CLINICAL MICROBIOLOGY REVIEWS, Jan. 2009, p. 65–75 Vol. 22, No. 10893-8512/09/$08.00�0 doi:10.1128/CMR.00029-08Copyright © 2009, American Society for Microbiology. All Rights Reserved.

Host-Microbe Interactions and Defense Mechanisms in theDevelopment of Amoebic Liver Abscesses

Julien Santi-Rocca,1,2 Marie-Christine Rigothier,3 and Nancy Guillen1,2*Institut Pasteur, Unite Biologie Cellulaire du Parasitisme, Paris, France1; INSERM U 786, Paris, France2; and Universite Paris-Sud,

Faculte de Pharmacie, Laboratoire de Biologie et Controle des Organismes Parasites, Chatenay-Malabry, France3

INTRODUCTION .........................................................................................................................................................65THE LIVER: VASCULAR AND SINUSOIDAL STRUCTURE ..............................................................................65

Liver Sinusoidal Endothelial Cells and Stellate Cells ........................................................................................66Immune System Cells in the Liver.........................................................................................................................66

PATHOPHYSIOLOGICAL FEATURES OF CROSS TALK BETWEEN E. HISTOLYTICA AND THELIVER.....................................................................................................................................................................67

Histopathology of Amoebic Liver Abscess Development in Animal Models ....................................................67Entamoeba histolytica in Blood and Endovascular Compartments.....................................................................68

Amoebic resistance to complement ....................................................................................................................68Blood pressure, parasite migration, and adhesion to LSECs ........................................................................69

The Host’s Innate and Inflammatory Responses .................................................................................................69Amoebic resistance to reactive oxygen and nitrogen species..........................................................................69Immune deregulation by PGE2 ...........................................................................................................................70

Host Cell Death by Apoptosis .................................................................................................................................70Adaptive Immune Response ....................................................................................................................................71

KEY AMOEBIC FACTORS IN THE DEVELOPMENT OF LIVER ABSCESSES.............................................71Virulence-Linked Factors.........................................................................................................................................71Amoebic Lytic Factors..............................................................................................................................................72Gal/GalNAc-Inhibitable Lectin ...............................................................................................................................72KERP1........................................................................................................................................................................73

CONCLUSIONS ...........................................................................................................................................................73ACKNOWLEDGMENTS .............................................................................................................................................73REFERENCES ..............................................................................................................................................................73

INTRODUCTION

Entamoeba histolytica (the etiological agent of human amoe-biasis) leads to several distinct clinical manifestations: diar-rhea, dysentery, and hepatic liver abscess. Invasion starts when,for hitherto unknown reasons, trophozoites residing in thecolon lyse, deplete the mucus, interact with exposed entero-cytes, dismantle cell junctions, and lyse host cells. Once theepithelial layer is disrupted, amoebae cross the basal laminaand the extracellular matrix (ECM); this is a process involvingparasite adherence, motility, and cytotoxicity toward host cells.Some trophozoites that reside in the colon penetrate the portalsystem, resist the stress provoked by this new environment, andfollow the bloodstream to the hepatic portal veinule and sinu-soids. The latter are the main structures where amoebae crossthe endothelium to reach the parenchyma, with a concomitantinitiation of inflammatory foci and abscesses. The typicalamoebic liver abscess (ALA) is due to necrotic lysis of the livertissue, which varies in size from a few centimeters to a largelesion. It is often single, usually in the posterior, superior areaof the right lobe. ALA is characterized, in its acute manifes-tation, by fever (temperature greater than 39°C), pain in the

right hypochondrium, and liver tenderness (43). Although theyalso present pain upon palpation of the liver, patients withchronic ALA exhibit different manifestations: mild or no fever,hepatomegaly, and rales or rhonchi. Upon auscultation, theclinical features of ALA allow one to suspect the amoebicetiology, which, however, needs further exploration, such asultrasound echography, to be confirmed.

Most patients with ALA are young adult males, although themale/female ratio is equal to 1 in children and infants. Patientswith ALA infrequently present concurrent colitis, althoughthey sometimes have a history of dysentery within the last year(1). Once diagnosed, ALA is managed by the administration ofmetronidazole or tinidazole, followed by treatment with a lu-minal amoebicide, i.e., paromomycine or diloxanide furoate(35). Responses to amoebic infection depend on the organ’scell types and architecture; here, we shall describe the interplaybetween the parasite and the hepatic tissue, which operatesduring liver abscess formation by E. histolytica.

THE LIVER: VASCULAR ANDSINUSOIDAL STRUCTURE

The liver parenchyma includes endothelial cells, hepato-cytes, and stellate cells, which together constitute a very com-plex, three-dimensional tissue that bathes in blood. Importantinflammatory cells (such as Kupffer cells [KCs], dendritic cells[DCs], lymphocytes, and leukocytes) are present and support

* Corresponding author. Mailing address: Unite Biologie Cellulairedu Parasitisme, Institut Pasteur, 28 Rue du Dr. Roux, 75724 ParisCedex 15, France. Phone: 331-45688675. Fax: 331-45688674. E-mail:nguillen@pasteur.fr.

65

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

immune and homeostasis features. Indeed, besides the liver’smetabolic and amphicrine secretion activities, the organ clearsmolecules and particles carried by the afferent blood; this no-tably includes senescent red blood cells, metabolites, and orallyingested antigens that are brought by the hepatic artery fromthe aorta and by the portal vein from the mesenteric tract.Incoming blood flows centripetally through the hepatic sinu-soids and subsequently reaches the efferent centrolobular vein.Figure 1 illustrates the afferent and efferent microvascularconnections to the sinusoids within a single hepatic lobule, andFig. 2 shows the structure of a sinusoid.

Liver Sinusoidal Endothelial Cells and Stellate Cells

The interface between the sinusoid lumen and the perisinu-soidal space (also known as the space of Disse [DS]), intowhich the microvilli of the parenchymal cell (lined up in rib-bons) project, is composed mainly of liver sinusoid endothelialcells (LSECs). These are fenestrated endothelial cells that lacktight junctions and a basal membrane, which enables particlesless than 12 nm in diameter to cross into the DS (44). LSECshave an enhanced endocytic capacity (with glycoprotein li-gands, ECM components, and immunoglobulin G [IgG] Fcreceptor immune complexes) and can process antigens forpresentation on class I or II major histocompatibility complex(MHC) molecules (27). In addition, they take part in the se-cretion of cytokines, eicosanoids (derived from arachidonic

acid), local inflammation mediators, nitric monoxide (NO),and ECM components. Hepatic stellate cells (SCs) (formerlyknown as hepatic lipocytes) are found in the DS and fulfillseveral functions: they can (i) regulate sinusoid blood flow bycontraction (performed via smooth muscle �-actin and whichcan be induced by NO), (ii) store or release retinoids (i.e.,vitamin A), and (iii) control the secretion of ECM. Liver injurycauses SCs to transform into myofibroblasts that participate inthe development of fibrosis. These cells can amplify the inflam-matory response by producing chemokines, which, in turn,induce the infiltration of mono- and polymorphonuclear leu-kocytes (PMNs).

Immune System Cells in the Liver

KCs are monocytes that reside in the liver sinusoids andstraddle the endothelium. KCs exert three main functions (an-tigen presentation, phagocytosis, and cytotoxicity) and thusplay a role in clearing the blood of senescent red blood cellsand opsonized or complement-coated particles (37). Followingphagocytosis of target particles, KCs can secrete potent inflam-matory mediators, such as reactive oxygen species (ROS), ei-cosanoids, NO, tumor necrosis factor (TNF), and other cyto-kines. TNF activates the presentation of ICAM-1, VCAM-1,and P- and E-selectin by LSECs (56); these proteins are ligandsfor (among others) natural killer cells (NKCs), monocytes, andPMNs. The latter cell types are involved in the inflammatory

FIG. 1. Vascular structure of a hepatic lobule. The liver is composed of elementary functional units called lobules. A single lobule is depicted.Each lobule is made of parenchymal cells, mainly hepatocytes (HCs), which account for up to 80% of the total liver volume, and nonparenchymalcells (6.5% of the total volume) located in the sinusoidal compartment. Inside the lobule, hepatocytes are arranged in irregular radiating columnsbathed by blood-carrying spaces called sinusoids. Lobules are prisms with a hexagonal cross section, the vertices of which are marked by the portaltriad: the hepatic arteriole (HA), portal veinule (PV), and bile duct (BD). Around 75% of the blood entering the sinusoids comes from the portalveinule. The remainder enters the sinusoids principally through branches of the hepatic arterioles via the arteriosinus twigs. Sometimes, directconnections (arterioportal anastomoses [APA]) are observed within the terminal portal veinules. Blood then leaves the sinusoids through theefferent centrolobular vein (CV). Please note that for reasons of clarity, this schematic view does not respect the exact scale and numbers of cellswithin a lobule.

