Effect of L-form Streptococcus pyogenes and of lipoteichoic acid on human cells in tissue culture

Lactobacilli reduce cell cytotoxicity caused by Streptococcus pyogenes by producing lactic acid that degrades the toxic component lipoteichoic acid

Lactobacilli are known to prevent colonization by many pathogens; nevertheless, the mechanisms of their protective effect are largely unknown. In this work, we investigated the role of lactobacilli during infection of epithelial cells with group A streptococci (GAS). GAS cause a variety of illnesses ranging from noninvasive disease to more severe invasive infections, such as necrotizing fasciitis and toxic shock-like syndrome. Invasion of deeper tissues is facilitated by GAS-induced apoptosis and cell death. We found that lactobacilli inhibit GAS-induced host cell cytotoxicity and shedding of the complement regulator CD46. Further, survival assays demonstrated that lactic acid secreted by lactobacilli is highly bactericidal toward GAS. In addition, lactic acid treatment of GAS, but not heat killing, prior to infection abolishes the cytotoxic effects against human cells. Since lipoteichoic acid (LTA) of GAS is heat resistant and cytotoxic, we explored the effects of lactic acid on LTA. By applying such an approach, we demonstrate that lactic acid reduces epithelial cell damage caused by GAS by degrading both secreted and cell-bound LTA. Taken together, our experiments reveal a mechanism by which lactobacilli prevent pathogen-induced host cell damage.

Tissue reactions to a component of root canal system bacteria: lipoteichoic acid

Lipoteichoic acid (LTA), present in Gram-positive microorganisms, has physiochemical characteristics that allow it to act as an immunogen. Due to polymicrobial characteristics of root canal infections, LTA can participate in the development of periapical disease. The reaction of the rat subcutaneous tissue to Teflon tube implants, filled with Fibrinol soaked in lipoteichoic acid (concentration of 150 microg/ml), was observed. Lipoteichoic acid provoked an inflammatory tissue reaction.

Evidence for altered proliferative ability of progenitors of urothelial cells in interstitial cystitis

Secondary cultures of basal urothelial cells isolated from patients with stress incontinence (7 patients), neurogenic bladder (2 patients), interstitial cystitis (IC) (27 patients), bladder rupture (1 patient) and bacterial cystitis (3 patients) grew under growth restricting conditions. All groups displayed reproducible colony size distribution, reflecting the proliferative potential distribution in the population of progenitor cells seeded. The percentage of large colonies (> 6 cells/colony), progeny of basal cells with high proliferative potential, was low in cultures from control patients with stress incontinence, neurogenic bladder or bladder rupture. Exposure of cultures from control patients with stress incontinence to lipoteichoic acid from Streptococcus faecalis, in vitro, increased the percentage of large colonies to levels statistically indistinguishable from those in untreated IC cultures. This supported the possibility that exposure of progenitors of urothelial cells to infection in vivo may cause the persistent increase in the percentage of large colonies in 80% of the IC patients tested. Given these findings, it was not surprising that the percentage of large colonies was also high in cultures from patients with acute bacterial cystitis. In conclusion, the present findings support the theoretical model for the etiology of IC we proposed based on our studies in normal urothelial cells (Elgavish et al., Journal of Cellular Physiology 169: 42-51, 52-65, 66-77, 1996): (1) The proliferative ability of a subpopulation of progenitors of urothelial cells is increased in IC; and (2) This change may be the result of recurrent exposure of progenitors of urothelial cells to injury due, possibly but not exclusively, to infection and chronic inflammation. We propose to use this change as a diagnostic tool for IC.

