S403
Immunity to Acanthamoeba Antonio Ferrante
From the Department of Immunology and University Department of Paediatrics, Adelaide Children's Hospital, Adelaide, South Australia, Australia
Human serum contains antibodies, mainly of the IgM and IgG isotypes, to pathogenic species of Acanthamoeba. This, as well as the capacity of these amebas to activate complement via the alternative pathway, may bea first-line defense against acanthamoeba infections in humans. Both antibody and complement appear to be important in promoting recognition of these amebas by phagocytic cells such as neutrophils. However, killing of amebas by neutrophils is dependent on lymphokine/monokine priming of the neutrophil. This priming augments the respiratoryburst activity and release of lysosomal enzymes of neutrophils in their response to the ameba. The products of the oxygen-dependent respiratory burst appear to be of prime importance in the killing of this free-living ameba. Antibodies also may prevent tissue invasion by Acanthamoeba by inhibiting its adherence, phagocytic activity, and migration and by neutralizing cytopathogenic amebic agents. Studies on experimental Acanthamoeba infections in mice showed marked species and strain specificity with regard to induction of protection with amebic antigens. Immune compromise or, alternatively, invasion at unique body sites in healthy individuals may form the basis for human infection with Acanthamoeba. The potential of Acanthamoeba species to cause infections in humans was recognized by Culbertson et a1. [1] some 30 years ago. However, studies with another free-living ameba, Naegleriafowleri, clearly established that free-living amebas were pathogenic to humans [2, 3]. More recently, Acanthamoeba species have been documented as causative agents in human amebic infections [4]. The diseases caused by freeliving amebas appear to fall into three well-defined disease entities. N. fowleri causes ~ meningoencephalitis that resembles fulminant bacterial meningitis, whereas Acanthamoeba species cause either chronic amebic encephalitis or chronic keratitis. Acanthamoebas exist in cyst and trophozoite stages in the environment. The trophozoite stage is the infective and invasive form of the organism. Patients with chronic amebic encephalitis have been suspected of having infections at primary sites such as the skin and respiratory tract [4, 5], and the encephalitic form of the disease appears to occur most frequently in debilitated and chronically ill individuals, particularly those who are receiving immunosuppressive therapy. In contrast, healthy individuals develop acanthamoeba keratitis, usually after minor trauma to the eye that involves contaminated water or soft contact lens wear.
Complement Activation Acanthamoeba species have been shown to activate complement by the alternative pathway [6]. When incubated in Reprints and correspondence: Dr. A. Ferrante, Department of Immunology, Adelaide Children's Hospital, Adelaide, South Australia 5006, Australia. Financial support: The National Health and Medical Research Council of Australia. Reviews of Infectious Diseases 1991;13(Suppl 5):8403-9 © 1991 by The University of Chicago. All rights reserved. 0162-0886/91/1302-0071$02.00
human serum, Acanthamoeba culbertsoni A-I, A. culbertsoni A-5, and Acanthamoeba rhysodes (HN-3) depleted the serum of complement hemolytic activity. All detectable complement activity was removed with concentrations of 1 x 107 trophozoites/mL of serum with each of the three species. Complement activation in vitro led to the rapid lysis of these amebas. This lysis also occurred after activation of complement by the alternative pathway, which occurred independently of antibody since adsorption of human serum at 4°C to remove antibodies did not remove the lytic activity for all species of Acanthamoeba [6]. Sialic acid on plasma membranes prevents activation of the alternative pathway of complement by increasing the affinity of factor ,BIH for C3b relative to that of factor B. This change in affinity prevents formation of the alternative pathway convertase (C3b,Bb), so that organisms with sialic acid may be spared this natural host defense mechanism. The capacity of Acanthamoeba to activate the alternative pathway of complement is consistent with findings that membranes of Acanthamoeba lack sialic acid [7, 8]. It has been demonstrated that the glycoprotein coat on the plasma membrane of African trypanosomes prevents the plasma membrane of these trypanosomes, which lacks sialic acid, from activating the alternative complement pathway [9]. Acanthamoeba species lack a coat or capsule [10]. It appears that in vitro lysis of Acanthamoeba occurs as a result of the concerted action of C5, C6, 0, C8, and C9 [6], which leads to membrane damage [6]. However, there is no evidence that this type of mechanism operates in vivo. For example, studies with trypanosomes have shown that these parasites are cleared by the mononuclear phagocytic system as nonlysed organisms [11]. Although serum can lyse Trypanosoma musculi, when phagocytic cells are present, no lysis of parasites occurs even in vitro [12]. Similar observations were
S404
made with A. culbertsoni (author's unpublished observations). It thus appears that the main function of complement activation is to generate opsonic factors such as C3b for recognition of the ameba by phagocytic cells. Activation of complement also leads to the generation of mediators of inflammation such as the anaphylatoxins, C3a and C5a, which contribute to the pathogenesis of parasitic diseases. The uncontrolled activation of complement by components of Acanthamoeba may explain in part the edema and damage to blood vessel walls characteristic of acanthamoeba infection.
Antibodies in Human Serum Human serum contains antibodies to both pathogenic Naegleria and Acanthamoeba species (table 1). Titers of antibody to Acanthamoeba between 1:20and 1:80were observed in adults in New Zealand [13]. These antibodies also were present in cord blood, a finding that suggests that antibodies to Acanthamoeba are transferred placentally. The antibodies were mainly of the IgM and IgG isotypes, as IgA was barely detectable. The results also showed that in human serum the concentration of antibodies to Acanthamoeba was much higher than the concentration of antibodies to Naegleria. "Natural" antibodies to protozoan parasites havebeen widely observed. For example, African trypanosomes that lack a glycoprotein surface coat are agglutinated by serum from a range of animal hosts that have not had contact with the organism [14]. Other studies have demonstrated that some trypanosome species, such as Trypanosoma lewisi and T. musculi, have a high host specificity and infect the rat and mouse, respectively. Sera from other animals have been shown to contain "natural" antibodies to these trypanosome species [15, 16]. The origin of these antibodies in the serum of animals that have not been exposed to these organisms is not known, but it has been suggested that they arise as a result of exogenous stimulation by cross-reacting antigens of bacterial, viral, or plant origin [14]. It is likely that in the case of free-living amebas, however, stimulation with amebic antigens is responsible for the presence of antibodies in human serum. Unlike trypanosomes, Naegleria and Acanthamoeba species are ubiquitous in the environment and thus could be a source of constant antigenic stimulation. This view is supported by the finding that antibodies to all trypanosome species that have been studied are predominantly in the IgM class [14-16], while those to Naegleria and Acanthamoeba species are distributed evenly in the IgM and IgG classes [13].
