The hepatitis B virus and the host response

Sa3

The hepatitis B virus and the host response

The clearance of hepatitis B virus infected cells from the liver is probably dependent on an interplay between the interferon system and the cellular limb of the host immune response. Although the importance of tbe nucleocapsid proteins as targets for sensitized cytotoxic T cells is estabkkd in chronic hepatitis B infection, tinther studies are txeded during the early phase of acute infstian. Tbe relative importance of pre-S sequen~s as inducers and targets of the virus neutralizing bumoral immune response is bernming euablisbed but their precise place will await t’be devclopmeot of in vitro models of hepadna virus infection and precise deiinition of the mechanism of viral uptake. In adult life, deficient pmduction of ainterferon and suppression of the ability of the host to respond to interferon are probably important factors underlying the development of cbmnic infection. In the neonate, however, specific suppression of Ihe cell-mediated immutte ~espcmse may be involved. Tbe presence of a tmttatxxt in the pre-core region of some virus isolates has recently been described. Hepatocytes infected with this virus cannot produce hepatitis B e antigen and the course of the Liver disease is fairly rapid. Whether tbis mutant causes liver damage in the same way as the wild virus or is directly cytopatbtc needs farther study. In adult-acquired chronic hepatitis B virus infection, a-interferon produces hepatitis B e antigen clearance in 26-W% of cases and is followed by resolution of the hepatic inflammation. Results in nwnataJly squired infection are less impressive and prednisolone priming followed by interferon may be needed. The value of interferon in the treatment of patients infected with the preare

mutant remains to be established.

ItthwJadkltt Recovery from hepatitis 13virus (HBV) infection is dependent on the integrated activity of the patient’s interferon (EN) and immune system. Chronic HBV infections arise because of defects in these defaces.

HBV is not directly cytopatbic and the liver damage is caused by immune lysis of infected bepatcqtes, an essential part of the recovery process. Analysis of the intlam-

matmy infiltrate demlmslrateP the presence of natural killer and cytotoxic T cells (1). Viral antigens are seen cm the surface of the bepatoeyte (2) and these, in asscciaticm with the class I MHC proteins, make the cell a target for cytotoxic T cell lysis (3). Studies ia pnients with cbmnic HBV infection suggest that the nucleocapsid antigens (hepatitis B core antigen (HBcAg) and hepatitis B e antigen (HBeAg), a cleavage product of HBcAg) are an impottant target (4,5). Hepatocytes usually express very little MHC class I &copmtein (6) but in the early stage of acute HBV infect@ following the production of a-EN, MHC expression on beqatocyta increase.9 and, coincidentally, tr msamfnaxs rise (Fig. I) (7). Recent okr-

H.C. THOMAS

HBsAg Anti-t-R Anti-HBs

HlA-1 in

10

liver 5

3a

Serum interferon

,5 Q3

-IL--

4

0

8

16

24

32

Fir. 3. The large, middle and major proteins of the envelope of HBV have the same carboxyl-terminalsequence. (Reproduced by Wd prcmieionfrom Tiohis P, et al. In: Viral Hepatitis and Liver Di case. New York: ,4an R Liss Inc, 1988; 295-300.)

Weeks Wg. 1. AcuteHiBV infection in the chimpanzee. Markers of virus infection appear early, coinciding with a detectable pulse of tirculating This is folfowed by enhanced display of MHC class I antigen on hepatocytes. Liver damage, as indicated by rising strum ASTIevek, coincides with the first appearance of host humoral immunity to HBV (anti-HBc). AST = aspartate aminotransferase (SOT). adFN.