66 SANTI-ROCCA ET AL. CLIN. MICROBIOL. REV.

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

response, and their early recruitment to the nascent inflamma-tory foci allows the continuation and control of innate immunedefenses.

Hepatic lymphocytes consist of different subsets, with a ma-jority of T cells (35%), followed by natural killer T cells (NKTCs)(30%), NKCs or pit cells (20%), and B cells (BCs) (10%) (25).Of these, NKTCs, NKCs, and BCs can be activated directly bypathogen-associated molecular patterns (PAMPs) withoutprior processing and presentation of antigens by antigen-pre-senting cells. Thus, NKTCs, NKCs, and BCs have an un-doubted role as fast-responding, potent cytotoxic effectors ormodulators of cytokine responses in the early stages of liverinfection.

DCs are present in the periportal and sinusoidal spaces (86).Whether DCs are true residents of the sinusoidal compartmentis subject to debate, since the presence of N-acetylgalac-tosamine ligands at the cell surface may be crucial for DCrecruitment to liver sinusoids (55). However, increased trafficand the blood-lymph translocation of DCs during infectionhighlight the cells’ potent role during the initiation and imple-mentation of an adaptive immune response (55). The involve-ment of DCs in the adaptive immune response via efficientantigen uptake, processing, and presentation is complementedby their role in innate immunity as lymphocyte-activated or-activating cells. Recent publications have described the crosstalk between DCs and NKCs, which triggers inflammatory cy-tokine production by both cell types and the enhancement ofthe cytotoxic potential of NKCs (85).

Lastly, circulating or lymph-borne leukocytes (such as lym-phocytes and PMNs) are cell types that flow through the liverand may be recruited during the various stages of the liver’sreaction to infection. Neutrophil PMNs, lymphocytes, andmonocytes can all trigger a so-called oxidative burst. Leukocyte

trafficking from the blood vessels to the site of infection occursin several steps. Following activation by chemokines and cyto-kines, leukocytes arrive at and adhere to the endothelial cellsurface and then transmigrate into the underlying tissue via (i)small gaps at the junctions between endothelial cells and (ii)transcellular migration across the endothelium (28).

PATHOPHYSIOLOGICAL FEATURES OF CROSS TALKBETWEEN E. HISTOLYTICA AND THE LIVER

Although the host deploys a massive inflammatory responseagainst E. histolytica, the parasite manages to survive withinthis environment. Virulence can be defined as a two-step strat-egy: invade and survive. The clonal nature of the trophozoitepopulations used for experimental infection cannot explain thefact that some resist and some succumb in the same microen-vironment during host infection. This underlines a potentialmechanism of local and individual adaptation, which wouldhypothetically lead to subpopulation specialization. However,overall virulence traits need to be well defined and can besummed up as the trophozoite’s ability to resist the comple-ment system and oxidative attacks and, concomitantly, invadethe host tissues by dismantling the latter’s structure and trig-gering cell death, processes sustained by parasite cytotoxicity,motility, and phagocytosis.

Histopathology of Amoebic Liver Abscess Developmentin Animal Models

Trophozoite virulence can be assessed in vivo by using ani-mal models for ALA formation; to this aim, three species ofrodent animal models are used. The first model features theintrahepatic (17) or intraportal (83) injection of trophozoites

FIG. 2. Entamoeba histolytica trophozoite within a sinusoid. The sinusoidal wall is made of LSECs and, with hepatocytes (HCs), delimitates theDS, in which SCs and components of the ECM are found. Leukocytes (LCs) are present in the sinusoidal lumen (SL), as are KCs, whichadditionally straddle the endothelium. The diagram presents the major factors involved in the interaction between E. histolytica and the liverendothelium. Within the hepatic sinusoids, trophozoites undergo complement attack and oxidative stress provoked by the high-oxygen partialpressure. Amoebae resist these threats by using Gal/GalNAc lectin (blue ellipses), lytic factors such as cysteine proteases (white ellipses), reducingenzymes (purple ellipse), and potentially other molecules of as-yet-unknown function (black ellipses). Trophozoites adhere to the endothelium andcause the apoptosis of LSECs. Amoebae can then migrate to the parenchyma (yellow arrow) through the newly created breach or between LSECs.

VOL. 22, 2009 AMOEBIC ABSCESS DEVELOPMENT 67

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

into Mongolian gerbils. Pathophysiological studies of ALA de-velopment in this animal (17) revealed characteristics similarto those of human ALA: the granulomata have a central,necrotic zone (containing inflammatory cells and lysed hepa-tocytes) surrounded by a ring of trophozoites and inflamma-tory cells that delimit the abscess from the rest of the hepatictissue.

In the murine model, amoebae are directly inoculated intothe liver of immunodeficient SCID mice (19) or immunocom-petent wild-type mice (49). Dead hepatocytes are detected at12 h postinfection in inflammatory foci surrounded by tropho-zoites; 24 h after inoculation, both sizes and numbers of in-flammatory foci are seen to have increased. Unlike othermodel rodents, wild-type mice defeat the infection after 2weeks.

The third model used in ALA development studies is theSyrian hamster. After intraportal inoculation of trophozoites,the histological features of infected livers are found to besimilar to those in humans and thus enable the development ofhepatic amoebiasis to be studied. Amoebic liver abscesses inhumans and hamsters have a common, characteristic structure,as described above and presented in Fig. 3 (69, 82). Threehours after intraportal trophozoite inoculation, the hamsterliver harbors a large number of inflammatory foci formed bythe recruitment of neutrophils and, later on, macrophages thatsurround a single trophozoite; at 12 h, a majority of inflam-matory foci are trophozoite free. This massive trophozoitekilling is attributed to the acute inflammatory response medi-ated by host immune cells. After this critical point, any surviv-ing amoebae divide, and the inflammatory foci develop intoabscesses. New inflammatory foci can result from the meta-

static spreading of amoebae throughout the liver tissue, a phe-nomenon that requires trophozoite motility. After 3 days ofinfection, abscesses start to coalesce until the formation ofnecrotic areas resembling those seen in humans (69). E. histo-lytica virulence then relies on the parasite’s ability to resist thehost’s multifaceted response to invasion.

Entamoeba histolytica in Blood andEndovascular Compartments

Prior to reaching the liver, trophozoites reside in the endo-vascular compartment, bathing in the blood, where they en-counter soluble components, circulating PMNs, and endothe-lial cells (Fig. 2). A first line of host defense is constitutivelymaintained by the integrity of the endovascular system. Thetrophozoites navigate in an environment in which the oxygenpartial tension is relatively high (oxygen saturation of 98% �1%) (33), along with other factors that may contribute to theirdeath (such as complement system components).

Amoebic resistance to complement. Complement is an evo-lutionarily conserved system composed of circulating zymogensthat undergo cascade activation by proteolytic cleavage uponantibody- or lectin-mediated binding to the surface of patho-gens. In vitro, E. histolytica trophozoites are able to activate thecomplement system in several ways (66), which leads to thelysis of susceptible parasites by the formation of membraneattack complexes (MACs). Pathogenic strains do still activatethe complement pathway but are resistant to killing by MACs(68). Two mechanisms sustaining complement resistance havebeen identified. On the one hand, galactose/N-acetylgalac-tosamine (Gal/GalNAc) lectin inhibits the assembly of the

FIG. 3. Amoebic liver abscess. Seven days after intraportal injection, the typical histological presentation of amoebic liver abscess in hamstersresembles a centripetral trophozoite layer (indicated by an arrow) stained with antiamoebic antibodies (in red) surrounding a central necrotic zone(indicated by “N”) and facing an external area containing large numbers of inflammatory cells (indicated by an arrowhead and stained withhematoxylin [in blue]). Scale bar, 50 �m.

68 SANTI-ROCCA ET AL. CLIN. MICROBIOL. REV.

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

C5b-9 complex and, thus, lysis by MAC (in a way similar to thatseen with CD59, a molecule with which the Gal/GalNAc lectinhas antigen cross-reactivity and sequence similarity) (12). Onthe other hand, cysteine proteases (CPs) interfere with theimmune response by degrading anaphylatoxin C3a. Moreover,the cleavage of the �-chain of C3b by the extracellular 56-kDaneutral CP leads to the formation of a C3b-like protein, whichprevents MAC formation (67). In protozoan parasites, phos-phoglycans in the glycocalyx may also act as a physical barrierthat protects cells from MAC attack. Interestingly, it has beensuggested that the proteophosphoglycans (PPGs) in the amoe-bic glycocalyx may participate in E. histolytica pathogenicity,since the closely related, nonpathogenic Entamoeba disparlacks a significant glycocalyx surface layer (8); furthermore,antibodies that bind to PPGs neutralize liver abscess formation(54). PPGs are anchored into the outer leaflet of the parasitecell membrane by a glycosylphosphatidylinositol (GPI) moiety.Synthesis of the GPI anchor requires a cascade of enzymes,including mannosyl transferase 1 (PIG-M1, the enzyme thattransfers the first mannose residue onto GPI). The blockage ofPIG-M1 reduces GPI synthesis and thus reduces PPG levels introphozoites (87). The PIG-M1-deficient parasites are highlysensitive to complement and are unable to provoke ALA de-velopment in hamsters, highlighting the importance of theseGPI-anchored molecules for trophozoite pathogenicity and re-sistance against complement attack.

