Correlation of in vitro properties of Rhodococcus (Corynebacterium) equi with virulence for mice

To study the virulence of Rhodococcus (Corynebacterium) equi, seven ATCC strains of different serotypes were tested for their LD50 in mice, clearance of the organism from the lungs and spleen following intravenous or intratracheal inoculation, and in vitro interaction with murine peritoneal macrophages. Strains ATCC 33704 and 33705 were virulent for mice and multiplied in the lungs and spleen, resulting in death of the animal in 5 days. The other five strains were avirulent for mice. The number of bacteria in the lungs and spleen of mice given these five strains decreased immediately. Pulmonary clearance of strains ATCC 33703, 33706, and 33707 was significantly more rapid than that of the virulent strains ATCC 33704 and 33705 12 hr after inoculation. Complete clearance of the avirulent strain ATCC 33707 occurred by day 14, while that of virulent ATCC 33704 and 33705 strains occurred by day 30. The virulent strains ATCC 33704 and 33705 were resistant not only to phagocytosis but also to intracellular killing by macrophages. Strains ATCC 33702 and 33706 were rapidly killed by macrophages although they were rather resistant to phagocytosis. Strain ATCC 33703 was easily phagocytized though resistant to killing by macrophages. The most avirulent strains, ATCC 33707 and 6939, were easily phagocytized and rapidly killed by macrophages. These results indicate that virulence appeared to be related to the ability of the organisms to resist clearance from the lungs and spleen and to resist phagocytosis and intracellular killing by macrophages.     

Interaction of virulent and non-virulent Rhodococcus equi human isolates with phagocytes, fibroblast- and epithelial-derived cells

Abstract Rhodococcus equi is a facultative, intracellular, Gram-positive coccobacillus, increasingly reported in pneumonia of AIDS-infected patients. We investigated killing resistance properties of human R. equi virulent and avirulent human strains. Avirulent β-lactam-susceptible strains had lower intracellular colony forming units after 45 min incubation in murine macrophages J774 and human monocyte-macrophage TPH-1 than those of virulent strains. Only virulent β-lactam-resistant strains persisted within macrophages for at least 18 min only. A β-lactam-resistant mutant was obtained from a β-lactam-susceptible strain after selection in a penicillin G-containing culture medium. This mutant strain, like the natural virulent strains, persisted within macrophages, harboured cell-associated appendages, produced phage-like particles and induced, after its intravenous inoculation, a chronic infection in BALB/c nude mice. Supernatant culture of virulent strains transferred partial macrophage-killing resistance properties to avirulent strains. The same supernatant was toxic for L-929, HeLa and Vero cell cultures. These supernatant effects were heat-inactivated, trypsin-inactivated and did not seem to be linked to phage-like particle presence. These data argue that virulence, β-lactam-resistance, and macrophage-killing resistance are associated in human R. equi isolates. Moreover, only virulent strains produced uncharacterized toxic factors.     

IL-12 induction of resistance to pulmonary blastomycosis

BACKGROUND: Why severity varies in blastomycosis outbreaks remains unresolved. In experimental pulmonary blastomycosis, susceptibility varied in mouse strains. In susceptible BALB/c the response is Th-2, in immunized, resistance is associated with Th-1. Can susceptibility be redirected by IL-12? Methods, results: BALB/c bronchoalveolar and peritoneal macrophages (PM) were shown deficient in IL-12 production in response to IFN-gamma+LPS. High dose IL-12 (1 microg, subcutaneously) treatment of BALB/c infected intranasally with Blastomyces resulted in enhanced survival (P<0.008). Since IL-12 was poorly tolerated, a new protocol for infected mice, IL-12 0.1 or 0.3 microg, every other day, resulted in minimal toxicity; almost all treated mice survived (P<0.002 vs. controls). When lungs of surviving mice were cultured, the 0.1 microg regimen resulted in fewer (P<0.02) cfu. For weeks after treatment, in vitro IFN-gamma treatment enabled PM Blastomyces killing. After infection spleen cells from IL-12 treated mice produced 4-fold more IFN-gamma and 3-fold less IL-10 in response to Blastomyces. IL-10 abrogated activation of macrophages by IFN-gamma for enhanced Blastomyces killing. Conclusions: A proper IL-12 treatment protocol induces resistance (survival and decreased growth in lungs), low toxicity, macrophage responsiveness to IFN-gamma for killing Blastomyces, up-regulation of IFN-gamma and down-regulation of IL-10 production.     

