Reviews in Medical Virology REVIEW Immunity and …iah.iec.pa.gov.br/iah/fulltext/pc/artigos/2013/...to biting mosquitoes. The period of infection is followed by a period of remission,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Rev. Med. Virol. 2013; 23: 305–318.Published online 22 July 2013 in Wiley Online Library
(wileyonlinelibrary.com)DOI: 10.1002/rmv.1752
R E V I E W
Immunity and immpathologic changes
Reviews in Medical Virology
Correspondnce to: Drniversidade Federal d-mail: juarez@pesquis*Correspondence to: Dnanindeua, Pará, Bra-mail: pedrovasconcelReferences for this revrticles published fromyellow fever virus”, “tosis”, “histopathologypathophysiology”, “dsepsis”, “viral hemorrublished between 1890arches in the authors'nd relevant referencesshed in English, Fren
une response, pathology and: progress and challenges in
the immunopathology of yellow fever†Juarez A. S. Quaresma1,4*, Carla Pagliari2, Daniele B. A. Medeiros3,Maria I. S. Duarte2 and Pedro F. C. Vasconcelos3,4**1Núcleo de Medicina Tropical, Universidade Federal do Pará, Belém, Pará, Brazil2Departamento de Patologia, Universidade de São Paulo, São Paulo, Brazil3Departamento de Arbovirologia e Febres Hemoprrágicas, Instituto Evandro Chagas, SVS/Ministério daSaúde, Ananindeua, Pará, Brasil4Departamento de Patologia, Universidade do Estado do Pará, Belém, Pará, Brazil
Received: 27 December 2012; Revised: 21 May 2013; Accepted: 28 May 2013
INTRODUCTIONYellow fever is an infectious disease,which producesa broad spectrum of clinical manifestations ranging
. J. A. S. Quaresma, Núcleo de Medicina Tropical,o Pará, Belém, Pará, Brazil.ador.cnpq.brr. P. F. C. Vasconcelos, Instituto Evandro Chagas,[email protected] were identified through searches of PubMed for1983 to 2011, by use of the terms “yellow fever”,yellow fever vaccine”, “Councilman bodies”, “apo-”, “immunohistochemistry”, “immunopathology”,engue”, “dengue hemorrhagic fever”, “hepatitis”,hagic fevers”, and “immune cells”. Relevant articlesand 1928, 1951, and 1973 were identified throughpersonal files. Articles resulting from these searchescited in those articles were reviewed. Articles pub-ch, and Portuguese were included.
from mild or unapparent infection to severe formswith hemorrhagic diathesis and acute hepatic and/or renal failures [1]. The disease affects approximately200 000 individuals and causes approximately 30 000deaths annually in the tropical regions of Africa andSouth America, but recent evidence suggests lowernumbers due highly under notification [2]. The dis-ease is caused by the yellow fever virus (YFV),the prototype species of the Flavivirus genus ofthe Flaviviridae family [3].
Yellow fever virus is maintained in distinctepidemiologic cycles in Africa and South America(Figure 1). In the urban cycle, YFV is transmittedbetween humans by the bite of peridomesticmosquitoes, Aedes aegypti. Urban transmission is nowonly sporadically reported in SouthAmerica but is stillfrequently described in Africa [2]. In the jungle cycle,YFV is maintained among monkeys by the bites ofseveral mosquito species of the genus Aedes in Africa,mainly Aedes africanus and Aedes furcifer, and severalspecies of theHaemagogus andSabethesgenera in South
Figure 1. Yellow fever maintenance cycles in the Americas (A) and Africa (B). The intermediate sylvatic cycle is only found in Africa