Blood pressure, parasite migration, and adhesion to LSECs.In the portal vein, the blood flow is 1.4 liters/min, the bloodpressure is 12 to 15 mm Hg, and erythrocyte velocity is 8 to 18cm/s. In comparison, sinusoidal blood flow is extremely low(3.4 � 0.16 ml/min), as is red blood cell velocity (0.1 mm/s)(65). The forces exerted on a parasite that adheres to theendothelium are thus much lower in the sinusoids and maypartly explain why the parasite crosses the endothelium withinthese structures. However, the low mean diameter of the sinu-soid lumen (6 �m) implies that the latter has to dilate and/orthat the parasite must change its morphology in order to crossthe endothelium. It is clear that the parasite and host cytoskel-etons play a key role at this invasive stage, as does cross talkbetween E. histolytica and sinusoid-resident cells. Nevertheless,little is known about the transmigration of E. histolyticathrough the hepatic endothelium. In particular, we do notknow whether the parasite uses specific adhesive molecules tointeract with LSECs. The onset of apoptosis in LSECs is ob-served just 1 h after the intraportal inoculation of parasites inthe hamster model of ALA. Twenty-four hours after inocula-tion, some PMN leukocytes recruited to inflammatory focithen also become apoptotic (9). However, the lack of tightjunctions in the sinusoidal endothelium can facilitate crossingby the parasite; the latter may well take advantage of LSECdeath (creating a larger breach) when reaching the hepaticparenchyma.

The Host’s Innate and Inflammatory Responses

During liver invasion by E. histolytica, the host will bothsequentially and simultaneously deploy a number of mecha-nisms to kill the parasite. The first line of tissue defense iscomposed of innate immune system cells that recognizePAMPs and trigger an inflammatory response. It was previ-

ously reported that female and male mice differ in their earlyresponses to E. histolytica liver invasion (49); females rapidlycleared the parasites, recruited higher numbers of NKTCs tothe infection site, and produced higher levels of gamma inter-feron (IFN-�). NKTC-deficient females or IFN-� neutraliza-tion led to a greater parasite survival level. IFN-� is a keyregulator of inflammation initiation, since it can activate mac-rophage production of TNF, which in turn promotes NO syn-thesis by neutrophil PMNs and the macrophages themselves.In in vitro models, the effects of IFN-� can be bypassed by therecognition of PPGs of E. histolytica by Toll-like receptor 2(TLR2) and TLR4, which results in direct TNF, interleukin-12p40 (IL-12p40), and IL-8 production (52). This proinflam-matory cytokine profile underlines the importance of the earlyrecognition of PAMPs (such as PPGs) and appropriate inflam-matory cell recruitment and activation during ALA onset (39).

Amoebic resistance to reactive oxygen and nitrogen species.During liver invasion, E. histolytica is exposed to a higheroxygen partial pressure and has to eliminate toxic metabolites,such as ROS produced by activated phagocytes during therespiratory burst (Fig. 4). The role of monocytes in liver infec-tion by E. histolytica has not yet been studied in detail. How-ever, it is clear that amoebic factors potently activate perito-neal macrophages and circulating monocytes, leading tocytokine and ROS production and then parasite and tissuedestruction. This self-attack during the oxidative burst is sub-ject to debate, since leukopenic hamsters challenged with wild-type amoebae did not develop liver abscesses (62). Two main(and potentially nonexclusive) hypotheses can be formulated:either PMNs produce an impaired respiratory burst in theALA context or they are killed by amoebae, thus releasing thecontent of their cytotoxic granules in a nondirected fashion. Ineither case, the role of ROS in the development of ALAs inanimal models and humans is important and needs furtherinvestigation.

In phagocytes, the activation of NADP(H) oxidase generatessuperoxide (O2

��) anions (SAs), which spontaneously combinewith water to form efficient oxidants (such as the hydroxylradical�OH� or hydrogen peroxide [H2O2]), which are toxic toE. histolytica (31). In addition, macrophages produce NO bythe action of NO synthase on arginine. The combination of NOwith SA produces peroxynitrites (ONOO�) that are very toxicfor trophozoites (47, 75, 77). Evidence for the role of ROS inhuman resistance to ALA formation was provided by correlat-ing recurrent onsets of hepatic infection by E. histolytica in apatient lacking an oxidative-burst mononuclear phagocyte re-sponse (57).

Although intensive work has been undertaken, the chronol-ogy of the early events leading to ROS production in amoebi-asis remains unclear. However, the production of NO and SAis elicited by TNF stimulation in most types of cells, mainlythrough an NF-�B activation pathway resulting in enhancedNADP(H) oxidase expression (30). TNF production during E.histolytica infection is thus suspected to be an important fea-ture of ALA genesis. E. histolytica can resist both reactiveoxygen and reactive nitrogen intermediates by using specificenzymes; SA is rapidly dismutated into H2O2 and O2 by iron-containing superoxide dismutase (15). An NADPH:flavin oxi-doreductase (p34 thioredoxin reductase) converts O2 to H2O2

(14, 18). Hydrogen peroxide could be detoxified by a rubre-

VOL. 22, 2009 AMOEBIC ABSCESS DEVELOPMENT 69

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

doxin/rubrerythrin complex (22), whose activity has not beenevidenced yet, although the two genes exist in E. histolytica.Furthermore, H2O2 and peroxynitrites are detoxified by athioredoxin redox system consisting of three enzymes: perox-iredoxin, thioredoxin, and p34 thioredoxin reductase. Perox-iredoxin is a 29-kDa surface molecule that functions as a per-oxidase and a thiol-specific antioxidant (64). These activitiesare performed by two cysteine residues (cysteine 47 and cys-teine 170) at the catalytic site; the N-terminal sequence (withnumerous cysteines) may constitute a source of free thiolgroups. The peroxidase activity leads to a reduction of hydro-gen peroxide to H2O. Thioredoxin exhibits antioxidant activityand can reduce peroxiredoxin. This activity is coupled to thatof a third enzyme, p34 thioredoxin reductase, which reducesthioredoxin using NADPH as a cofactor.

Amoebae may use different strategies to overcome the pro-duction of peroxynitrites. Virulent trophozoites can interferedirectly with NO production by host cells: the parasite pos-sesses a gene coding for an arginase that, when expressed byand purified from the bacterium Escherichia coli, efficientlyhydrolyzes arginine into urea and L-ornithine and thus com-petes for substrate with the inducible NO synthase enzymesproduced by activated phagocytes (26). Furthermore, L-orni-thine directly inhibits NO synthesis in macrophages. Usingthese complementary effects that still need to be demonstratedduring infection, amoebae may competently impair NO pro-duction in macrophages.

Immune deregulation by PGE2. Prostaglandin E2 (PGE2) isan endogenous, anti-inflammatory mediator that inhibits theproduction of cytokines and class II MHC proteins and thusinterferes with antigen-presenting functions (36). In mamma-lian cells, PGE2 is synthesized by the cyclooxygenase (COX)-catalyzed conversion of arachidonic acid. Two COX isoforms

exist: COX-1 is produced constitutively in many cells, andinducible COX-2 is generally produced following exposure toproinflammatory molecules (79). The regulatory actions ofPGE2 are mediated by two G-protein-coupled receptors,whose activation in T lymphocytes inhibits cell proliferation(by increasing the intracellular cyclic AMP concentration andprotein kinase activity) and reduces class II MHC expression.During the development of ALA in hamsters, E. histolyticainduces greater PGE2 and COX-2 activity in PMNs and acti-vated macrophages (34). Moreover, a COX-2-like protein hasbeen purified from trophozoites, and a protein that interactswith an anti-mammalian COX-2 antibody has been detected inexperimental ALA infection. This COX-2-like protein presentslittle homology with the mammalian enzymes but is able toproduce PGE2 from arachidonic acid; it may thus be respon-sible for the COX-2-like activity detected during colonic amoe-biasis (23, 70). As liver infection progresses, T-lymphocyteactivity and cytokine (IL-1, -2, and-8 and TNF) and chemokine(monocyte chemoattractant protein 1 and macrophage inflam-matory proteins � and �) production continue to fall, suggest-ing a role for PGE2 in immunosuppression during ALA for-mation.