vations (personal) indicate that the X gene product may activate cellular genes, including those for MHC class I

proteins. the 2-5A

origoadenylace sjmthetaselendonucleasu systo inhibition of viral protein synthesis (Fig. 2). This would be expected to produce an antiviral

tern ‘, which leads

Enhancement nl HI-A-1 syntkx

2. Effect of 1FNs on the virus-infected cell. a- and/%lFNs acting through a common mtor and )“IFN acting through a separate receptor activate (+) 2.5A oligoadenylate synthetase. This catalyses production of oligoadenylares, which activate endogenaus tibonucleasl: leading to do3vage of viral RNA. IFN also cwses enhanced (4-1 exprtssian of MHC (WA) da= I proteins on the hepatqte surface, facilitating recognition of virus-infected ~41s by th+zceeliular immune mechanisms of ihe host. HBe and WBc ant&s are the target of this component uI the host immune response. Tc, Ts. TH: cyUcrtoxic.supprcaor and helper T cells. (+) and (-) denote enhancemen1 dnd suppression, respectWy.

m

state in uninfected regenerating liver cells, preventing infection during the lysis of infected hepatwytes. The virus neutrahzing

antibodies

are directed

re-

to epi-

tapes on the envelope of the virus, This is commd of three polypeptides, each with the sanw carboxy (Fig. 3); these xc designated the large, middle and major envelop proteins. The amino-terminal 120 amin aEid region of the large protein, which is not present in the middle and major proteins, is designated pr&I. This by-

tmims

that t$ egion of 21-47 amino acids is capable of &ding Q to the ‘i~.r.,mbrans of the hqmtoqte (8) and is, tlwrefore, probably involved in virus uptake. The pre-S2 region of 55 amino acids, present in the large and middle proteins, binds &merized human serum albumin (pHSA) (9) and, &awe this protein also binds selectively to the hep atocytej of species that are susceptible to HBV infection and not to non-susceptible species (IO), it has been proposed that pHSA acts as a bridge between the pre-S2 regionof the viralenvelqx and the hepatocyte, facilitating uptake and infection of the cell, MOE recent observations showing that pHSA is not present in serum and that native albumin in physiological concentrations t!ocks the bind= ingof pHSA to hepatitis B surface antigen (HBsAg) (11) make this hypothetical uptake system unlikely to occur in viva in man. The antibodies found in cowabsent serum bind predominantly to the epitopes of the HBs gene-encoded region (X2), the carbxykrminal region of which is prent in aU the eov&pe

proteins, Two hydrophilic regions of amino acids 124-147 (Fig. 4). form of intramoIecular disulphide bridges and are the binding region for the majority (NO%) of

this polypeptidc, loops

the=

becaux

ma1 subjects immunized with plasma-derived and recombinant DNA-produced HB~vaccine (14). Although patients who have recovered horn HBV infection or have been vaccinated with HBsAg are ptotected from further iofetiion. there has teen considerable debate on whether the middle pre-S2-bearing and large pre-Sl-bearing plypeptides should be added to existing vnscines. During natoral infection, antiindies to preS1 and pre-S2 regions appear before antibodies to the HBs region (15,161 and if pre-Sl is the region of the envelope of the virus that binds to the hepatocyte during iofection, antibodies to this region might be virus neutralizing; inclosion of the pre-S region io a vaccine, therefore. might be desirable. Forthemtote, 15% of “omtal individualsdo not teipond

to HBs small pulypeptide

and, by recruiting addi-

tional T-helper ceik. btcbtsion of pte-S2 facilitates the develooment of amibodv to the HBs eoitooes 1171. More redent data indicate that the io&si”” HBsAe further enhances the humoral epit”p& (18) (Fig. 5). the actigen-bblding

lapachy of c”nv&scent

of HBcAg with resoonse to HBs

seta. using

mott”cl”ttal tmtib”dii binding to tbw regiottt., it has bee” possible to show that administration to cbiipatuees of antibody to the regkm X24-136 prevents infection (13). Ardibodies to this region, as well as to other epitopes 0” the S geoe-etwded polypeptide, ate present in the xmtn of patients coovale~ceot from HBV infection and in nor-

cilrcttk

lbepatilis B htfm

It is likely that there are maoy different teasoos why patients develop the chmoic can&r state of HBV, which exists in several forms. Chronic HBe onrigenwnicpmiem