306 J. Quaresma et al.
America, especially Haemagogus janthinomys [4,5]. Athird peri-urban or rural cycle has been described insmallAfrican communities living near forests inwhichthe mosquito Aedes simpsoni is mainly responsible fortransmission [4].Three clinical stages of illness have been described
in humans in the classical picture of yellow fever,
3–6days after the infecting bite. The disease beginswith the acute onset of fever (approximately 39°C),headache, malaise, photophobia, backache, myalgia,irritability, restlessness, nausea, and vomiting [1,6,7].This symptomatology lasts for 3–5days and isknown as the period of infection (i.e. period ofviremia). During this period, the blood is infectious
to biting mosquitoes. The period of infection isfollowed by a period of remission, which lastsapproximately 12h–2days. During this stage, thepatient feels better, with a sensation of recovery, butsuddenly, some patients become severely ill with typ-ical signs of liver and renal failure, which character-izes the third stage, the period of intoxication(Figure 2). During this stage, patients develop severehemorrhagic fever and multi-organ dysfunctionaccompanied by jaundice, oliguria or anuria, car-diovascular instability, and hemorrhagic diathesisduring which episodes of hematemesis, melena,petechiae, ecchymosis, and other hemorrhagicmanifestations are common [1,6,7].In this final stage, thrombocytopenia (a platelet
count of less than 50 000 per mililiter), prolongedclotting and prothrombin times, and lower levelsof liver clotting factors are observed. Laboratoryclinical tests indicate diminished fibrinogen andfactor VIII and elevated fibrin split products, whichare characteristics of disseminated intravascularcoagulation [8].Immediately after a YFV inoculation by an infected
mosquito, the virus first replicates in the local lymphnodes and then disseminates to many organs,causing lesions, either as a consequence of the directviral cytopathic effect or due to alterations secondaryto the immune response of the host [9–11].
Figure 2. The three different yellow fever clinical stages (periodof infection or viremia, period of remission, and period of intoxi-cation or localization)
In yellow fever pathogenesis, the virus elicits twodistinct and separate patterns of injury, viscerotropismand neurotropism. YFV is neurotropic in mice, andthey succumb to an encephalitic infection. YFV isviscerotropic in primates, including humans, inwhomit causes lesions in multiple organs, such as the liver,spleen, heart, and kidneys [12]. The liver is theprimary target organ for YFV. Indeed, YFV induceshepatocellular injury characterized by eosinophilicnecrosis, apoptosis of hepatic cells, andmacro steatosisand micro steatosis, mainly located in the midzone; ascarce portal inflammatory infiltrate is also found(Figure 3) [13–15]. In this review, we address severalaspects of the pathogenesis and immunopathologyof yellow fever.
GENERAL PATTERN OF LIVER INJURIESThe pattern of liver involvement in yellow fever isintense, primarily in the midzone region (zone 2)when compared with the acini zones 1 and 3. Thisinvolvement has been frequently reported in bothhuman cases and experimental monkey andhamster models. There is questionable evidencethat explains the preferential lesion patternobserved for yellow fever and other arbovirusinfections. This pattern shows a course of intensevascular involvement and more severe midzonallesions [15–18]. Some authors have proposed thatthe YFV has higher tropism to zone 2 cells, in whicha greater concentration of yellow fever antigens hasbeen observed [19]. However, it should be noted thatthe same midzone lesion pattern is also observed inother arboviral diseases, such as dengue and RiftValley fever [20]. Similar midzone injury has beendescribed elsewhere [21,22]. The pathophysiology ofyellow fever and dengue clearly is of infectiousvasculopathies, characterized by increased vasculardamagewhich results in low-flowhypoxia. Therefore,the hemodynamic changes in connection with theliver tropism of both viruses explain the preferentialobserved midzonal focus in the course of dengue,yellow fever, and other hepatotropic flaviviruses [22].
In yellow fever, in all liver tissue (Figure 3),swelling of the hepatocytes is present, but it is moresevere in zones 1 and 3. The common finding inother viral hepatitis infections such as ballooningdegeneration is not observed during yellow fever.The presence of Councilman bodies is typical ofyellow fever, and their morphology is impressivelywide, with pleomorphic hepatocytes ranging fromwell-defined to bizarre cells distributed in all liver
Figure 3. Photomicrographs of liver in fatal yellow fever using histopathology (A–E) and an immunohistochemical assay (F). Panels A–Dwerestainedwith hematoxylin and eosin (HE); panel Ewas reticulin stained; panel Fwas stained by immunohistochemical assay. (A) Typicalmidzonallesions in the liver (arrows); (B) Different patterns of Councilman bodies (apoptotic hepatocytes) in the hepatic acini (arrows); (C) Areas showinglytic necrosis of hepatocytes (black arrow) andmicrogoticular steatosis (dark blue arrow); (D) Portal space showsmoderate inflammatory infiltratemainly composed of mononuclear cells disproportionate to the intense acinar lesion (arrows); (E) Reticulin trabeculae show an absence of septalfibrotic tissue and preservation of the typical liver architecture; (F) Immunohistochemical assay shows yellow fever antigen in the portal spacestained with an anti-yellow fever monoclonal antibody