Host Cell Death by Apoptosis

In liver abscesses in the SCID mouse and hamster models,E. histolytica induces apoptosis in LSECs, hepatocytes, andinflammatory cells (9, 61). Apoptosis is divided into extrinsicand intrinsic pathways. The extrinsic pathway is triggered bythe interaction of specific ligands with cell surface Fas andTNF receptor family members. The intrinsic pathway leads tothe breakdown of mitochondrial outer membrane integrity,followed by the release of cytochrome c and other intermem-

FIG. 4. Control of host responses by Entamoeba histolytica. Following stimulation by Gal/GalNAc lectin (blue ellipses) and (probably) otheramoebic surface factors (black ellipses), macrophages produce TNF, which promotes the production of reactive oxygen intermediates (ROI) orreactive nitrogen intermediates (RNI), with the latter including NO. ROI and RNI are detoxified by the parasite’s reducing enzymes (purpleellipses): among them are peroxiredoxin, p34 thioredoxin reductase (p34), and an iron-containing superoxide dismutase (Fe-SOD). Trophozoitescan also prevent NO production by converting arginine to ornithine. Amoebae can also synthesize (and make host cells synthesize) PGE2, whichreduces TNF and class II MHC (MHC II) production. The above-described mechanisms that E. histolytica uses to subvert the inflammatoryresponse appear to be essential for ALA development, along with other amoebic features supported by the Gal/GalNAc lectin, lytic proteins likeamoebapores and CPs (white ellipses), and molecules of unknown function (black ellipses), such as the virulence factor KERP1.

70 SANTI-ROCCA ET AL. CLIN. MICROBIOL. REV.

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

brane space components. Both pathways lead to the activationof caspase-3 and then apoptotic cell death, characterized bymembrane blebbing, DNA fragmentation, and exposure ofphosphatidylserine on the external membrane (73). The exactmechanism of caspase-3 activation during ALA formation hasnot been fully resolved. Previous studies reported that liverabscess formation can occur in the absence of either Fas orTNFR1 (78). E. histolytica causes apoptosis of cells that lackcaspase-8 or have been treated with caspase-9 inhibitors (38),suggesting that classic upstream activation of caspase-8 or -9 iseither an accessory for or not involved in the activation ofcaspase-3. However, upon coincubation with trophozoites, hu-man lymphocytes (Jurkat T cells) exhibited a higher intracel-lular Ca2� concentration and calpain proteolytic activity, whichled (via a mitochondrion-independent mechanism) tocaspase-3 cleavage and activation (45). The observed in vitrorelevance of amoebiasis in this mechanism needs to be con-firmed in vivo during ALA development.

Recent microarray data on SCID mice hepatically inocu-lated with wild-type (WT) trophozoites evidenced significantlyhigher levels of expression of genes linked to the two above-mentioned apoptotic pathways (61). After 24 h of infection, thetranscription of the genes encoding the Fas receptor,TNFR1/2, and ASK1 was upregulated. The latter is a proteinthat activates the JNK signaling pathway (upon binding to Junand the death domain-associated protein DAXX) and ulti-mately leads to the blockage of Bcl-2 by phosphorylation. Theinvolvement of the mitochondrial apoptotic pathway was alsoevidenced by the transcription upregulation of the proapop-totic factors Bim, Noxa, Bid, Puma, and tBid.

Transcriptional analysis using infected whole liver tissue alsohighlighted the upregulation of genes that regulate the cellcycle and its progression (cyclin-dependent kinase inhibitorp21 and cyclin D2). Likewise, the expression of genes linked tohepatic regeneration (such as various transcriptional factors),which are usually activated by IL-6 and hepatic growth factor(HGF) (61), was upregulated. Although liver infection with E.histolytica triggers host cell death, the mechanisms of regener-ation probably engaged by the host and evidenced at themRNA level may help to balance tissue destruction and slowdown the evolution of the disease, thus preserving the organfunctions.

Adaptive Immune Response

Trophozoites can infect the human liver for several monthsor years before abscesses are diagnosed. Expectedly, a hostimmune response is triggered, leading to the production ofcirculating immunoglobulins. The major immunodominant an-tigens located at the parasite surface are the Gal/GalNAc-inhibitable lectin and the serine-rich protein SREHP. Impres-sively, 80% of patients with ALAs have circulating IgGs thatrecognize these antigens but do not cure the infection (63); thisobservation suggests a possible involvement of a defense mech-anism against the antibody-dependent control of amoebic in-vasiveness, such as the concentration of surface-bound anti-bodies to the posterior edge of the parasite by surface receptorcapping, followed by their release via the extrusion of uroids(16).

Another parameter to consider for ALA development is the

susceptibility of the host. Few studies to identify host geneticpredispositions for amoebic infection have been made. It isnoteworthy that the HLA-DR3 haplotype as well as the SCO1complotype have been associated with a greater susceptibilityfor ALA formation in a Mexican population (6, 7). A higherprevalence of ALA was also linked to HLA-DR5 and HLA-DR6 haplotypes but exclusively in infants. The influence of thegenotype on the infection outcome is undoubted; however, aslong as the mechanisms sustaining these phenotype traits arenot unraveled, it is impossible to state if the different alleles ofclass II MHC genes are directly responsible for the lowerability to arrest amoebic infection or if they evidence an allo-type without being involved in the susceptibility to ALA.

KEY AMOEBIC FACTORS IN THE DEVELOPMENTOF LIVER ABSCESSES

Pioneering studies have demonstrated the impact of E. his-tolytica genotypes on the outcome of infection (2). Further-more, the purification of trophozoites from different organs ofthe same patients revealed that their tropism was linked todifferent genotypes (3). These observations suggest either ad-aptation by E. histolytica to the environment encountered uponinvasion or the selection of invasion-prone parasites during thedevelopment of pathology. The difference at the genetic levelbetween trophozoites able to cause intestinal or hepatic dis-ease would partly explain the rare occurrence of intestinalsymptoms in patients that present ALA (1). Trophozoites frompatients can be purified and cultured in vitro, during whichtime their ability to form ALA in animals decreases (60, 81).However, recurrent passage into animals maintains parasitevirulence (59) and involves mechanisms that are still underinvestigation.

Virulence-Linked Factors

The publication of the genome of E. histolytica (48) naturallyfacilitated transcriptome studies of this parasite. Microarrayshave been successfully used to compare highly virulent para-sites (recently recovered from liver abscesses) with avirulenttrophozoites (unable to form liver abscesses) from the samestrain (71). This transcriptome comparison highlighted someoverexpressed genes that were previously directly linked tovirulence and to stress response in invasion-prone amoebae.Among them, the stress-inducible gene ssp1, the asparagine-rich protein-encoding gene ariel1, and the 20-kDa factor-en-coding gene (all of unknown function) were overexpressed.Interestingly, the expression of the peroxiredoxin andrubrerythrin genes was also upregulated. Peroxiredoxin is in-volved in protection against ROS in general (18) and H2O2 inparticular, and the expression of the gene increases virulencein the nonpathogenic Rahman strain of E. histolytica (61).Rubrerythrin is involved in H2O2 detoxification and exhibits anNAD(P)H oxidase activity in Desulfovibrio vulgaris (22); how-ever, its function remains to be proven in E. histolytica. Theoverexpression of the two latter genes in virulent parasites is ofinterest for ALA development, since inflammatory cell recruit-ment and the subsequent setup of inflammation with the pro-duction of toxic effectors, like ROS, are early features of liverinfection by E. histolytica.

VOL. 22, 2009 AMOEBIC ABSCESS DEVELOPMENT 71

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

Amoebic Lytic Factors

A hallmark of tissue invasion by Entamoeba histolytica re-sides in its ability to directly lyse human cells on contact and todestroy ECM during invasion (51, 80). The molecules poten-tially involved in cytotoxicity and cytolysis were identified asbeing CPs (50) and amoebapores (46). Sequencing of the E.histolytica genome has revealed that CPs compose a family ofat least 44 genes (20). Most of the latter are not expressed incultured parasites, in which 90% of the total protease activityis due to CP-A1, CP-A2, and CP-A5 proteins (13). To date,CPs have been identified in intracytoplasmic vesicles (29),bound to the cell surface (40), and in secreted forms (90).Direct evidence for the involvement of CPs in the formation ofALA has been obtained by inhibiting their function with E-64(90) but also by interfering with cp-a5 gene expression usingantisense technology (5) and epigenetic gene silencing (90).Furthermore, ICP proteins, in addition to their activity of CPinhibitors, were shown to downregulate secretion and thus therelease of CPs in the extracellular medium; this double effecton the inhibition of CP activity and the impact on the secretionof other factors is thought to repress trophozoite virulence(72).

