Site-specific invasion of the basal ganglia by Nocardia asteroides GUH-2

Nocardia asteroides GUH-2 (GUH-2) invades the nigrostriatal region of the brain in mice [15]. Selective dopaminergic neuronal dropout in the substantia nigra results in parkinsonian changes characterized by movement disorders responsive to L-dopa [15]. This is the only reported example of an experimental bacterial model for parkinsonism. Following i.v. inoculation of GUH-2 into the non-human primate Macaca fasicularius, the nocardiae preferentially invaded and grew within the basal ganglia (substantia nigra, caudate, putamen, and globus pallidus) often without inducing apparent inflammation. Reduced, limited growth of nocardiae occurred in the white matter of the cerebral cortex, medulla, and hippocampus, whereas neither significant adherence to nor growth within the meninges was observed. Twenty-four hours after injection, nocardial cells were found within capillary endothelial cells, the basal lamina, neurons, astroglia and in axonal extensions. The bacteria, in endothelial cells, were surrounded by a unit membrane, but in the basal lamina they appeared to be free and not membrane bound. After the organisms passed into the brain parenchyma, the nocardiae once again became surrounded by membrane, often being encapsulated by numerous layers with the innermost layer tightly adherence to the bacterial surface. There was a propensity for nocardial growth within and along myelinated axons, either with or without disruption to the surrounding myelin sheath. There was electron microscopic evidence that the nocardiae induced a neurodegenerative response especially in the substantia nigra region. Thus, the early interactions of GUH-2 within the primate brain appeared to be similar to those reported in the mouse. Received: 23 September 1999     

Nocardia asteroides strain GUH-2 induces proteasome inhibition and apoptotic death of cultured cells

Many bacterial pathogens have the ability to induce apoptosis in their hosts. It was previously shown that Nocardia asteroides strain GUH-2, a Gram-positive facultatively intracellular pathogen, is capable of inducing the apoptotic death of dopaminergic cells in the murine brain and in PC12 cells, a rat cell line. In this study, the apoptosis-inducing potential of N. asteroides GUH-2 was further explored using HeLa cells, a human epithelial cell line. HeLa cells were incubated for 5h with live nocardiae, heat-killed bacteria, or unconcentrated nocardial culture filtrate, and changes to the cells were monitored. Consistent with the previous studies, N. asteroides GUH-2 induced DNA fragmentation and apoptosis in HeLa cells. Caspase activation and disruption of the mitochondrial membrane potential were also investigated to determine their roles in the induction of cell death. In all these experiments, significant changes were only induced by live nocardiae. A recent publication demonstrated that systemic administration of proteasome inhibitors can induce a Parkinsonian syndrome in rats that includes intraneuronal inclusions and characteristic behavioral alterations. Similar effects have been observed in mice and monkeys infected with N. asteroides GUH-2. In addition, some reports have shown that proteasome inhibition causes apoptotic death of affected cells. We therefore investigated the ability of N. asteroides GUH-2 to inhibit proteasome activity. Proteasome activity was significantly reduced, suggesting that this mechanism may be involved in the induction of apoptosis by these bacteria.     

Characterization of dopamine-depleting activity of Nocardia asteroides strain GUH-2 culture filtrate on PC12 cells

Experimental infection of BALB/c mice with the Gram-positive bacterium Nocardia asteroides (strain GUH-2) results in life-long movement abnormalities including head shaking and spinning when held by the tail. The head shaking is temporarily inhibited by treatment with dopamine's precursor levodopa, suggesting that abnormalities in dopaminergic neurotransmission may be involved in these movement abnormalities. Cell-free filtrates from N. asteroides cultures induce > 70% dopamine depletion in rat pheochromocytoma PC12 cells, suggesting that Nocardia's effects on dopamine neurons may result in part from secreted factors. The nature of this dopamine-depleting activity was examined in the present study. Dopamine-depleting activity in N. asteroides culture filtrate was resistant to heat (100 degrees C x 30 min), proteases, and chloroform extraction, and was present in a low molecular mass (< 3 kDa) fraction. It was partially inhibited by decreasing (to 4.0) or increasing (to 10.0) the filtrate pH. GUH-2 filtrate increased cellular lactate dehydrogenase release by only 2%, and induced apoptotic morphology in only 11% of PC12 cells, suggesting that dopamine-depleting activity was not due to either cell injury or induction of apoptosis. These results suggest that a protease-resistant, low molecular mass substance secreted by N. asteroides may be responsible for its dopamine-depleting effects.     