Studies on the interactions of immunostimulated macrophages and Yersinia enterocolitica O:8

Immunological and electron microscopy investigations of the phagocytic and killing activities of peritoneal macrophages from rats and mice against Yersinia enterocolitica serotype O:8 cells were performed. The effect of in vivo application of cytoplasmic membranes (CM) from the stable Escherichia coli WF+ L-form on macrophage activity was also studied. It was established that rat macrophages more actively phagocytosed the plasmidless pYV(-) Y. enterocolitica cells, compared to the plasmid-bearing pYV(+) Y. enterocolitica cells. The killing ability against both variants of the Y. enterocolitica strain was significantly enhanced in macrophages from CM-treated rats after 2 h, 4 h, and 24 h incubation. The CM treatment enhanced the phagocytic activity of the macrophages. The in vitro interaction of normal and immunostimulated rat macrophages with both pYV(+) and pYV(-) variants of Y. enterocolitica did not lead to any additional apoptotic and necrotic changes in macrophages compared to control macrophages, which were cultivated without Y. enterocolitica. Electron-microscopic investigation showed that mouse macrophages eliminated Y. enterocolitica pYV(+) cells in vivo after 24 h. No engulfed or digested bacterial cells were observed. Activation of cell surfaces and vacuolization of macrophage cytoplasm, both of CM-treated non-infected and infected mice, were observed. The experimental results showed that Y. enterocolitica pYV(+) cells could be eliminated by peritoneal macrophages.     

ATP-induced killing of virulent Mycobacterium tuberculosis within human macrophages requires phospholipase D

The global dissemination of antibiotic-resistant Mycobacterium tuberculosis has underscored the urgent need to understand the molecular mechanisms of immunity to this pathogen. Use of biological immunomodulatory compounds to enhance antituberculous therapy has been hampered by the limited efficacy of these agents toward infected human macrophages and lack of information regarding their mechanisms of activity. We tested the hypotheses that extracellular ATP (ATPe) promotes killing of virulent M. tuberculosis within human macrophages, and that activation of a specific macrophage enzyme, phospholipase D (PLD), functions in this response. ATPe treatment of infected monocyte-derived macrophages resulted in 3.5-log reduction in the viability of three different virulent strains of M. tuberculosis. Stimulation of macrophage P2X7 purinergic receptors was necessary, but not sufficient, for maximal killing by primary macrophages or human THP-1 promonocytes differentiated to a macrophage phenotype. Induction of tuberculocidal activity by ATPe was accompanied by marked stimulation of PLD activity, and two mechanistically distinct inhibitors of PLD produced dose-dependent reductions in ATPe-induced killing of intracellular bacilli. Purified PLD restored control levels of mycobacterial killing to inhibitor-treated cells, and potentiated ATPe-dependent tuberculocidal activity in control macrophages. These results demonstrate that ATPe promotes killing of virulent M. tuberculosis within infected human macrophages and strongly suggest that activation of PLD plays a key role in this process.     

Comparison of the responses of peritoneal macrophages from Japanese flounder (Paralichthys olivaceus) against high virulent and low virulent strains of Edwardsiella tarda

In vivo infection studies in Japanese flounder (Paralichthys olivaceus) demonstrated that the number of viable cells of the virulent strain (NUF251) of Edwardsiella tarda increased gradually in kidney and hepato-pancreas after intraperitoneal injection, but the low virulent strain (NUF194) did not. To gain insight into the virulence factors of E. tarda, in vitro responses of Japanese flounder (P. olivaceus) peritoneal macrophages to these strains were compared in terms of phagocytosis, bactericidal activity, and reactive oxygen species (ROS) generation as measured by chemiluminescence (CL) responses. Microscopic observation revealed that these two strains of E. tarda were phagocytosed by the peritoneal macrophages, and there was no significant difference in the mean numbers of ingested bacteria per macrophage between these strains. A gradual increase in the number of viable cells of the highly virulent strain within macrophages was observed during 9h post-phagocytosis, whereas no significant replication of the low virulent strain within macrophages was detected. These results suggest that the virulent strain of E. tarda has an ability to survive and replicate within macrophages, while the low virulent strain has no such ability. When the peritoneal macrophages were exposed to the opsonized low virulent E. tarda strain, a rapid increase in CL response was induced. However, the highly virulent strain caused only background level of CL response. By the subsequent stimulation with phorbol myristate acetate, the macrophages exposed to the virulent E. tarda strain showed extremely higher CL response than that of the one exposed to the low virulent E. tarda strain. These results suggest that the virulent E. tarda prevents the activation of ROS generation system during phagocytosis, though the system is still capable of responding to other stimulation. The virulent strain significantly reduced the CL response induced by xanthine/xanthine oxidase system, while the low virulent strain had almost no effect. Furthermore, the virulent strain showed greater resistance to H(2)O(2) than the low virulent strain. Our results suggest that the virulent strain of E. tarda is highly resistant to ROS, and such ability might allow the organism to survive and multiply within phagocytes, and may serve to disseminate E. tarda throughout the host during in vivo infection.     