Stimulation of Immunity in Experimental Animals Mice infected intranasally with trophozoites of A. culbertsoni A-I develop fatal meningoencephalitis, with an approximate mean survival time of 7 days [17, 18]. Immunization of mice with sonicated antigen of A. culbertsoni A-I induced pro-
RID 1991;13 (Suppl 5)
Ferrante
Table 1. Human serum antibodies to free-living amebas. Genus, species Naegleria fowleri gruberi Acanthamoeba culbertsoni castellanii
Titer of IIF antibody *
Antibody isotype
Presence in cord blood
1: 5-1 :20
Mainly IgM and IgG
+
1:20-1 :80
Mainly IgM and IgG
+
* IIF = indirect immunofluorescence.
tection against a lethal intranasal challenge with this ameba. Significant protection was induced with one immunizing dose of the antigen [18]. After three immunizing doses, almost all animals were protected [18]. These findings contrast with those for a similar antigen preparation from N. fowleri, which induced very little protection against N. fowleri in mice even when repeated immunization was used [19]. The protection induced in the case of A. culbertsoni A-I was highly specific for this Acanthamoeba. Immunization with similar antigenic preparations from either A. culbertsoni A-5, A. rhysodes (HN-3), or Acanthamoeba polyphaga (MRK-IA) failed to induce any protection against a lethal challenge with A. culbertsoni A-I [18]. When the culture fluid from A. culbertsoni A-I was used as a source of antigen for immunizing mice, the animals were significantly protected from infection by the ameba [18]. This finding is similar to results obtained with N. fowleri [19].
The Role of Phagocytic Cells Neutrophils appear to play an important role in immunity to pathogenic free-living amebas [20-27]. Studies with N. fowleri showed that an important characteristic of immune mice is the rapid deployment of neutrophils at sites of amebic challenge [20, 21, 25]. Abrogation of neutrophil function and the depletion of neutrophils in immune mice significantly reduces the immunity to amebas [25]. Studies on both Naegleria and Acanthamoeba have shown that lymphocytes and macrophages from human peripheral blood do not act as effector cells against these amebas [22, 23], even in the presence of antibodies. Human neutrophils also fail to kill both types of ameba [22, 23]. But the same leukocytes, when treated with a conditioned medium that contained lymphokines from phytohaemagglutinin-activatedmononuclear leukocytes, were capable of killing both N. fowleri and A. culbertsoni A-I. Killing by lymphokinealtered neutrophils was found to require the presence of either antibody or complement [22, 23]. Usually a number of neutrophils attacked the amebas, which is characteristic of the neutrophil extracellular killing mechanism. Studies with N. fowleri have demonstrated that the neutrophil-killing mechanism involves both the oxidative respiratory system and the
Immunity to Acanthamoeba
RID 1991;13 (Suppl 5)
Figure 1. Diagramatic representation of possible interactions of neutrophils with free-living amebas. Neutrophils recognize and attach to amebas via Fc receptors (FcR) and/or complement receptors (CR). Other lectin-like receptors may also be involved. These interactions alone are not adequate to promote killing of amebas or to mobilize the relevant antimicrobial biochemical responses of the neutrophil. In contrast, pretreatment of neutrophils with polypeptide cytokines such as TNFa leads to significant stimulation of the respiratory burst (NADPH oxidase activation), which results in the production of oxygen-derived reactive species (ODRS) and the release of constituents of specific (spec.) and azurophil (azur.) granules. This positive response requires involvement of either FcR or CR. Granule components alone appear not to be amebicidal. In the case of N. fowleri, the ameba is highly sensitive to HOCI produced by the myeloperoxidase-H202halide system. In contrast, Acanthamoeba is resistant to HOCI but is sensitive to H202. Other components shown: the hexose monophosphate shunt (HMP) is closely linked to the activity of the NADPH oxidase; catalase (Cat), superoxide dismutase (SOD), and reduced glutathione (GSH) are involved in scavenging ODRS and in protecting the cell from auto-oxidation. BPI = bacterial permeability increasing protein; MPO = myeloperoxidase.
B 12 binding
protein
Lactoferrin Lysozyme BPI
{3 -glucuronidase Collagenase Elastase Lysozyme Myeloperoxidase BPI
: :l~: : ;:{~ l: : :~ : : : : {:/~f: /
S405
S406
Ferrante
35
Antiamebic Properties of Antibodies
30
25
:> g w
20
U Z
w
U
(J)
W
Z
15
~
::::>
...J
~
W
I
RID 1991;13 (Suppl 5)
10
U
5
0 0
4
6
8
10
12
TIME (min)
Figure 2. Kinetics of the lucigenin-dependent neutrophil chemiluminescence response and the effects of TNFa. Neutrophils (PMN) were pretreated for 30 minutes with 100 U of TN Fa and then tested for response in the presence or absence of A. culbertsoni A-I treated with human serum and fixed with formalin. Neutrophils not treated with TNFa (controls) were incubated with similar additions. The lucigenin-dependent chemiluminescence was totally inhibitable by superoxide dismutase, showing that the assay measures superoxide produced by neutrophils. Measurements were made in a luminometer.
enzyme found in azurophil granules, myeloperoxidase [24] (figure 1). While the amebas were sensitive to H20 2 , most sensitivity was to hypochlorite (HOCI), a product of the myeloperoxidaseH20 2-halide system [24]. Myeloperoxidase-deficient neutrophils lacked amebicidal activity. In contrast, A. culbertsoni A-I was much more resistant than N. fowleri to HOCI but appeared to be more sensitive to H 20 2 (author's unpublished observations) (figure 1). Direct addition of material extracted from granules, which was highly effective for killing Salmonella minnesota R595 and in damaging human and bovine cartilage in vitro [24, 28], had no effect on these amebas [24]. A number of T cell and macrophage cytokines have been shown to augment both the respiratory burst and release of lysosomal enzyme by neutrophils [27, 29-37], and this is probably the primary mechanism by which lymphokines enhance the killing of amebas by neutrophils. Studies with A. culbertsoni A-I showed that the macrophage cytokine, tumor necrosis factor ex (TNF ex), augmented the respiratory response to this ameba (figure 2).