Virulence of Nocardia asteroides during its growth cycle.

Cells of Nocardia asteroides undergo structural and chemical changes, especially in the cell wall, during growth in brain heart infusion broth. Experiments were devised to determine whether these changes affected the virulence of Nocardia for mice. It took, on the average, 1,380 times the number of colony-forming units at the stationary phase to achieve the same mortality induced by the log-phase cells. Cells in either the lag phase or early stationary phase of growth were intermediate in the numbers of colony forming units required to kill mice. Dry-weight determinations at different stages of growth demonstrated that the log-phase organisms were approximately 10 times heavier than stationary-phase cells. Thus, on the basis of dry-weight (micrograms) values, the average colony-forming unit of log phase is approximately 130 times more virulent than in stationary-phase cultures. Therefore, the stage of growth affects greatly the virulence of N. asteroides.     

Site-specific growth of Nocardia asteroides in the murine brain.

The growth of Nocardia asteroides GUH-2 and two mutants (NG-49 and I-38-syn) in regions of the brains of BALB/c mice was determined by microdissection and viable counting. GUH-2 grew throughout the murine brain but at different growth rates that depended on the specific location. The rate of increase in total CFU per brain during GUH-2 infection was unaffected by the inoculum size; however, in five of eight brain regions, an alteration in the inoculum size resulted in altered nocardial growth rates. Mutant NG-49 showed a significantly slower rate of increase in total CFU per brain than did the parental strain, GUH-2, and significantly decreased growth rates in seven brain regions. Mutant I-38-syn showed a rate of increase in total CFU per brain similar to that of the parental strain; however, this mutant grew significantly faster in the cerebellum and pons-medulla. Growth appeared to be a necessary precursor to the cellular damage that resulted in the variety of neurological disorders observed in mice infected with N. asteroides GUH-2, because mutant NG-49 exhibited a decreased ability to grow in specific regions of the brain and did not induce signs of neurological damage. In contrast, mutant I-38-syn induced neurological signs in a larger percentage of the infected animals than did parental strain GUH-2 and grew better in certain regions of the brain than did the parental strain. Furthermore, there appeared to be a relationship between the growth of N. asteroides in the substantia nigra and the induction of an L-dopa-responsive head shake that was observed in some of the mice following a sublethal intravenous injection of N. asteroides GUH-2.     

Nocardia asteroides and Nocardia brasiliensis infections in mice.

A model for Nocardia asteroides and Nocardia brasiliensis infections in Swiss white mice has been established without the addition to the inocula of any form of adjuvant. Serial histopathological studies revealed that these two actinomycetes cause lesions that are quite different in their features. An acute suppurative abscess characterizes the lesions of N. asteroides. In the case of N. brasiliensis infections a granuloma is produced in which a striking feature is the presence of large numbers of foam-laden macrophages, although occasional exceptions to this pattern were noted. Electron microscopic studies demonstrated that these macrophages contain within their cytoplasm organisms in varying stages of degeneration. Repeated mortality studies in mice failed to demonstrate differences in mortality rates produced by N. asteroides and N. brasiliensis. Thus, despite relatively trivial biochemical and antigenic differences between these two species of Nocardia, the local pathogenic response is quite different. The presence in the "brasiliensis lesion" of foamy macrophages with intracellular organisms is reminiscent of the histopathological features of lepromatous leprosy and of disseminated Myocobacterium bovis infection when this occurs in the immune suppressed situation. It is possible that N. brasiliensis infection produces a depression of cellular immunity that modifies the local host response to the organism.     