A subpopulation of human urothelial cells is stimulated to proliferate by treatment in vitro with lipoteichoic acid, a cell wall component of Streptococcus faecalis

Urinary tract infection with gram-positive bacteria is common. Avenues for ingress of bacteria into the bladder include luminal and suburothelial infection. Terminally differentiated superficial urothelial cells lining the lumen of the bladder are often shed in response to infection. In contrast, infection-induced altered function of progenitors of urothelial cells residing in the basal layer of the urothelium is likely to have long lasting effects on the structure and function of the urothelium. The main objective of the present studies was to investigate in vitro the possibility that exposure to lipoteichoic acid, a cell wall component of the gram-positive Streptococcus faecalis (LT-2), stimulates basal urothelial cells to proliferate. To simulate conditions that restrict proliferation and inhibit terminal differentiation of urothelial cells in the basal layer, secondary cultures of urothelial cells (UT) were grown on collagen or fibronectin coated substrate in medium containing low levels of Ca2+ (0.2 mM) and growth factors (0.005% bovine pituitary extract [BPE]). Under these conditions, UT cultures displayed a highly reproducible colony size distribution, possibly due to the fact that colonies were progeny of basal cells with various proliferative potentials, retained in vitro. In cultures grown under growth-restricting conditions the majority of progenitors appeared to be quiescent, just like stem cells in the basal layer of the urothelium. Thus, the population of large colonies (more than six cells/colony), was small when a steady state of growth was achieved, 3-7 days after seeding. Growth factors (0.005-0.5% BPE) caused a dose-dependent increase in this population of large colonies. Moreover, treatment of UT grown under growth-restricting conditions (0.005% BPE) with LT-2 increased steady-state levels of the population of large colonies to levels obtained in cultures growing under optimal conditions with respect to growth factors. These results indicated that the subpopulation of progenitors, quiescent under normal conditions, could be stimulated to proliferate. Two lines of evidence were consistent with the possibility that treatment with LT-2 stimulated proliferation of the subpopulation of progenitors and that large colonies were the progeny of this subpopulation of single cells: (1) treatment with LT-2 increased the percentage of single cells that incorporated bromodeoxyuridine (i.e., proliferated) in a time-dependent manner. (2) An increase in the percentage of large colonies was found following LT-2-triggered proliferation of single cells. We propose that, under normal conditions, cells produced in response to LT-2-triggered proliferation of stem cells are removed from the system due to an increased rate of differentiation followed by apoptosis. Recurrent infection and inflammation may not allow these processes to proceed effectively, resulting in chronic injury to the bladder. Moreover, under conditions in which stem cells accumulate mutations that incapacitate their progeny to undergo apoptosis, LT-triggered proliferation could be a contributing factor to tumorigenesis.

Streptococcal cysteine protease augments lung injury induced by products of group A streptococci

Streptococcus pyogenes infections in humans may be associated with severe clinical manifestations, including adult respiratory distress syndrome and a toxic shock-like syndrome. These observations have led to the investigation of products of group A streptococci that may contribute to increased virulence. Streptococcal pyrogenic exotoxin B is a highly conserved precursor of an extracellular cysteine protease that is secreted by S. pyogenes. We investigated the ability of this streptococcal cysteine protease (SCP) to act synergistically with either streptococcal cell wall antigen (SCW) or streptolysin-O (SLO) to augment lung injury in rats. Intratracheal administration of either SCW or SLO alone caused lung injury, as measured by pulmonary vascular leak. Bronchoalveolar lavage (BAL) fluid analysis showed that SCW induced neutrophil accumulation and appearance of interleukin-1beta and tumor necrosis factor alpha. In contrast, SLO induced neither neutrophil influx nor significant cytokine elevations in BAL fluids. Intratracheal administration of SCP with either SCW or SLO resulted in synergistic augmentation of lung vascular permeability and accumulation of BAL neutrophils. The synergy was reduced when SCP was either heat inactivated or coinstilled with a peptide inhibitor of the protease. SCP in the presence of SCW resulted in a significant increase in BAL fluid tumor necrosis factor alpha content but not in immunoreactive interleukin-1beta. Moreover, the copresence of SAP with SAW resulted in increased BAL fluid nitrite-nitrate levels, indicative of nitric oxide production. These data demonstrate that SCP acts synergistically with other S. pyogenes products (SCW or SLO) to increase tissue injury and provide additional evidence that SCP may function as an important virulence factor in group A streptococcal infections.