Amebic invasion and destruction of tissues is dependent on a number of properties of free-living amebas. These properties include their capacity to adhere to mucosal surfaces; to migrate through tissues; to phagocytose; and to release oxygen radicals and a variety of cytopathogenic agents/enzymes, such as elastase, which degrade a variety of connective tissues [38]. The migratory characteristics of Acanthamoeba species are shown in figures 3 and 4. Migration of the pathogenic A. culbertsoni A-I and of the nonpathogenic A. culbertsoni A-5 were similar at 37°C (figure 3) and 28°C (figure 4). A pathogenic ameba, A. polyphaga (MRK-IA), was similarly active at both temperatures. However, A. rhysodes (HN-3) migrated poorly at these temperatures. It is interesting that N. fowleri migrated poorly at the lower temperature [39], a finding that indicates that some species of Acanthamoeba have greater tissue-penetrating properties relative to those of species of N. fowleri at lower temperatures that may be encountered in skin lesions, the nasal passages, and eyes. Antibodies may function in various ways to limit the invasiveness of Acanthamoeba. Addition of antibodies to freeliving amebas prevents their adhesion and spreading. The cells "round" and show reduced migratory properties [40]. Phagocytosis may be inhibited by antibodies, and it is most likely that antibodies neutralize the amebic cytopathogenic agents [13]. Finally, antibodies, particularly of the IgG isotypes, promote neutrophil-mediated killing of amebas. These mechanisms may operate collaboratively to prevent invasion by Acanthamoeba and to promote its elimination from body tissues.
Concluding Remarks Although understanding of the mechanisms of immunity to N. fowleri has progressed in the last 10 years, a concerted effort needs to be made to understand the defense mechanisms that operate against Acanthamoeba species. This presents a major task because of the different types of diseases caused by Acanthamoeba that involve a number of different species. Recent findings suggest that immunity to Acanthamoeba species is deployed as depicted in figure 5. Amebas that invade from skin lesion are opsonized by antibody and complement. Antibody may control invasiveness of the organism by neutralizing amebic cytopathogenic substances and by inhibiting phagocytosis of amebas and their adherence to tissue components. Phagocytic cells such as neutrophils are attracted to the infection site by chemotactic agents likely to be generated when complement is activated. Evidence suggests that sensitized T cells and macrophages playa functional role by releasing lymphokines that activate neutrophils, an essential step in the process by which leukocytes confine and kill the amebas.
A. polyphaga (MRK-1A) A. culbertsoni (A-1)
30
A. culbertsoni (A5)
E E
20
Z
Figure 3. Migration of Acanthamoeba
o
species under agaroseat 37°C. Resultsare presented as the mean ± SE of six experiments.
~
« II:
C)
~ 10
A. rhysodes (HN-3)
o-+--~=-J.--""-===~:::::::::':"----r-----r----
o
2
3
4
5
TIME (hrs)
3'5
3·0
2·5
E E
z
Migration of Acanthamoeba species under agaroseat 28°C. Resultsare presentedas the mean of four experiments. Figure 4.
a
2·0
i=
-c a:
C)
~
1·5
1·0
0·5
(HN-3)
o-+o-~~-...- -...-...===f=:::::::::==~-----r---
o
3
TIME (hrs)
4
RID 1991;13 (Suppl 5)
Ferrante
5408
wound
mast cell
••
vasoactive
complement
.: ,
- -
~
_ _ _ _ _ _...:.b::..:.lo=-:od vessel
==================Figure S. Possible host responses to Acanthamoeba species. Penetration of ameba at mucosal surfaces, skin, or other tissues may be dealt with immediately by a collaborative action of different components of the immune system. Evidence at present indicates a requirement for activation of phagocytic cells by lymphokines, antibody, and complement - the activated cells acting as recognition factors - and a requirement for antibody to prevent normal amebic phagocytic activity.
Persistent acanthamoeba infection may be a consequence of immune depression, a likely occurrence in chronically iII and debilitated individuals who are receiving immunosuppressive therapy. Reduced resistance probably is the result of decreased antibody production, lowered complement levels, a depressed rate of accumulation of leukocytes, inhibited leukocyte function, and impaired Iymphokine production. Infection in healthy individuals also may occur if a site where levels of antibodies and complement are low, such as the eye, is involved. In view of the unavailability of effective chemotherapy for acanthamoeba infections, immunochemotherapy may be an important consideration for treatment of these infections.
6.
7. 8. 9.
10. 11.
12.
References 1. Culbertson CG, Smith JW, Minner JR. Acanthamoeba: observations on animal pathogenicity [letter]. Science 1958;127:1506 2. Fowler M, Carter RF. Acute pyogenic meningitis probably due to Acanthamoeba sp.: a preliminary report. BMJ 1965;2:740-2 3. Carter RF. Description of Naegleria sp. isolated from two cases of primary amoebic meningoencephalitis and of the experimental pathological changes induced by it. J Pathol 1970;100:217-44 4. Martinez AJ, ed. Free-living amebas: natural history, prevention, diagnosis, pathology and treatment of disease. Boca Raton, FL: CRC Press 1985 5. Thong YH, Ferrante A. Experimental pharmacology. In: Rondanelli EG, ed. Amphizoic amoebae: human pathology. Infectious diseases
13.
14. 15.
16.