Monoclonal antibodies to Nocardia asteroides and Nocardia brasiliensis antigens.

Nocardia asteroides and Nocardia brasiliensis whole-cell extracts were used as antigens to generate monoclonal antibodies (MAbs). Six stable hybrid cell lines secreting anti-Nocardia spp. MAbs were obtained. These were characterized by enzyme-linked immunosorbent assay, Western blot (immunoblot), and immunofluorescence assay. Although all the MAbs exhibited different degrees of cross-reactivity with N. asteroides and N. brasiliensis antigens as well as with culture-filtrate antigens from Mycobacteria spp., they have the potential for use as reagents in the purification of Nocardia antigens.     

Relationship between cell surface composition, adherence, and virulence of Candida albicans.

A comparison was made of the adherence to acrylic and to human buccal epithelial cells of seven strains of Candida albicans isolated from active infections (I strains) and two strains obtained from asymptomatic carriers (C strains). After growth in defined medium containing a relatively low concentration (50 mM) of glucose as the carbon source, the adherence of I and C strains to either surface was similar and all strains were sensitive to spheroplast formation with Zymolyase 5000. Growth in medium containing a high concentration (500 mM) of sucrose or galactose enhanced the adherence of I strains up to 5- and 11-fold, respectively, and there were corresponding increases in resistance to spheroplast formation. Sucrose- or galactose-grown C strains showed only small increases in adherence and remained relatively sensitive to spheroplast formation. When inoculated intravenously into mice, I strains grown in 500 mM sucrose were up to five times more virulent than organisms grown in 50 mM glucose, while I strains grown in 500 mM galactose showed a 5- to 24-fold increase in virulence. Fifty percent lethal doses obtained for C strains were similar after growth on all three carbon sources. We conclude that I strains are able to modify their surface composition in response to high concentrations of certain sugars in the growth environment. Such modification can enhance both their ability to adhere to surfaces and their virulence. C strains lack this capability, or possess it to a lower degree, and may therefore have a lower pathogenic potential.     

Relationship of macrophages to cell-mediated immunity in experimental Nocardia asteroides infection.

Marked in vivo intracellular killing of Nocardia asteroides occurred in the peritoneal macrophages obtained 72 h after an intraperitoneal challenge with N. asteroides, in guinea pigs either actively immunized with ribonucleic acid protein or passively immunized by immune spleen cell transfer from actively immunized donor guinea pigs. This specific killing of N. asteroides in immune macrophages persisted for at least up to 60 days. Administration of antimacrophage sera before intravenous challenge with N. asteroides in the immune guinea pigs produced an early death of the animals, and the total tissue counts of N. asteroides in the liver, spleen, lungs, and heart remained the same in them as in unimmunized controls.     

Variations in properties of Nocardia asteroides resulting from growth in the cell wall-deficient state

Several revertants possessing a cell wall were obtained from L-forms of Nocardia asteroides 10905. These L-form revertants differed from the parent in respect to colonial and cellular morphology, pigmentation, metabolic, capacities, cell wall structure, and mycolic acid composition. These data suggest a possible mechanism for the taxonomic heterogeneity and phenotypic diversity observed among N. asteroides strains.     

Differences in the interactions of Nocardia asteroides with macrophage, endothelial, and astrocytoma cell lines.