308 J. Quaresma et al.
areas but clearly found more intensely in themidzone. These aspects may correspond to thedifferent apoptotic stages observed in the yellowfever-infected liver [14,15].Quantitative microscopic and immunohisto-
chemical analyses have confirmed the distributionand importantly, that the apoptosis representedby Councilman bodies predominates over lyticnecrosis [10,11,15,23]. It is worthy of note that thispattern of apoptotic cells has been described onspecimens from humans but also from non-humanprimates. The characteristics of cells and theirultrastructural aspects confirm apoptosis as themain cell death mechanism during severe yellowfever [10,11,15,23]. It is important to emphasize thata similar pattern of liver injury is also observed inviral hepatitis [24,25].In parallel with the apoptosis, especially in the
midzone, lytic necrosis of hepatocytes has beenfound in well-limited areas by minimal inflamma-tion (predominantly lymphocytes and neutrophils).
These two mechanisms of hepatic cell death, plusswelling and steatosis, are clearly associated withthe liver failure classically described during thecourse of severe yellow fever and explain the highlevels of both alanine and aspartate aminotransfer-ase in the blood [26].
In golden hamsters, the proportion of hepatocytelyses are related to the increase of apoptosis as thedisease progressed [16,17]. However, quantifyingthe contribution of each of these components to theinduction of liver failure requires further studies.The focal areas of lytic necrosis are primarily foundwithin zone 2, associated with a discretelymphomononuclear inflammatory infiltrate [14,15].
Both steatosis and necrosis occur moreintensely in the midzone area; the first is charac-terized by wide morphological variability,including microdroplets or a morula-like aspect(“morula cells”) that apparently results fromaccumulation of large amounts of vesicles in thehepatocyte cytoplasm, called macrodroplet
patterns [15,27]. Steatosis is one of the markedaspects of the yellow fever liver, and its intensitycoincides with the intensity of apoptosis (Figure 3).The macro steatosis and micro steatosis commonlyfound in yellow fever may correspond to differentstages of the reversible injury process [15]. Twostudies on chronic hepatitis have correlated steatosiswith a poor prognosis [28,29]. It is noted that thispattern of steatosis has been also described duringdelta virus hepatitis and hepatitis B [28].During severe yellow fever, acinar mononuclear
inflammatory infiltrate is a common finding butusually discrete and disproportionate to the severedegree of hepatic damage. Indeed, in the areas oflytic necrosis, the polymorphonuclear infiltrate isscarce. In the acini, the inflammatory infiltrate ismore evident in zone 2, similar to the hepatocyteinjury [14,15,17]. The edema is discrete in the portalspaces, and the presence of lymphomononuclear(lymphocytes and macrophages) cells is quite discreteand slightly more concentrated than that observed inthe other acini areas and independent of the degreeof parenchymal involvement [10,11,15]. The portaland acinar inflammatory infiltrates in humansseverely infected by yellow fever and also inexperimental models of yellow fever consist oflymphomononuclear cells. [10,11,14,17]. Importantly,near the necrotic areas, a neutrophilic infiltrate iscommonly observed.[15]. In contrast, the extensiveapoptosis seems to be responsible for thedisproportionate parenchymatous injury versusinflammation which results in a paucity cellularinflammatory response [10,15,23] (Table 1).
APOPTOSIS AS A PREDOMINANT CELLDEATH MECHANISM IN YELLOW FEVERApoptosis is a cardinal feature of liver injury inmany infections. In the liver, two pathways ofapoptosis have been observed: extrinsic and intrin-sic pathways that are dependent and independentof death receptors, respectively. Death receptorsare major mediators of the apoptosis pathway inthe liver and have been implicated in the pathogen-esis of viral hepatitis and other liver diseases [30].Mechanisms of liver injury are complex andinvolve the interaction of cytokines, reactiveoxygen species, Kupffer cells, and immune cells[31]. Some studies have demonstrated that
apoptosis is the predominant mechanism of celldeath in some viral hemorrhagic fevers [10].
In yellow fever, both intrinsic and extrinsicpathways may be involved in the induction ofapoptosis. In the intrinsic pathway, oxidative stressand the mediators released in the endotheliallesions with microvascular alterations cause mito-chondrial lesions with release of members of theBcl-2 protein family that interact with the caspasecascade to induce cell death by apoptosis [32].
