Search

Each of five monoclonal antibodies (mAbs) prepared against the type 1 fimbriae of Actinomyces viscosus T14V reacted with a 54 kDa cloned protein previously identified as a fimbrial subunit. This purified protein completely inhibited the reaction of a specific anti-type-1-fimbria rabbit antibody with A. viscosus whole cells. Maximum values for the number of antibody molecules bound per bacterial cell ranged from 7 x 10(3) to 1.2 x 10(4) for the different 125I-labelled mAbs and was approximately 7 x 10(4) for 125I-labelled rabbit IgG or Fab against either type 1 fimbriae or the 54 kDa cloned protein. Although the different mAbs, either individually or as a mixture, failed to inhibit the type-1-fimbria-mediated adherence of A. viscosus T14V to saliva-treated hydroxyapatite, each rabbit antibody gave 50% inhibition of adherence when approximately 5 x 10(4) molecules of IgG were bound per cell. However, binding of each corresponding rabbit Fab had no significant effect on bacterial attachment unless much higher concentrations were used. These findings suggest that antibodies directed solely against the 54 kDa fimbrial subunit do not react with the putative receptor binding sites of A. viscosus T14V type 1 fimbriae. Instead, inhibition of attachment by the polyclonal antibodies may depend on an indirect effect of antibody binding that prevents the fimbria-receptor interaction.

1) Fimbriae from A. viscosus T14V may be similar to those found on other bacteria. 2) The antigenic difference between virulent and avirulent A. viscosus T14 appears to be of a quantitative rather than a qualitative nature and is related to fimbriae and not to the cell wall polysaccharide. 3) Coaggregation between A. viscosus T14V and S. sanguis 34 is mediated by fimbriae on the former which have specificity for beta-linked galactosyl residues.

Concomitant with glucose-induced insulin release, there occurred an increase of ATP from 4.40 plus or minus 0.21 to 23.16 plus or minus 0.52 pmol/100 islets/min (P less than 0.001) in the effluent from perifused rat islets. There is a linear relationship between circulating ATP and insulin levels both in the stimulated and basal state (r = 0.689, P less than 0.01). Islets incubated with labeled adenine for a short period of time (37.5 min) showed no release of radioactivity upon subsequent glucose-induced insulin release. Islets incubated for a prolonged interval with labeled adenine (150 min) showed an increase in acid soluble radioactivity in the effluent during glucose-induced insulin release. Following incubation of the islets with labeled adenine for 150 min, approximately 5% of the homogenate radioactivity was found in the secretory granules. Using column chromatography to separate the adenine nucleotides, the distribution of radioactivity among the various nucleotides in the secretory granule fraction was found to be: AMP 54.42 plus or minus 4.96%, ADP 14.20 plus or minus 1.63%, ATP 15.39 plus or minus 3.84%, and cAMP 16.07 plus or minus 2.11%. The distribution of radioactivity in the effluent adenine nucleotides after glucose-induced insulin release was: AMP 32.83 plus or minus 4.62%, ADP 24.52 plus or minus 2.77%, ATP 28.13 plus or minus 5.45%, and cAMP 26.01 plus or minus 3.34%. The absolute levels of adenine nucleotides in the secretory granules were ATP 4.19 plus or minus 0.88, ATP madp 7.94 plus or minus 2.20 and cAMP 4.46 plus or minus 1.74 pmol/ug prot. The levels in the islet effluent were ATP, 15.30 plus or minus 2.70, ATP qDP, 29.43 plus or minus 3.49 and cAMP 7.66 plus or minus 1.93 pmol/100 islets/min for the first ten min of glucose-stimulated insulin release. Thereafter there was a rapid decline in effluent cAMP while ATP and ADP remained in essentially equivalent amounts. The distribution of radioactivity and absolute levels of the adenine nucleotides in the effluent reflects that found in the secretory granules, confirming previous observations that insulin release is occurring by exocytosis.

Recombinant human gamma interferon (rIFN-gamma) was examined for its ability to activate human peripheral blood monocyte-derived macrophages to kill tumor cells and to affect the replication of two phylogenetically distinct intracellular pathogens, Mycobacterium tuberculosis and Leishmania donovani. Macrophages preincubated overnight with doses of rIFN-gamma from 5 to 500 U/ml killed [3H]thymidine-labeled mouse L929 tumor targets, as measured by the release of [3H]thymidine into the supernatant after 48 h. Counts of macrophages initially infected with leishmania promastigotes showed that rIFN-gamma-pretreated macrophages could both inhibit the replication of and kill the resulting intramacrophage amastigotes over a 7-day period. However, rIFN-gamma pretreatment of macrophages actually enhanced mycobacterial replication over a 5- to 7-day period, as assessed by (i) counting acid-fast bacilli or (ii) lysing macrophages to release bacteria and determining the numbers of viable units. Mycobacterial growth was not affected by rIFN-gamma in the absence of macrophages. rIFN-gamma pretreatment had opposite effects on the uptake of mycobacteria and leishmania. As many as 80% fewer activated macrophages ingested mycobacteria compared with controls, whereas 50% more activated macrophages were infected with leishmania. These results suggest that rIFN-gamma may interfere with the immune destruction of intracellular tubercle bacilli and that the mechanisms of immunity against mycobacteria and leishmania may differ.

