JOURNAL OF CLINICAL MICROBIOLOGY, Apr. 1985, p. 593-5980095-1137/85/040593-06$02.00/0Copyright C 1985, American Society for Microbiology

Quantitation of Hepatitis B Virus (HBV) Core Antigen in Serum inthe Presence of Antibodies to HBV Core Antigen: Comparison with

Assays of Serum HBV DNA, DNA Polymerase, and HBV e

AntigenREINHARD BREDEHORST,' HINRIK VON WULFFEN,1* AND CELSO GRANATO2

Institute fur Medizinische Mikrobiologie und Immunologie, Universitat Hamburg, Martinistrasse 52, 2000 Hamburg 20,Federal Republic of Germany,l and Clînica de Moléstias do Hospital das Clinicas da Faculdade de Medicina da

Universidade de Sâo Paulo, CP 8191, Sao Paulo, Brazil2

Received 6 September 1984/Accepted 12 December 1984

A quantitation procedure for hepatitis B core antigen (HBcAg) in serum without prior removal of antibodiesto HBcAg is described. The virus nucleoprotein core was released from hepatitis B virus (HBV) particles bytreatment with Nonidet P-40 detergent and allowed to form immune complexes with homologous antibodies toHBcAg present in the sera of HBV-infected individuals. After precipitation with 2.0% polyethyleneglycol-1.5% Tween 20, the HBcAg immune complexes were dissociated by treatment with 3 M KSCN and thenadsorbed onto polystyrene beads in the presence of the SCN- ions. Thereby, HBcAg and antibodies to HBcAgwere linked independently of each other to the matrix, and the core antigen could be quantitated directly byincubation of the beads with 2-51-labeled anti-HBc. Even in the presence of an excess of antibodies to HBcAgin the polyethylene glycol precipitates, HBcAg could be detected without appreciably affecting the sensitivity.The assay proved to be specific for core determinants and exhibited excellent reproducibility. The applicationof the HBcAg assay in 185 hepatitis B e antigen-positive sera revealed HBc antigenemia in 99% of the sera

containing hepatitis B e antigen at titers of .1:256 and 43% of the sera with lower hepatitis B e antigen levels.However, only in 6 of the 34 HBcAg-negative sera could HBV DNA be detected by blot hybridization. Whencorrelated with HBV-associated DNA polymerase (DNAP) activity, HBc antigenemia was found in allDNAP-positive sera (n = 95) and in 39% of the hepatitis B e antigen-positive sera without detectable DNAPactivity (n = 44). Of the DNAP-negative sera with HBc antigenemia, 94% contained HBV DNA, whereas in theabsence of HBcAg, HBV DNA could be detected only in 3 of 27 DNAP-negative sera. With regard to sensitivity,the HBcAg assay appeared to be less sensitive than the hybridization technique, but more sensitive than theDNAP assay.

The hepatitis B virus (HBV), also referred to as Daneparticle, consists of an outer lipoprotein coat (hepatitis Bsurface antigen [HBsAg]) and an internal core containinghepatitis B core antigen (HBcAg) and hepatitis B e antigen(HBeAg) (15, 23). In serum HBcAg is also found as smallparticles (22-nm spheres and filamentous particles 22 nm incross-section and of varying length); HBeAg is found asfree, soluble protein.Although HBsAg and soluble HBeAg can now be speci-

fied via radioimmunoassay (RIA) routinely by commercialkits, determinations of serum HBcAg are still difficult ortime consuming. Consequently, these methods cannot beregarded as suitable for the routine screening of sera. Thereis, however, a need for a simple and rapid serological test,since a quantitative determination of HBcAg in serum mayprovide clinically useful information in the absence of liverbiopsies. The development of chronic liver disease in HBsAgcarriers appears to be linked to the persistent synthesis ofthe HBcAg (1, 5, 18), and HBc antigenemia has been shownto correlate positively with HBcAg in the liver (19). Further-more, serum HBcAg has been described as a marker ofproductive HBV infection that is more sensitive and specificthan HBV-associated DNA polymerase activity (DNAP; 19)and more reliable with respect to infectivity than HBeAg(21).

* Corresponding author.

In this paper, a simple and reliable technique with highsensitivity is described to detect HBcAg in serum. Theprocedure is based on the observation of Neurath et al. (14)that proteins including HBcAg are attached to polystyreneeven when dissolved in chaotropic ion solutions of highmolarity, thereby allowing the detection of core antigen inthe presence of homologous antibodies to HBcAg. Themethod was applied to the analysis of serum samples fromHBeAg-positive patients and compared with assays ofHBeAg, DNAP, and serum HBV DNA by blot hybridiza-tion.

MATERIALS AND METHODS

HBeAg, anti-HBe, anti-HBc, and anti-HBs were tested bycommercial RIAs (ABBOTT-HBe, CORAB, and AUSAB;Abbott Laboratories, North Chicago, 111.). HBsAg was de-termined either by an enzyme immunoassay (AUSZYME Il;Abbott Laboratories) or by quantitative immunoelectropho-resis on agarose gel containing antibody, as described byLaurell (4, 11). Determinations of serum HBV DNA byhybridization with 32P-labeled cloned HBV DNA were kindlyperformed by W. H. Gerlich (Hygiene-Institut, Gottingen,Federal Republic of Germany). Briefly, serum samples were

digested with proteinase K and extracted with phenol andether, and the aqueous extract was passed through a mem-

brane filter (Gene Screen; New England Nuclear Corp.,Boston, Mass.). The membrane with the adsorbed DNAsamples was hybridized to 32P-labeled cloned HBV DNA.