In Vitro Studies on the Mechanism of Acquired Resistance to Tuberculous Infection

The relationship between lymphocytes and macrophages in cellular immunity against tuberculous infection was studied by means of an in vitro cell culture system without addition of streptomycin. The peritoneal macrophages were obtained from normal mice or mice immunized with heat-killed tubercle bacilli in paraffin oil, boosted with live BCG and infected with H37Rv cells in vitro. The infected monolayers of macrophages were cultivated for 48 hr with immune lymphoid cells obtained from immunized mice. The intracellular growth of H37Rv cells 3,5 and 7 days after infection was examined by counting tubercle bacilli within infected macrophages under a microscope. 1) The increase of bacilli within macrophages derived from immunized mice was slightly smaller than that in normal macrophages. 2) The addition of immune lymph node cells to the macrophage monolayers resulted in a marked decrease in the number of bacilli within both normal and “immune” macrophages. Conversely, normal lymph node cells exhibited an enhancing effect on the intracellular bacillary growth. 3) Immune lymph node cells showed a higher capacity to cause macrophages to suppress intracellular growth of bacilli than that of splenic lymphoid cells or thymocytes after addition to macrophage monolayers. 4) The treatment of lymphoid cells with inhibitors of protein synthesis, cycloheximide or streptovitacin A, resulted in a remarkable reduction of the ability of sensitized lymphocytes to cause macrophages to suppress multiplication of intracellular bacilli.     

In vivo administration of recombinant IFN-gamma induces macrophage activation, and prevents acute disease, immune suppression, and death in experimental Trypanosoma cruzi infections

Recombinant murine IFN-gamma (rMu-IFN-gamma) was demonstrated to be a potent in vivo activator of mouse peritoneal macrophages to kill Trypanosoma cruzi in vitro and to be capable of conferring protection against death from acute T. cruzi infection. Following i.p. injections of rMu-IFN-gamma, resident peritoneal macrophages were cultured and infected with T. cruzi in vitro. Numbers of intracellular parasites were determined at different times thereafter. Ten or 100 micrograms (1 microgram = 6.5 X 10(5) U) of Mu-IFN-gamma, injected both 24 and 4 h before macrophage harvest, induced up to 99% inhibition of T. cruzi. One microgram of rMu-IFN-gamma was not effective under these conditions. In vitro inhibition of T. cruzi by peritoneal macrophages occurred by 24 h after infection and continued until at least 120 h after infection. There were no significant differences in initial parasite uptake by macrophages from IFN-gamma-treated or control mice, indicating that the rMu-IFN-gamma induced parasite killing. One i.p. dose of 10 micrograms was as effective as two doses if the single injection was given 24 h before macrophage harvest. In subsequent experiments, mice were given multiple injections of 10 micrograms rMu-IFN-gamma beginning 24 h before or 2 h after infection with virulent T. cruzi. Mice treated with rMu-IFN-gamma had significantly lower parasitemias and decreased morbidity compared with control mice. Proliferative responses to Con A and antibody responses to SRBC were not significantly lowered in IFN-gamma-treated mice, in contrast to untreated infected controls. All of the IFN-gamma-treated mice survived acute T. cruzi infection, whereas 100% of saline-treated infected mice died. It was demonstrated in this study that rMu-IFN-gamma activated mouse macrophages in vivo to kill T. cruzi and that rMu-IFN-gamma significantly reduced morbidity and immune suppression, and eliminated mortality resulting from acute infection with this parasite.     