17.
color atlas monographs. Padua, Italy: Piccin Nuova Libraria, 1987: 251-72. Ferrante A, Rowan-Kelly B. Activationof the alternative pathwayof complement by Acanthamoeba culbertsoni. Clin Exp Immunol 1983; 54:477-85 Korn ED, Olivecrona T. Composition of an amoeba plasma membrane. Biochem Biophys Res Commun 1971;45:90-7 Stevens AR. Biochemical studies of pathogenic free-living amoebae. Giornale di Malatte Infettive e Parassitarie 1977;29:690-6 Ferrante A, Allison AC. Alternative pathway activation of complement by African trypanosomes lacking a glycoprotein coat. Parasite Immunol 1983;5:491-8 Bowers B, Korn ED. The fine structure of Acanthamoeba castellanii. I. The trophozoite. J Cell Bioi 1968;39:95-111 Ferrante A, Jenkins CR. Evidence implicating the mononuclear phagocytic system of the rat in immunity to infection with Trypanosoma lewisi. Aust J Exp Bioi Med Sci 1978;56:201-9 Ferrante A. The role of the macrophage in immunity to Trypanosoma musculi. Parasite Immunol 1986;8:117-27 Cursons RTM, Brown TJ, Keys EA, Moriarty KM, Till D. Immunity to pathogenic free-living amoebae: role of humoral antibody. Infect Immun 1980;29:401-7 Ferrante A, Allison AC. Natural agglutinins to African trypanosomes. Parasite Immunol 1983;5:539-46 Ferrante A, The role of natural agglutinins and trypanolytic activity in host specificity to Trypanosoma musculi. Parasite ImmunoI1984;6: 509-17 Ferrante A. Trypanolytic activity, agglutinins, and opsonins in sera from animals refractory to Trypanosoma lewisi. Infect Immun 1985; 49:378-82 Culbertson CG. The pathogenicity of soil amebas. Annu Rev Microbiol 1971;25:231-54
RID 1991;13 (Suppl 5)
Immunity to Acanthamoeba
18. Rowan-Kelly B, Ferrante A. Immunization with killed Acanthamoeba culbertsoni antigen and amoeba culture supernatant antigen in experimental acanthamoeba meningoencephalitis.Trans R Soc Trop Med Hyg 1984;78:179-82 19. Thong YH, Ferrante A, Rowan-Kelly B, O'Keefe D. Immunization with live amoebae, amoebic lysate and culture supernatant in experimental Naegleria meningoencephalitis. Trans R Soc Trop Med Hyg 1980; 74:570-6 20. Ferrante A, Thong YH, Unique phagocyticprocess in neutrophil-mediated killing of Naegleria fowleri. Immunol Lett 1980;2:37-41 21. Thong YH, Carter RF, Ferrante A. Rowan-Kelly B. Site of expression of immunity of Naegleria fowleri in immunized mice. Parasite Immunol 1983;5:67-76 22. Ferrante A, Mocatta TJ. Human neutrophils require activation by mononuclear leucocyte conditioned medium to kill the pathogenic free-living amoeba, Naegleria fowleri. Clin Exp Immunol 1984; 56:559-66 23. Ferrante A, Abell TJ. Conditioned medium from stimulated mononuclear leukocytes augments human neutrophil-mediated killing of a virulent Acanthamoeba sp. Infect Immun 1986;51:607-17 24. Ferrante A, Hill N, Abell TJ, Pruul H. Role of myeloperoxidase in the killing of Naegleriafowleri by lymphokine-altered human neutrophils. Infect Immun 1987;55:1047-50 25. Ferrante A, Carter RF, Lopez AF, Rowan-KellyB, Hill NL, Vadas MA. Depression of immunity to Naegleria fowleri in mice by selectivedepletion of neutrophils with monoclonal antibody. Infect Immun 1988;56:2286-91 26. Ferrante A, Hill NL, Goh DHB. Kumaratilake L. Altered neutrophils in mice immune to experimental Naegleria amoebic meningoencephalitis. Immunol Lett 1989;22:301-6 27. Ferrante A. Augmentation of the neutrophil response to Naegleria fowleri by tumour necrosis factor alpha. Infect Immun 1989;57:3110-5 28. Kowanko IC, Bates EJ. Ferrante A. Mechanisms of human neutrophilmediated cartilage damage in vitro: the role of lysosomal enzymes, hydrogen peroxide and hypochlorous acid. Immunol Cell Bioi 1989; 67:321-9 29. Weisbart RH, Golde OW, Clark SC, Wong GG, Gasson rc. Human
30.
31.
32.
33.
34.
35.
36.
37.
38. 39. 40.
S409
granulocyte-macrophage colony-stimulating factor is a neutrophil activator. Nature 1985;314:361-2 Klebanoff SJ, Vadas MA, Harlan JM, Sparks LH, Gamble JR, Agosti JM, Waltersdorph AM. Stimulation ofneutrophils by tumour necrosis factor. J Immunol 1986;136:4220-5 Shalaby MR, Palladino MA Jr, Hirabayashi SE, Eessalu TE, Lewis GO, Shepard HM, Aggarwal BB. Receptor binding and activation of polymorphonuclear neutrophils by tumor necrosis factor-alpha. J Leukocyte Bioi 1987;41:196-204 Kowanko IC, Ferrante A. Stimulation of neutrophil respiratory burst and lysosomal enzyme release by human interferon-gamma. Immunology 1987;62:149-51 Kowanko IC, Ferrante A. Interleukin-2 inhibits migration and stimulates respiratory burst and degranulation of human neutrophils in vitro. Immunol Lett 1987;15:285-9 Ferrante A, Nandoskar M, Bates EJ, Goh ORB, Beard U. Thmour necrosis factor beta (lymphotoxin) inhibits locomotion and stimulates the respiratory burst and degranulation of neutrophils. Immunology 1988;63:507-12 Ferrante A, Nandoskar M, Walz A, Goh DHB, Kowanko rc Effects of tumour necrosis factor alpha and interleukin-l alpha and beta on human neutrophil migration, respiratory burst and degranulation. Int Arch Allergy Appl Immunol 1988;86:82-91 Nathan CF. Neutrophil activationon biological surfuces. Massive secretion of hydrogen peroxide in response to products of macrophages and lymphocytes. J Clin Invest 1987;80:1550-60 Ferrante A. Tumor necrosis factor alpha potentiates neutrophil antimicrobial activity: increased fungicidal activity against Torulopsis glabrata and Candida albicans and associated increases in oxygen radical production and lysosomal enzyme release. Infect Immunol 1989;57:2115-22 Ferrante A, Bates EJ. Elastase in the pathogenic free-living amoebae Naegleria and Acanthamoeba spp. Infect Immun 1988;56:3320-1 Thong YH, Ferrante A. Migration patterns of pathogenic and nonpathogenic Naegleria spp. Infect Immun 1986;51:177-80 Ferrante A, Thong YH. Antibody induced capping and endocytosis of surface antigens in Naegleriafowleri. Int J ParasitoI1979;9:599-601
Immunity to Acanthamoeba Antonio Ferrante
From the Department of Immunology and University Department of Paediatrics, Adelaide Children's Hospital, Adelaide, South Australia, Australia