The role of the thymocyte in its species-specific binding to macrophages has been explored. Although formalin treatment of macrophages resulted in loss of binding to thymocytes, formalin treatment of thymocytes did not have this effect. However, two differences between living and formalin-treated thymocytes were noted. Formalin-treated thymocytes bound to macrophages of any species whereas the binding of living thymocytes was species specific. Living thymocytes attained maximum binding in approximately 1 hr and then the fraction bound gradually diminished. Formalin-treated thymocytes remained bound to the macrophage and appeared to be phagocytized. Released thymocytes did not bind to fresh macrophages, but released macrophages bound to fresh thymocytes. The results suggest that the binding of thymocytes to macrophages results in maturation of thymocytes.

Monospecific antisera against the Ag1 and Ag2 fimbrial components of Actinomyces viscosus T14V were produced by immunizing rabbits with immune precipitates of each antigen harvested from crossed-immunoelectrophoresis plates. The Fab fragments prepared from these sera were used as specific reagents in immunoelectron microscopy to identify each fimbrial component on the bacterial surface and also were assayed for their abilities to prevent the coaggregation of A. viscosus T14V with Streptococcus sanguis 34, an interaction that is lactose sensitive. Each Fab preparation appeared to react with different fibrillar structures present on the actinomycete, and only the Ag2-specific Fabs blocked coaggregation. These results provide strong support for the presence of distinct types of fimbriae on A. viscosus T14V and indicate the exclusive involvement of Ag2 fimbriae in lactose-sensitive adherence.

Surface antigens of Actinomyces viscosus T14V were released from cell walls by digestion with lysozyme. These were separated by ion-exchange and gel filtration chromatography into fractions rich in carbohydrate or protein. The former contained a polysaccharide high in 6-deoxytalose, along with a peptide fragment from the cell wall. In the protein-rich fractions, material of high molecular weight was present, which contained some carbohydrate and up to 14.3% nitrogen. Aspartic acid, threonine, glutamic acid, lysine, alanine, and glycine were detected in these fractions, along with smaller amounts of 10 other amino acids. Most of the alanine was present as the L isomer and thus was not from peptidoglycan. Electron microscopy of the high-molecular-weight material revealed long fibrils, 3.5 to 4.5 nm in diameter, which resembled those seen on bacterial cells. V-specific antiserum, prepared by absorbing anti-A. viscosus T14V serum with cell walls of the avirulent strain (A. viscosus T14AV), did not react with the 6-deoxytalose polysaccharide but reacted well with isolated fibrils, and this was not inhibited by 6-deoxytalose.

The simple apolar C-mycosides, i.e., structurally well-defined hydrophobic glycopeptidolipids of several Mycobacterium species (see diagram below), were earlier shown to behave as receptors for adsorption of mycobacteriophage D4. This phage is usually virulent for Mycobacterium smegmatis. More complex, polar C-mycosides with additional carbohydrate substituents attached solely to the deoxytalose have recently been described. They are the highly specific serotyping antigens discovered by W. B. Schaefer--lipids which characterize members of the Mycobacterium avium-Mycobacterium intracellulare-Mycobacterium scrofulaceum (MAIS) complex. Both kinds are depicted in the structure below: (Formula: see text) where X equals H (for simple, apolar C-mycosides) and X equals small oligosaccharides (for antigenic forms; more complex, polar C-mycosides). The present investigations showed that the purified polar antigenic lipids exhibit considerably less adsorptive activity for D4 than do the apolar C-mycosides. Thus, the haptenic oligosaccharides are believed to shield the site in the molecule that the phage recognizes, and the blocking is reinforced by the specific antibodies that the antigens elicit. Although the MAIS serovars usually also produce the phage-reactive apolar C-mycosides, they are not permissive hosts for D4, nor do whole cells adsorb the phage. We suggest that in these species the apolar forms are probably "covered" at the cell surface by the antigenic lipids. Therefore, these antigenic mycosides may play a putative role in virulence of the MAIS members by protecting these mycobacteria from their own potential pathogen. The results of chemical transformations at specific sites of the mycoside core coupled with studies of simple synthetic lipid glycosides indicated that the principal phage receptor activity resides in the terminal methylated rhamnose (see diagram). It is this sugar which is evidently masked by the (seemingly remote) haptenic oligosaccharides.

Various natural and synthetic substances classified as polyanionics have been implicated in antagonizing phagosome-lysosome fusion in cultured macrophages. The phenomenon has been judged by comparing the transfer of selected markers from secondary lysosomes to phagosomes in control and in "polyanion" cells. Our earlier studies showed that use of one of the markers, the membrane-permeating acridine orange, was plagued with artifacts that were especially misleading in the presence of polyanionic agents. We now question the validity of data obtained by the alternative technique, electron microscopy. Our present evidence shows that nonionic hydrocolloids of sufficiently high molecular weight prevent the transfer of various colloidal electron-opaque markers from lysosomes to phagosomes in the same manner as does the powerful polyanionic "fusion inhibitor" dextran sulfate. Both kinds of hydrocolloids, however, allow delivery of lysosomal, low-molecular-weight highly charged non-permeant fluorescent markers to phagosomes, probably by a fusion process. We propose that neither type of hydrocolloid inhibits fusion; instead, when sufficiently concentrated, they trap particulate electron-opaque markers in a gelatinous matrix, which may move only slowly out of lysosomes. The polyanionics trap the electron-opaque markers physically and acridine orange ionically. Hence, the semblance of "fusion inhibition."