Role for intracellular platelet-activating factor in the circulatory failure in a model of gram-positive shock

1. This study investigates the effects of two structurally different antagonists of platelet-activating factor (PAF), BN52021 and WEB2086, on the circulatory and renal failure elicited by lipoteichoic acid (LTA) from Staphylococcus aureus (an organism without endotoxin) in anaesthetized rats. 2. Administration of LTA (10 mg kg-1, i.v.) caused hypotension and vascular hyporeactivity to noradrenaline (1 microgram kg-1, i.v.) WEB2086 (5 mg kg-1, i.v., 20 min before and 150 min after LTA) inhibited the delayed fall in mean arterial blood pressure (at 300 min: 99 +/- 6 mmHg vs. 75 +/- 6 mmHg, P < 0.01) and prevented the decrease in pressor response to noradrenaline (at 300 min: 36 +/- 5 mmHg min vs. 17 +/- 5 mmHg min, P < 0.01). Surprisingly, BN52021 (20 mg kg-1, i.v., 20 min before and 150 min after LTA) neither prevented the hypotension (74 +/- 6 mmHg) nor the vascular hyporeactivity (21 +/- 5 mmHg min). However, BN52021 inhibited the hypotension to injections of PAF as well as the circulatory failure elicited by lipopolysaccharides (10 mg kg-1, i.v.). 3. LTA caused an increase in plasma concentration of creatinine from 39 +/- 5 microM (sham-operated) to 70 +/- 8 microM and urea from 4.7 +/- 0.1 to 13.1 +/- 1.6 mM. The renal failure elicited by LTA was significantly inhibited by WEB2086 (creatinine: 45 +/- 4 microM and urea: 5.7 +/- 0.7 mM), but not by BN52021. 4. The induction of nitric oxide synthase activity in lungs by LTA was attenuated by WEB2086 from 98 +/- 17 to 40 +/- 15 pmol L-citrulline 30 min-1 mg-1 protein (P < 0.01), but not by BN52021 (148 +/- 21 pmol L-citrulline 30 min-1 mg-1 protein). Similarly, WEB2086, but not BN52021, inhibited the increase in plasma nitrite concentration associated with the delayed circulatory failure caused by LTA. The release of tumour necrosis factor-alpha (TNF-alpha) after injection of LTA was not attenuated by WEB2086. 5. The induction of nitrite release by cultured macrophages activated with LTA (10 micrograms ml-1 for 24 h) was inhibited by 74 +/- 4% by WEB2086 (3 x 10(-4) M), but not by BN52021, indicating that only WEB2086 acts on intracellular PAF receptors. 6. Thus, the intracellular release of PAF contributes to the circulatory and renal failure and induction of nitric oxide synthase elicited by LTA in anaesthetized rats. The difference between the two structurally different PAF antagonists in our septic shock models using either LTA or lipopolysaccharide (LPS), shows the importance of models for Gram-positive sepsis in the elucidation of the pathophysiology of septic shock and for the evaluation of potential drugs.

Integrin-mediated adhesive properties of uroepithelial cells are inhibited by treatment with bacterial toxins

Gram-negative bacteria are a dominant cause of urinary tract infection, and their ability to produce toxins is an important virulence attribute. Cellular mechanisms triggered by the production of toxins in the lower urinary tract have not been completely defined. Ureteral epithelial cells (UT; A. Elgavish, Infect. Immun. 61: 3304-3312, 1993) have served as an in vitro model to explore the possibility that bacterial toxins act on UT by affecting integrin-mediated adhesive properties. The effect of treatment with lipopolysaccharides (LPS) from three strains of the gram-negative Escherichia coli [055:B5 (LPS-1), 0111:B4 (LPS-4), and 0127:B8 (LPS-5)] and lipoteichoic acids from two gram-positive bacteria, Streptococcus faecalis (LT-2) and Bacillus subtilis (LT-3), were examined. LPS-5 inhibited markedly UT attachment to collagen and fibronectin. LPS-4 had no effect, whereas LPS-1 inhibited UT attachment to collagen but not to fibronectin. The fact that LPS-5 and LT-2 inhibited an Arg-Gly-Asp sequence-sensitive component of UT attachment to fibronectin is consistent with the possibility that these toxins acted via a mechanism involving typical fibronectin receptors. UT spreading was inhibited markedly by LPS-1, LT-2, and LT-3, whereas LPS-4 and LPS-5 had no effect. Because clustering of integrins is a crucial step in integrin-mediated signal transduction, the possibility that toxins inhibited spreading by affecting clustering was tested. Treatment with LT-2, which inhibited spreading dramatically, abolished completely a UT cell population containing more than five to eight beta 1- or beta 4-subunit-containing integrin clusters.(ABSTRACT TRUNCATED AT 250 WORDS)