This is the first phase of chronic infection and may arise after exposure at any stage of life. (i) Chronic HLW infewionfollowing erpmure in neonatd life. Ninety-five per cent of babies born to HBeAg

H.C. THOMAS

Macharism

Mabmal blowI

passesacross the placentaand maysupprtissthe deve?op mentof the cellular immune response to the nucleocapsid proteins which are the target during immuneclearanceof infected hepatocytes (Fig. 7). (it,? Fulminan~ heparitir following exposure in nemaid life_ Rarely, children bum to HBsAg/anti-HBe @tive mothers develop fulminant hepatitis. The mechanism is, once again, unknown. One hypothesis Is ?hae maternal anti-H& and anti-H& pass across the placenta and initially mod&e the lysis of infectedhepatocyks by the cell-mediated immune response of the child(5); the virus

F&. 7. Neonatal HBV infection: postulated m&an& of viral persistence. ir is propased that HBeAg crose5 the placenta from the maternal blood and induces tolerance to H&kg, which is OIE of the

targetsof the cellular immuneresponse. Maternal IgG anti-HBc also ccusses the pIacenta into the neonatal circulation. Persistence of Hl3V infected cells in the foetal liver ia fwilitatd as maternal IgG blocks recognition of virus-infectedcells by cytotoxic T cells (43, Early exposure to soluble virus protein (HBeAgj may induct a state of antigenic to!erancc to the virus with spedfic suppressor cells in-

hibiting the host defense mechanism.

positive mothers become infected,

and over

95% of these

(20). HBeAg, a 100~ molecular mass (13-15 kDa) soIuble protein, derived km the pre-core/NBC polypqtide by cleavage of both the amino- and carboxyl-terminal ends,

Hepatocyteo supporting HBV replication delay immune lysis by suppression of HlA-I HqatOCyteS

cantaining Integrated HBV do not express nucleocapsid proteins

HBeAglAb conversion

. m .. m m

birth, only 5-10% of subjects infeckd after 9s perid develop chronic infection (Fig. 6). There are probably several different reasons for the development of chronic infection at this stage of life. One defect, which has receutiy been dawmented,

develop a chronic carrier state (Fig. 6) (19). These infants probably receive a large i~~oc~&rn of viws, buth from maternal blood More or during birth, and from close contact with secr&ons soon after birth. The reason for these infants failing to clear the virus is unknown but likely to relate to the immaturity of the neonata! immune system

HBeAg -we

would thereforespread throughoutthe Iiverand, as the matemai antibdies disappearat 3-6 months, cytotoxic T cells sensitizedto HEeAg andHBcAgwould thenrapidly destroythe infectedcells,resultingin hepa-tic failure. (iii) Chronic HB V infectiunfduwing exposure ufter the neonara/period, In marked contrast to the situation at

is the production of subnormal

quantities of a-1FN by peripheral bloorl mononuclear cells @l-24). These inditiduls atsu have evidence of at+ normal activation of their hepatoqtes by IFN: levels of hepatic Z-SA synthetase are only minimally elevated (2S) and MHC class I proteins are present in very Iow ckn&y on infected hepatocytes (26), whereasIeveIs of MA synthetase and MHC class I displayare markedly increased in peripheral blood lymphocytes of chronic HBV tiers (27). These data suggest activationof peripheralblood

Anti-H BeAb tve Rearrangement of integrated HBY causes tray3formaPon

l

l

HBV particles; Gradual clearance of hepatwytes supporting HBV replication

l

membrane HEc autigcn; nr, membrane HBc; membrane MHC class I protein;

Cells containing integra& HBV proliferate

\

-NMHr101

/

?I& 8. During the HBtAg positive pbas (Mt side), hepatocyttscontainingHBV show mpprcsion of MHC class I (HLA-1) antigen displayand awid T octl lysis. During the anti-We phase (right tide), hepatocytes containing integrated HBV sequences avoidT cell lysis because they dOnot express nuckcapsid

proteins. HCC: hepatocellular cticoma.

