Efficiency of cationic lipid-mediated transfection of polarized and differentiated airway epithelial cells in vitro and in vivo.

Systematic analysis of a large number of different cationic lipids has led to the identification of novel structures (GL-67) and formulations of cationic lipid:plasmid DNA (pDNA) complexes that facilitate high levels of gene expression in lungs of mice. However, despite significant improvement in gene transfer activity, we show here that the efficiency of GL-67-mediated gene transduction of intact airway epithelia is still relatively low. Administration of GL-67:pCF1-CFTR (encoding the cystic fibrosis transmembrane conductance regulator) complexes into the nasal epithelium of cystic fibrosis (CF) transgenic mice resulted only in marginal correction of the ion transport defects. Measurements of nasal potential differences (PD) showed no correction of the sodium (Na+) transport defect, and only partial restitution of the chloride (Cl-) transport defect was achieved in a small proportion of the animals after perfusion of the nasal epithelium with the complexes. Furthermore, in contrast to results obtained following instillation of GL-67:pDNA complexes into the lungs of mice, perfusion of GL-67:pDNA into the nasal epithelium resulted only in a moderate enhancement of gene transduction activity relative to that attained with naked pDNA alone. To determine the basis for this low efficiency of transfection, a series of studies was conducted to identify some of the barriers governing cationic lipid-mediated gene transfer to the airway epithelium. We show here that the transfection activity of GL-67 was affected by the polarization, differentiation, and proliferative state of the cells. Diminished transfection activity was observed with nonmitotic, highly polarized and differentiated airway epithelial cells. This observed reduction in gene expression with nonmitotic cells was determined to be due in part to inefficient nuclear translocation of the pDNA from the cytoplasm. Together these data indicate that much improvement in the ability of cationic lipids to transfect polarized and differentiated airway epithelial cells is a necessary prerequisite for effective cationic lipid-mediated gene therapy of airway diseases such as CF.

Nucleotide sequence of the gene encoding the 35-kDa protein of Mycobacterium tuberculosis.

A 35-kilodalton (kDa) protein of Mycobacterium tuberculosis is expressed by recombinant Escherichia coli which possess a plasmid that contains a 2.4-kilobase fragment of M. tuberculosis chromosomal DNA. The nucleotide sequence of this fragment was determined by the dideoxynucleotide chain-termination method. Analysis of the sequence revealed four open reading frames that could encode proteins greater than 250 amino acids in length. The reading frame for the 35-kDa protein was identified by subcloning DNA fragments into expression vectors pTTQ18 and pTTQ19, and assaying for production of the 35-kDa protein by Western blotting. A protein with a primary structure of 270 amino acids and a predicted molecular weight of 29,260 daltons was deduced from the nucleotide sequence. A computer-aided search of nucleic and amino acid sequence databases did not identify any proteins with significant sequence similarity to this protein. The organization of the gene encoding this protein was compared with other mycobacterial genes that have been sequenced. Information obtained from the investigation of this protein may aid in the development of reagents to diagnose and control mycobacterial disease.

Localization of the streptococcal C5a peptidase to the surface of group A streptococci.

Immunofluorescent staining was used to determine that the streptococcal C5a peptidase (SCP) exists as a cell surface antigen on group A streptococci. The ability of hyperimmune serum to neutralize cell-associated SCP activity provided further evidence for the location of SCP. Quantification of SCP during growth in vitro by indirect enzyme-linked immunosorbent assay showed that approximately 90% of the measurable antigen is cell bound.

Phylogenetic relationship of Gemella morbillorum to Gemella haemolysans.

Partial 16S rRNA gene sequences (16S rDNA) of Gemella morbillorum and Gemella haemolysans were determined by sequencing polymerase chain reaction-amplified DNA. A phylogenetic analysis of 53 eubacterial 16S rRNA sequences grouped the gemellae on a distinct branch separate from the 18 members of the genus Streptococcus. DNA-DNA hybridization results indicate that the two gemellae are related at the genus level but are not a single species.

In vivo Streptococcus pyogenes C5a peptidase activity: analysis using transposon- and nitrosoguanidine-induced mutants.