An in vitro model for studying host cell interactions with Nocardia asteroides was developed. Thus, macrophage cell lines J774A.1 and P388D1, pulmonary artery endothelium cell line CPAE, rat glial tumor cell line C6, and human astrocytoma cell lines CCF-STTG1 and U-373 MG were infected with either log- or stationary-phase cells of N. asteroides GUH-2, and the host cell-nocardia interactions were determined by light microscopy and electron microscopy. Polyclonal antinocardial antibody did not enhance uptake of nocardiae by any of these cell lines; however, log-phase cells of GUH-2 infected a higher percentage of J774A.1 and P388D1 than did stationary-phase organisms. When cells infected with stationary-phase GUH-2 were incubated for 6 h, filaments developed, which indicated that nocardial growth had occurred. In J774A.1 and P388D1, only 31 to 57% of the total stationary-phase coccobacillary cells that were phagocytized formed filaments within 6 h. This indicated that there was some inhibition of growth of the phagocytized nocardiae within these macrophage cell lines; however, the nocardiae grew within the endothelial (> 87% filaments) and astrocytoma (100% filaments) cell lines. Microfilament inhibitor cytochalasin B inhibited uptake of GUH-2 by macrophages and other cell lines, except that there was no effect on uptake of nocardial cells by astrocytoma cell line U-373 MG. Scanning and transmission electron microscopy showed phagocytosis of GUH-2 by the different cell lines. In cytochalasin B-treated cells, nocardiae were shown to penetrate through the cell surface and become internalized in a manner distinct from typical phagocytosis, suggesting that filamentous forms of this organism have a phagocytosis-independent invasion factor. The extent of this cytochalasin-resistant cellular penetration by the nocardiae differed in the different cell lines.     

Ultrastructural analysis of growth of Nocardia asteroides during invasion of the murine brain.

BALB/c mice were infected with 10(6) CFU of log-phase cells of Nocardia asteroides GUH-2 by tail vein injection (at this lethal inoculum dose, approximately 800 to 1,000 CFU becomes deposited in the brain). At 24 h after infection, the ultrastructural interactions of the nocardiae during growth within the murine brain were investigated. The nocardiae grew perivascularly in the pons, substantia nigra, hypothalamus, and thalamus portions of the brain, where they were either within or associated with most brain cell types. There appeared to be a propensity for growth within the soma of neurons and their axonal extensions. The nocardial cells were surrounded by 1 to 30 layers of membrane, and the innermost membrane was usually tightly adherent to the cell wall. This compartmentalization of nocardiae within brain cells could contribute to their failure to induce an inflammatory response or a cytopathic effect. Nevertheless, the bacteria were able to obtain adequate nutrients from the host to grow within the brain. The nocardiae were not completely inert, since some of the brain cells showed signs of degeneration. The myelin sheaths of axons were the most strongly affected, and there was evidence of demyelinization and axonal degeneration. Nocardiae growing within brain cells were phagocytized by compact, dense cells that were probably microglia. There was no ultrastructural evidence that the nocardiae were damaged by these phagocytes 24 h after infection; nevertheless, these cells may be important for the elimination of nocardiae from the brain during a nonlethal infection.     

Experimental infection of chickens with Nocardia asteroides and Nocardia transvalensis

Ten-day-old cockerels were infected with N. asteroides or N. transvalensis by the oral or intraperitoneal routes. Clinical signs included depression, gasping and emaciation. Grey nodules or foci were observed in the lungs, air sacs, liver and breast muscles. These organs showed necrosis, granulomas, lymphoid follicles and infiltration of lymphocytes, heterophils and macrophages. Intraperitoneal infection resulted in a more severe disease with either organism. N. transvalensis appeared more pathogenic than N. asteroides intraperitoneally and vice versa with the oral route.     

DNA probes for the identification of Nocardia asteroides.

DNA probes for the rapid identification of Nocardia asteroides were obtained by constructing a genomic library of strain GUH-2 in the lambda cloning vector EMBL3. Of 50 recombinant clones tested, 2 were identified that hybridized with 31% of the N. asteroides strains in a reference collection without cross-hybridization with related members of the Actinomycetales. Additional libraries were then generated from selected strains of N. asteroides that had failed to hybridize with any of the GUH-2 clones. Four additional clones were obtained from these strains which, when pooled, provided DNA probes specific for all of the N. asteroides strains tested.     