HBV AN:, THE IIOST RESPONSE lymphocytes liut not of infected hepatmytes, and raise the possibility that HBV within the hepatoqte has setetiively *switched off the responsiveness of the hepatocyte to IFNs. In vitro transfection of HBV genomes into cells in tissue culture makes them specifically non-responsive to IFN: they remain sumptible to lysis by Sindbis virus and MI-K induction by IFN daes not occur (28). Thus, during I-IBV reptication, it is hgpothesieed that T ~11s cannot 1~ inkcted 4s &cause the virus suppresses the expression of MHC dass I pm&ins (Fig. 8). (iv) HBeAghti-HBe scroconversion. hype= thesized that, as the level of HBV replication within hepatocytes falls, the suppression of MHC protein display

It is

the

is reduced and infected cells are Iysed by cytotoxic T cells probably sensitized to nucleocapsid epitopes; the result is lobular hepatitis. Hepatocytes HBV sequences do not express teins bea* the preferred site genome is witbin the regulatory tig for the protein (2!3), These immune tysis (I?%_8)_ (v)

Pre-core mutation stq~ tt~ x3cretM al HBe an@en and albws transladan d prz-ccw @aa) and core (183aa) peptides

ppec-

mj

MG

II

--19 Tmnslation of the pre-umhrc 0

1W

.-p19

-virus

b%mmipt may snrt at either AUG sites. Wka trazrslationSW nt ibe 6mt AUG, the pre-

that contain integrated MLV nucleocapsid proof integration ia the viral region of rhe gene encodcells will -fore avoid

Anti-HBe positive patien fi HrirhIU) evidence ofHBV

replicatim. The sera of these patients do not contain virus par?icLs detectable by dot/blot hybritition. However, in some cases, low IeveIs of viraemia can !x detectsd by the polyrnerase chain reactti. HBsAg in these patients is ened by iategrated HBV DNA sequences. Ultimately> it is argued that, during cell division, the integrated sequences rearrange (30) and cause malignant transformation of tie hcpatocyte. Clinically evident hcpato&lular carcinoma then appears within mouths and the patients qidly die. (vi) Ckmnic anG?Be positive patients with’ pemistens HBV rcpliccuion (pr+core HBV mutant). An HBV mutant, with a Point mutation generating a novel stop codon at tk end of the preare sequence, has been described (31). Patients infected with this virus have signifimt viraemia but cannot secrete HBeAg, which is &rived from tht prc~rckore translation pruduti (Fig* 91, This nudeoc;H@d prottin is not produced in the presence cf the stop CO&I but translation does occur from the second AUG at the tqinning of HJ3c. This 19 klla HBc is all that is required for virion nucl~apsid and virus partick production. These patients develop rapidly progressive I&r disease and are common in the Mediterranean basin (32).

Mechhanism ofactionof intqferon The mechanism of action of IFN in inducing HBeAg/ anti-HIle sf5roconversion is unknown. Several changes have ken noted both in the immune system and in the infected hepatme (Fig- 10). An increase in MHC class I

protein

display (7) and an alteration in viral protein ex-

pression (33) in the be@ocyte occur within 24 b of the start of IFN therapy. An increase in the he~per/cytMoxic suppressor cell r& mrs at 6-8 weeks, whta hepatitis associated with sercxonversion starts, and is not seen in those that do not req~M. Changes in humoral immunity also occw. Patients who respond to IFN therapy either have IgM anti-H& present before treatment or develop 2 during the course of tberapy (34). Changes in cell-mediated immunity may a& occur but have not yet been documented. with WReAg/ Inhibition of viral re#&on,

cz?Y<ed

ltdliiof HsVBpMdM

H.C. THOMAS

s88

ever, inflammatory liver disease will ameliorate in these patients if long-term inhibition of HBV replicaticln is

achieved, Several investigations have shown improvement in alanine aminotransferase levels following clearance of fIBeAg and HBV DNA tirn the strum. A review of 50 treated and 25 untreated antrol patients has shown a reTmn

duction in inflammatory necrosis of liver cek anda diminished rate of development of cirrhu;is in patients respond-