The streptococcal C5a peptidase removes a six-amino-acid fragment from human C5a and thereby inactivates this chemotaxin. We used transposon and chemical mutagenesis to generate mutants of Streptococcus pyogenes that did not produce C5a peptidase. These mutants showed no alteration in expression of capsule, M protein, streptolysins O and S, or pyrogenic exotoxin C. Serial passage of a peptidase-producing strain in vivo resulted in a 100-fold increase in production of C5a peptidase. The presence of C5a peptidase delayed the accumulation of polymorphonuclear leukocytes (PMNLs) in the peritoneal cavities of mice after intraperitoneal challenge. However, there was no difference in virulence (as evaluated by LD50) between strains that produced and those that lacked C5a peptidase. Although C5a peptidase is expressed on the cell surface, antibody to this enzyme did not opsonize streptococci for phagocytosis in vitro. These studies show that C5a peptidase alters the normal host inflammatory response by delaying the accumulation of PMNLs at the foci of streptococcal infection.

16S rRNA sequences of Bartonella bacilliformis and cat scratch disease bacillus reveal phylogenetic relationships with the alpha-2 subgroup of the class Proteobacteria.

The primary structures of 16S rRNAs of Bartonella bacilliformis, an isolate of the cat scratch disease (CSD) bacillus, and a strain phenotypically similar to the CSD bacillus were determined by reverse transcriptase sequencing. These microorganisms were found to be members of the alpha-2 subgroup of the class Proteobacteria. The sequence from B. bacilliformis was most closely related to the rRNA of Rochalimaea quintana (91.7% homology), the etiologic agent of trench fever. The sequence from the isolate of the CSD bacillus showed the greatest homology with Brucella abortus (89.7%) and, when compared with oligonucleotide catalog data, formed a cluster with Rhodopseudomonas palustris, Pseudomonas carboxidovorans, Nitrobacter species, and Bradyrhizobium species. The 16S rRNA sequence was also determined for the Cleveland Clinic isolate, which was previously shown to be phenotypically similar to and approximately 30% related, by DNA hybridization, to the CSD bacillus. The Cleveland Clinic isolate was isolated from a patient not diagnosed with CSD. The rRNAs from these bacteria exhibited 98.2% homology, confirming that this isolate is a second species in the same genus as the CSD bacillus. Our data suggest that neither B. bacilliformis nor the CSD bacillus is the etiologic agent of bacillary epithelioid angiomatosis.

Proposals to unify the genera Bartonella and Rochalimaea, with descriptions of Bartonella quintana comb. nov., Bartonella vinsonii comb. nov., Bartonella henselae comb. nov., and Bartonella elizabethae comb. nov., and to remove the family Bartonellaceae from the order Rickettsiales.

DNA hybridization data (hydroxyapatite method, 50 to 70 degrees C) indicate that Rickettsia prowazekii, the type species of the type genus of the family Rickettsiaceae, is substantially less closely related to Rochalimaea species than was previously thought. The levels of relatedness of Rickettsia prowazekii to Rochalimaea species and to Bartonella bacilliformis under optimal conditions for DNA reassociation were 0 to 14%, with 25.5% or greater divergence in related sequences. When stringent reassociation criteria were used, the levels of relatedness were 0 to 2%. The genera Bartonella and Rochalimaea are currently classified in different families (the Bartonellaceae and the Rickettsiaceae) in the order Rickettsiales. On the basis of DNA relatedness data, previous 16S rRNA sequence data, guanine-plus-cytosine contents, and phenotypic characteristics, neither Bartonella bacilliformis nor Rochalimaea species are closely related to other organisms currently classified in the order Rickettsiales. In fact, the closest relative of these organisms is Brucella abortus. It is therefore proposed that the family Bartonellaceae should be removed from the order Rickettsiales. Previous 16S rRNA sequence data and DNA hybridization data revealed high levels of relatedness between Bartonella bacilliformis and the four Rochalimaea species, indicating that these species are members of a single genus. It is proposed that the genus Rochalimaea should be united with the genus Bartonella in the family Bartonellaceae. The name Bartonella is retained as the genus name since it has nomenclatural priority over the name Rochalimaea. This means that new combinations for the Rochalimaea species must be created. Proposals are therefore made for the creation of Bartonella quintana comb. nov., Bartonella vinsonii comb. nov., Bartonella henselae comb. nov., and Bartonella elizabethae comb. nov.

Cross-reactivity of genetic probe for detection of Mycobacterium tuberculosis with newly described species Mycobacterium celatum.

An acridinium ester-labeled DNA probe (AccuProbe; Gen-Probe Inc., San Diego, Calif.) for the identification of the Mycobacterium tuberculosis complex gave discrepant results with the newly described species M. celatum. Examination of 20 strains of M. celatum showed that 8 were positive with the probe; the remaining 12 were negative.