Division and death rates of Salmonella typhimurium inside macrophages: use of penicillin as a probe.

In mouse peritoneal macrophages infected in vitro with Salmonella typhimurium the number of viable bacteria and the number of stainable bacteria detected by light microscopy both increased at similar rates with a doubling time of more than 1 h. Antibiotics were not present; instead extracellular bacteria were removed by frequently rinsing the cells. The bacterial doubling time in the same medium in the absence of macrophages was about 20 min. Penicillin added to macrophage monolayers rapidly entered the macrophages, reaching a diffusion equilibrium. The penicillin-induced bacterial death rate appeared to depend on the bacterial division rate as well as on the penicillin concentration. These properties of penicillin were used to monitor intracellular bacterial division and death rates. The results indicated that intracellular killing, with the disappearance of stainable bacteria, did not contribute to the extended doubling time in macrophages. It was concluded that the intracellular environment of the bacteria was probably growth inhibitory but not bactericidal.     

Macrophages in resistance to rickettsial infection: macrophage activation in vitro for killing of Rickettsia tsutsugamushi.

Rickettsia tsutsugamushi, strain Gilliam, replicates in cultures of resident peritoneal macrophages from BALB/c mice. Macrophage cultures treated with culture supernatants of spleen cells from rickettsial-infected mice stimulated with heat-killed rickettsiae markedly suppressed macrophage infection by rickettsiae. Rickettsiacidal activity of activated macrophages was dependent upon both lymphokine concentration and time of incubation in lymphokines. Treatment of macrophage cultures with lymphokines before exposure to viable rickettsiae resulted in an immediate decrease in percent macrophages infected and numbers of viable intracellular rickettsiae. In these cultures, enhanced intracellular killing was also apparent with further incubation (24 hr). The immediate effect of lymphokine-pretreated macrophages was dissociated from intracellular killing by infecting macrophage cultures first and adding lymphokines after infection. In these cultures, both percent macrophages infected and titers of viable intracellular rickettsiae were dramatically reduced as well.     

Differences in initial rate of intracellular killing of Salmonella typhimurium by resident peritoneal macrophages from various mouse strains

To determine the underlining mechanism of the difference in innate susceptibility of mouse strains to infection by Salmonella typhimurium, the ingestion and in vitro intracellular killing of S. typhimurium by resident peritoneal macrophages of mouse strains that differ in natural resistance to this microorganism has been studied. The results revealed that the rate constants of in vitro phagocytosis (Kph) in the presence of inactivated rabbit immune serum did not differ between macrophages of susceptible C57BL/10 and resistant CBA mice (for both strains: Kph = 0.021 min-1). The rate constant of in vitro intracellular killing (Kk) was determined 1) after in vivo phagocytosis (CBA, Kk = 0.055 min-1; C57BL/10, Kk = 0.031 min-1), 2) after in vitro phagocytosis of preopsonized bacteria (CBA, Kk = 0.020 min-1; C57BL/10, Kk = 0.012 min-1), and 3) during continuous phagocytosis in vitro (CBA, Kk = 0.029 min-1; C57BL/10, Kk = 0.013 min-1). With all three approaches, the initial rate of intracellular killing by normal macrophages of Salmonella-resistant CBA mice amounted to about 1.7 times the value found for macrophages of susceptible C57BL/10 mice (p less than 0.01). This trait difference was independent of the previous way of ingestion of the bacteria, unaffected by the kind of opsonization, and specific for S. typhimurium, because Staphylococcus aureus and Listeria monocytogenes were killed by macrophages of these mouse strains with equal efficiency (p greater than 0.50). These findings indicate that a difference in genetic background expressed in the efficacy of intracellular killing by resident peritoneal macrophages immediately upon ingestion of S. typhimurium is relevant for the innate resistance of mice against S. typhimurium.     