Death receptors can also regulate apoptosis byexternal killers binding or ligating cell surfacereceptors. For example, Fas, TNF receptor 1,TNF-related apoptosis-inducing ligand receptor 1(TRAIL), and TRAIL receptor 2 are common deathcell receptors expressed in the human liver [33].During fatal yellow fever, ApopTag-labeled cells(Millipore, Billerica, MA, USA) have been foundin all regions of the hepatic lobule [10,15]. Ligandsfor these receptors are expressed in human and ro-dent liver tissues by immunohistochemistry andmolecular assays. More intense immunolabelingfor apoptotic markers and these ligands in themidzone were observed during the course of yel-low fever (Figure 4) [11].
Recently, it has been demonstrated that theimmunoexpression for of FasL and apoptosiscoincide and are more intense in the midzone,indicating that this receptor may play a role in thegenesis of liver injury in yellow fever [11,23].Interestingly, this same aspect is a characteristicfinding of liver lesions in severe dengue cases.
It is a common observation that apoptotic hepaticcells correspond to the Councilman bodies in humansand to cells under apoptotic conditions in rodents[15,17]. Ultrastructural analysis has confirmed thatCouncilman bodies correspond to hepatocytes asdescribed elsewhere [14]. Quantitative analysisoriginally demonstrated that the intense apoptoticcomponent represented by the Councilman bodiesupon lightmicroscopy is confirmed by immunohisto-chemistry and electron microscopy reported for bothyellow fever-infected humans and monkey tissues(Figure 4) [15,34].
Several mechanisms are implicated in theapoptosis pathway following yellow fever andinclude the direct cytopathic effect of YFV, animportant mitochondrial dysfunction due tolow-flow hypoxia, and the influence of immunefactors in the in situ immune response on theliver [22]. This cell death process in the liver
Figure 4. Microphotographs obtained from immunohistochemical assays for apoptosis and for characterizing the immune responsein the liver of fatal yellow fever. Arrows in the panels (A–F) indicate apoptotic cells (A); CD8+ T lymphocytes (B); CD4+ T lympho-cytes (C); CD20+ B lymphocytes (D); CD68+ macrophages (E); and S100+ antigen cells (F). Panels G–I represent the more prominentcytokine profile in the liver during fatal yellow fever. Arrows in the panels indicate those cells that express different cytokines.Cells are immunolabeled for (G) INF-.yscp;; (H) TNF-α; and (I) TGF-β
311Immunopathology of yellow fever
following yellow fever with minimal inflamma-tory infiltrate is explained because apoptosisdoes not induce an inflammatory response [35].The main consequence of this death
phenomenon is the possibility of developing atherapeutic approach to protect the liver in se-vere cases of disease. The therapeutic approachshould block the mechanisms that trigger this
process and, consequently, inhibit theimmunopathological events involved in thepathogenesis of yellow fever.
THE PATTERN OF CELLULAR IMMUNERESPONSE IN YELLOW FEVERThe immune response pattern is significant in theevolution of tissue lesions inherent to differentinfectious diseases [36]. Viral liver infections, suchas hepatitis B and C, and the cytotoxic cellularresponse, contribute to disease evolution whilecontrolling viral replication [37]. In addition, theexistence of asymptomatic patients chronicallyinfected with hepatitis C virus, a member of theFlaviviridae family, suggests a role for immuneresponses in the genesis of tissue lesions and theevolution of liver injury during infections inducedby flaviviruses [37,38].Dengue and yellow fever are noted because of their
unique pathogenesis, which is characterized byintense hepatic involvement [15,39]. Some authorshave characterized the phenotypic pattern of thecellular immune response in the yellow fever liver;by immunohistochemistry assays, the local (in situ)immune response demonstrates positive reactivity toseveral inflammatory cells, which exhibited surfacereceptors, such as CD45RO, CD4, CD8, CD20, S100,CD68, and CD57 [10,11,23]. Additionally, CD45RO-positive T cells have been observed both in the aciniand portal tract. In the acini, immunolabeling is morefrequently observed than in themidzone. This patternconcurswithmore intense liver injuries in this zone inwhich immunolabeled T lymphocytes aggregate,whereas phagocytic cells (neutrophils and macro-phages) have been found in the surrounding areasof hepatic injury [15].Regarding the portal tract, the immunohisto-
chemical pattern observed during fatal yellowfever is of a higher density of immunolabeled Tlymphocytes when compared with the acini.Therefore, the flow of migratory immune cellsduring yellow fever elicits an immune responsethat may be initiated prior to the response in theportal tract and then in the acini. Interestingly, inthe infiltrates of the acini and portal tract, CD4+ Tcells are more frequent than CD8+ Tcells (Figure 4).The predominance of T lymphocytes, mainly CD4+and CD8+ T cells, is a common finding in severalother hepatotropic viral infections, and perhaps,they may be involved in the viral clearance process
during the course of yellow fever, as has beendescribed in hepatitis B and C [40].