Cloning and nucleotide sequence analysis of psaA, the Streptococcus pneumoniae gene encoding a 37-kilodalton protein homologous to previously reported Streptococcus sp. adhesins.

Gene psaA, which encodes the Streptococcus pneumoniae 37-kDa protein, was cloned in Escherichia coli, and its complete nucleotide sequence was determined. Analysis of the sequence of the 2.4-kb cloned fragment revealed three open reading frames (ORFs). ORF2, which is 933 bp long, was identified as psaA. The two other ORFs identified flank psaA. ORF1, located upstream of psaA, is 836 nucleotides long and encodes a protein with a calculated molecular mass of 29,843 Da. The sequence for ORF3, located downstream of psaA, was only partially determined. Northern (RNA) blot analysis of pneumococcal RNA suggests that psaA is transcribed as part of a polycistronic message. Analysis of the primary structure of the protein encoded by this gene indicated significant similarity to two previously reported streptococcal proteins, SsaB (80% similarity) and FimA (92.3% similarity), from S. sanguis and S. parasanguis, respectively. These two homologous proteins have been shown to be associated with bacterial adhesion, and the possibility of a similar role for PsaA is hypothesized.

The human antibody response to streptococcal C5a peptidase.

An ELISA was developed to measure antibody, both IgG and IgA, against the streptococcal C5a peptidase (SCP), in human sera and saliva. Generally, sera and saliva from young, uninfected children lacked antibody to SCP. In contrast, most sera and saliva specimens from healthy adults had measurable levels of anti-SCP IgG and SCP-specific secretory IgA (anti-SCP sIgA). Paired acute and convalescent sera from patients with streptococcal pharyngitis possessed significantly higher levels of anti-SCP IgG than did sera from healthy individuals. Sera containing high concentrations of anti-SCP immunoglobulin were capable of neutralizing SCP activity. A survey of healthy adults and children also showed that the latter were significantly less likely to have anti-SCP sIgA in their saliva. Detection of this antibody in greater than 90% of the saliva specimens obtained from children who had recently experienced streptococcal pharyngitis demonstrated that children can produce a secretory response. This is thought to be the first report of a secretory IgA response in humans to a somatic antigen of Streptococcus pyogenes.

Molecular characterization and proposal of a neotype strain for Bartonella bacilliformis.

Bartonella bacilliformis, the etiologic agent of bartonellosis, was characterized biochemically and by DNA hybridization, guanine-plus-cytosine content, genome size, and 16S rRNA sequencing. DNAs from the two strains in our collection exhibited 97% relatedness in hydroxyapatite reactions done at 55 degrees C (optimal reassociation criterion) and 100% relatedness in reactions done at 70 degrees C (stringent reassociation criterion). There was no evidence of divergence within the related sequences. B. bacilliformis DNA showed no relatedness to the cat scratch disease bacillus or to a strain of a second species in the same genus as the cat scratch disease bacillus in hybridization reactions done at 65 degrees C. The guanine-plus-cytosine contents of DNAs from the two B. bacilliformis strains were 39 and 40 mol%. Time course reassociation, done by determining spectrophotometrically the time required for one-half of the denatured DNA to form duplexes, indicated that B. bacilliformis has a genome size of approximately 4 x 10(8). The 16S rRNA sequence analysis indicated that B. bacilliformis is in the alpha-2 subgroup of the purple bacteria, class Proteobacteria, and that its closest relatives are Rochalimaea quintana and Brucella abortus. Strain KC583 (= Herrer 020/F12,63 = ATCC 35685) is proposed as the type strain of B. bacilliformis.

Neisseria weaveri sp. nov., formerly CDC group M-5, a gram-negative bacterium associated with dog bite wounds.