The role of IL-10 in promoting disease progression in leishmaniasis

To determine the role of IL-10 in cutaneous leishmaniasis, we examined lesion development following Leishmania major infection of genetically susceptible BALB/c mice lacking IL-10. Whereas normal BALB/c mice developed progressive nonhealing lesions with numerous parasites within them, IL-10(-/-) BALB/c mice controlled disease progression, and had relatively small lesions with 1000-fold fewer parasites within them by the fifth week of infection. We also examined a mechanism whereby Leishmania induced the production of IL-10 from macrophages. We show that surface IgG on Leishmania amastigotes allows them to ligate Fc gamma receptors on inflammatory macrophages to preferentially induce the production of high amounts of IL-10. The IL-10 produced by infected macrophages prevented macrophage activation and diminished their production of IL-12 and TNF-alpha. In vitro survival assays confirmed the importance of IL-10 in preventing parasite killing by activated macrophages. Pretreatment of monolayers with either rIL-10 or supernatants from amastigote-infected macrophages resulted in a dramatic enhancement in parasite intracellular survival. These studies indicate that amastigotes of Leishmania use an unusual and unexpected virulence factor, host IgG. This IgG allows amastigotes to exploit the antiinflammatory effects of Fc gamma R ligation to induce the production of IL-10, which renders macrophages refractory to the activating effects of IFN-gamma.     

Differential contribution of sodC1 and sodC2 to intracellular survival and pathogenicity of Salmonella enterica serovar Choleraesuis

Several of the most virulent Salmonella enterica strains possess two genes encoding periplasmic Cu,Zn superoxide dismutase, sodC1 and sodC2, located on a lambdoid prophage and on the chromosome, respectively. These genes contribute to Salmonella virulence by protecting bacteria from superoxide generated by the host's phagocytes. To investigate the respective contributions of sodC1 and sodC2 to the virulence of a clinical isolate of Salmonella enterica serovar Choleraesuis (S. choleraesuis), we have analyzed both the intracellular survival of wild type and sodC mutant strains within J774 macrophages and Caco-2 cells, and their ability to proliferate in intraperitoneally-infected mice in competition assays. In agreement with previous studies, mutant strains lacking one or both sodC genes were equally impaired in their ability to survive within activated macrophages. However, when macrophage killing experiments were carried out with non-opsonized bacteria, sodC2 contributed to intracellular survival more than sodC1, indicating that changes in the pathways of bacterial uptake can modify the relative role of the two sodC genes. More unexpectedly, we have found that the ability of S. choleraesuis to survive within Caco-2 cells was severely affected by inactivation of sodC genes, sodC2 being more important than sodC1. As Caco-2 cells actively produce superoxide, this suggests that oxygen radical production by colonic cells has a role in controlling proliferation of facultative intracellular bacteria. Mouse infection studies confirmed that, in the S. choleraesuis strain under investigation, both sodC genes are required to confer full virulence, sodC2 contributing slightly more than sodC1 to Salmonella pathogenesis. Our findings contrast with the results of other studies carried out in S. enterica serovar Typhimurium and suggest that the relative contributions of sodC1 and sodC2 to host-pathogen interactive biology may vary depending on the Salmonella serovar or strain.     

Candida albicans isolates with different genomic backgrounds display a differential response to macrophage infection

Few human pathogens possess the ability exhibited by Candida albicans to colonize and cause symptomatic infections at different body sites. The host immune system is the major factor determining whether this opportunistic yeast behaves as a commensal or as a pathogen, since C. albicans strains appear capable of expressing similar virulence factors in response to specific body-district cues. This report provides evidence showing that C. albicans isolates with diverse genomic backgrounds (b and c karyotypes) differently modulate their pathogenic potential when assayed in cocultures with human monocytic derived macrophages (THP-1 cells). Striking differences were observed in the ability to undergo bud-hypha transition, a relevant C. albicans virulence factor, between b and c karyotypes (P<0.0001) upon their internalization by macrophages. All c types were able to develop hyphal forms, resist intracellular killing, replicate, and escape from macrophages. The b type isolates, which were shown to be more efficiently ingested by THP-1 cells than the c type strains (P=0.013), were susceptible to intracellular killing and predominantly found as blastoconidia inside macrophages. Despite their different intracellular disposition, both b and c type isolates were equally able to undergo morphogenesis and to express NRG1 and HWP1 genes, markers of the bud-hypha transition program, during in vitro propagation. Since macrophages play a critical role in the host resistance to C. albicans, the different response of b and c isolates to macrophage infection suggests that the c type strains are better suited to behave as a more virulent strain cluster.     