Human serum contains antibodies, mainly of the IgM and IgG isotypes, to pathogenic species of Acanthamoeba. This, as well as the capacity of these amebas to activate complement via the alternative pathway, may bea first-line defense against acanthamoeba infections in humans. Both antibody and complement appear to be important in promoting recognition of these amebas by phagocytic cells such as neutrophils. However, killing of amebas by neutrophils is dependent on lymphokine/monokine priming of the neutrophil. This priming augments the respiratoryburst activity and release of lysosomal enzymes of neutrophils in their response to the ameba. The products of the oxygen-dependent respiratory burst appear to be of prime importance in the killing of this free-living ameba. Antibodies also may prevent tissue invasion by Acanthamoeba by inhibiting its adherence, phagocytic activity, and migration and by neutralizing cytopathogenic amebic agents. Studies on experimental Acanthamoeba infections in mice showed marked species and strain specificity with regard to induction of protection with amebic antigens. Immune compromise or, alternatively, invasion at unique body sites in healthy individuals may form the basis for human infection with Acanthamoeba. The potential of Acanthamoeba species to cause infections in humans was recognized by Culbertson et a1. [1] some 30 years ago. However, studies with another free-living ameba, Naegleriafowleri, clearly established that free-living amebas were pathogenic to humans [2, 3]. More recently, Acanthamoeba species have been documented as causative agents in human amebic infections [4]. The diseases caused by freeliving amebas appear to fall into three well-defined disease entities. N. fowleri causes ~ meningoencephalitis that resembles fulminant bacterial meningitis, whereas Acanthamoeba species cause either chronic amebic encephalitis or chronic keratitis. Acanthamoebas exist in cyst and trophozoite stages in the environment. The trophozoite stage is the infective and invasive form of the organism. Patients with chronic amebic encephalitis have been suspected of having infections at primary sites such as the skin and respiratory tract [4, 5], and the encephalitic form of the disease appears to occur most frequently in debilitated and chronically ill individuals, particularly those who are receiving immunosuppressive therapy. In contrast, healthy individuals develop acanthamoeba keratitis, usually after minor trauma to the eye that involves contaminated water or soft contact lens wear.
Complement Activation Acanthamoeba species have been shown to activate complement by the alternative pathway [6]. When incubated in Reprints and correspondence: Dr. A. Ferrante, Department of Immunology, Adelaide Children's Hospital, Adelaide, South Australia 5006, Australia. Financial support: The National Health and Medical Research Council of Australia. Reviews of Infectious Diseases 1991;13(Suppl 5):8403-9 © 1991 by The University of Chicago. All rights reserved. 0162-0886/91/1302-0071$02.00
human serum, Acanthamoeba culbertsoni A-I, A. culbertsoni A-5, and Acanthamoeba rhysodes (HN-3) depleted the serum of complement hemolytic activity. All detectable complement activity was removed with concentrations of 1 x 107 trophozoites/mL of serum with each of the three species. Complement activation in vitro led to the rapid lysis of these amebas. This lysis also occurred after activation of complement by the alternative pathway, which occurred independently of antibody since adsorption of human serum at 4°C to remove antibodies did not remove the lytic activity for all species of Acanthamoeba [6]. Sialic acid on plasma membranes prevents activation of the alternative pathway of complement by increasing the affinity of factor ,BIH for C3b relative to that of factor B. This change in affinity prevents formation of the alternative pathway convertase (C3b,Bb), so that organisms with sialic acid may be spared this natural host defense mechanism. The capacity of Acanthamoeba to activate the alternative pathway of complement is consistent with findings that membranes of Acanthamoeba lack sialic acid [7, 8]. It has been demonstrated that the glycoprotein coat on the plasma membrane of African trypanosomes prevents the plasma membrane of these trypanosomes, which lacks sialic acid, from activating the alternative complement pathway [9]. Acanthamoeba species lack a coat or capsule [10]. It appears that in vitro lysis of Acanthamoeba occurs as a result of the concerted action of C5, C6, 0, C8, and C9 [6], which leads to membrane damage [6]. However, there is no evidence that this type of mechanism operates in vivo. For example, studies with trypanosomes have shown that these parasites are cleared by the mononuclear phagocytic system as nonlysed organisms [11]. Although serum can lyse Trypanosoma musculi, when phagocytic cells are present, no lysis of parasites occurs even in vitro [12]. Similar observations were
S404
made with A. culbertsoni (author's unpublished observations). It thus appears that the main function of complement activation is to generate opsonic factors such as C3b for recognition of the ameba by phagocytic cells. Activation of complement also leads to the generation of mediators of inflammation such as the anaphylatoxins, C3a and C5a, which contribute to the pathogenesis of parasitic diseases. The uncontrolled activation of complement by components of Acanthamoeba may explain in part the edema and damage to blood vessel walls characteristic of acanthamoeba infection.
Antibodies in Human Serum Human serum contains antibodies to both pathogenic Naegleria and Acanthamoeba species (table 1). Titers of antibody to Acanthamoeba between 1:20and 1:80were observed in adults in New Zealand [13]. These antibodies also were present in cord blood, a finding that suggests that antibodies to Acanthamoeba are transferred placentally. The antibodies were mainly of the IgM and IgG isotypes, as IgA was barely detectable. The results also showed that in human serum the concentration of antibodies to Acanthamoeba was much higher than the concentration of antibodies to Naegleria. "Natural" antibodies to protozoan parasites havebeen widely observed. For example, African trypanosomes that lack a glycoprotein surface coat are agglutinated by serum from a range of animal hosts that have not had contact with the organism [14]. Other studies have demonstrated that some trypanosome species, such as Trypanosoma lewisi and T. musculi, have a high host specificity and infect the rat and mouse, respectively. Sera from other animals have been shown to contain "natural" antibodies to these trypanosome species [15, 16]. The origin of these antibodies in the serum of animals that have not been exposed to these organisms is not known, but it has been suggested that they arise as a result of exogenous stimulation by cross-reacting antigens of bacterial, viral, or plant origin [14]. It is likely that in the case of free-living amebas, however, stimulation with amebic antigens is responsible for the presence of antibodies in human serum. Unlike trypanosomes, Naegleria and Acanthamoeba species are ubiquitous in the environment and thus could be a source of constant antigenic stimulation. This view is supported by the finding that antibodies to all trypanosome species that have been studied are predominantly in the IgM class [14-16], while those to Naegleria and Acanthamoeba species are distributed evenly in the IgM and IgG classes [13].