Macrophage response to bacteria: induction of marked secretory and cellular activities by lipoteichoic acids

Lipoteichoic acids (LTAs) from various bacterial species, including Staphylococcus aureus, Streptococcus pyogenes, Streptococcus pneumoniae, Enterococcus faecalis, and Listeria monocytogenes, were examined for the ability to induce secretory and cellular responses in a pure population of bone marrow-derived mononuclear phagocytes. Some of the highly purified LTAs, in particular LTAs from Bacillus subtilis, S. pyogenes, E. faecalis, and Enterococcus hirae, were able to affect each of the macrophage parameters measured, i.e., reductive capacity, secretion of tumor necrosis factor and nitrite, and tumoricidal activity. As after stimulation with whole organisms or other bacterial products, secretion of tumor necrosis factor induced by these LTAs reached its maximum within the first few hours of the interaction, while secretion of nitrite and tumoricidal activity required 24 to 36 h for full expression. Other purified LTAs, i.e., LTAs from Streptococcus sanguis, S. pneumoniae, and L. monocytogenes, as well as lipomannan from Micrococcus luteus affected only some of these parameters, while native LTA from S. aureus was inactive. There was no obvious correlation between biological activity and chain length, kind of glycosyl substituents, glycolipid structures, or fatty acid composition of LTAs. Deacylation of LTAs resulted in a complete loss of activity, and deacylated LTAs did not impair the activity of their acylated counterparts, suggesting that acyl chains may be essential for binding of LTA to the cell surface. The results demonstrate that some LTA species are potent inducers of macrophage secretory and cellular activities.

Deposition of circulating streptococcal lipoteichoic acid in mouse tissues

The tissue binding properties of streptococcal lipoteichoic acid (LTA) were studied using normal and passively immunized BALB/c mice. After intraperitoneal injection in non-immunized mice, 3H-LTA concentrations in blood, heart, kidney and liver were highest between 24 and 30 h post-injection. LTA deposits in heart remained high for the next 24 h, whereas other tissue levels decreased. Constant amounts of 3H-LTA were detected in urine throughout the 48 h period. In passively immunized mice, the amount of tissue deposition of 3H-LTA was inversely proportional to the ratio of antibodies to LTA. Autoradiography revealed focal deposits of 3H-LTA in heart, kidney and liver. These observations indicate that LTA, released by streptococci growing at remote body sites, can be carried by the blood to internal organs where it can accumulate and participate in pathogenesis.

Effects of growth factors, hormones, bacterial lipopolysaccharides, and lipotechoic acids on the clonal growth of normal ureteral epithelial cells in serum-free culture