Immunization of NMRI mice against virulent Toxoplasma gondii. Differing efficacy of eleven cyst-forming Toxoplasma strains

Mice were infected with eleven cyst-forming Toxoplasma strains of varying virulence and challenged 1 month later with highly virulent BK strain parasites. The early cellular reaction in vivo was estimated by collecting the peritoneal exudates 24 h after challenge. This consisted of 54.34--77.61% lymphocytes and 0-8.88% infected macrophages in the eleven immunized groups in contrast to 27.17% lymphocytes and 18.64% infected macrophages in the control group. The peritoneal exudate 72 h after challenge comprised of 55.88-73.86% lymphocytes and 0-6.97% infected macrophages compared to 25.88% lymphocytes and 92% infected macrophages in the control group. Following the virulent challenge, deaths occurred in seven of the eleven groups immunized with live strains and ranged 5-42% at the end of a 6-week observation period. The most virulent of the Toxoplasma strains used - Alt and Gail - gave higher mortality while the least virulent ones - 558, 1070, K8 and KSU - were solidly resistant. The significantly different mortality obtained with strains Alt, Gail and Witting in a similar experiment performed at an interval of 7 months is discussed.     

Variation in the virulence of strains of Mycoplasma pulmonis related to susceptibility to killing by macrophages in vivo.

The virulence of five strains of Mycoplasma pulmonis, as judged by their ability to survive in the respiratory tract and induce pneumonia in CBA mice, was related to the ability of viable organisms to persist in the peritoneal cavity. This appeared to be the result of differences in the ability of the strains to resist killing by peritoneal macrophages in vivo. It is suggested that resistance to phagocytosis by macrophages is an important determinant of virulence for M. pulmonis.     

Experimental Salmonellosis

When mononuclear phagocytes (macrophages), were infected with Salmonella enteritidis and confined in a diffusion chamber to incubate with normal macrophages, they confer on the normal macrophages cellular immunity, as detected by inhibition to intracellular multiplication of a virulent strain 116-54 and resistance to cellular degeneration caused by phagocytosis of bacteria. An immune ribonucleic acid (RNA) was extracted from the peritoneal macrophages maintained in tissue culture bottles in a homogeneous cell population which had been infected with strains of 5. enteritidis. When peritoneal macrophages, cultured in a homogeneous cell population, were treated in vitro with this agent, they developed cellular immunity and cellular antibody. The RNA preparation was not inactived by treatment with deoxyribonuclease, with pronase or with antibodies to a virulent strain 116-54 of S. enteritidis. These facts suggest that the macrophages constitute a cell line responsible for active antibody formation.     

Models for the study of the toxic action of Legionella pneumophila

L. pneumophila virulent culture and the filtrate of this culture disintegrated with ultrasound were shown to be toxic for guinea-pig peritoneal macrophages in vitro and for AKR mice. The virulent culture of this infective agent and the filtrate of the supernatant fluid obtained from the washings of its virulent culture had no toxic effect on a macrophage monolayer culture and on mice. The use of these models for characterizing Legionella intracellular toxic activity, as well as for characterizing Legionella strains isolated from patients and the environment, is proposed.     

Increased sensitivity of immune macrophages to the cytotoxic action of virulent shigellae

The in vitro interaction of live bacteria belonging to virulent and avirulent Shigella and Salmonella strains with peritoneal macrophages obtained from mice immunized by the intragastric administration of these bacteria has been studied. In contrast to Salmonella-activated macrophages capable of resisting the intracellular proliferation and the cytopathic action of homologous bacteria, Shigella-activated macrophages become more sensitive to the cytopathic action of virulent shigellae. The ability of shigellae to render an aggravating cytopathic effect on the activated macrophages correlates with the virulence of dysentery bacilli and is practically absent in avirulent strains, including S. flexneri 2a No. 516 M vaccine strain.     

Effects of long-term in vitro cultivation on Leishmania donovani promastigotes

Promastigotes of Leishmania donovani that had been subcultured in modified Tobie's medium for 2 to 3 years showed decreased infectivity and lack of virulence for hamsters and mice compared to newly transformed promastigotes. Amastigotes derived from these long-term promastigote cultures decreased in number rapidly in hamsters, but only slightly in mice, over a 48-day period. In cultured mouse and hamster macrophages infected in vitro, amastigotes derived from long-term cultures rapidly decreased to low numbers, which persisted. The same pattern was seen in macrophages treated with catalase, an inhibitor of the oxygen-dependent killing mechanism of the macrophage. Promastigotes from long-term cultures also differed from virulent first-passage promastigotes in size, growth patterns in Tobie's medium, and in the quantities of some of their antigens.