Amoebapores are intravesicular 77-amino-acid proteins (4)bearing the saposin-like protein (SAPLIP) domain encoun-tered in saposin B, a so-called “physiological detergent” (58).Other proteins exhibiting SAPLIP domains (such as NK-lysinand granulysin) have pore-forming activity. Of the 16 SAPLIPsidentified in E. histolytica, only three amoebapores are pre-dicted to have pore-forming activity by sequence analysis (48).Amoebapores are essential for bacterial phagocytosis in E.histolytica and E. dispar. Their cytotoxicity toward eukaryotic(46) and prokaryotic (4) cells adds evidence for a role inpathogenesis, as demonstrated by the absence of ALA forma-tion following hamster or SCID mouse infection with tropho-zoites whose amoebapore A gene expression has been down-regulated by antisense technology (11) or epigenetic silencing(89).

In the transcriptome of virulent parasites, we also identifiedsome so-called virulence factors whose expression was down-regulated, such as CP-A6 and amoebapores A and C (71). It isnoteworthy that CP-A6 was overexpressed during experimen-tal intestinal infection in CBA/J mice (32). These observationshighlight the need for extensive and differential analyses of therole of E. histolytica virulence factors to better understand thepathophysiology of both intestinal invasion and ALA develop-ment.

Gal/GalNAc-Inhibitable Lectin

The Gal/GalNAc lectin is a protein complex comprising a170-kDa heavy subunit (HgL) and a 30- to 35-kDa light subunit(LgL). On the amoebic surface, this complex binds to a 150-kDa intermediate subunit (IgL). The Gal/GalNAc lectin’s im-portance in amoebic virulence has been demonstrated by theimpaired development of experimental ALA infection follow-ing the disruption of lectin function in engineered parasites byblocking signaling through either HgL or LgL synthesis (5, 9,42, 69, 84).

It has been suggested that the Gal/GalNAc lectin plays a

role in initiating the immune response against E. histolytica.Pioneering work by Seguin et al. demonstrated the potentactivation of bone marrow-derived macrophages (BMDMs) bythe cysteine-rich region of the 170-kDa Gal/GalNAc lectinheavy subunit (74). This activation leads to TNF mRNA up-regulation and the production of consistent amounts of TNF.Upon priming with IFN-� and stimulation by the Gal/GalNAclectin of E. histolytica, BMDMs produce TNF and NO andexhibit amoebicidal activity (75). Priming alone or blockage ofthe native lectin’s TNF-stimulating region induced neither theupregulation of inducible NO synthase and TNF gene expres-sion nor the production of the corresponding proteins.

The cysteine-rich region of HgL has sequence homology toCD59; blockage by monoclonal antibodies revealed its involve-ment in complement resistance (12, 53). Other regions of theHgL subunit have been blocked by antibodies; trophozoiteadherence and cytotoxicity were inhibited by the 8C12 mono-clonal antibody directed against the carbohydrate recognitiondomain (53), which shares homologies with HGF’s receptor-binding domain (24). Moreover, the carbohydrate recognitiondomain competed in vitro with HGF for binding to the c-Metreceptor and may thus enable trophozoites to adhere to hepa-tocytes in vivo (24). The upregulation of TLR2 gene expressionwas observed in vitro in murine macrophages exposed to Gal/GalNAc lectin, resulting in proinflammatory cytokine produc-tion (41); these effects were confirmed in vivo in mouse ALAs(61).

Gal/GalNAc lectin’s contribution to inflammatory cell re-cruitment and cytokine production has been carefully studiedin the hamster model (10). HGL-2 trophozoites (engineered todisrupt the Gal/GalNAc heavy chain lectin function) (21) mod-ified the pathophysiology of ALA; HGL-2 parasites producedmore inflammatory foci than did the WT strain, but the fociwere smaller and contained twice as many trophozoites. Thishigher survival rate may be a consequence of reduced immunecell activation, a feature correlated with a reduction in the flowof macrophages from the portal veinules and sinusoids to in-flammatory foci. TNF production was not detected after 24 hof infection with HGL-2 parasites, corresponding to the ab-sence of macrophage/KC activation; this observation corre-lated with the poor recruitment of PMNs and monocytes to theinflammatory foci, which were reduced in size. Whereas WTtrophozoites induced LSEC apoptosis after 1 h of infection andhepatocyte and PMN apoptosis within 24 h, HGL-2 parasitesdisplayed diminished, delayed proapoptotic activity. In the in-flammatory foci produced by HGL-2 trophozoites, apoptoticactivity was 3.5 times lower than that in the inflammatory fociformed by WT amoebae. This suggests a role for the Gal/GalNAc lectin in the signaling pathway leading to apoptosisand the involvement of amoeba-triggered cell death in viru-lence. HGL-2 parasites also displayed lower motility, whichprompted abortive spreading into the tissue (9, 21). Thus,macrophage priming and the subsequent contact with the Gal/GalNAc lectin of E. histolytica may be involved in the in vivokilling of trophozoites by BMDMs. However, the first step inthe response against E. histolytica infection has not yet beendetermined. A role for other amoebic molecules and/or a co-operative response involving other host inflammatory cells isplausible. For instance, neutrophil recruitment is observed,along with a dramatic trophozoite death; the amoebicidal ac-

72 SANTI-ROCCA ET AL. CLIN. MICROBIOL. REV.

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

tivity of macrophages may thus depend on their molecularcross talk with neutrophils.

KERP1

As previously cited, invasion-prone parasites overexpress agroup of genes that encode lysine-rich factors that we namedKRiPs and lysine- and glutamic acid-rich proteins (KERPs), allof them of unknown function (71). Of these, special attentionwas paid to KERP1, a KERP thought to be involved in host-parasite interactions (76). In addition to this prominent mem-ber, six other krip genes encoding hypothetical proteins ofas-yet-unknown function were overexpressed in virulent para-sites.

Further studies confirmed that KERP1 was involved in ALAformation in animal models (71): at the mRNA level, kerp1gene expression was downregulated in the first 3 days of infec-tion and then rose to above initial levels when the abscessesreached confluence; this correlated with a progressive, corticalaccumulation of KERP1 protein as the infection progressed.The subtle regulation of kerp1 gene expression during thecourse of infection evidenced the potential involvement ofKERP1 in ALA development. This potential role in virulencewas reinforced by the progressive decrease in KERP1 abun-dance in cultured trophozoites (i.e., those losing the ability toform ALAs in hamsters). Lastly, the hitherto-unknown role ofKERP1 in ALA formation was assessed in vivo: the downregu-lation of kerp1 gene expression in virulent trophozoites (usingantisense technology) consistently reduced abscess numberand size in hamsters, raising the still-open question ofKERP1’s role in early hepatic amoebiasis: is KERP1 involvedin (i) parasite resistance to blood components and/or the in-flammatory response, (ii) trophozoite interaction with hostcells, for instance, adherence, which may lead to host cell deathand phagocytosis, and (iii) triggering inflammation duringALA development?

CONCLUSIONS

The development of ALAs is a fatal feature of infection byEntamoeba histolytica and is the most common extraintestinalform of invasive amoebiasis. Indeed, it is estimated that around100,000 people succumb to ALA each year (88). The host andparasite factors leading to liver infection remain largely unde-scribed. However, the virulence of E. histolytica isolates isdefined by their ability to provoke ALA in animal models, andthus, a number of factors required for virulence and ALAdevelopment have been identified, including those necessaryfor complement resistance (PPGs), ROS resistance (peroxire-doxin), lysis (CPs and amoebapores), and cell adherence (no-tably, KERP1 and the Gal/GalNAc lectin).

KERP1 stands out since (i) it is specific to E. histolytica and(ii) kerp1 gene expression is increased during ALA develop-ment and correlates with the onset of the host inflammatoryresponse. Indeed, analysis of the stress and antioxidative re-sponses of virulent parasites (in which kerp1 gene expression isupregulated) will help decipher the trophozoite’s adaptation toan infectious lifestyle and identify the factors needed for liverinvasion. The conclusions of this work will shed new light on

the pathophysiology, treatment, and future prophylaxis of he-patic amoebiasis.

ACKNOWLEDGMENTS

The work of the BCP Unit is supported by grants from the FrenchMinistry of National Education’s PRFMMIP Program, from the Pas-teur-Weizmann Research Council, and from the Institut Pasteur (Pro-gramme Transversal de Recherche 179) and by an INCO-DEV grantunder the European Union’s Fifth Framework Program. J.S.-R. re-ceived Ph.D. fellowships from the Fondation pour la Recherche Medi-cale and from the Regime Social des Independants.

We thank Daniela Faust for advice and discussion on the manuscriptand the other members of the BCP Unit for their assistance.

REFERENCES

1. Adams, E. B., and I. N. MacLeod. 1977. Invasive amebiasis. II. Amebic liverabscess and its complications. Medicine (Baltimore) 56:325–334.

2. Ali, I. K., U. Mondal, S. Roy, R. Haque, W. A. Petri, Jr., and C. G. Clark.2007. Evidence for a link between parasite genotype and outcome of infec-tion with Entamoeba histolytica. J. Clin. Microbiol. 45:285–289.