An important way of eliminating infected cells isthe neutralization of circulating virus particles bythe immune system thereby blocking the intracellularviral cycle [41,42]. This mechanism may be regulatedby cytokine expression that has synergistic effects inthe immune network [42]. The role of severalcytokines in the liver during the immune responsehas been described in human and experimentalmodels of yellow fever, and their effects may not belimited to the liver. These cytokines can also beimplicated in the intense vasculopathy observed inthe severe cases of yellow fever and perhaps indengue fever and dengue hemorrhagic fever/dengueshock syndrome (DHF/DSS) [10,11,43].
The cellular immune response in yellow fever iscomplex and not only involves the participationof CD8+ or cytotoxic T cells but also the action ofCD4+ T cells, macrophages, polymorphonuclearcells, antigen-presenting cells (S100), and NK cells[10,11]. The pattern of cells that elicit this orches-trated immune response in yellow fever has beencharacterized as mainly CD4+ T lymphocytesand, to a lesser extent, CD8+ T lymphocytes.Associated with the T-cell population, antigen-presenting S100+ cells, macrophages CD68+ and NKCD57+ cells are found in the acini areas, but theirconcentrations are higher in the midzone (Figure 4)[23]. Despite the limited quantity of these cells whencompared with the intense tissue injury, their partici-pation appears to be important in the determinationof disease evolution and outcome. This fact isconfirmed by the presence of large numbers ofcytotoxic cells in the midzone region [10,11,23].However, the exact role and mechanism of action ofthese cells in the immunopathogenesis of yellow feverrequire additional studies.
The quantitative presence of CD4+ T cellsdemonstrates a pivotal role of these cells in theimmunopathogenesis of yellow fever [10]. CD4+ Tcells display cytolytic activity in several infections,and during the clinical course of dengue andyellow fever, they may have a critical role in thecellular response against these viruses, either bycytolytic actions or by interactions with the cytokinenetwork [44]. In addition, the role of activatedmacro-phages can be confirmed by the release of severalcytokines such as oxygen-derived free radicals, nitricoxide, and TNF-α. The role of TNF-α is critical in thepathogenesis of shock and tissue injury during severe
yellow fever [1]. The activation of cytotoxic CD8+ Tlymphocytes, either by cytokines or by immuneprocessing of the professional S100 cells, areincriminated by expression ofMHC class Imolecules,and they will interact with infected hepatocytes toplay an important role in the induction of lyticnecrosis and apoptotic cell death [10,11,23].Cytolysis is an important mechanism for viral
clearance, but this process does not always elimi-nate the etiological agent, as described for hepatitisB and C, possibly representing a mechanism toescape the action of the immune system. [14,45].During severe yellow fever, the lysis of infected
cells by CD8+ T lymphocytes can release viralparticles that will be exposed to the action ofantibodies produced by B CD20+ cells. These Blymphocytes are often found in the portal tract aswell as in the acinar region, and therefore, mayplay a role in the mechanism of antigen processingduring the immune response. In addition, theproduction of specific antibodies after differentia-tion of plasma cells will also act synergistically.[10,11,18] Finally, Treg and TH17 lymphocytes havebeen demonstrated to be important factors in otherinfections [46,47], and perhaps, they may play arole in the evolution and outcome of yellow fever.
