers were as follows: for ragA, CGCTATTCTTCCTTTGCTTGCT and TTAC CATCCGCATCGACTTGA; for ragB, AATACTGAAAAT CCACGA and TAGGGGCTGCGACAAAAA; and for PG0183, GTGAACAAGCAGATTGGGG and CATAAGA GAGACGGAAACGAG). Prior to ligation with erm, the ragA product was digested with SacI plus PstI and the PG0183 product by SphI plus EcoRI. The DNA products were introduced into P. gingivalis by electroporation and transformants selected on media containing clindamycin, using methods described previously (16); replacement of the wild-type alleles was confirmed by demonstrating the expected length change for ragA or ragB by PCR and the expected restriction fragment changes by Southern blotting with ragA or ragB probes. In initial experiments, we were unsuccessful in the mutation of ragA or ragB in W50 but obtained mutants with mutations in both genes with the alternative ...

Periodontal infections by Porphyromonas gingivalis are associated with a sustained systemic IgG antibody response and elevations in local antibody synthesis to this organism. One of the targets of this response is a protease, RgpAcat, which is an important virulence determinant of this organism. Recently, we demonstrated that this molecule is glycosylated and that the glycan chains are immunologically related to P. gingivalis lipopolysaccharide (LPS) (Curtis et al., Infect Immun 1999;62:3816-3823). In the present study, we examined the role of these glycan additions in the immune recognition of RgpAcat, by sera from adult periodontal patients (n = 25). Serum IgG antibody levels to P. gingivalis W50, RgpAcat and LPS and to recombinant RgpA were determined by enzyme-linked immunosorbant assay (ELISA). No correlation was observed between the antibody levels to RgpAcat from P. gingivalis and the recombinant form of this enzyme expressed in Escherichia coli. However, a strong association was found between the recognition of LPS and the wild-type enzyme (R = 0.8926; p = 0.0005). Incorporation of LPS into the ELISA led to a significant reduction (mean 25%; range 0.8-43%, SD = 15; p < 0.05) in the recognition of RgpAcat, but had no effect on the recognition of control antigens. Deglycosylation of RgpAcat led to the abolition of immune recognition by patient serum IgG, which suggests that the glycan additions to this molecule are the principal targets of the immune response. Therefore, glycosylation of the RgpAcat protease may play an important role in immune evasion by shielding the primary structure from immune recognition.