In vitro tissue culture techniques were employed to study the effects of bacterial endotoxins on the growth of normal epithelial cells from the human ureter (NHU). Primary cultures of NHU cells were initiated from explant outgrowth cultures of human ureteral tissue and cultured on collagen gel in F-12* medium containing 1% fetal calf serum (FCS). Optimal clonal growth of secondary cultures of NHU cells seeded at relatively low seeding cell densities, directly on plastic dishes, was achieved in F-12* medium containing bovine pituitary extract (0.5% BPE) and 0.05% BSA. Results indicated that insulin in the F-12* medium could be replaced by three orders of magnitude less IGF-1. Further clonal growth experiments demonstrated that PGE1 is growth stimulatory and can replace BPE as a growth factor requirement. This finding was in agreement with the fact that BPE growth requirement could be replaced by cholera toxin or dibutyryl cAMP. These results suggested that both BPE and cholera toxin operated by activation of a cAMP-dependent mitogenic pathway. Seven gram-negative bacterial lipopolysaccharides (LPS) and three gram-positive bacterial lipotechoic acids (LT) were tested for their effects on NHU clonal growth. Three out of the five LPS derived from Escherichia coli (strains 055:B5, 0128:B12, and 0127:B8), LPS from Klebsiella pneumoniae, and LPS from Pseudomonas aeruginosa all showed significant growth inhibitory effects at minimally effective doses ranging from 5 to 25 micrograms/ml. LPS derived from E. coli strain (0111:B4) had no growth effects at the highest concentration tested (100 micrograms/ml). In contrast, LT derived from Streptococcus pyogenes, S. faecalis, Staphylococcus aureas, and Bacillus subtilis all markedly enhanced clonal growth at concentrations ranging from 1 microgram/ml less than [LT] less than 50 micrograms/ml. LT from Strep. pyogenes was inhibitory to clonal growth at 100 micrograms/ml. The growth inhibitory effects of LPS were shown to be sensitive to the presence of hydrocortisone in the growth medium, indicating that LPS effects on growth are mediated via the arachidonic acid cascade. We speculate that these results indicate a link between the susceptibility of uroepithelial tissue to the pathogenic microflora seen in urinary tract diseases and the differential sensitivity of proliferation-competent uroepithelial cells to growth inhibition by LPS produced by gram-negative bacteria. However, further studies with uropathogenic serotypes will be necessary to corroborate this possibility. The growth-stimulating activity of LTs produced by gram-positive bacteria may be due to their ability to bind to cell-associated fibronectin and to activate the fibronectin receptor as part of ligand receptor-induced mitogenic transmembrane signalling pathway.

Streptococcus pyogenes clinical isolates and lipoteichoic acid

Minimally subcultured clinical isolates of virulent nephritogenic and nonnephritogenic Streptococcus pyogenes of the same serotype showed major differences in lipoteichoic acid (LTA) production, secretion, and structure. These were related to changes in coccal adherence to and destruction of growing human skin cell monolayers in vitro. A possible relationship between cellular LTA content and group A streptococcal surface hydrophobicity was also investigated. Nephritogenic S. pyogenes M18 produced twice as much total (i.e., cellular and secretory) LTA as did the virulent, serologically identical, but nonnephritogenic isolate. Also, the LTAs from these organisms differed markedly. The polyglycerol phosphate chain of LTA from the nephritogenic isolate was longer (1.6 times) than was that from the nonnephritogenic isolate. Likewise, both LTAs indicated the presence of alanine and the absence of glucose. Amino sugars were found in LTA from only nephritogenic S. pyogenes. Teichoic acid, as a cellular component or secretory product, was not detected. The adherence of two different nephritogenic group A streptococcal serotypes (M18 and M2) exceeded that of the serologically identical but nonnephritogenic isolates (by about five times), indicating a correlation between virulent strains causing acute glomerulonephritis and adherence to human skin cell monolayers. Likewise, LTA from nephritogenic S. pyogenes M18 was more cytotoxic (1.5 times) than was that from the nonnephritogenic isolate for human skin cells, as determined by protein release. This difference was not perceptible by the more sensitive dye exclusion method (i.e., requiring less LTA), which emphasizes changes in host cell morphology and death. Also, the secretion of LTA by only virulent nephritogenic S. pyogenes M18 was exacerbated by penicillin (a maximum of four times). Finally, while the adherence of nephritogenic S. pyogenes M18 decreased markedly after continued subculturing in vitro, the surface hydrophobicity did not.

Adherence of Streptococcus agalactiae to synchronously growing human cell monolayers without lipoteichoic acid involvement