Monocytes were isolated by counterflow contrifugation from human peripheral mononuclear cells obtained from whole blood by the Ficoll-Hypaque technique. Examination of volume spectra of the monocytes showed a bimodal distribution with peaks at 330 and 370 micron3. The upper peak contained 61.4 +/- 3.3% of the total monocyte population, whereas the remainder were in the lower peak which overlaps the lymphocyte spectrum. The purity of the monocyte preparations from six normal adult donors, identified by peroxisomes observed by electron microscopy, was 90.4 +/- 1.6%.

Phagocytosis of Actinomyces viscosus T14V and A. naeslundii WVU45 by human polymorphonuclear leukocytes in the absence of antibody or complement was mediated by the lectin associated with the type 2 fimbriae of these bacteria. This effect was markedly enhanced by exogenous sialidase, an enzyme also secreted by these actinomyces. Since sialidase treatment of the bacteria did not result in increased phagocytosis, this enzyme presumably acts by unmasking receptors for the fimbrial lectin on phagocytic cells. The viability of A. viscosus T14V, which possesses type 1 and type 2 fimbriae (1+ 2+), and A. naeslundii WVU45, which possesses only type 2 fimbriae (2+), was decreased by at least 98% following incubation with polymorphonuclear leukocytes in the presence of sialidase. Entirely analogous findings were obtained with a 1- 2+ mutant of A. viscosus T14V. In contrast, the phagocytosis of 1+ 2- and 1- 2- mutants of A. viscosus T14V and a 2- mutant of A. naeslundii WVU45 was minimal or absent. Lactose and beta-methylgalactoside inhibited the destruction of the bacteria, whereas cellobiose and alpha-methylgalactoside were ineffective. Thus, the type 2 fimbriae of the oral actinomyces recognize galactose-containing receptors on polymorphonuclear leukocytes which have been exposed by the removal of sialic acid, an interaction that is followed by internalization and subsequent killing of the bacteria.

A highly specific aggregation factor for Streptococcus sanguis H1 (AFH1) was obtained by lysozyme treatment of Actinomyces viscosus T14V. At 1 micrograms/ml, AFH1 aggregated a suspension of S. sanguis H1, with which A. viscosus T14V coaggregates by a mechanism not inhibited by lactose: even at much higher levels AFH1 caused little or no aggregation of streptococci from other coaggregation groups (J. O. Cisar et al., Infect. Immun. 24:742-752, 1979). The most active fraction of AFH1 obtained by gel chromatography (near the void volume of Bio-Gel A1.5 m) reacted as a single antigen with anti-A. viscosus T14V serum and was unrelated to the fimbrial antigens of A. viscosus T14V. Smaller molecular fractions, at high levels, inhibited aggregation of S. sanguis H1 by high-molecular-weight AFH1 as well as coaggregation of S. sanguis H1 with A. viscosus T14V. The AFH1 fraction with high aggregating activity was composed of approximately 53% cell wall components (alanine, glutamine, lysine, N-acetylglucosamine, and N-acetylmuramic acid). 40% polysaccharide (N-acetylgalactosamine, rhamnose, and 6-deoxytalose), and 7% protein; teichoic acid was not detected. The fraction which inhibited aggregation and coaggregation contained much less of the cell wall constituents and more of the polysaccharide than the fraction with potent aggregating activity. Aggregation was completely prevented either by treating AFH1 with 0.01 M periodate at 25 degrees C for 4 h or by treating S. sanguis H1 with heat or pronase. A role for electrostatic forces in the aggregation was indicated by: (i) NaCl inhibition of aggregation, and (ii) a great decrease in aggregation potency as a result of chemical modification of either cationic or anionic groups of AFH1. On the other hand, NaCl reversed the aggregation only very weakly. The overall data suggest that a carbohydrate-protein interaction may be dominant in the aggregation of S. sanguis H1 by AFH1 and in the coaggregation of S. sanguis H1 with A. viscosus T14V.

Macrophages synthesize many secretory products in vitro but the stimuli for their production and their pathophysiologic significance in vivo are largely unknown. In the present investigation, we found that hyperoxia damaged rabbit alveolar macrophages (AM) in vitro as manifested by decreased cell numbers, increased lactate dehydrogenase (LDH) release, and the development of ultrastructural abnormalities that resembled those seen in AM in situ or lavaged from lungs of rabbits exposed to hyperoxia in vivo. Hyperoxia also stimulated cultured rabbit AM to release chemotaxins for polymorphonuclear leukocytes (PMN) that were similar in molecular weight to chemotaxins obtained from lung lavages of rabbits exposed to hyperoxia in vivo. Our results suggest that alveolar macrophage secretory products may play a physiologically relevant role in recruitment of PMN to the lungs in pulmonary oxygen toxicity.