3. Ali, I. K., S. Solaymani-Mohammadi, J. Akhter, S. Roy, C. Gorrini, A.Calderaro, S. K. Parker, R. Haque, W. A. Petri, and C. G. Clark. 2008. Tissueinvasion by Entamoeba histolytica: evidence of genetic selection and/or DNAreorganization events in organ tropism. PLoS Negl. Trop. Dis. 2:e219.

4. Andra, J., R. Herbst, and M. Leippe. 2003. Amoebapores, archaic effectorpeptides of protozoan origin, are discharged into phagosomes and kill bac-teria by permeabilizing their membranes. Dev. Comp. Immunol. 27:291–304.

5. Ankri, S., T. Stolarsky, R. Bracha, F. Padilla-Vaca, and D. Mirelman. 1999.Antisense inhibition of expression of cysteine proteinases affects Entamoebahistolytica-induced formation of liver abscess in hamsters. Infect. Immun.67:421–422.

6. Arellano, J., J. Granados, E. Perez, C. Felix, and R. R. Kretschmer. 1991.Increased frequency of HLA-DR3 and complotype SC01 in Mexican mestizopatients with amoebic abscess of the liver. Parasite Immunol. 13:23–29.

7. Arellano, J., M. Perez-Rodriguez, M. Lopez-Osuna, J. R. Velazquez, J.Granados, N. Justiniani, J. I. Santos, A. Madrazo, L. Munoz, and R.Kretschmer. 1996. Increased frequency of HLA-DR3 and complotype SCO1in Mexican mestizo children with amoebic abscess of the liver. ParasiteImmunol. 18:491–498.

8. Bhattacharya, A., R. Arya, C. G. Clark, and J. P. Ackers. 2000. Absence oflipophosphoglycan-like glycoconjugates in Entamoeba dispar. Parasitology120:31–35.

9. Blazquez, S., M. C. Rigothier, M. Huerre, and N. Guillen. 2007. Initiation ofinflammation and cell death during liver abscess formation by Entamoebahistolytica depends on activity of the galactose/N-acetyl-D-galactosaminelectin. Int. J. Parasitol. 37:425–433.

10. Blazquez, S., C. Zimmer, G. Guigon, J. C. Olivo-Marin, N. Guillen, and E.Labruyere. 2006. Human tumor necrosis factor is a chemoattractant for theparasite Entamoeba histolytica. Infect. Immun. 74:1407–1411.

11. Bracha, R., Y. Nuchamowitz, M. Leippe, and D. Mirelman. 1999. Antisenseinhibition in Entamoeba histolytica causes a decrease in amoebic virulence.Mol. Microbiol. 34:463–472.

12. Braga, L. L., H. Ninomiya, J. J. McCoy, S. Eacker, T. Wiedmer, C. Pham, S.Wood, P. J. Sims, and W. A. Petri, Jr. 1992. Inhibition of the complementmembrane attack complex by the galactose-specific adhesion of Entamoebahistolytica. J. Clin. Investig. 90:1131–1137.

13. Bruchhaus, I., T. Jacobs, M. Leippe, and E. Tannich. 1996. Entamoebahistolytica and Entamoeba dispar: differences in numbers and expression ofcysteine proteinase genes. Mol. Microbiol. 22:255–263.

14. Bruchhaus, I., S. Richter, and E. Tannich. 1998. Recombinant expressionand biochemical characterization of an NADPH:flavin oxidoreductase fromEntamoeba histolytica. Biochem. J. 330:1217–1221.

15. Bruchhaus, I., S. Richter, and E. Tannich. 1997. Removal of hydrogenperoxide by the 29 kDa protein of Entamoeba histolytica. Biochem. J. 326:785–789.

16. Calderon, J., M. de Lourdes Munoz, and H. M. Acosta. 1980. Surfaceredistribution and release of antibody-induced caps in entamoebae. J. Exp.Med. 151:184–193.

17. Chadee, K., and E. Meerovitch. 1984. The pathogenesis of experimentallyinduced amebic liver abscess in the gerbil (Meriones unguiculatus). Am. J.Pathol. 117:71–80.

18. Choi, M. H., D. Sajed, L. Poole, K. Hirata, S. Herdman, B. E. Torian, andS. L. Reed. 2005. An unusual surface peroxiredoxin protects invasive Ent-amoeba histolytica from oxidant attack. Mol. Biochem. Parasitol. 143:80–89.

19. Cieslak, P. R., H. W. Virgin IV, and S. L. Stanley, Jr. 1992. A severecombined immunodeficient (SCID) mouse model for infection with Entam-oeba histolytica. J. Exp. Med. 176:1605–1609.

20. Clark, C. G., U. C. Alsmark, M. Tazreiter, Y. Saito-Nakano, V. Ali, S.

VOL. 22, 2009 AMOEBIC ABSCESS DEVELOPMENT 73

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

Marion, C. Weber, C. Mukherjee, I. Bruchhaus, E. Tannich, M. Leippe, T.Sicheritz-Ponten, P. G. Foster, J. Samuelson, C. J. Noel, R. P. Hirt, T. M.Embley, C. A. Gilchrist, B. J. Mann, U. Singh, J. P. Ackers, S. Bhattacharya,A. Bhattacharya, A. Lohia, N. Guillen, M. Duchene, T. Nozaki, and N. Hall.2007. Structure and content of the Entamoeba histolytica genome. Adv.Parasitol. 65:51–190.

21. Coudrier, E., F. Amblard, C. Zimmer, P. Roux, J. C. Olivo-Marin, M. C.Rigothier, and N. Guillen. 2005. Myosin II and the Gal-GalNAc lectin playa crucial role in tissue invasion by Entamoeba histolytica. Cell. Microbiol.7:19–27.

22. Coulter, E. D., and D. M. Kurtz, Jr. 2001. A role for rubredoxin in oxidativestress protection in Desulfovibrio vulgaris: catalytic electron transfer torubrerythrin and two-iron superoxide reductase. Arch. Biochem. Biophys.394:76–86.

23. Dey, I., K. Keller, A. Belley, and K. Chadee. 2003. Identification and char-acterization of a cyclooxygenase-like enzyme from Entamoeba histolytica.Proc. Natl. Acad. Sci. USA 100:13561–13566.

24. Dodson, J. M., P. W. Lenkowski, Jr., A. C. Eubanks, T. F. Jackson, J.Napodano, D. M. Lyerly, L. A. Lockhart, B. J. Mann, and W. A. Petri, Jr.1999. Infection and immunity mediated by the carbohydrate recognitiondomain of the Entamoeba histolytica Gal/GalNAc lectin. J. Infect. Dis. 179:460–466.

25. Dong, Z., H. Wei, R. Sun, and Z. Tian. 2007. The roles of innate immunecells in liver injury and regeneration. Cell. Mol. Immunol. 4:241–252.

26. Elnekave, K., R. Siman-Tov, and S. Ankri. 2003. Consumption of L-argininemediated by Entamoeba histolytica L-arginase (EhArg) inhibits amoebicidalactivity and nitric oxide production by activated macrophages. Parasite Im-munol. 25:597–608.

27. Elvevold, K. H., G. I. Nedredal, A. Revhaug, and B. Smedsrod. 2004. Scav-enger properties of cultivated pig liver endothelial cells. Comp. Hepatol. 3:4.

28. Engelhardt, B., and H. Wolburg. 2004. Transendothelial migration of leu-kocytes: through the front door or around the side of the house? Eur.J. Immunol. 34:2955–2963.

29. Garcia-Rivera, G., M. A. Rodriguez, R. Ocadiz, M. C. Martinez-Lopez, R.Arroyo, A. Gonzalez-Robles, and E. Orozco. 1999. Entamoeba histolytica: anovel cysteine protease and an adhesin form the 112 kDa surface protein.Mol. Microbiol. 33:556–568.

30. Gauss, K. A., L. K. Nelson-Overton, D. W. Siemsen, Y. Gao, F. R. DeLeo, andM. T. Quinn. 2007. Role of NF-kappaB in transcriptional regulation of thephagocyte NADPH oxidase by tumor necrosis factor-alpha. J. Leukoc. Biol.82:729–741.

31. Ghadirian, E., S. D. Somerfield, and P. A. Kongshavn. 1986. Susceptibility ofEntamoeba histolytica to oxidants. Infect. Immun. 51:263–267.

32. Gilchrist, C. A., E. Houpt, N. Trapaidze, Z. Fei, O. Crasta, A. Asgharpour,C. Evans, S. Martino-Catt, D. J. Baba, S. Stroup, S. Hamano, G. Ehren-kaufer, M. Okada, U. Singh, T. Nozaki, B. J. Mann, and W. A. Petri, Jr. 2006.Impact of intestinal colonization and invasion on the Entamoeba histolyticatranscriptome. Mol. Biochem. Parasitol. 147:163–176.