Freshly isolated virulent and nonvirulent strains of Streptococcus agalactiae type III were used to study differences in coccal adherence to synchronously dividing, subconfluent human embryonic amnion and fetal lung monolayers in vitro. The adherence frequency by virulent isolates of mid-logarithmically growing cocci to amnion cells varied markedly with host cell age, being highest shortly after eucaryotic cell division. This variation was not observed with lung cell monolayers, suggesting that cyclic production or exposure of coccal receptor sites on the eucaryotic cell surface with age is not a common property of all primary human cells in vitro. However, and regardless of age, not all cells within these synchronously dividing populations bound virulent cocci, indicating that a very small segment of a population may always be unresponsive to host cell interactions with a coccal pathogen. By comparison, adherence of nonvirulent coccal isolates to amnion and lung cells remained constant and of a very low order, regardless of host cell age. Maximal adherence of virulent S. agalactiae to young host cells occurred at early and mid-logarithmic phases of growth. However, at the late stationary growth phase, adherence was reduced to almost that of nonvirulent isolates. Pretreatment of virulent S. agalactiae with anti-lipoteichoic acid (LTA) serum failed to inhibit coccal adherence to these different host cells. Heat negated adherence. Group B coccal LTA was cytotoxic for these host cells. However, pretreatment of amnion and lung cells with nontoxic levels of this amphiphile did not prevent attachment of virulent cocci. Finally, coccal pretreatment with pronase abrogated adherence to either host cell even though surface-exposed LTA was uneffected, as observed by the indirect fluorescent-antibody procedure. Likewise, no observable difference in surface LTA was detected when fresh isolates of virulent and nonvirulent coccal strains were compared by this procedure. These studies suggest that protein involvement, rather than LTA, is primarily responsible for mediating virulent S. agalactiae type III attachment to these synchronously growing, subconfluent eucaryotic monolayers in vitro.

Effect of streptococcal lipoteichoic acid on prolyl hydroxylase activity as related to collagen formation in mouse fibroblast monolayers

Dried and wet mouse fibroblast monolayers with labeled collagenous substrate were used to study the effects of lipoteichoic acid (LTA) on cellular prolyl hydroxylase activity. LTA is a scavenger of cations, and Fe2+ is essential for prolyl hydroxylase activity. Surprisingly, addition of LTA to dried monolayers resulted in increased prolyl hydroxylase activity, whereas preincubation of Fe2+ with LTA only negated this increase. However, significant inhibition of enzyme activity by wet monolayers occurred whether LTA was added directly to the test system or whether it was used after preincubation with Fe2+. These data suggest that LTA causes membrane perturbations. Also, that the binding of LTA to the membrane of dried and wet monolayers appears to be decidedly different when based on the subsequent availability of Fe2+ for cellular prolyl hydroxylase activity. The ability of LTA to act as a cationic exchanger and the presence of intracellular Fe2+ inaccessible to LTA probably accounted for the lack of complete inhibition of prolyl hydroxylase activity by this amphiphile in the wet cell system. Considerably less iron was needed to negate the partial inhibition of prolyl hydroxylase activity by LTA in viable cells than was needed to restore the increased enzyme activity by this amphiphile in equivalent dried preparations. These and other results showed that, although LTA does not affect collagen polypeptide chain formation in wet monolayers, its involvement at the molecular level does result in a marked decrease in the hydroxylation of collagenous peptidyl prolyl residues through LTA interaction with Fe2+. This reduction in prolyl hydroxylase activity equaled the reduction in hydroxylation of collagenous protein in fibroblast monolayers caused by LTA reported earlier (O. Leon and C. Panos, Infect. Immun. 40:785-794, 1983). Therefore, these data suggest that partial inhibition of prolyl hydroxylase activity is directly related to the synthesis of defective collagen by wet fibroblast monolayers exposed to minute amounts of group A, type 12 streptococcal LTA. Use of LTA also showed that complete inhibition of hydroxyproline formation is not required for the continued formation and accumulation of defective collagenous protein by these monolayers.

Binding of Streptococcus mutans antigens to heart and kidney basement membranes

Using indirect immunofluorescence, alkali-extracted components of Streptococcus mutans were found to bind in vitro to capillary walls and sarcolemmal sheaths of monkey cardiac muscle and to glomerular and tubular basement membranes of monkey kidney. Adsorption of S. mutans components to tissue fragments was also detected by indirect radioimmunoassay and immunoblotting on nitrocellulose paper. Antibodies did not bind to untreated, control tissues in these experiments, proving that antigens shared by S. mutans and tissue components were not involved. Rabbit and monkey heart and kidney components bound S. mutans antigens of 24,000, 35,000, and 65,000 Mr. Monkey heart also bound molecules of 90,000 and 120,000 Mr. Rabbits immunized by intravenous injection of disrupted S. mutans cells developed severe nephritis that was characterized by the deposition of immunoglobulins, complement component C3, and S. mutans antigens in the glomeruli. Immunoglobulin G eluted from nephritic kidneys reacted in immunoblots with the 24,000, 35,000, and 65,000 Mr components of S. mutans extract, indicating that the antigens that bound to tissue in vitro also bound in vivo and reacted with antibodies in situ. Antibodies to other S. mutans antigens were not detected in the kidney eluate, although they were present in the serum of the same rabbit.