CYTOKINES AND THE IMMUNE RESPONSEThe pattern of cytokine expression in response toyellow fever has been described both in humanand experimental infection models [10,11]. Amarked expression of TNF-α, IFN-γ, and intenseimmunolabelling for TGF-β has been demonstratedin fatal human cases (Figure 4). Other cytokines arenot as highly expressed but are observed in all ofthe hepatic acini. Those cytokines show a greaterexpression in the midzone and portal tract andoccur less frequently in other areas. Liverexpression of IL1-α/β, IL-4, IL-8, and IL-10 is lowerin the lobule areas and portal tract, and there is nopreference for any area of the hepatic lobule.The role of cytokines in the process of viral
clearance has already been described in hepatitisB and C [48]. TNF-α and IFN-γ are of majorimportance because both cytokines can induceapoptosis in hepatocytes by binding to theirrespective receptors and activating the death do-main, which results in cell death [48,49]. The sin-gular antiviral effect of IFN-γ is well known, andthis cytokine is associated with macrophage acti-vation as well as the increase in the expression of
the molecules related to MHC types I and II. Inthe human host, TNF-α is incriminated withpresentation of antigens to lymphomononuclearcells and particularly in binding of CD8+ T-cellsto the surface of the hepatocytes, an importantmechanism leading to cell death. The pathogenicaction of TNF-α on the course of viral hepatitis Band C is not completely understood, but the induc-tion of apoptotic death by TNF-α may be associatedwith the persistence of infection in the host [49].
In the course of severe yellow fever, theactions of IFN-γ and TNF-α remain to be defined.However, Quaresma and colleagues [10,11,23]demonstrated that when immunolabellinghepatic cells with these cytokines to characterizea Th1 immune response, the immunolabellingwas more intense in the midzone, and thesecytokines may have been directly or indirectlyimplicated in both viral clearance and theinduction of cell death, both by apoptosis andnecrosis. Therefore, it is clear that they may playa role during the process that results in apoptosisof liver cells [50]. It is also clear that receptorssuch as Fas, which is especially expressed infatal/severe yellow fever cases, may act as possi-ble mediators for cell death by activating deathreceptors and inducing apoptosis. The intenseimmunolabelling for TGF-β in the yellow feverliver is an important characteristic of thisinfection [10,22]. TGF-β is a cytokine with strongimmunosuppressor and pro-apoptotic effects;this fact can elicit two more histopathologicalfindings in the yellow fever liver: the intenseapoptosis and the poorly inflammatory infiltratedisproportional to the intense tissue injury(Figure 4). It was demonstrated that TGF-β playsan important role in fatal human yellow feverpossibly by inducing an intense tissue damage[10,22]. Schuster and Krieglstein [51] havesuggested that the effects of TGF-β and TNF-αon tissue are closely associated with theirinteraction with the targeted cells. Others havedemonstrated that both cytokines are prominent andconcomitantly expressed in all lobule areas but aremuch more intensively expressed in the midzone[10]. The proposed diagram in Figure 5 shows the im-mune response and pathophysiology mechanismduring yellow fever. The coordinated immunologicresponses are driven by the CD4+ lymphocytes viaMHC II and result in activation of several immunecells and cytokines, especially in the liver.
Figure 5. (A) Schematic of the proposed mechanism of infection and pathogenesis during yellow fever based on fatal human cases.Following the infecting bite, yellow fever virus is initially recognized by antigen presenting cells (dendritic cells) and presented in thelymph nodes to the CD4+ lymphocytes, in which the virus replicates and spreads to other organs by viremia. In the liver (and in otherorgans), the immune response, coordinated by those lymphocytes and the liberation of several cytokines (especially TNF-α and INF-γ),will determine the midzonal apoptosis of hepatocytes and several pathologic changes (apoptosis in heart and acute tubular necrosis inthe renal tissues), which culminates with multi-organ failure and hemorrhagic diathesis (vasculopathy). (B) Proposed immune responseand pathophysiology mechanism during yellow fever infection. The coordinated immunologic responses are driven by CD4+ lympho-cytes via MHC II and activate several immune cells and cytokines, especially in the liver. The humoral response produces specificantibodies against yellow fever virus by B lymphocytes, and the cellular response results in apoptosis and hepatocytic necrosis, whichis accompanied by discrete inflammatory infiltrate due mainly to TGF-β action, which is an apoptotic inductor and inhibitor of theinflammatory response and the cytokine storm in the liver