33. Grap, M. J., L. Pettrey, and D. Thornby. 1997. Hemodynamic monitoring: acomparison of research and practice. Am. J. Crit. Care 6:452–456.

34. Gutierrez-Alarcon, A., M. Moguel-Torres, O. Mata-Leyva, G. Cuellar-Nevarez,T. Siqueiros-Cendon, G. Erosa, E. Ramos-Martinez, P. Talamas-Rohana, andB. Sanchez-Ramirez. 2006. Entamoeba histolytica: inflammatory process duringamoebic liver abscess formation involves cyclooxygenase-2 expression in mac-rophages and trophozoites. Exp. Parasitol. 114:154–159.

35. Haque, R., C. D. Huston, M. Hughes, E. Houpt, and W. A. Petri, Jr. 2003.Amebiasis. N. Engl. J. Med. 348:1565–1573.

36. Harizi, H., M. Juzan, C. Grosset, M. Rashedi, and N. Gualde. 2001. Den-dritic cells issued in vitro from bone marrow produce PGE(2) that contrib-utes to the immunomodulation induced by antigen-presenting cells. Cell.Immunol. 209:19–28.

37. Helmy, K. Y., K. J. Katschke, Jr., N. N. Gorgani, N. M. Kljavin, J. M. Elliott,L. Diehl, S. J. Scales, N. Ghilardi, and M. van Lookeren Campagne. 2006.CRIg: a macrophage complement receptor required for phagocytosis ofcirculating pathogens. Cell 124:915–927.

38. Huston, C. D., E. R. Houpt, B. J. Mann, C. S. Hahn, and W. A. Petri, Jr.2000. Caspase 3-dependent killing of host cells by the parasite Entamoebahistolytica. Cell. Microbiol. 2:617–625.

39. Ivory, C. P., and K. Chadee. 2007. Activation of dendritic cells by theGal-lectin of Entamoeba histolytica drives Th1 responses in vitro and in vivo.Eur. J. Immunol. 37:385–394.

40. Jacobs, T., I. Bruchhaus, T. Dandekar, E. Tannich, and M. Leippe. 1998.Isolation and molecular characterization of a surface-bound proteinase ofEntamoeba histolytica. Mol. Microbiol. 27:269–276.

41. Kammanadiminti, S. J., B. J. Mann, L. Dutil, and K. Chadee. 2004. Regu-lation of Toll-like receptor-2 expression by the Gal-lectin of Entamoebahistolytica. FASEB J. 18:155–157.

42. Katz, U., S. Ankri, T. Stolarsky, Y. Nuchamowitz, and D. Mirelman. 2002.Entamoeba histolytica expressing a dominant negative N-truncated light sub-unit of its gal-lectin are less virulent. Mol. Biol. Cell 13:4256–4265.

43. Katzenstein, D., V. Rickerson, and A. Braude. 1982. New concepts of amebicliver abscess derived from hepatic imaging, serodiagnosis, and hepatic en-

zymes in 67 consecutive cases in San Diego. Medicine (Baltimore) 61:237–246.

44. Kempka, G., and V. Kolb-Bachofen. 1988. Binding, uptake, and transcytosisof ligands for mannose-specific receptors in rat liver: an electron microscopicstudy. Exp. Cell Res. 176:38–48.

45. Kim, K. A., Y. A. Lee, and M. H. Shin. 2007. Calpain-dependent calpastatincleavage regulates caspase-3 activation during apoptosis of Jurkat T cellsinduced by Entamoeba histolytica. Int. J. Parasitol. 37:1209–1219.

46. Leippe, M., J. Andra, R. Nickel, E. Tannich, and H. J. Muller-Eberhard.1994. Amoebapores, a family of membranolytic peptides from cytoplasmicgranules of Entamoeba histolytica: isolation, primary structure, and poreformation in bacterial cytoplasmic membranes. Mol. Microbiol. 14:895–904.

47. Lin, J. Y., and K. Chadee. 1992. Macrophage cytotoxicity against Entamoebahistolytica trophozoites is mediated by nitric oxide from L-arginine. J. Im-munol. 148:3999–4005.

48. Loftus, B., I. Anderson, R. Davies, U. C. Alsmark, J. Samuelson, P. Amedeo,P. Roncaglia, M. Berriman, R. P. Hirt, B. J. Mann, T. Nozaki, B. Suh, M.Pop, M. Duchene, J. Ackers, E. Tannich, M. Leippe, M. Hofer, I. Bruchhaus,U. Willhoeft, A. Bhattacharya, T. Chillingworth, C. Churcher, Z. Hance, B.Harris, D. Harris, K. Jagels, S. Moule, K. Mungall, D. Ormond, R. Squares,S. Whitehead, M. A. Quail, E. Rabbinowitsch, H. Norbertczak, C. Price, Z.Wang, N. Guillen, C. Gilchrist, S. E. Stroup, S. Bhattacharya, A. Lohia, P. G.Foster, T. Sicheritz-Ponten, C. Weber, U. Singh, C. Mukherjee, N. M. El-Sayed, W. A. Petri, Jr., C. G. Clark, T. M. Embley, B. Barrell, C. M. Fraser,and N. Hall. 2005. The genome of the protist parasite Entamoeba histolytica.Nature 433:865–868.

49. Lotter, H., T. Jacobs, I. Gaworski, and E. Tannich. 2006. Sexual dimorphismin the control of amebic liver abscess in a mouse model of disease. Infect.Immun. 74:118–124.

50. Lushbaugh, W. B., A. F. Hofbauer, and F. E. Pittman. 1985. Entamoebahistolytica: purification of cathepsin B. Exp. Parasitol. 59:328–336.

51. Lushbaugh, W. B., A. B. Kairalia, A. F. Hofbauer, J. R. Cantey, and F. E.Pittman. 1978. Cytotoxic activity of a cell-free extract of Entamoeba histo-lytica. Trans. R. Soc. Trop. Med. Hyg. 72:105–106.

52. Maldonado-Bernal, C., C. J. Kirschning, Y. Rosenstein, L. M. Rocha, N.Rios-Sarabia, M. Espinosa-Cantellano, I. Becker, I. Estrada, R. M. Salazar-Gonzalez, C. Lopez-Macias, H. Wagner, J. Sanchez, and A. Isibasi. 2005.The innate immune response to Entamoeba histolytica lipopeptidophospho-glycan is mediated by Toll-like receptors 2 and 4. Parasite Immunol. 27:127–137.

53. Mann, B. J., C. Y. Chung, J. M. Dodson, L. S. Ashley, L. L. Braga, and T. L.Snodgrass. 1993. Neutralizing monoclonal antibody epitopes of the Entam-oeba histolytica galactose adhesin map to the cysteine-rich extracellular do-main of the 170-kilodalton subunit. Infect. Immun. 61:1772–1778.

54. Marinets, A., T. Zhang, N. Guillen, P. Gounon, B. Bohle, U. Vollmann, O.Scheiner, G. Wiedermann, S. L. Stanley, and M. Duchene. 1997. Protectionagainst invasive amebiasis by a single monoclonal antibody directed againsta lipophosphoglycan antigen localized on the surface of Entamoeba histo-lytica. J. Exp. Med. 186:1557–1565.

55. Matsuno, K., H. Nomiyama, H. Yoneyama, and R. Uwatoku. 2002. Kupffercell-mediated recruitment of dendritic cells to the liver crucial for a hostdefense. Dev. Immunol. 9:143–149.

56. Maurus, C. F., M. K. Schneider, D. Schmidt, G. Zund, and J. D. Seebach.2006. Activation of human microvascular endothelial cells with TNF-alphaand hypoxia/reoxygenation enhances NK-cell adhesion, but not NK-cytotox-icity. Transplantation 81:1204–1211.

57. Moran, P., G. Rico, M. Ramiro, H. Olvera, F. Ramos, E. Gonzalez, A.Valadez, O. Curiel, E. I. Melendro, and C. Ximenez. 2002. Defective pro-duction of reactive oxygen intermediates (ROI) in a patient with recurrentamebic liver abscess. Am. J. Trop. Med. Hyg. 67:632–635.

58. Munford, R. S., P. O. Sheppard, and P. J. O’Hara. 1995. Saposin-likeproteins (SAPLIP) carry out diverse functions on a common backbone struc-ture. J. Lipid Res. 36:1653–1663.

59. Neal, R. A., and P. Vincent. 1956. Strain variation in Entamoeba histolytica.II. The effect of serial liver passage on the virulence. Parasitology 46:173–182.

60. Olivos, A., E. Ramos, M. Nequiz, C. Barba, E. Tello, G. Castanon, A.Gonzalez, R. D. Martinez, I. Montfort, and R. Perez-Tamayo. 2005. Ent-amoeba histolytica: mechanism of decrease of virulence of axenic culturesmaintained for prolonged periods. Exp. Parasitol. 110:309–312.