Teichoic acids of Streptococcus agalactiae: chemistry, cytotoxicity, and effect on bacterial adherence to human cells in tissue culture

The ratio of teichoic acid to lipoteichoic acid (LTA) in a strain of Streptococcus agalactiae type III was found to be 8:1, with the total amount of LTA being 0.1% of the dry weight of the organism. Purified teichoic acid contained D-alanine and possibly a small amount of D-glucose and was approximately 22 glycerol phosphate units in length. The linkage between each of these units was 1-3. In addition, LTA contained a complex lipid, more glucose, and an unusually high content of a short-chain fatty acid, tridecanoic acid. This LTA was cytotoxic for a variety of human cell monolayers in tissue culture, including one derived from the human central nervous system. Established human cells were more sensitive than primary cell monolayers to this LTA, with as little as 12.5 micrograms of LTA per ml being cytotoxic for HeLa cells. Teichoic acid (250 micrograms/ml) was nontoxic under identical conditions. These cytotoxicity results suggest an LTA involvement in group B streptococcal pathogenesis. Also, the first model system for the study of group B streptococcal adherence to primary human embryonic amnion cells in tissue culture is detailed. This system was used to quantitate pronounced differences in tissue tropism between S. agalactiae and Streptococcus pyogenes and showed enhanced binding by this group A coccus over that of S. agalactiae for amnion cell monolayers. The adherence of both streptococcal species to only a portion (40%) of these amnion cells suggested that host cell receptor expression may vary for primary cells in vitro. Finally, this strain of S. agalactiae was shown to adhere to amnion cells by a non-LTA-mediated mechanism. The possibility of an LTA-mediated versus a protein-mediated adherence mechanism for host cells that is related to the virulence of S. agalactiae is discussed.

Morphological changes and pathology of mouse glomeruli infected with a streptococcal L-form or exposed to lipoteichoic acid

The morphology and pathology of cultured mouse glomeruli were examined at the cellular and subcellular levels after infection with a physiological isotonic L-form of Streptococcus pyogenes type 12 or exposure to streptococcal lipoteichoic acid. These changes, as viewed by light microscopy, were identical regardless of the method used to induce glomerular cytotoxicity. They were characterized by an initial reduction in the outgrowth of cells, some cellular granulation, and later, destruction of the confluent monolayer. Once initiated, cytotoxicity could not be reversed by refeedings, and complete glomerular destruction resulted after 2 weeks. Electron microscope studies revealed that the basement membrane of intact glomeruli exposed to streptococcal lipoteichoic acid had become greatly thickened (two- to fourfold) and electron dense. Our recent biochemical findings have shown that streptococcal lipoteichoic acid increases the amount of collagen formed and retained by mouse fibroblasts in tissue culture as well as causing a reduction in the hydroxylation of proline in both intracellular and secreted collagenous material (Leon and Panos, Infect. Immun. 40:785-794, 1983). These results, together with the present findings, suggest that the thickening of the glomerular basement membrane may be due to defective collagen biosynthesis as a result of streptococcal lipoteichoic acid. The use of cultured glomeruli as a model system for studying the earliest basement membrane alterations in the absence of an immune response as a result of streptococcal lipoteichoic acid is suggested.

Cytotoxicity and inhibition of normal collagen synthesis in mouse fibroblasts by lipoteichoic acid from Streptococcus pyogenes type 12