Lactose-sensitive fimbriae were identified as the only fimbriae present on Actinomyces naeslundii WVU45 (ATCC 12104). A single antigen reactive with antiserum against WVU45 cells was detected by cross immunoelectrophoresis of isolated fimbriae, and a monospecific antiserum against this antigen reacted with all fimbriae observed on the bacterial surface by immunoelectron microscopy. Moreover, the loss of one cell surface antigen by a spontaneous mutant of A. naeslundii WVU45 (WVU45M), isolated by its failure to react with a monospecific antibody against the fimbriae, was associated with the loss of all fimbriae. The functional involvement of the fimbriae in lactose-sensitive bacterial adherence was demonstrated by the ability of WVU45, but not WVU45M, cells to agglutinate neuraminidase-treated erythrocytes and by the lactose-sensitive hemagglutinating activity of immune complexes formed with isolated fimbriae and monospecific antibody. Bacterial agglutination assays with different monospecific antibodies revealed an antigenic similarity between the fimbriae of A. naeslundii WVU45 and the lactose-sensitive fimbriae (type 2) of Actinomyces viscosus T14V. In contrast, cross-reactivity was not observed between the WVU45 fimbriae and type 1 fimbriae, the structures involved in lactose-resistant adherence of strain T14V to saliva-treated hydroxyapatite. Functional differences between the fimbriae of A. naeslundii and A. viscosus strains may be correlated with well-established differences in the in vivo distribution of these organisms: namely, the preference of typical A. naeslundii for epithelial surfaces and of A. viscosus for tooth surfaces.

Lamellar bodies of type II alveolar epithelial cells are the intracellular storage sites of lung surfactant, while tubular myelin figures are an extracellular surfactant form found in the alveolar fluid. A refined procedure was used to isolate intact lamellar bodies from rat lung homogenates in a fraction of high purity and yield. The stability of isolated lamellar bodies under various conditions was determined by electron microscopy. Lamellar bodies were completely disrupted after incubation at 37, 24, and 0 degrees C for 0.6, 2, and 12 hr, respectively, in 0.33 M sucrose, 0.01 M HEPES (pH 7.4). In addition, they were completely disrupted after incubation for 1 hr in 0.33 M sucrose, 1 mM EGTA at 0 degrees C or 0.154 M NaCl or 0.10 M sodium phosphate (pH 7.4) at 24 degrees C. Incubation of isolated lamellar body fractions in medium containing 5 mM Ca++ or Mg++ at 37 degrees C for 1 hr resulted in the appearance of tubular myelin figures. A procedure is also presented for the isolation of tubular myelin figures from rat lung lavage fluid.

Actinomyces viscosus T14V is virulent (V) for monoinfected rats, causing periodontal disease and bone loss, whereas, A. viscosus T14AV, a mutant strain, is avirulent (AV). Surface antigens from the T14V and T14AV strains were prepared by lysozyme digestion of cell walls and were compared by immunodiffusion against antisera to T14V and T14AV whole cells. The V-associated antigen (V-antigen) was detected readily in the T14V, but not readily in the T14AV cell wall extract. Antiserum specific for the V-antigen was prepared by absorbing anti-A. viscosus T14V serum with cell walls from the T14AV strain. This antiserum was used in the indirect peroxidase-labeled antibody technique to localize the V-antigen on the bacterial cell surface at the ultrastructural level. With whole bacterial cells, the V-antigen was found on fine fibrils and was detected in both the T14V and T14AV strains. The presence of V-antigen on the AV strain was supported by the demonstration of antibodies against the V-antigen in anti-A. viscosus T14AV serum. Examination of isolated bacterial cell walls revealed a greater amount of fibrils and V-antigen on the T14V cell wall than on the T14AV cell wall. The data suggest that the presence of V-antigen represents a quantitative rather than a qualitative difference between the V and the AV strains of A. viscosus T14. Samples of human plaque were examined, and the V-antigen was found to be a specific marker for the fibril-containing layer of certain plaque bacteria, which are probably strains of A. viscosus or A. naeslundii.

An abhesive or 'anti-glue' function for lung surfactant is proposed which is reconcilable with the known unfolding of alveolar walls at low lung volumes. The theory is developed based upon the fact that the work required to part wetted surfaces is directly proportional to the surface tension. Application of the concept of the spreading coefficient from surface physical chemistry S = gammata-(gamma lla + gamma lle), where the gamma's refer to the surface (interfacial) tensions among the tissue (t), air (a) and lining layer (ll) surface, is shown to explain qualitatively physiological data in lungs subjected to maneuvers with working fluids varying widely in surface tension and spreading properties.