CDC group M-5 is a rod-shaped, gram-negative, nonmotile bacterium associated with dog bite wounds. DNA-DNA relatedness and biochemical and growth characteristics were studied for 54 strains from the collection at the Centers for Disease Control and Prevention. One typical M-5 strain, 8142, was further studied by 16S rRNA sequencing. DNA from 40 of 53 strains showed 82 to 100% relatedness (hydroxyapatite method) to labeled DNA from strain 8142. The guanine-plus-cytosine (G + C) content in 8 of the 41 highly related M-5 strains was 50.5 to 52 mol%. These 41 strains were oxidase and catalase positive, nonfermentative, nitrite positive, nitrate negative, weakly phenylalanine deaminase positive, aerobic, and alpha-hemolytic (sheep blood). DNA from the 13 remaining strains showed only 7 to 46% DNA relatedness to strain 8142. These 13 non-M-5 strains differed from the M-5 strains in G + C content, growth characteristics, and biochemical profiles. DNA from M-5 strain 8142 was most closely related to DNA from groups EF-4b (47%) and EF-4a (45%). 16S rRNA sequence analysis placed M-5 strain 8142 in the Neisseriaceae cluster of the beta-3 subgroup of the class Proteobacteria. It was most homologous (98.4 to 98.8%) to Neisseria animalis, Neisseria flavescens, Neisseria canis, and Neisseria elongata. All data are consistent with M-5 being a new species of Neisseria, for which we propose the name Neisseria weaveri.

Comparison of arbitrarily primed polymerase chain reaction, ribotyping, and monoclonal antibody analysis for subtyping Legionella pneumophila serogroup 1.

Arbitrarily primed polymerase chain reaction (AP-PCR) was used to characterize Legionella pneumophila serogroup 1. Cells from a single colony could be subtyped by AP-PCR within a few hours. The discrimination between strains of L. pneumophila serogroup 1 by AP-PCR was equivalent to that by monoclonal antibody analysis and ribotyping. Four strains representing the monoclonal antibody pattern most frequently associated with outbreaks all yielded unique amplicon patterns by AP-PCR.

Nucleotide sequence of htpB, the Legionella pneumophila gene encoding the 58-kilodalton (kDa) common antigen, formerly designated the 60-kDa common antigen.

Gene htpB, which encodes the 58-kilodalton protein of Legionella pneumophila, was cloned in Escherichia coli and its complete nucleotide sequence was determined. Analysis of this sequence revealed an open reading frame of 1,644 nucleotides encoding a protein with a predicted molecular mass of 57,952 daltons. Data obtained by amino-terminal sequencing of the purified 58-kilodalton protein agreed, except for one amino acid residue, with the predicted amino acid sequence, identifying this open reading frame as htpB. A comparison of the primary structure of this protein to other proteins of similar molecular weights from E. coli, Mycobacterium leprae, M. tuberculosis, and Coxiella burnetii revealed significant regions of sequence similarity, which are discussed.

Identification of Bordetella pertussis infection by shared-primer PCR.

A shared-primer PCR method for the detection of infection was developed by using primers derived from DNA sequences upstream of the structural genes for the porin proteins of Bordetella pertussis and Bordetella parapertussis. This method resulted in a 159-bp PCR product specific for B. pertussis and a 121-bp DNA fragment specific for B. parapertussis and allowed for the simultaneous detection of these pathogens. The PCR procedure was shown to be very specific since no PCR product was obtained from 36 non-Bordetella bacterial DNAs. Nasopharyngeal aspirates (NPAs) from children suspected of having pertussis were evaluated by the PCR method, culture, and the Chinese hamster ovary (CHO) cell assay, which detects pertussis toxin. B. pertussis was cultured from 119 of 205 NPAs assayed, and the presence of pertussis toxin was detected in 69 of the NPAs by the CHO cell assay. When ethidium bromide staining was used to detect PCR products, 100 NPAs gave positive results by shared-primer PCR; 94 of these NPAs were also positive by culture. The result indicated a sensitivity of 79% for PCR when culture was used as the standard. The sensitivity of PCR was increased to 95% when a digoxigenin immunoblot system was used. An additional 20 NPAs from patients with suspected pertussis that were culture negative also gave positive results by PCR. The specific and sensitive PCR method described here should be useful for both the clinical diagnosis of pertussis and case identification in vaccine trials.

Bordetella holmesii sp. nov., a new gram-negative species associated with septicemia.

CDC nonoxidizer group 2 (NO-2) currently consists of 15 gram-negative, rod-shaped, oxidase-negative, asaccharolytic, brown soluble pigment-producing strains isolated from blood cultures, usually from young adults. On the basis of their cellular fatty acid profiles, NO-2 strains formed a single group that was identical with the profile of Bordetella avium. 16S rRNA sequencing of one NO-2 strain and the type strains of B. pertussis, B. parapertussis, B. bronchiseptica, and B. avium showed a high degree of homology (> or = 98% over 1,525 bases). The NO-2 guanine-plus-cytosine content (61.5 to 62.3 mol%) and major ubiquinone analysis (ubiquinone-8) results were both consistent with those for the genus Bordetella. DNA relatedness studies (hydroxyapatite method) confirmed a close relatedness between NO-2 and Bordetella species and demonstrated that NO-2 strains were a single new species. The name B. holmesii sp. nov. is proposed for CDC group NO-2.

