Effect of Antimicrobial Compounds on Balamuthia mandrillaris Encystment and Human Brain Microvascular Endothelial Cell Cytopathogenicity

Abstract

Cycloheximide, ketoconazole, or preexposure of organisms to cytochalasin D prevented Balamuthia mandrillaris-associated cytopathogenicity in human brain microvascular endothelial cells, which constitute the blood-brain barrier. In an assay for inhibition of cyst production, these three agents prevented the production of cysts, suggesting that the biosynthesis of proteins and ergosterol and the polymerization of actin are important in cytopathogenicity and encystment.

Balamuthia mandrillaris is an emerging protozoan pathogen with the ability to produce central nervous system infections (6, 7, 19). The current treatment regimen involves a mixture of drugs to provide additive or synergistic effects, but even so mortality remains very high (∼98%) (4, 19). This may be due to difficulties in diagnosing Balamuthia amoebic encephalitis (BAE), resulting in a delay in the initiation of chemotherapy; poor penetration of antimicrobial compounds across the blood-brain barrier; and perhaps the ability of B. mandrillaris to switch its phenotype into the resistant cyst form. For the latter, it has been shown that B. mandrillaris cysts develop a relatively impermeable triple-walled structure (9, 15). In addition to possible drug resistance, cysts may also reactivate following antimicrobial chemotherapy leading to recurrence of infection. Thus, a complete understanding of B. mandrillaris encystment and identification of compounds that can interfere with the encystment process should be of value in the improved treatment of BAE. Given that B. mandrillaris is a close relative of Acanthamoeba (2), it may contain similar membrane sterols, i.e., ergosterol and its precursor cycloartenol and ergosterol-like sterols (11, 17). This is supported by findings that ketoconazole, a preferential inhibitor of ergosterol biosynthesis (5), exhibits amoebastatic effects on B. mandrillaris in vitro (14) and BAE patients showed some response to this compound (3). Here, we examined the roles of ergosterol biosynthesis, cytoskeletal rearrangements, and protein synthesis in B. mandrillaris encystment and determined whether inhibiting these pathways would block amoeba-mediated cytopathogenicity in cultured human brain microvascular endothelial cells (HBMEC).

B. mandrillaris ATCC 50209, isolated from the brain of a mandrill baboon, was obtained from the American Type Culture Collection and routinely cultured on host cell monolayers as feeder layers as previously described (6). For cytopathogenicity assays, primary HBMEC were isolated from human tissue and grown in RPMI 1640 medium containing 10% fetal bovine serum, 10% NuSerum, 2 mM glutamine, 1 mM pyruvate, penicillin (100 U/ml), streptomycin (100 U/ml), nonessential amino acids, and vitamins as previously described (1, 18).

RPMI 1640 medium induces optimal B. mandrillaris encystment at 37°C.

To determine the optimal conditions to induce encystment in B. mandrillaris, various parameters were used. Briefly, B. mandrillaris (>95% trophozoites at a cell density of 0.5 × 10⁵ to 5 × 10⁵/ml) were suspended in RPMI 1640 medium alone or containing glucose-NaCl (to achieve up to 500 mosmol) at temperatures ranging from 4°C to 42°C. Plates were incubated for up to 48 h, followed by the addition of sodium dodecyl sulfate (0.5% final concentration) to lyse the remaining trophozoites. Counts were performed with a hemocytometer, both before and after sodium dodecyl sulfate (SDS) treatment. To quantify encystment, the percentage of B. mandrillaris amoebae that transformed into cysts was determined as follows: % encystment = (no. of amoebae after SDS treatment/no. of amoebae before SDS treatment) × 100. Data are presented as the mean ± the standard error. To determine their viability, cysts prepared in the presence or absence of drugs but prior to SDS treatment were inoculated onto HBMEC monolayers and incubated for up to 7 days and periodically observed for the emergence of trophozoites. Our results demonstrated that B. mandrillaris suspended in RPMI 1640 medium alone at a cell density of 2 × 10⁵/ml and incubated at 37°C for 48 h exhibited optimal encystment (Table 1). Of note, the presence of MgCl₂ induced large clustering of amoebae, which presented problems in counting and thus MgCl₂ was omitted.