Mycobacterium avium is a cause of nontuberculous chronic granulomatous infections which is attracting increased attention as a frequent opportunistic pathogen in acquired immunodeficiency syndrome. Some important aspects of its human pathogenicity were investigated by using cultured human macrophages infected with it. The uptake and replication of various strains of M. avium in the macrophages could be measured by CFU counts of the bacteria in samples of lysed, sonicated macrophages. Microscopic counts of acid-fast bacilli were not useful because the bacteria multiplying in the macrophages were usually not acid fast. Electron microscopy showed the intracellular bacilli to multiply by transverse fission, to be surrounded in individual vacuoles by a broad electronlucent zone, and to have thinner cell walls than extracellularly grown M. avium. Fifteen strains, including examples of serovars 1, 2, 4, 8, and 9, were studied for uptake and rate of replication in cultured macrophages from three normal subjects. The strains were isolates from patients with nontuberculous granulomatous infection, acquired immunodeficiency syndrome, or unrelated problems, or they were laboratory reference cultures. There were no differences among them in phagocytosis, but there were differences in intracellular replication. Laboratory strains tended to be avirulent, that is, they did not replicate in the macrophages. Patient isolates usually were virulent and could be compared for virulence by intracellular replication rates. Virulence correlated with flat, transparent bacterial colony morphology on nutrient agar but not with serovar or kind of patient from whom the bacteria were isolated. However, among strains of transparent colony morphology there were wide differences in virulence. A virulent bacilli generally produced domed, opalescent colonies on nutrient agar. A virulent bacilli predominated in populations of M. avium conditioned to growth in bacteriologic culture medium. Bacilli of virulent colony morphology predominated in populations passaged through cultured macrophages. The model described here presents a new approach to the investigation of the pathogenicity of M. avium for human subjects and may be more patient relevant than animal models.

The survival of some intracellular pathogens within macrophages may be aided by an ability of the organism to antagonize, from within the entrapping phagosome, its fusion with lysosomes. On the other hand, certain polyanionic agents have been implicated in imposing a similar block to fusion from the lysosomal domain--because the transfer of various foreign markers from lysosomes to newly formed phagosomes is remarkably inhibited in these polyanion-containing cells. Based on an analysis of various observations and our own recent data, we propose that the polyanionics do not, in fact, prevent phagosome-lysosome fusion but, instead, physically entrap the usual markers in a gelatinous matrix within the lysosomes. This view accounts for many paradoxical consequences of polyanionic accumulation and for the curiously normal behavior of macrophages that are presumed to be suffering from such a crucial intracellular dysfunction.

A mechanism is suggested by which the membranes of lamellar bodies are converted to tubular myelin (TM) in the lung. It is argued that a simple corrugation of the membranous sheets can produce the TM formation. Such corrugation would occur in response to simple stresses acting on the lamellar body membranes. The intersections of the tubular figures are formed by fusion of adjacent corners in the corrugations. This results in a more stable hydrophobic bonding of phospholipid molecules. Strong supportive evidence for the mechanism is given by electron micrographs of TM formations.

A cell-associated lectin activity that mediates lactose-inhibitable adherence of Actinomyces viscosus T14V has been localized to a specific population of fimbriae by the use of monoclonal antibodies. Nine monoclonal antibodies were produced that reacted with only 1 of 2 immunoelectrophoretically distinct fimbrial components on T14V. The fibrillar morphology of this component was revealed by the immunoelectronmicroscopic examination of bacteria incubated with the monoclonal antibodies. The lectin activity associated with these structures was detected when isolated fimbriae were cross-linked with monoclonal antibodies to form immune complexes with agglutination activity for neuraminidase-treated human erythrocytes, a reaction that was inhibited by lactose. Although the 9 monoclonal antibodies differed in their fine specificities, they reacted only with strains of A. viscosus and A. naeslundii that exhibited lactose-inhibitable adherence. These findings indicate that the lectin activity common to these bacteria resides on fimbriae that are antigenically related to those of T14V.

Damage to alveolar macrophages (AM) from hyperoxia (95% O2) is associated with release of factors that recruit and activate neutrophils, but the mechanisms underlying injury to AM from hyperoxia are unknown. We hypothesized that damage to AM from hyperoxia involves generation of highly reactive toxic oxygen derivatives, and we tested this premise by exposing cultured rabbit AM to hyperoxia in the presence of scavengers that inactivate various reactive oxygen species. We found that either dimethyl thiourea, a scavenger of hydroxyl radical, or catalase, a scavenger of H2O2, protected cultured rabbit AM against hyperoxic damage, which suggests that H2O2 or an H2O2-derived product, such as hydroxyl radical, contribute to damage to AM from hyperoxia.