Mycobacterium celatum sp. nov.

A new slowly growing nonphotochromogenic Mycobacterium species of clinical importance is described. The biochemical characteristics of this organism were similar to those of Mycobacterium xenopi and members of the Mycobacterium avium complex. However, none of the strains reacted with commercially available genetic probes for the M. avium complex. The strains were resistant to most antituberculosis drugs. Multilocus enzyme electrophoresis revealed two original electrophoretic types, which was suggestive of new species. The strains contained alpha-, keto-, and dicarboxylic mycolates, as determined by thin-layer chromatography. A mycolic acid analysis by high-performance liquid chromatography revealed a chromatographic pattern similar to that of M. xenopi, but distinct from the patterns of previously described Mycobacterium species. Hexadecanoic and tuberculostearic acids were identified as the major cell wall fatty acids by gas-liquid chromatographic analysis; hexacosanoic acid was the major mycolic acid cleavage product, and 2-eicosanol was the major alcohol. Evaluation of the 16S rRNA sequence confirmed the phylogenetic position of the organism among the slowly growing Mycobacterium species. Cultures representing this new species have been deposited in the American Type Culture Collection as strains ATCC 51130 and ATCC 51131T (T = type strain). The name Mycobacterium celatum is proposed.

Rochalimaea elizabethae sp. nov. isolated from a patient with endocarditis.

A Rochalimaea-like organism (strain F9251) was isolated from a patient with endocarditis after blood drawn for culture before antimicrobial therapy was subcultured onto blood and chocolate agars and incubated for 2 weeks in 5% CO2. The strain was phenotypically similar to known Rochalimaea species. The cellular fatty acid composition of strain F9251 was close to but distinct from those of the three known Rochalimaea species and was most similar to that of R. vinsonii. Labeled DNA from strain F9251 was 59 to 67% related to DNAs from type strains of the three described Rochalimaea species, and its 16S rRNA gene sequence was 98.9% or more homologous to their 16S rRNA gene sequences. These findings support classification of F9251 as a new Rochalimaea species, for which the name Rochalimaea elizabethae sp. nov. is proposed. The patient infected with the organism had large bacterial vegetations on his aortic valve and was cured with antibiotics and valve-replacement surgery. Recognition of the procedures required to identify this and other Rochalimaea species suggests that clinical laboratories should prolong the incubation times of cultures of blood and tissue from patients with suspected endocarditis, patients with fever of unknown origin, and immunocompromised patients with fever so that the full spectrum of disease caused by these organisms can be recognized.

Ability of adenovirus vectors containing different CFTR transcriptional cassettes to correct ion transport defects in CF cells.

Cystic fibrosis (CF) airway epithelial cells exhibit defective adenosine 3',5'-cyclic monophosphate (cAMP)-mediated chloride (Cl) secretion, abnormal hyperabsorption of sodium (Na+), and aberrant fluid transport. Adenovirus-mediated transduction of cystic fibrosis transmembrane conductance regulator (CFTR) corrects these ion and fluid transport abnormalities in CF cells. However, several challenges remain pertaining to the use of adenovirus vectors for gene delivery, including the efficiency of gene transfer and the host response to the vector. To improve the efficacy of adenovirus-mediated gene transfer, we have constructed a series of recombinant adenoviruses containing different CFTR transcriptional units, and we have evaluated their relative ability to correct electrolyte and fluid transport in polarized CF airway epithelial cells. The ability of the vectors to correct the CF Cl- transport defects was greatest when the human cytomegalovirus promoter was used. The E1a and phosphoglycerate kinase promoters resulted in the greatest persistence of functional CFTR expression. Efficacy of gene expression by recombinant adenoviruses improved as the cells were treated with increasing multiplicities of infection, as the duration of viral contact with the target cells was lengthened, and when the virus concentration was increased. Transduction of functional CFTR Cl- channel activity reversed the abnormal Na+ hyperabsorption observed in CF cells in a dose-dependent manner, suggesting that Na+ channel activity is downregulated by CFTR. Although efficient correction of both cAMP-mediated Cl- transport and fluid secretion could be achieved readily with these vectors, normalization of the Na+ absorption required vector administration at high multiplicities of infection.