YELLOW FEVER: A MODEL OF INFECTIOUSVASCULOPATHYEndothelial cells occupy a pivotal role in theinflammatory response that occurs during infec-tions of viral, fungal, and bacterial origin. Thisresponse is driven by inflammatory cytokines thatare overproduced and target the microvascularendothelial cells, as has been well-demonstratedin sepsis [52]. These cytokines can induce direct andindirect effects on endothelial cells, which can berelevant for the pathogenesis of septic shock andhemorrhagic events in viral infections such as severedengue and yellow fever [20]. An overproduction ofinflammatory cytokines, such as TNF-α, IL-1, IL-6,IL-8, and IFN-γ, is found during the clinical courseof septicemia and other infectious diseases [53].After the infectious bite, the YFV is initially
recognized by dendritic cells, which act asantigen-presenting cells and carry the virus to lymphnodes. In the lymph nodes, YFV is presented to CD4+ lymphocytes, which trigger an orchestratedimmune response; this response mainly targets theliver, which will be stricken by severe hepatitis(midzonal necrosis/apoptosis), but also targets theheart (myocarditis), the kidneys (acute renal tubularnecrosis), and other organs that are responsible fordisease severity and high case fatality rates observedin severe yellow fever (Figure 5A). However, it isunknown why some people develop severe yellowfever, whereas others present asymptomatic oroligosymptomatic infections [1,7].Hemorrhage is a common clinical manifesta-
tion in patients with classic yellow fever andsevere DHF/DSS. The process and mechanismof this phenomenon are poorly understood, buta key finding is an endothelial cell damage inthe capillary net [54].Several studies on humans and experimental
animal models have described the basic pathogenicevents of these endothelial alterations for severedengue, which resulted in the classification ofhuman infections into dengue fever, DHF/DSS,and other less frequent clinical presentations [55–57,33,58]. For YFV, the events driving the immuneresponse at the endothelial level have yet to be fullydescribed. Additionally, the immune responsesinvolving T lymphocyte activation, cytokine orchemokine production, and complement cascadeactivation are considered important for the patho-genesis of virtually all of the viral hemorrhagicfevers [20].
The production of TNF-α, instead of IFN-γ, byactivated T cells may play a central role in thepathogenesis of hemorrhagic events in severe casesof yellow fever [10]. In contrast, patients with DHF/DSS have subnormal serum levels of complementcomponents C3, C4, andC5 and an increased comple-ment metabolism [33]. These observations demon-strate that both innate immunity and adaptiveimmunity are involved in the severe disease form;however, the key factor that induces hemorrhagicevents in viral hemorrhagic fevers is still unknown.
Additionally, studies on direct and indirectinteractions between endothelial cells and dengue vi-rus have become a central focus in the understandingof the occurrence of hemorrhagic diathesis in thepathogenesis of DHF/DSS. Dengue virus antigenshave been detected in the vascular endothelium inbiopsy tissues from DHF/DSS patients [56]. More-over, the presence of dengue virus in the bloodstreamresults in a severe inflammatory response withendothelial activation and dysfunction, culminatingin alterations of coagulation factors, apoptosis, anddetachment of endothelial cells with disruption oftheir barrier function [56,33].
In the experimental golden hamster model ofyellow fever, it was demonstrated that the infectedendothelial cells of liver vessels are positive for viralantigens. These data indicate that, like dengue virus,YFV demonstrates tropism for endothelial cells[16,17]. These same observations were described forfatal yellow fever cases in humans [15,23]. Midzonallesions can occur during hypoxia following drug-in-duced hepatic injury, and the same alterations in thehepatic microvasculature may contribute to thepathologic picture characteristic of the disease [21].Hepatic tissue includes the unique cell-type knownas sinusoidal endothelial cells. This endothelial cellphenotype is maintained by vascular endothelialgrowth factor, which is released by hepatocytes andstellate cells in the space of Disse and a downstreamautocrine loop through nitric oxide release [59].