Adherence of Nocardia asteroides within the murine brain.

Nonlethal infection of BALB/c mice with Nocardia asteroides GUH-2 (GUH-2) produces a variety of neurological signs, including an L-dopa-responsive movement disorder in 10 to 15% of the infected population. To study nocardial interactions with the brain, we characterized the attachment of GUH-2 within specific regions through the use of microdissection. Following an intravenous injection of a single-cell suspension of log-phase GUH-2, viable cells were recovered from all regions of the brain, and the distribution of the nocardiae was independent of the size of the inoculum. In addition, two mutants of GUH-2 were found to possess significantly altered binding characteristics with regard to both the percentage of the inoculum bound per brain and the relative distribution of adherence to regions of the brain, when compared with the parental strain. These results indicated that GUH-2 bound throughout the murine brain and suggested that GUH-2 utilized specific receptors to facilitate this attachment.     

Distribution of serotypes of Nocardia asteroides from animal, human, and environmental sources.

The antigenic types of 129 isolates of Nocardia asteroides from diverse clinical, environmental, and geographic origins were determined. The majority of the isolates studied were of bovine (56) or human (44) origin; 11 were derived from six species of animals other than cattle, and 10 were isolated from environmental sources; the source of 8 strains could not be determined. Testing culture filtrate antigens against four standard reference sera in a gel diffusion precipitin test established the antigenic type of 95.3% of the isolates. After excluding strains that weighted the data because of common infection, the distribution of serotypes was examined according to the origin of the isolate. Type I was the most frequently encountered serotype (31.9%); types III (15.0%) and IV (20.4%) were also observed frequently, as was the antigenic mixture III + IV (14.2%). There was an apparent difference in frequency of type III and IV antigens among isolates of bovine and human origin; type III made up 20.0% of the bovine isolates and 13.6% of the human isolates, whereas type IV constituted 10.0% of bovine and 27.3% of human isolates.     

Interaction of Nocardia asteroides with cultured rabbit alveolar macrophages.

The interaction between virulent and less virulent strains of Nocardia asteroides and cultured rabbit alveolar macrophages was studied. It was shown that cells of the less virulent strain (N. asteroides 10905) were rapidly phagocytized and destroyed. However, some cells were able to avoid being killed, and they persisted within the macrophage in an altered, gram-negative form. These variants apparently increased in numbers after several days within the macrophage population, so that at 9 days postinfection more colony-forming units per macrophage were recovered than at 3 h. Little or no extracellular growth was observed in the tissue culture medium. During the increase at 9 days, both transitional-phase variants and L-forms of N. asteroides were isolated from the macrophages but not from the medium. Gram-positive bacterial cells were never observed in 9-day infected macrophages. In contrast, cells of the more virulent strain (N. asteroides 14759) were not destroyed after being ingested. After 6 h postinfection, it was observed that the number of colony-forming units per macrophage had increased significantly. There was no corresponding increase in extracellular organisms observed in the culture medium. Therefore, cells of N. asteroides 14759 were able to grow rapidly within cultured rabbit alveolar macrophages. Upon continued incubation of the infected cells (24 h postinfection), it was shown that this strain of Nocardia grew out of the macrophages as acid-fast branching filaments. From these data, it is clear that the initial interaction between N. asteroides and unstimulated, nonimmune alveolar macrophages depends upon the relative virulence of the nocardial strain.     

Method for improved extraction of DNA from Nocardia asteroides.

In a variation of standard DNA extraction methods, Nocardia asteroides was repeatedly exposed to sodium dodecyl sulfate at 60 degrees C for 30 min; each extraction was followed by centrifugation, removal of the nucleic acid-rich supernatant, and suspension of the cell pellet in fresh sodium dodecyl sulfate. The pooled supernatants contained a substantially higher amount of DNA than the first supernatant alone. The possible implications of this procedure on the development of DNA probes are discussed.