TABLE 1.

Encystment of B. mandrillaris at 37°C

Treatment	Mean % encystment ± SE
RPMI 1640 medium (2 × 105 amoebae/ml at 37°C)	45.5 ± 3.5
Osmolarity (2 × 105 amoebae/ml of RPMI 1640 medium at 37°C)
8% glucose	25 ± 1.4
16% glucose	21.5 ± 9.2
200 mM NaCl	23 ± 5.6
400 mM NaCl	25 ± 4.2
Tempb
4°C	10 ± 0.4
10°C	12.5 ± 3.5
20°C	16.5 ± 0.7
30°C	26.5 ± 2.1
37°C	53.3 ± 6.1
42°C	40 ± 7
No. of amoebae/ml of RPMI 1640 medium/well
5 × 104	29 ± 1.4
2 × 105	46.5 ± 3.6
5 × 105	15 ± 6.5
Cytochalasin Db
10 μM	26.5 ± 2.1 (42)a
50 μM	16.5 ± 0.7 (64)
100 μM	11 ± 2.8 (76)
Cycloheximideb
100 μM	3 ± 1.4 (94)
200 μM	2.5 ± 0.7 (95)
500 μM	0.75 ± 0.3 (98.4)
Clindamycinb
5 μg/ml	44 ± 4.2 (4)
10 μg/ml	46 ± 5.6 (0)
Flucytosineb
5 μg/ml	26 ± 2.8 (43)
10 μg/ml	20 ± 2.6 (56)
Artemisininb
5 μg/ml	28.5 ± 3.5 (37)
10 μg/ml	22 ± 4.2 (52)
Ketoconazoleb
5 μg/ml	58 ± 3.8 (0)
10 μg/ml	47.8 ± 4.2 (0)

^aValues in parentheses indicate percent inhibition of encystment.

^b2 × 10⁵ amoebae/ml of RPMI 1640 medium.

Protein synthesis is crucial in B. mandrillaris encystment and amoeba-mediated cytopathogenicity.

The eukaryotic translational inhibitor cycloheximide blocked B. mandrillaris encystment (more than 90%) at micromolar concentrations (Table 1), as measured by increased SDS lability. In addition, flucytosine, which affects RNA and DNA synthesis and inducs abnormal protein synthesis, partially inhibited B. mandrillaris encystment (Table 1), further supporting the cycloheximide findings, while artemisinin, an antimalarial endoperoxide from Artemisia annua (Chinese sweet wormwood) (13), partially inhibited B. mandrillaris encystment (Table 1). In contrast, clindamycin, an inhibitor of bacterial protein synthesis that binds to the 50S subunit and inhibits peptidyl transferase activity, had no significant effect on B. mandrillaris encystment (Table 1).

Next, we determined whether B. mandrillaris-mediated HBMEC death involves protein synthesis. Amoebae were treated with cycloheximide for 45 min in serum-free medium and then added to HBMEC monolayers in the presence of inhibitor in 24-well plates (10⁵ amoebae/well). Plates were incubated at 37°C in a 5% CO₂ incubator for up to 24 h. After this incubation, supernatants were collected and cytotoxicity was determined by measuring lactate dehydrogenase release with a cytotoxicity detection kit (Roche Applied Science) as previously described (16). The results revealed that cycloheximide, flucytosine, and artemisinin partially inhibited parasite-mediated HBMEC cytopathogenicity (Table 2). In contrast, clindamycin had no effect on B. mandrillaris-mediated HBMEC cytopathogenicity (Table 2). Of note, 200 μM cycloheximide alone exhibited significant HBMEC cytopathogenicity (data not shown).

TABLE 2.

