Coxiella burnetii is a gram-variable obligate intracellular bacterium which carries out its development cycle in the phagolysosome of eucaryotic cells. Ultrastructural analysis of C. burnetii, in situ and after Renografin purification, by transmission electron microscopy of lead-stained thin sections has revealed extreme pleomorphism as demonstrated by two morphological cell types, a large cell variant (LCV) and a small cell variant (SCV). Potassium permanganate staining of purified rickettsiae revealed a number of differences in the internal structures of the cell variants. (i) The outer membrane of the SCV and LCV were comparable; however, the underlying dense layer of the SCV was much wider and more prominent than that of the LCV. The periplasmic space of the SCV was not readily visualized, whereas the periplasmic space of the LCV was apparent and resembled that of other gram-negative bacteria. (ii) Complex internal membranous intrusions which appeared to originate from the cytoplasmic membrane were observed in the SCV. The LCV did not harbor an extensive membranous system. (iii) Some LCVs contained a dense body in the periplasmic space. This endogenous structure appeared to arise in one pole of the LCV as an electrondense "cap" formation with the progressive development of a dense body approximately 130 to 170 nm in diameter which was eventually surrounded by a coat of at least four layers. Our observations suggest that the morphogenesis of C. burnetii is comparable, although not identical, to cellular differentiation of endospore formation. A developmental cycle consisting of vegetative and sporogenic differentiation is proposed.

Coxiella burnetii, phase I and II, cells cultivated in the yolk sac of chicken embryos were separated from host cell components by two cycles of isopycnic Renografin gradient centrifugation. Initial steps in the purification of viable C. burnetii involved differential centrifugation and sedimentation through an aqueous solution of 30% sucrose and 7.6% Renografin. After the first, but not the second, cycle of Renografin gradient centrifugation, the cells were passed through microfilter glass filters which facilitated the removal of host components. The integrity of morphologically different cell variants was maintained during purification procedures by suspending highly purified C. burnetii in phosphate-buffered saline-sucrose solutions. C. burnetii, phases I and II, obtained by these methods appeared to be free from host cell components by serological methods while retaining morphological integrity and infectivity for yolk sacs and experimental animals. Average yields of C. burnetii were 2.83, 1.5, and 0.84 mg (dry weight) per yolk sac of the Ohio strain (phase I), 9 Mile strain (phase I), and 9 Mile strain (phase II), respectively. Recovery of phase I cells averaged about 70%, whereas the recovery of phage II cells was approximately 40%. The temporal sequence of phase I and II antibody response was demonstrated in infected and vaccinated animals. Also, no antibody response in mice and guinea pigs to yolk sac antigens was detectable after two injections of vaccine or viable cells. Importantly, this is the first report of the separation of viable phase II cells of C. burnetii free of host components.

The opportunistic pathogen Pseudomonas aeruginosa produces type 4 fimbriae which promote adhesion to epithelial cells and are associated with a form of surface translocation called twitching motility. Transposon mutagenesis was used to identify loci required for fimbrial assembly or function by screening for mutants that lack the spreading colony morphology characteristic of twitching motility. Six mutants were isolated that contain transposon insertions upstream of the previously characterized gene pilQ. This region contains four genes: pilM-P, which encode proteins with predicted sizes of 37.9, 22.2, 22.8 and 19.0 kDa, respectively. pilM-P appear to form an operon and to be expressed from a promoter in the intergenic region between pilM and the divergently transcribed upstream gene ponA. PilM-P were found to be required for fimbrial biogenesis by complementation studies using twitching motility and sensitivity to fimbrial-specific phage as indicators of the presence of functional fimbriae. This was confirmed by electron microscopy. PilO and PilP did not have homologues in the sequence databases, but the predicted PilN amino acid sequence displayed similarity to XpsL from Xanthamonas campestris, a protein required for protein secretion. PilP contained a hydrophobic leader sequence characteristic of lipoproteins, while PilN and PilO have long internal hydrophobic domains which may serve to localize them to the cytoplasmic membrane. PilM has shared sequence motifs with the cell division protein FtsA from Bacillus subtilis and Escherichia coli, as well as the rod-shape-determining protein MreB from E. coli. These motifs are also conserved in eukaryotic actin, in which they are involved in forming an ATPase domain. Deletion mutants of pilM and pilQ displayed a dominant negative phenotype when transformed into wild-type cells, suggesting that these genes encode proteins involved in multimeric structures.