Coaggregation between Actinomyces viscosus T14V and Streptococcus sanguis 34 depends on interaction of a lectin on A. viscosus T14V with a cell surface carbohydrate on S. sanguis 34. This carbohydrate was isolated, and its chemical makeup was established. The carbohydrate remained attached to S. sanguis 34 cells through extraction with Triton X-100 and treatment with pronase. It was cleaved from the cell residue by autoclaving and purified by differential centrifugation and column chromatography on DEAE-Sephacel and Sephadex G-75. The polysaccharide contained phosphate which was neither inorganic nor monoester. Treatment with NaOH-NaBH4, followed by Escherichia coli alkaline phosphatase, or with 48% HF at 4 degrees C, followed by NaBH4, yielded inorganic phosphate and oligosaccharide alditols. Therefore, the polysaccharide is composed of oligosaccharide units joined together by phosphodiester bridges. The structure and stereochemistry of the main oligosaccharide alditol was established previously (F. C. McIntire, C. A. Bush, S.-S. Wu, S.-C. Li, Y.-T. Li, M. McNeil, S. Tjoa, and P. V. Fennessey, Carbohydr. Res. 166:133-143). Permethylation analysis, 1H and 31P nuclear magnetic resonance studies on the whole polysaccharide revealed the position of the phosphodiester linkages. The polysaccharide is mainly a polymer of (6) GalNAc(alpha 1-3)Rha(beta 1-4)Glc(beta 1-6)Galf(beta 1-6)GalNAc(beta 1- 3)Gal(alpha 1)-OPO3. It reacted as a single antigen with antiserum to S. sanguis 34 cells and was a potent inhibitor of coaggregation between A. viscosus T14V and S. sanguis 34. Quantitative inhibition of precipitation assays with oligosaccharides, O-allyl N-acetylgalactosaminides, and simple sugars indicated that specific antibodies were directed to the GalNAc end of the hexasaccharide unit. In contrast, coaggregation was inhibited much more effectively by saccharides containing betaGalNAc. Thus, the specificity of the A. viscosus T14V lectin is strikingly different from that of antibodies directed against the S. sanguis 34 polysaccharide.

Mutants of Actinomyces viscosus T14V lacking type 1 or type 2 fimbriae or both were selected by their failure to react with rabbit antibodies against either or both fimbrial antigens. Immunospecific double labeling with iron dextran and ferritin-conjugated antibodies showed two types of fimbriae on individual cells of the parent organism, a single type on mutant strains with type 1+2- and type 1-2+ fimbriae and no labeled or unlabeled fimbriae on a type 1-2- fimbria-deficient strain. The mutational loss of one fimbrial antigen did not appear to affect expression of the other, since bacteria with one or two types of fimbriae bound similar amounts of a monoclonal antibody directed against the fimbrial antigen present on both bacterial phenotypes. The strong adsorption of strains with type 1+2+ or 1+2- fimbriae to saliva-treated hydroxyapatite and weak adsorption of those with type 1-2+ or no fimbriae was consistent with the known involvement of type 1 fimbriae in this attachment process. Similarly, the A. viscosus lectin was clearly associated with the expression of type 2 fimbriae, since only the strains with type 1+2+ or 1-2+ fimbriae participated in lactose-sensitive coaggregations with Streptococcus sanguis 34. Further studies using the fimbria-deficient mutant strains showed that aggregation of A. viscosus T14V in the presence of sialidase-treated human saliva involved both types of fimbriae, whereas neither type was required for the lactose-resistant coaggregation of the organism with certain streptococcal strains.

Neutrophils obtained from most of 13 healthy young women during menstruation and those obtained from the same women on nonmenstrual days killed comparable numbers of Staphylococcus aureus strain 502A, produced comparable amounts of superoxide anion, and had comparable lysozyme levels. In contrast, neutrophils obtained from a few women during menstruation exhibited decreased function. In particular, neutrophils from one healthy woman developed transient menstrual period-related abnormalities in bactericidal function, superoxide anion production, and lysozyme activity and release; these abnormalities occurred during each of three menstrual periods tested but not during three menstrual periods following the ingestion of aspirin for 14 days. The results suggest that menstrual period-related dysfunction of neutrophils may occur in some healthy women, sometimes rendering them more susceptible to menstrual period-related infections.

Coaggregation between Actinomyces viscosus T14V (T14V) and Streptococcus sanguis 34 (Ss34) depends upon specific reaction between lectin on T14V and carbohydrate on Ss34. Studies on coaggregation inhibition by sugars related to D-galactose, beta-galactosides, and amphipathic molecules revealed: (i) D-fucose, D-talose approximately equal to D-galactose, which was 0.2 potency of lactose. No other hexoses or pentoses inhibited at 0.1 M. (ii) Gal beta (1 leads to 3)GalNAc alpha OCH2C6H5 was the most potent beta-galactoside inhibitor; it had 20 times the potency of lactose. (iii) Anionic nonaromatic amphipathic compounds were good inhibitors; sodium deoxycholate (I) was equal to lactose; sodium dodecyl sulfate (II) had 15 times the potency of lactose; there was 90 to 100% irreversible inhibition when T14V was treated with 0.005 M (II). Treatment of Ss34 with II had no effect. (iv) Synergism of inhibition was observed between lactose and I or lactose and II, e.g., inhibition by 0.01 M lactose = 5%; inhibition by 0.01 M I = 9%; inhibition by 0.01 M lactose + 0.01 M I = 87%. (v) The irreversible inhibition by II was prevented when 0.25 M lactose or 0.25 M I was present during treatment of T14V with 0.005 M II. (vi) Synergism and prevention by lactose or by I of irreversible inhibition by II suggest that all three react at the same site on T14V lectin. We hypothesize that the T14V lectin combining site for Ss34 carbohydrate has specific affinity for beta-galactosides and for anionic nonaromatic amphipathic molecules. This site can be saturated by either kind of reagent to exclude the other reagent or to inhibit coaggregation.