The toxicity of lipoteichoic acid (LTA) from Streptococcus pyogenes type 12 was investigated by using mouse fibroblasts in culture in the absence of serum. Morphologically, while low concentrations of LTA elicited a subtle effect characterized by progressive cellular degeneration with practically no release of protein, larger concentrations (greater than 50 micrograms/ml) of this amphiphile resulted in rapid death of cell monolayers. Metabolic studies utilized a concentration of LTA (17.5 micrograms/ml) which caused the smallest change in cell morphology in the least number of mouse fibroblast cells per monolayer. Under these conditions, cell monolayers showed an increase of 450% in their content of collagenous protein after exposure to LTA. However, the amount of such material secreted remained unchanged. Also, changes in the type of collagenous protein formed were observed after exposure to LTA. Collagenous protein accumulating intracellularly was found to be practically hydroxyproline-free. However, collagenous protein secreted by this cell line showed a significantly reduced content of hydroxyproline as compared with control cells unexposed to this coccal membrane component. Column chromatographic studies confirmed that the collagenous protein secreted by monolayers exposed to LTA was defective (under hydroxylated). It was concluded that LTA does not affect the amount of collagenous protein secreted. However, it does increase the amount of this protein formed and retained by this cell line as well as causing a reduction in the hydroxylation of proline in both intracellular and secreted collagenous material. A possible relationship between abnormal basement membrane morphology and disturbed collagen synthesis in post-streptococcal glomerulonephritis as related to LTA is discussed.

Isolation of heart- and kidney-binding protein from group A streptococci

Tritium-labeled, water-soluble components of Streptococcus pyogenes type M6 absorbed to cardiac tissue in vitro. Tissue binding was time dependent, saturable, and reversible. Chromatography of the crude bacterial extract on Bio-Gel P-300 indicated a molecular weight greater than 300,000 for the heart-binding component. Sodium dodecyl sulfate (SDS) dissociated this aggregate into a protein of 18,000 to 20,000 daltons as determined by Sephacryl S-200 chromatography and SDS-polyacrylamide disc gel electrophoresis. The tissue-binding protein was also purified from streptococcal extracts by absorption to immobilized heart components. SDS-polyacrylamide gel electrophoresis of the protein desorbed from tissue revealed a radioactive band of 19,000 daltons. Indirect immunofluorescence tests on cardiac tissue treated with streptococcal extract showed an accumulation of a bacterial antigen on the sarcolemmal sheaths. Streptococcal components also adsorbed to basement membranes of kidney. Antisera prepared to isolated cytoplasmic membranes and water-soluble extracts of S. pyogenes type M6 were the most sensitive reagents for the detection of bacterial components bound to tissue. Antisera prepared to isolated cell walls and to intact bacteria were weakly reactive in these assays.

Attachment of bacteria to mammalian surfaces

Much attention has been devoted to the study of bacterial adherence to mammalian surfaces in vitro during the past several years. Some in vivo evidence also suggests that this process may indeed be an integral part of the pathogenesis of colonization and certain infections. The biochemical basis of attachment and definition of the actual receptor sites involved are just starting to become known and seem to be different amongst individual bacteria genera. However, pili may mediate attachment of a variety of gram-negative organisms to receptor cells, and streptococcal lipoteichoic acids probably serve a similar function. Some recent study methods and results in this field are reviewed.

Binding of streptococcal antigens to muscle tissue in vitro

Antigens extracted from cells of Streptococcus pyogenes T6 and Streptococcus mutans strains AHT, BHT, 10449, OMZ175, and K1R adsorbed to the sarcolemmal sheath of cardiac muscle cells in vitro. Similar preparations from S. salivarius, S. sanguis, Staphylococcus aureus, and Lactobacillus casei had weak or negligible tissue-binding activity. Tissue-bound bacterial antigens were detected with homologous rabbit antisera with both indirect immunofluorescence tests and an indirect radioimmunoassay. Serological cross-reactivity was observed between the tissue-binding factors of S. pyogenes and S. mutans cells but not between the bacteria and muscle tissue. In a comparative study of extraction procedures, the greatest yield of tissue-binding factors was obtained from group A streptococci by cell disruption in buffer at 4 degrees C. Hot aqueous phenol and hot water extracts were inactive. Antibodies specific for the tissue-binding factor(s) were readily adsorbed from rabbit anti-S. pyogenes serum by a preparation of isolated cytoplasmic membranes but not by a suspension of cell wall fragments. The heart-binding component of S. pyogenes cell extracts was inactivated by protease digestion and heat treatment and to a lesser extent by periodic acid oxidation. The capacity of heart cell components to adsorb streptococcal antigens was reduced by protease treatment but not by the action of neuraminidase, hyaluronidase, organic solvents, or detergents.