Stimulation of Immunity in Experimental Animals Mice infected intranasally with trophozoites of A. culbertsoni A-I develop fatal meningoencephalitis, with an approximate mean survival time of 7 days [17, 18]. Immunization of mice with sonicated antigen of A. culbertsoni A-I induced pro-
RID 1991;13 (Suppl 5)
Ferrante
Table 1. Human serum antibodies to free-living amebas. Genus, species Naegleria fowleri gruberi Acanthamoeba culbertsoni castellanii
Titer of IIF antibody *
Antibody isotype
Presence in cord blood
1: 5-1 :20
Mainly IgM and IgG
+
1:20-1 :80
Mainly IgM and IgG
+
* IIF = indirect immunofluorescence.
tection against a lethal intranasal challenge with this ameba. Significant protection was induced with one immunizing dose of the antigen [18]. After three immunizing doses, almost all animals were protected [18]. These findings contrast with those for a similar antigen preparation from N. fowleri, which induced very little protection against N. fowleri in mice even when repeated immunization was used [19]. The protection induced in the case of A. culbertsoni A-I was highly specific for this Acanthamoeba. Immunization with similar antigenic preparations from either A. culbertsoni A-5, A. rhysodes (HN-3), or Acanthamoeba polyphaga (MRK-IA) failed to induce any protection against a lethal challenge with A. culbertsoni A-I [18]. When the culture fluid from A. culbertsoni A-I was used as a source of antigen for immunizing mice, the animals were significantly protected from infection by the ameba [18]. This finding is similar to results obtained with N. fowleri [19].
The Role of Phagocytic Cells Neutrophils appear to play an important role in immunity to pathogenic free-living amebas [20-27]. Studies with N. fowleri showed that an important characteristic of immune mice is the rapid deployment of neutrophils at sites of amebic challenge [20, 21, 25]. Abrogation of neutrophil function and the depletion of neutrophils in immune mice significantly reduces the immunity to amebas [25]. Studies on both Naegleria and Acanthamoeba have shown that lymphocytes and macrophages from human peripheral blood do not act as effector cells against these amebas [22, 23], even in the presence of antibodies. Human neutrophils also fail to kill both types of ameba [22, 23]. But the same leukocytes, when treated with a conditioned medium that contained lymphokines from phytohaemagglutinin-activatedmononuclear leukocytes, were capable of killing both N. fowleri and A. culbertsoni A-I. Killing by lymphokinealtered neutrophils was found to require the presence of either antibody or complement [22, 23]. Usually a number of neutrophils attacked the amebas, which is characteristic of the neutrophil extracellular killing mechanism. Studies with N. fowleri have demonstrated that the neutrophil-killing mechanism involves both the oxidative respiratory system and the
Immunity to Acanthamoeba
RID 1991;13 (Suppl 5)
Figure 1. Diagramatic representation of possible interactions of neutrophils with free-living amebas. Neutrophils recognize and attach to amebas via Fc receptors (FcR) and/or complement receptors (CR). Other lectin-like receptors may also be involved. These interactions alone are not adequate to promote killing of amebas or to mobilize the relevant antimicrobial biochemical responses of the neutrophil. In contrast, pretreatment of neutrophils with polypeptide cytokines such as TNFa leads to significant stimulation of the respiratory burst (NADPH oxidase activation), which results in the production of oxygen-derived reactive species (ODRS) and the release of constituents of specific (spec.) and azurophil (azur.) granules. This positive response requires involvement of either FcR or CR. Granule components alone appear not to be amebicidal. In the case of N. fowleri, the ameba is highly sensitive to HOCI produced by the myeloperoxidase-H202halide system. In contrast, Acanthamoeba is resistant to HOCI but is sensitive to H202. Other components shown: the hexose monophosphate shunt (HMP) is closely linked to the activity of the NADPH oxidase; catalase (Cat), superoxide dismutase (SOD), and reduced glutathione (GSH) are involved in scavenging ODRS and in protecting the cell from auto-oxidation. BPI = bacterial permeability increasing protein; MPO = myeloperoxidase.
B 12 binding
protein
Lactoferrin Lysozyme BPI
{3 -glucuronidase Collagenase Elastase Lysozyme Myeloperoxidase BPI
: :l~: : ;:{~ l: : :~ : : : : {:/~f: /
S405
S406
Ferrante
35
Antiamebic Properties of Antibodies
30
25
:> g w
20
U Z
w
U
(J)
W
Z
15
~
::::>
...J
~
W
I
RID 1991;13 (Suppl 5)
10
U
5
0 0
4
6
8
10
12
TIME (min)
Figure 2. Kinetics of the lucigenin-dependent neutrophil chemiluminescence response and the effects of TNFa. Neutrophils (PMN) were pretreated for 30 minutes with 100 U of TN Fa and then tested for response in the presence or absence of A. culbertsoni A-I treated with human serum and fixed with formalin. Neutrophils not treated with TNFa (controls) were incubated with similar additions. The lucigenin-dependent chemiluminescence was totally inhibitable by superoxide dismutase, showing that the assay measures superoxide produced by neutrophils. Measurements were made in a luminometer.
enzyme found in azurophil granules, myeloperoxidase [24] (figure 1). While the amebas were sensitive to H20 2 , most sensitivity was to hypochlorite (HOCI), a product of the myeloperoxidaseH20 2-halide system [24]. Myeloperoxidase-deficient neutrophils lacked amebicidal activity. In contrast, A. culbertsoni A-I was much more resistant than N. fowleri to HOCI but appeared to be more sensitive to H 20 2 (author's unpublished observations) (figure 1). Direct addition of material extracted from granules, which was highly effective for killing Salmonella minnesota R595 and in damaging human and bovine cartilage in vitro [24, 28], had no effect on these amebas [24]. A number of T cell and macrophage cytokines have been shown to augment both the respiratory burst and release of lysosomal enzyme by neutrophils [27, 29-37], and this is probably the primary mechanism by which lymphokines enhance the killing of amebas by neutrophils. Studies with A. culbertsoni A-I showed that the macrophage cytokine, tumor necrosis factor ex (TNF ex), augmented the respiratory response to this ameba (figure 2).