Experimental studies of other liver disease modelssuggest that the loss of the differentiated sinusoidalendothelial cell phenotype in capillarization is dueto the loss of hepatocyte-dependent and stellatecell-dependent vascular endothelial growth factorproduction, which consequentially causes a loss ofnitric oxide production by the endothelial cells [59].Sinusoidal endothelium porosity and the lack of anorganized basement membrane are important foroxygen diffusion to hepatocytes. In almost all yellow
fever liver cells, an intense vascular damage has beendescribed especially in fatal cases; interestingly, nosimilar lesions are described in the controls of studiesby Quaresma and colleagues [10,11,15,23,22]. Theseauthors reported an association between biochemicalchanges in the hepatocytes in all acini areas, and theoutcome to this cellular injury is hypoxia, particularlyin the midzone, a well-known liver area with thepoorest blood supply [15,22]. Nonetheless, the roleof this characteristic pattern in the immunopathologyof yellow fever remains to be defined. The highTNF-α and IFN-γ expression levels in the livertissue of human fatal yellow fever cases mayindicate the participation of these cytokines inthe endothelial damage induced by yellow fever.Finally, it is important to highlight the role ofdirect YFV action in the liver cells supported byintense labeling of cell by YFV antigens in fatalcases and experimental studies, especially onhamsters [1,15,23,22]. Figure 5B shows the im-mune response and pathophysiology mechanismsin the liver during yellow fever infection. Indeed,the CD4+ lymphocytes via MHC II coordinatedthe responses of several immune cells andcytokines.
FUTURE CHALLENGES IN OBTAINING ATHERAPEUTIC APPROACH FORYELLOW FEVERMolecular biologic techniques have been used toobtain an experimental treatment of hepatitis Cand HIV. The objectives of these approacheswere to inhibit genes related to the pathogenesisof these viral agents [60]. Therefore, consideringthe recent data obtained in the immunopathol-ogy of yellow fever, an interesting approachwould be to inhibit the critical events involvedin yellow fever pathogenesis, such as caspase 3inhibition in the liver, to determine and quantifythe influence of necrosis and apoptosis oninduced liver failure observed during the courseof severe yellow fever. Additionally, immuno-logic procedures that inhibit pro-apoptotic andantiviral responses of the immune system to aYFV infection should be considered a possibletherapeutic target for managing the disease,especially techniques that promote the controlof the cytokine network responses with theobjective of generating cytokine profiles with
more efficient antiviral actions and that minimizeliver tissue injury. Finally, the inhibition of TGF-β andTNF-α effects and the blockade of the apoptosiscascade in the course of yellow fever represent aninteresting possibility for obtaining an immunologictherapeutic approach that can confer innumerablebenefits for patients, mainly in severe clinical sce-narios including the serious viscerotropic diseaseinduced by the 17D yellow fever vaccine.
CONCLUSIONIn this review, we have addressed the mostimportant immunopathological findings that occurduring the course of severe yellow fever. Wedescribe the immune response and pathophysio-logical mechanisms that occur in severe yellowfever to explain the pathogenesis of the disease.Immediately after the infecting bite, the virus isdetected in subcutaneous tissues by dendritic cellsthat carry the virus to lymph nodes. There, YFVreplicates and is presented to CD4+ lymphocytes,which coordinate both the innate and adaptiveimmunologic responses. After replication, YFV isreleased by cells in the lymphatic ducts and theninto the blood vessels (viremia). Following viremia,YFV, using its tropism, infects several organsincluding the heart, thymus, kidney, and liver. Inthe liver, Kupffer cells and hepatocytes becomeinfected; especially apoptosis but also lytic necrosisis responsible for many of the clinical symptoms.Significant lesions, which are mainly apoptotic,are observed in other tissues. Three importantcytokines are associated and released, TNF-α, IFN-γ,and TGF-β, due to these apoptotic conditions. Thesecytokines, acting under the coordination of CD4+cells and CD8+ cytotoxic lymphocytes, are related toapoptosis and several of the symptoms observed insevere yellow fever, such as vascular leak syndrome,thrombocytopenia, altered aspartate aminotransfer-ase, alanine aminotransferase, blood urea nitrogenand creatinine, icterus, vomiting, and hemorrhagicdiathesis.
CONFLICT OF INTERESTThe authors have no competing interest
ACKNOWLEDGEMENTSThe authors thank Thomas Monath and RobertTesh for their critical review of the manuscript. Thiswork had the financial support of the National
Institute for Science and Technology for ViralHemorrhagic Fevers by National Council for Scien-tific and Technological Development (CNPq)/Coordenação de Aperfeiçoamento de Pessoal deNivel Superior/Fundação de Amparo à Pesquisa
do Estado do Pará grant number 573739/2008-0 and CNPq 301641/2010-2. The authors discloseany role of the sponsors on the study design,analysis, interpretation, writing, and the decisionto submit the paper for publication.