Effects of various drugs on B. mandrillaris-mediated cytopathogenicity in HBMEC

Treatment	Mean % cytopathogenicity in HBMEC ± SE
None (amoebae alone)	74 ± 4.5
Cytochalasin D
10 μM	64.7 ± 4.5 (13)a
50 μM	43 ± 5.5 (42)
Cycloheximide
100 μM	41.3 ± 4 (44)
200 μM	81.4 ± 3.7 (0)
Clindamycin
5 μg/ml	62.2 ± 5.9 (16)
10 μg/ml	66.4 ± 2.1 (10)
Flucytosine
5 μg/ml	59.5 ± 3.5 (20)
10 μg/ml	29 ± 2.8 (61)
Artemisinin
5 μg/ml	48.6 ± 8.5 (34)
10 μg/ml	40.4 ± 2.8 (45)
Ketoconazole
5 μg/ml	64 ± 2.8 (14)
10 μg/ml	8.5 ± 2.7 (88.5)

^aValues in parentheses indicate percent inhibition of cytopathogenicity in HBMEC.

Both encystment in B. mandrillaris and amoeba-mediated HBMEC cytopathogenicity involve cytoskeletal rearrangements.

Given that encystment involves the acquisition of a triple-walled structure, we hypothesized that cytoskeletal rearrangements may play an important role in B. mandrillaris encystment. Encystment assays performed in the presence of cytochalasin D, an inhibitor of actin polymerization, blocked amoeba encystment (Table 1). To determine the role of phagocytosis in amoeba-mediated HBMEC death, parasites were treated with cytochalasin D for 45 min and then cytopathogenicity assays were performed. As shown in Table 2, cytochalasin D partially inhibited B. mandrillaris-mediated HBMEC death, suggesting that B. mandrillaris uses phagocytic processes to produce HBMEC damage. Because host cell destruction is a result of contact (8, 10), the term cytopathogenicity is used. In support, parasite-conditioned medium did not produce HBMEC destruction.

Ergosterol biosynthesis may play an important role in the encystment and cytopathogenicity of B. mandrillaris.

Previous studies have shown that ketoconazole has amoebastatic effects on B. mandrillaris (14), suggesting that ergosterol may be an important constituent of amoeba membranes. Here, we observed that cysts developing in the presence of ketoconazole (10 μg/ml) were only partially formed and appeared somewhat hollow. More importantly, these cysts were not viable, as determined by their inoculation on HBMEC monolayers as feeder layers and incubation for up to 7 days (data not shown). These findings led us to conclude that ketoconazole blocked encystment, suggesting a role for ergosterol biosynthesis in amoeba encystment (Table 1). Furthermore, ketoconazole exhibited protective effects against amoeba-mediated HBMEC cytopathogenicity (Table 2). In the presence of ketoconazole (10 μg/ml), the HBMEC monolayers remained intact while lactate dehydrogenase release was minimal. In addition, ketoconazole alone had no significant toxic effects on the host cells (<5%; data not shown).

Ketoconazole is known to interfere with the biosynthesis of ergosterol, which is an important structural component of the membrane, thus leading to defective membranes, increased permeability, and leakage of ions from the cell (12). This is not a surprising finding, given that Balamuthia is a close relative of Acanthamoeba and ergosterol is known to be a major sterol membrane component of Acanthamoeba (11, 17). These studies suggested that the ergosterol biosynthesis pathway may be a potential target in the rational development of therapeutic interventions against B. mandrillaris infections. This is supported by the fact that ergosterol biosynthesis is limited to fungi and protozoa, while human cells contain cholesterol. Further characterization of the ergosterol biosynthesis pathway, including its regulation, may suggest improved combinations of sterol biosynthesis inhibitors to optimize their antiparasitic effects.

In conclusion, these studies suggested that protein synthesis, cytoskeletal rearrangements, and possibly ergosterol biosynthesis are important pathways in B. mandrillaris encystment and that inhibiting these pathways blocked amoeba-mediated cytopathogenicity in cultured primary HBMEC. Future studies of the molecular mechanisms associated with B. mandrillaris encystment should provide insights into the biology of these important organisms that may also be of potential value in the rational development of therapeutic interventions.