Coxiella burnetii morphological cell types were fractionated into large-cell variant cell walls, two fractions of small-cell variant cell walls, and one fraction of small-cell variant whole cells. Based on the contents of peptidoglycan (PG)-constituents and the yields of the sodium dodecyl sulfate-insoluble PG-protein complex (PG-PC) from cell walls, the fraction of large-cell variant cell walls contained significantly less PG than did the fraction of small-cell variant cell walls. The yields of PG-PC from the fractions of large-cell variant cell walls and small-cell variant cell walls were 2 and 32%, respectively. These results indicated that the PG of the largecell variant cell walls may be partially digested by PG-lytic enzymes or incompletely synthesized, whereas the small-cell variant cell walls appeared to have intact PG. Proteins associated with PG-PC were resistant to proteolysis by trypsin, protease VI, and proteinase K. Saturated and unsaturated fatty acids were detected in whole cells and cell walls but not in PG-PC, which contained a 3-deoxy-D-mannooctulosonic acid-like component that is also present in phase I lipopolysaccharide. Immunogenicity of the fractions was tested by measuring the temporal sequence of phase II and phase I antibody responses in vaccinated rabbits. Both phase II and phase I antibody responses were demonstrated with all fractions except the sodium dodecyl sulfate supernatant of the small-cell variant cell walls, whereas PG-PC elicited a pure phase II antibody response up to 29 days postvaccination. The immunogenicity of these fractions may reflect a quantitative difference in antigen concentration or may be due to a qualitative difference in phase II and I determinants.

The aim of this study was to investigate the antigenic structures of the morphologically distinct cells of the Coxiella burnetii developmental cycle. Postembedding immunoelectron microscopy with polyclonal antibodies produced in rabbits to (i) phase I cells, (ii) a chloroform-methanol residue fraction of cells, (iii) the cell walls (CW) of large and small cells and small dense cells (SDC), and (iv) the peptidoglycan-protein complexes of small cells and SDC labelled the continuum of morphologically distinct cells. But these antibodies did not distinguish between the antigenic structures of the various cells. Monoclonal antibodies to the phase I lipopolysaccharide labelled the CW of a majority of the smaller cells, but there was diminished reactivity to the larger cells. Although monoclonal antibodies to a 29.5-kDa outer membrane protein labelled the CW of the large mother cells, the large cells, and the small cells, a minority of the SDC with compact CW were not labelled. The endogenous spore within the mother cell was not labelled by the polyclonal or monoclonal antibodies to cellular components. A selected population of SDC was prepared by osmotic lysis of large cells, differential centrifugation, Renografin step-gradient fractionation, and breakage of the small cells in a French press at 20,000 lb/in2. The pressure-resistant SDC collected as fraction CL did not contain the 29.5-kDa protein, as evidenced by the lack of (i) Coomassie brilliant blue staining of protein in the 29.5-kDa region of sodium dodecyl sulfate-polyacrylamide gels and (ii) reactivity of the 29.5-kDa protein antigenic epitopes in immunoblotting with monoclonal antibodies to the protein. In contrast, CW of the pressure-sensitive small cells contained the 29.5-kDa protein. Therefore, the observed ultrastructural differences between large and small cells and SDC reflect differences in sensitivity to breakage by pressure treatment and in cell-associated antigens. Although the process of differentiation in C. burnetii remains an enigma, we have taken steps toward identifying cellular antigens as markers of differentiation. The pressure-resistant SDC in fraction CL that are devoid of the 29.5-kDa protein may be useful for answering questions about the physiological events required for triggering outgrowth and sequential regulation of the Coxiella developmental cycle.