(days)

ing to therapy

b

with loss of markersof HW replication

(35). Patient selection Inf&ivity and progressive liver disease are the major reasons for attempting therapy with eitheradeninearabinosidemonophosphate(ARA-AMP)or the a-IFNs. Var-

F&. 11, (a) Complete response: inhibition of HBV replication and clearance of HBsAg. {b) Incomplete response: inhibition of HBV replication associated with a transient exacerbation and then nclrmalizationof the liver function tests. Note that the patient re. mains HBsAg positive, presumably because of the presence of integrated HBV sequences. Benefits to the patient include reduced infectivity and cessation of inflammatory liver disease.

anti-HBe seroconversion, is usually long lasting and accompanied, after a transient exacerbation, by amelioration of the inflammatory liver disease (35) (Fig. 10). Reactivation may occur, however, particularly in kmo-

ious ran&x&d, controlkdstudiesusingeitherrecombinant or-IFN(X,37) or lymphoblastaidIFN (38~401, in doxs of 10 tillion unitsthriceweekly for 6 months,have producedHBeAg clearancein 2688% of cases, si@cantlymoreoftenthanoccurredspontaneouslyin the amrrol group. The respom rates were very variabk becauseof the heterqeneity of the patient populations with regard to severityof liverdiseaseat the start of therapy. Additional analy& of some studios shows that Asian carriers, presumably infected from birth, do not respond to therapy with IR: alone (41). In these patients, however, prednkolone withdrawal followed by ARA-AMP or QIFN may improve seroconvtion rates (42). Currently, IFNs are the satisfactory therapy,but

mast

sexuaipatients, with or without human immunodeficiency virus infection (36). Patients treated early in the COUTS~of the disease will lose HBsAg from their serum if cessation of HBV replication is achieved (Fig, Ila), If integration of HBV sequences into the host cell genome has occurred before starting treatment, HBs antigenaemia will continue after

not all patientsrespondand the side effects of IFN are substantial.For this reason,both clinicaland serological featuresthat predicta greaterchanceof response have been sought;these includemarkedinflammatory actitity~ high serumtransaminaselevels and, more recently,the presenceof IgManti-HBc(34) and low serum HBV DNA levels (37). Whether IFN will be of value in the treatment of chronic infectionwith the pre-core mutant HB virus re-

cessation of detectable HBV replication (Fig. 1 Lb), How-

mains to be determined.

References 1 EgginkHF,Ho~thoffJH,HuitemaS,GipsCH,Poppem~S.Cellular and humoral immune reactions in chronic active liver disease. Lymphocyte subsets in liver biopsies of patients with unlreated idiopathic auto-immune hepatitis, chronic active hep at& B and primary biliary cirrhosis. Clin Exp Immunol 1982; 50: 17-24. 2 Gudat F. Bianchi L, Sonnab& W. Pattwnof core and sur-

et4,

face expression in liver tissue refkctsstate of immune response in hepatitis. J Lab Invest 1975; 32: i-9. 3 Doherly PC, Zinkernqel RM. A biological role for major histocompatibilitv antigen. Lancet 1975: i: 140%9. 4 Eddleston AWLF, Mondelli M, Mieli-Vergani 43, Williams R. Lymphocyte cytotoxicity to autologous hcpatwytcs in chronic hepatitis B virus infection. Hepatology lN2; 2:122%X. 5 pignatelli M. Waters J, Lever AML, et al. Cytotoxic T-cell responses to the nucleocapsid proreinsof HBV in chronic hepatitis.

the