Amebic invasion and destruction of tissues is dependent on a number of properties of free-living amebas. These properties include their capacity to adhere to mucosal surfaces; to migrate through tissues; to phagocytose; and to release oxygen radicals and a variety of cytopathogenic agents/enzymes, such as elastase, which degrade a variety of connective tissues [38]. The migratory characteristics of Acanthamoeba species are shown in figures 3 and 4. Migration of the pathogenic A. culbertsoni A-I and of the nonpathogenic A. culbertsoni A-5 were similar at 37°C (figure 3) and 28°C (figure 4). A pathogenic ameba, A. polyphaga (MRK-IA), was similarly active at both temperatures. However, A. rhysodes (HN-3) migrated poorly at these temperatures. It is interesting that N. fowleri migrated poorly at the lower temperature [39], a finding that indicates that some species of Acanthamoeba have greater tissue-penetrating properties relative to those of species of N. fowleri at lower temperatures that may be encountered in skin lesions, the nasal passages, and eyes. Antibodies may function in various ways to limit the invasiveness of Acanthamoeba. Addition of antibodies to freeliving amebas prevents their adhesion and spreading. The cells "round" and show reduced migratory properties [40]. Phagocytosis may be inhibited by antibodies, and it is most likely that antibodies neutralize the amebic cytopathogenic agents [13]. Finally, antibodies, particularly of the IgG isotypes, promote neutrophil-mediated killing of amebas. These mechanisms may operate collaboratively to prevent invasion by Acanthamoeba and to promote its elimination from body tissues.
Concluding Remarks Although understanding of the mechanisms of immunity to N. fowleri has progressed in the last 10 years, a concerted effort needs to be made to understand the defense mechanisms that operate against Acanthamoeba species. This presents a major task because of the different types of diseases caused by Acanthamoeba that involve a number of different species. Recent findings suggest that immunity to Acanthamoeba species is deployed as depicted in figure 5. Amebas that invade from skin lesion are opsonized by antibody and complement. Antibody may control invasiveness of the organism by neutralizing amebic cytopathogenic substances and by inhibiting phagocytosis of amebas and their adherence to tissue components. Phagocytic cells such as neutrophils are attracted to the infection site by chemotactic agents likely to be generated when complement is activated. Evidence suggests that sensitized T cells and macrophages playa functional role by releasing lymphokines that activate neutrophils, an essential step in the process by which leukocytes confine and kill the amebas.
A. polyphaga (MRK-1A) A. culbertsoni (A-1)
30
A. culbertsoni (A5)
E E
20
Z
Figure 3. Migration of Acanthamoeba
o
species under agaroseat 37°C. Resultsare presented as the mean ± SE of six experiments.
~
« II:
C)
~ 10
A. rhysodes (HN-3)
o-+--~=-J.--""-===~:::::::::':"----r-----r----
o
2
3
4
5
TIME (hrs)
3'5
3·0
2·5
E E
z
Migration of Acanthamoeba species under agaroseat 28°C. Resultsare presentedas the mean of four experiments. Figure 4.
a
2·0
i=
-c a:
C)
~
1·5
1·0
0·5
(HN-3)
o-+o-~~-...- -...-...===f=:::::::::==~-----r---
o
3
TIME (hrs)
4
RID 1991;13 (Suppl 5)
Ferrante
5408
wound
mast cell
••
vasoactive
complement
.: ,
- -
~
_ _ _ _ _ _...:.b::..:.lo=-:od vessel
==================Figure S. Possible host responses to Acanthamoeba species. Penetration of ameba at mucosal surfaces, skin, or other tissues may be dealt with immediately by a collaborative action of different components of the immune system. Evidence at present indicates a requirement for activation of phagocytic cells by lymphokines, antibody, and complement - the activated cells acting as recognition factors - and a requirement for antibody to prevent normal amebic phagocytic activity.
Persistent acanthamoeba infection may be a consequence of immune depression, a likely occurrence in chronically iII and debilitated individuals who are receiving immunosuppressive therapy. Reduced resistance probably is the result of decreased antibody production, lowered complement levels, a depressed rate of accumulation of leukocytes, inhibited leukocyte function, and impaired Iymphokine production. Infection in healthy individuals also may occur if a site where levels of antibodies and complement are low, such as the eye, is involved. In view of the unavailability of effective chemotherapy for acanthamoeba infections, immunochemotherapy may be an important consideration for treatment of these infections.
6.
7. 8. 9.
10. 11.
12.
References 1. Culbertson CG, Smith JW, Minner JR. Acanthamoeba: observations on animal pathogenicity [letter]. Science 1958;127:1506 2. Fowler M, Carter RF. Acute pyogenic meningitis probably due to Acanthamoeba sp.: a preliminary report. BMJ 1965;2:740-2 3. Carter RF. Description of Naegleria sp. isolated from two cases of primary amoebic meningoencephalitis and of the experimental pathological changes induced by it. J Pathol 1970;100:217-44 4. Martinez AJ, ed. Free-living amebas: natural history, prevention, diagnosis, pathology and treatment of disease. Boca Raton, FL: CRC Press 1985 5. Thong YH, Ferrante A. Experimental pharmacology. In: Rondanelli EG, ed. Amphizoic amoebae: human pathology. Infectious diseases
13.
14. 15.
16.