61. Pelosof, L. C., P. H. Davis, Z. Zhang, X. Zhang, and S. L. Stanley, Jr. 2006.Co-ordinate but disproportionate activation of apoptotic, regenerative andinflammatory pathways characterizes the liver response to acute amebicinfection. Cell. Microbiol. 8:508–522.

62. Perez-Tamayo, R., R. D. Martinez, I. Montfort, I. Becker, E. Tello, and R.Perez-Montfort. 1991. Pathogenesis of acute experimental amebic liver ab-scess in hamsters. J. Parasitol. 77:982–988.

63. Petri, W. A., Jr., R. Haque, and B. J. Mann. 2002. The bittersweet interfaceof parasite and host: lectin-carbohydrate interactions during human invasionby the parasite Entamoeba histolytica. Annu. Rev. Microbiol. 56:39–64.

64. Poole, L. B., H. Z. Chae, B. M. Flores, S. L. Reed, S. G. Rhee, and B. E.

74 SANTI-ROCCA ET AL. CLIN. MICROBIOL. REV.

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from

Torian. 1997. Peroxidase activity of a TSA-like antioxidant protein from apathogenic amoeba. Free Radic. Biol. Med. 23:955–959.

65. Puhl, G., K. D. Schaser, B. Vollmar, M. D. Menger, and U. Settmacher. 2003.Noninvasive in vivo analysis of the human hepatic microcirculation usingorthogonal polarization spectral imaging. Transplantation 75:756–761.

66. Reed, S. L., J. G. Curd, I. Gigli, F. D. Gillin, and A. I. Braude. 1986.Activation of complement by pathogenic and nonpathogenic Entamoebahistolytica. J. Immunol. 136:2265–2270.

67. Reed, S. L., and I. Gigli. 1990. Lysis of complement-sensitive Entamoebahistolytica by activated terminal complement components. Initiation of com-plement activation by an extracellular neutral cysteine proteinase. J. Clin.Investig. 86:1815–1822.

68. Reed, S. L., P. G. Sargeaunt, and A. I. Braude. 1983. Resistance to lysis byhuman serum of pathogenic Entamoeba histolytica. Trans. R. Soc. Trop.Med. Hyg. 77:248–253.

69. Rigothier, M. C., H. Khun, P. Tavares, A. Cardona, M. Huerre, and N.Guillen. 2002. Fate of Entamoeba histolytica during establishment of amoe-bic liver abscess analyzed by quantitative radioimaging and histology. Infect.Immun. 70:3208–3215.

70. Sanchez-Ramirez, B., M. Ramirez-Gil, I. Vazquez-Moctezuma, E. Ramos-Martinez, and P. Talamas-Rohana. 2004. Entamoeba histolytica: induction ofcyclooxygenase-2 expression during amoebic liver abscess formation in ham-sters (Mesocricetus auratus). Exp. Parasitol. 106:119–125.

71. Santi-Rocca, J., C. Weber, G. Guigon, O. Sismeiro, J. Y. Coppee, and N.Guillen. 2008. The lysine- and glutamic acid-rich protein KERP1 plays a rolein Entamoeba histolytica liver abscess pathogenesis. Cell. Microbiol. 10:202–217.

72. Sato, D., K. Nakada-Tsukui, M. Okada, and T. Nozaki. 2006. Two cysteineprotease inhibitors, EhICP1 and 2, localized in distinct compartments, neg-atively regulate secretion in Entamoeba histolytica. FEBS Lett. 580:5306–5312.

73. Schultz, H., E. Csernok, K. Herlyn, P. H. Reichel, F. Moosig, O. A. Cornely,M. K. Fagerhol, and W. L. Gross. 2003. ANCA against bactericidal/perme-ability-increasing protein, azurocidin, calprotectin and defensins in rheu-matic and infectious diseases: prevalence and clinical associations. Clin. Exp.Rheumatol. 21:S117–S120.

74. Seguin, R., B. J. Mann, K. Keller, and K. Chadee. 1995. Identification of thegalactose-adherence lectin epitopes of Entamoeba histolytica that stimulatetumor necrosis factor-alpha production by macrophages. Proc. Natl. Acad.Sci. USA 92:12175–12179.

75. Seguin, R., B. J. Mann, K. Keller, and K. Chadee. 1997. The tumor necrosisfactor alpha-stimulating region of galactose-inhibitable lectin of Entamoebahistolytica activates gamma interferon-primed macrophages for amebicidalactivity mediated by nitric oxide. Infect. Immun. 65:2522–2527.

76. Seigneur, M., J. Mounier, M. C. Prevost, and N. Guillen. 2005. A lysine- andglutamic acid-rich protein, KERP1, from Entamoeba histolytica binds tohuman enterocytes. Cell. Microbiol. 7:569–579.

77. Seydel, K. B., S. J. Smith, and S. L. Stanley, Jr. 2000. Innate immunity to

amebic liver abscess is dependent on gamma interferon and nitric oxide in amurine model of disease. Infect. Immun. 68:400–402.

78. Seydel, K. B., and S. L. Stanley, Jr. 1998. Entamoeba histolytica induces hostcell death in amebic liver abscess by a non-Fas-dependent, non-tumor ne-crosis factor alpha-dependent pathway of apoptosis. Infect. Immun. 66:2980–2983.

79. Singer, I. I., D. W. Kawka, S. Schloemann, T. Tessner, T. Riehl, and W. F.Stenson. 1998. Cyclooxygenase 2 is induced in colonic epithelial cells ininflammatory bowel disease. Gastroenterology 115:297–306.

80. Talamas-Rohana, P., and I. Meza. 1988. Interaction between pathogenicamebas and fibronectin: substrate degradation and changes in cytoskeletonorganization. J. Cell Biol. 106:1787–1794.

81. Thompson, P. E., D. McCarthy, and J. W. Reinertson. 1954. Observations onthe virulence of Entamoeba histolytica during prolonged subcultivation.Am. J. Hyg. 59:249–261.

82. Tsutsumi, V., R. Mena-Lopez, F. Anaya-Velazquez, and A. Martinez-Palomo.1984. Cellular bases of experimental amebic liver abscess formation. Am. J.Pathol. 117:81–91.

83. Velazquez, J. R., P. Llaguno, J. Fernandez-Diez, M. Perez-Rodriguez, J.Arellano, M. Lopez-Osuna, and R. R. Kretschmer. 1995. Antigen inducedeosinophilia protects gerbils (Meriones unguiculatus) against experimentalamebic abscess of the liver. Arch. Med. Res. 26:S93–S98.

84. Vines, R. R., G. Ramakrishnan, J. B. Rogers, L. A. Lockhart, B. J. Mann,and W. A. Petri, Jr. 1998. Regulation of adherence and virulence by theEntamoeba histolytica lectin cytoplasmic domain, which contains a beta2integrin motif. Mol. Biol. Cell 9:2069–2079.

85. Walzer, T., M. Dalod, S. H. Robbins, L. Zitvogel, and E. Vivier. 2005.Natural-killer cells and dendritic cells: “l’union fait la force.” Blood 106:2252–2258.

86. Watanabe, T., H. Katsukura, T. Chiba, T. Kita, and Y. Wakatsuki. 2007.Periportal and sinusoidal liver dendritic cells suppressing T helper type1-mediated hepatitis. Gut 56:1445–1451.

87. Weber, C., S. Blazquez, S. Marion, C. Ausseur, D. Vats, M. Krzeminski,M. C. Rigothier, R. C. Maroun, A. Bhattacharya, and N. Guillen. 2008.Bioinformatics and functional analysis of an Entamoeba histolytica manno-syltransferase necessary for parasite complement resistance and hepaticalinfection. PLoS Negl. Trop. Dis. 2:e165.

88. WHO. 1997. WHO/PAHO/UNESCO report of a consultation of experts onamoebiasis. Wkly. Epidemiol. Rec. 72:97–100.

89. Zhang, X., Z. Zhang, D. Alexander, R. Bracha, D. Mirelman, and S. L.Stanley, Jr. 2004. Expression of amoebapores is required for full expressionof Entamoeba histolytica virulence in amebic liver abscess but is not necessaryfor the induction of inflammation or tissue damage in amebic colitis. Infect.Immun. 72:678–683.

90. Zhang, Z., S. Mahajan, X. Zhang, and S. L. Stanley, Jr. 2003. Tumornecrosis factor alpha is a key mediator of gut inflammation seen in amebiccolitis in human intestine in the SCID mouse-human intestinal xenograftmodel of disease. Infect. Immun. 71:5355–5359.

VOL. 22, 2009 AMOEBIC ABSCESS DEVELOPMENT 75

on October 17, 2020 by guest

http://cmr.asm

.org/D

ownloaded from