The coaggregation-defective (COG-) mutant Actinomyces viscosus T14V(PK455), which is unable to participate in lactose-sensitive adherence, and its COG+ parent were compared to structurally define the mutational loss of cell-associated lectin activity. Immunoelectrophoretic comparisons of crude extracts or isolated fimbriae from both strains demonstrated type 2 fimbriae (previously designated Ag2) in preparations obtained from the parent but none in those obtained from the mutant. This result was verified by the immunoelectron-microscopic identification of type 1 (previously designated Ag1 or VAl) and type 2 fimbriae on the parent organism but only type 1 fimbriae on the mutant. A comparison of the amounts of extractable fimbriae of each type and the capacity of the cells to bind 14C-labeled monoclonal antibodies specific for each fimbrial component showed that the mutational loss of type 2 fimbriae had no significant quantitative effect on fimbriation of the COG- mutant with type 1 structures. Cells of A. viscosus T14AV, a mutant with various adherence defects that include the COG- phenotype, displayed fimbriae of both types, but in greatly reduced amounts. Thus, the properties of mutant strain T14V(PK455) associated the lectin activity with type 2 fimbriae, whereas those of strain T14AV provided little insight into the mechanism of lactose-sensitive adherence. In addition, the precise nature of the cell surface modification displayed by strain T14V(PK455) provides clear evidence for the existence of distinct and independent fimbriae on A. viscosus T14V.

The adherence of Actinomyces naeslundii WVU45 to monolayer cultures of human epithelial cell lines was mediated by the lactose-sensitive fimbriae (type 2) of strain WVU45. The attachment of Actinomyces viscosus T14V, which has both types 1 and 2 fimbriae, was approximately half that of A. naeslundii, and only minimal attachment of A. naeslundii and A. viscosus mutants lacking type 2 fimbriae was detected. The adherence of strain WVU45 was enhanced two- to threefold by neuraminidase treatment of the epithelial cells. The Fab fragments of antibodies which recognize the type 2 fimbriae inhibited the adherence of A. naeslundii WVU45 to the epithelial cells. The bacterial interaction with epithelial cells was inhibited by lactose, methyl-beta-D-galactoside, and N-acetyl-D-galactosamine, but not by methyl-alpha-D-galactoside, cellobiose, N-acetyl-D-glucosamine, L-fucose, or D-mannose. To further characterize the epithelial cell receptors for the bacterial lectin, we utilized several plant and invertebrate lectins as potential inhibitors of bacterial adherence. Lectins from Bauhinia purpurea and Arachis hypogaea which recognize N-acetyl-D-galactosamine, D-galactose, and D-galactose-beta-(1----3)-N-acetyl-D-galactosamine inhibited bacterial attachment, and binding of these lectins to epithelial cells was enhanced by the addition of neuraminidase. Lectins reacting with alpha-linked D-galactose, alpha-linked N-acetyl-D-galactosamine, D-mannose, or sialic acid were not inhibitory. Under similar assay conditions, adherence of a mannose-sensitive strain of Escherichia coli was inhibited by concanavalin A but not by the lectin from Bauhinia purpurea. These results indicate that certain plant lectins have specificities similar to that of the actinomyces fimbrial lectin and are, therefore, useful probes for identifying receptors on epithelial cells for certain bacteria.

Mast cells were studied during the induction of chronic graft-vs-host disease (GVHD) induced in mice across minor histocompatibility barriers. B10.D2 spleen cells (or control BALB/c cells) were injected into irradiated (600 rad) BALB/c recipients. Serial skin biopsies were taken over 26 days, during which time changes occurred resembling scleroderma, namely, dermal fibrosis, a mononuclear cell infiltrate, and loss of fat and appendages. Mast cells, when stained with toluidine blue, "disappeared" from GVHD, but not from control skin. Ultrastructural analysis showed that mast cells in GVHD skin were indeed present but underwent degranulation. Some mast cells showed only pale expanded sacs, indicating granule depletion. Because these cells could not be seen by toluidine blue staining but were plainly present, we have called them "phantom mast cells." Cellular activation occurred in many GVHD mast cells as shown by increased cytoplasmic activity, with numerous Golgi complexes, ribosomes, granular endoplasmic reticulum, and small vesicles. No identifiable mast cells were seen after day 19. No significant changes were seen in the mast cells of syngeneic control mice. We believe that immunologic processes in chronic GVHD cause a slow release of mast cell granule contents, which is different from anaphylactic degranulation. The depleted mast cells (invisible by toluidine blue staining) are also activated, perhaps in an attempt to replete their stores of granule contents. We discuss the relation of mast cell changes to fibrosis.

Actinomyces viscosus T14V and Streptococcus sanguis 34 coaggregate by a mechanism which is not inhibited by 1 M NaCl, is dextran independent, requires calcium, is pH dependent with an optimum at pH 8.0 to 8.5, and appears to require the interaction of a protein or glycoprotein on A. viscosus with a carbohydrate on S. sanguis. The coaggregation is inhibited more than 80% by 0.01 M lactose, 0.02 M beta-methyl-D-galactoside, or 0.05 M D-galactose; inhibition of coaggregation was less than 10% in 0.1 M alpha-methyl-D-galactoside, melibiose, maltose, cellobiose, sucrose, and a number of monosaccharides. At very high concentrations of enzyme, protease from S. griseus destroyed the reactive site on A. viscosus but not on S. sanguis. Both were totally resistant to dextranase. Periodate (0.01 M; pH 4) inactivated both bacteria. The ability of S. sanguis to coaggregate with A. viscosus was not destroyed by phenol-water extraction at 65 degrees C for 15 min. When the bacteria were cultured under specified conditions, the coaggregation was highly reproducible. Under the same conditions, T14AV, the avirulent mutant of A. viscosus T14V, did not coaggregate with S. sanguis 34. Electron microscopic studies of coaggregates, labeled immunochemically with antibody to A. viscosus, indicated that fibrils on A. viscosus may be involved in the coaggregation.