17.
color atlas monographs. Padua, Italy: Piccin Nuova Libraria, 1987: 251-72. Ferrante A, Rowan-Kelly B. Activationof the alternative pathwayof complement by Acanthamoeba culbertsoni. Clin Exp Immunol 1983; 54:477-85 Korn ED, Olivecrona T. Composition of an amoeba plasma membrane. Biochem Biophys Res Commun 1971;45:90-7 Stevens AR. Biochemical studies of pathogenic free-living amoebae. Giornale di Malatte Infettive e Parassitarie 1977;29:690-6 Ferrante A, Allison AC. Alternative pathway activation of complement by African trypanosomes lacking a glycoprotein coat. Parasite Immunol 1983;5:491-8 Bowers B, Korn ED. The fine structure of Acanthamoeba castellanii. I. The trophozoite. J Cell Bioi 1968;39:95-111 Ferrante A, Jenkins CR. Evidence implicating the mononuclear phagocytic system of the rat in immunity to infection with Trypanosoma lewisi. Aust J Exp Bioi Med Sci 1978;56:201-9 Ferrante A. The role of the macrophage in immunity to Trypanosoma musculi. Parasite Immunol 1986;8:117-27 Cursons RTM, Brown TJ, Keys EA, Moriarty KM, Till D. Immunity to pathogenic free-living amoebae: role of humoral antibody. Infect Immun 1980;29:401-7 Ferrante A, Allison AC. Natural agglutinins to African trypanosomes. Parasite Immunol 1983;5:539-46 Ferrante A, The role of natural agglutinins and trypanolytic activity in host specificity to Trypanosoma musculi. Parasite ImmunoI1984;6: 509-17 Ferrante A. Trypanolytic activity, agglutinins, and opsonins in sera from animals refractory to Trypanosoma lewisi. Infect Immun 1985; 49:378-82 Culbertson CG. The pathogenicity of soil amebas. Annu Rev Microbiol 1971;25:231-54
RID 1991;13 (Suppl 5)
Immunity to Acanthamoeba
18. Rowan-Kelly B, Ferrante A. Immunization with killed Acanthamoeba culbertsoni antigen and amoeba culture supernatant antigen in experimental acanthamoeba meningoencephalitis.Trans R Soc Trop Med Hyg 1984;78:179-82 19. Thong YH, Ferrante A, Rowan-Kelly B, O'Keefe D. Immunization with live amoebae, amoebic lysate and culture supernatant in experimental Naegleria meningoencephalitis. Trans R Soc Trop Med Hyg 1980; 74:570-6 20. Ferrante A, Thong YH, Unique phagocyticprocess in neutrophil-mediated killing of Naegleria fowleri. Immunol Lett 1980;2:37-41 21. Thong YH, Carter RF, Ferrante A. Rowan-Kelly B. Site of expression of immunity of Naegleria fowleri in immunized mice. Parasite Immunol 1983;5:67-76 22. Ferrante A, Mocatta TJ. Human neutrophils require activation by mononuclear leucocyte conditioned medium to kill the pathogenic free-living amoeba, Naegleria fowleri. Clin Exp Immunol 1984; 56:559-66 23. Ferrante A, Abell TJ. Conditioned medium from stimulated mononuclear leukocytes augments human neutrophil-mediated killing of a virulent Acanthamoeba sp. Infect Immun 1986;51:607-17 24. Ferrante A, Hill N, Abell TJ, Pruul H. Role of myeloperoxidase in the killing of Naegleriafowleri by lymphokine-altered human neutrophils. Infect Immun 1987;55:1047-50 25. Ferrante A, Carter RF, Lopez AF, Rowan-KellyB, Hill NL, Vadas MA. Depression of immunity to Naegleria fowleri in mice by selectivedepletion of neutrophils with monoclonal antibody. Infect Immun 1988;56:2286-91 26. Ferrante A, Hill NL, Goh DHB. Kumaratilake L. Altered neutrophils in mice immune to experimental Naegleria amoebic meningoencephalitis. Immunol Lett 1989;22:301-6 27. Ferrante A. Augmentation of the neutrophil response to Naegleria fowleri by tumour necrosis factor alpha. Infect Immun 1989;57:3110-5 28. Kowanko IC, Bates EJ. Ferrante A. Mechanisms of human neutrophilmediated cartilage damage in vitro: the role of lysosomal enzymes, hydrogen peroxide and hypochlorous acid. Immunol Cell Bioi 1989; 67:321-9 29. Weisbart RH, Golde OW, Clark SC, Wong GG, Gasson rc. Human
30.
31.
32.
33.
34.
35.
36.
37.
38. 39. 40.
S409
granulocyte-macrophage colony-stimulating factor is a neutrophil activator. Nature 1985;314:361-2 Klebanoff SJ, Vadas MA, Harlan JM, Sparks LH, Gamble JR, Agosti JM, Waltersdorph AM. Stimulation ofneutrophils by tumour necrosis factor. J Immunol 1986;136:4220-5 Shalaby MR, Palladino MA Jr, Hirabayashi SE, Eessalu TE, Lewis GO, Shepard HM, Aggarwal BB. Receptor binding and activation of polymorphonuclear neutrophils by tumor necrosis factor-alpha. J Leukocyte Bioi 1987;41:196-204 Kowanko IC, Ferrante A. Stimulation of neutrophil respiratory burst and lysosomal enzyme release by human interferon-gamma. Immunology 1987;62:149-51 Kowanko IC, Ferrante A. Interleukin-2 inhibits migration and stimulates respiratory burst and degranulation of human neutrophils in vitro. Immunol Lett 1987;15:285-9 Ferrante A, Nandoskar M, Bates EJ, Goh ORB, Beard U. Thmour necrosis factor beta (lymphotoxin) inhibits locomotion and stimulates the respiratory burst and degranulation of neutrophils. Immunology 1988;63:507-12 Ferrante A, Nandoskar M, Walz A, Goh DHB, Kowanko rc Effects of tumour necrosis factor alpha and interleukin-l alpha and beta on human neutrophil migration, respiratory burst and degranulation. Int Arch Allergy Appl Immunol 1988;86:82-91 Nathan CF. Neutrophil activationon biological surfuces. Massive secretion of hydrogen peroxide in response to products of macrophages and lymphocytes. J Clin Invest 1987;80:1550-60 Ferrante A. Tumor necrosis factor alpha potentiates neutrophil antimicrobial activity: increased fungicidal activity against Torulopsis glabrata and Candida albicans and associated increases in oxygen radical production and lysosomal enzyme release. Infect Immunol 1989;57:2115-22 Ferrante A, Bates EJ. Elastase in the pathogenic free-living amoebae Naegleria and Acanthamoeba spp. Infect Immun 1988;56:3320-1 Thong YH, Ferrante A. Migration patterns of pathogenic and nonpathogenic Naegleria spp. Infect Immun 1986;51:177-80 Ferrante A, Thong YH. Antibody induced capping and endocytosis of surface antigens in Naegleriafowleri. Int J ParasitoI1979;9:599-601