Abram, Dinah (Purdue University, Lafayette, Ind.), Henry Koffler, and A. E. Vatter. Basal structure and attachment of flagella in cells of Proteus vulgaris. J. Bacteriol. 90:1337-1354. 1965.-The attachment of flagella to cells of Proteus vulgaris was studied electron microscopically with negatively stained and shadow-cast preparations of ghosts from standard cultures and from special cultures that produced "long forms." The flagellum, the basal portion of which is hooked, arises within the cell from a nearly spherical structure, 110 to 140 A in diameter. This structure appears to be associated with the cytoplasmic membrane; it may be a part of the membrane or a separate entity that lies just beneath the membrane. Flagella associated with cell walls free from cytoplasmic membrane frequently have larger bodies, 200 to 700 A in diameter, associated with their base. These structures probably consist at least partly of fragments of the cytoplasmic membrane, a portion of which folds around a smaller structure. Flagella in various stages of development were observed in long forms of P. vulgaris cells grown at low temperature. The basal structure of these flagella was similar to that of the long or "mature" flagella. Strands connecting the basal structures were observed in ghosts of long forms; these strands appear to be derived from the cytoplasmic membrane. Flagella were found to be attached to fragments of cell wall and to cytoplasmic membrane in a similar manner as they are attached to ghosts. In isolates of flagella that have been separated from the cells mechanically, the organelles often terminate in hooks which almost always appear naked, but have a different fine structure than the flagellum proper.

Abram, Dinah (Purdue University, Lafayette, Ind.), A. E. Vatter, and Henry Koffler. Attachment and structural features of flagella of certain bacilli. J. Bacteriol. 91:2045-2068. 1966.-The attachment of flagella to cells of various mesophilic and thermophilic strains of Bacillus was studied electron microscopically. Studies of ghost cells and membrane fragments indicate that flagella are connected to the cytoplasmic membrane. Flagella removed from cells mechanically, during autolysis, or by phage lysis, have attached to the base of their proximal hooks material that is heterogeneous in character. In part, this material consists of cytoplasmic membrane; its varied shape appears to be caused by the folding of the membrane around the proximal end of the flagellum at the site of attachment. It is uncertain whether this material represents a real structure or an artifact. Highresolution microscopy reveals differences in the fine structure of intact flagella of the various strains studied. The proximal hook and the flagellar filament are distinct in morphology and fine structure. A specialized structure is associated with the hook of flagella of B. brevis and B. circulans. The filament of flagella of B. stearothermophilus 2184 has two regions that show marked differences in the manner in which the subunits appear to be organized. No correlation was found between the site of origin of flagella and the location of reduced tellurite when the reduction of potassium tellurite was used to indicate the loci of enzymatic respiratory activities.

Purified and crude flagellar isolates from cells of Bacillus pumilus NRS 236 were treated with acid, alcohol, acid-alcohol, or heat, and were examined electron microscopically in negatively stained and shadow-cast preparations. Under certain conditions, each of these agents causes the flagella to break between the proximal hooks and the spiral filaments. In such preparations, filaments are seen in various stages of disintegration, whereas hooks of fairly constant length retain their integrity and morphological identity. When crude isolates of flagella are treated under these conditions, the hooks remain attached to membrane fragments or bear basal material. These findings substantiate previous structural observations that led to the view that the proximal hook is a distinct part of the bacterial flagellum and further confirm that the hook is tightly associated with basal material and the cytoplasmic membrane. It appears that the hook is a polarly oriented structure, and that the interactions between the hook and the basal material or the cytoplasmic membrane are different from those between the hook and the filamentous portion of the organelle. Moreover, both types of interaction apparently differ still from those by which the flagellin subunits are held together in the flagellar filament. Hooks were isolated by exploiting the differences in relative stability shown by the various morphological regions of the bacterial flagellum.

The cellular structure of Porphyridium cruentum was studied with both light and electron microscope. The photosynthetic plastid in this red alga was found to be structurally similar to that in the Chlorophyceae and higher green plants. The phycobilins, as well as the chlorophyll, seem to be associated with the lamellae of the plastid. The pyrenoid, a region of low lamellar density, contains no tubules, and does not appear to function in synthesis or storage of reserve material. Grains of floridean starch are located in the cytoplasm, outside the plastid. Typical mitochondrial organelles were not observed. The nucleus is eccentric, and contains a nucleolus located on the inner face of the nucleus, nearest the plastid. The schedule for staining the nucleus is given in detail. Other cell structures (sheath, dictyosomes, etc.) are described. Growing cells in light of intensity leads to disruption of the parallel arrangement of the lamellar characteristic of cells grown in moderate light.