593

Vol. 21, No. 4

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from

594 BREDEHORST, VON WULFFEN, AND GRANATO

onti HBc

Dane Porticle(f yHBsAg

Y ~~~ HBcAgcore1. Release of core particles by

Y I NP4O NPLO treatment\ NP 40

2. Formation of immune complexesI \ (ICs)

3. PEG-precipitationSCN- 4. Dissociation of ICs by

3M thiocyanate

SCN- 5. Adsorption under dissociatingj conditions

6. Determination of HBc Ag with1251-anti HBc

FIG. 1. Schematic outline ofthe standard quantitation procedureof HBcAg in serum.

Binding of DNA was detected by autoradiography. Thedetection limit was 10' genomes per ml (W. H. Gerlich andE. Zyzik, personal communication). Protein was quantitatedby the procedure of Lowry et al. (12).DNAP assay. DNAP activity was determined in gradient

fractions and pellet concentrates of clarified sera by themethod of Kaplan et al. (10). Pellet concentrates of sera(x20 by volume) were prepared by centrifugation (4 h at50,000 r.p.m.; Spinco SW-50.1 rotor; 4°C) through a discon-tinuous 10 to 30% (wt/wt) sucrose gradient in 20 mMTris-hydrochloride-130 mM NaCI (pH 7.4). DNAP resultswere expressed in picomoles of [3H]thymidine x 102/ml oftest sample in 5 h. As a control, DNAP activity was tested inparallel in the presence of 1% ethidium bromide (see refer-ence 7), and the assay was considered to be positive whenthe amount of incorporated radioactivity in the standardassay exceeded that of the ethidium bromide-inhibited reac-tion by a factor of at least 2.1.

Standard quantitation procedure of HBcAg in serum. Se-rum volumes of 400 ,ul (if Iess, then mixed with poolednormal human serum [NHS] to give a total volume of 400 pd)were incubated at 37°C for 60 min after the addition of 10 ,ulof phosphate-buffered saline [PBS] (pH 7.2) containing 5%(vol/vol) Nonidet P-40 (Particle Data Laboratories, Ltd.,Hahn St. Elmhurst) and 0.5% (vol/vol) 2-mercaptoethanol,diluted with 3.6 ml of PBS (pH 7.2), and incubated again for1 h at 37°C (Fig. 1). As described by Siersted et al. (22), themixture containing HBcAg immune complexes (ICs) wascentrifuged at 1,700 x g for 20 min without refrigeration toremove grossly aggregated material. The supernatant phasewas mixed with an equal volume of ice-cold PBS (pH 7.2)containing 40 g of polyethylene glycol-6000 (PEG-6000) and30 g of Tween 20 per liter. After incubation in an ice bath(2°C) overnight, the precipitate was centrifuged at 1,700 x g(20 min; 4°C), and the supernatant phase was removed bygentle aspiration with a Pasteur pipette connected to a filterpump. The precipitate was washed twice with 3 ml each ofPBS (pH 7.2) containing 25 g of PEG-6000 and 15 g of Tween

20 per liter, with mixing between the centrifugation steps(1,700 X g; 20 min; 4°C).The final PEG precipitate was dissolved in 220 ,ul of 3 M

KSCN (in distilled water) and incubated at 37°C for 30 min;200 pl was added to 0.6-cm-diameter polystyrene balls(Precision Plastic Ball Co., Chicago, Ill.). After adsorptionovernight at room temperature (with gentle shaking), thecoated polystyrene beads were washed four times withdistilled water (Pentawash equipment; Abbott Laboratories)and then incubated at 37°C for 3 h with 200 ,ul of '25I-labeledanti-HBc containing 100 ,ul of the reagent of CORAB (6 x105 to 7 x 105 cpm) and 100 ,ul of PBS (pH 7.2) with 4%(vol/vol) pooled NHS and 4% (vol/vol) fetal calf serum.Prolongation of incubation time resulted in only a minimalincrease of bound 1251-labeled anti-HBc. After eight washeswith distilled water, the balls were assayed in a gammacounter. All determinations were corrected for the relativelysmall nonspecific binding of 1251-labeled anti-HBc to poly-styrene balls coated with PEG precipitates from 400-plsamples of pooled NHS processed in parallel to the HBcAg-positive sample. When results are expressed as S/N ratios, Sis the counts per minute of a patient sample, and N is theaverage of counts per minute from three to five healthycontrols.

Effect of excess anti-HBc on the detection of HBcAg. HBVparticles, concentrated from 12 ml of serum from a patientwith chronic type B hepatitis and separated from anti-HBcas described by Hess et al. (6), were suspended in 200 pul of20 mM Tris-hydrochloride-130 mM NaCi (pH 7.4) andincubated for 60 min at 37°C after the addition of 2.5 pil of10% (vol/vol) Nonidet P-40 and 2.5 pl of 1% (vol/vol)2-mercaptoethanol. Ten-microliter samples of this reactionmixture were then mixed with various volumes of pooledNHS or anti-HBc serum (titer by RIA, 1:105; HBcAgnegative), made up to a final volume of 100 ,ul with PBS (pH7.2), and incubated again for 45 min at 37°C. After theaddition of 100 pl of 6 M KSCN, the samples were processedaccording to the standard quantitation procedure.

Preparation of 1251-labeled HBcAg ICs. HBV particles,concentrated from 2 ml of serum as described for the DNAPassay, were mixed with 400 pl of pooled NHS and 200 pul ofI251-labeled anti-HBc immunoglobulin G antibodies (1.25 x106 cpm/27.6 p.g of immunoglobulin G; Abbott Laboratories)which had been purified by DEAE-Sephacel chromatogra-phy as described by Mollison (13). The formation of HBcAgICs and subsequent PEG precipitation followed the standardquantitation procedure. The precipitate was analyzed forradioactivity, and results were corrected for the small amountof contaminating noncomplexed '251-labeled anti-HBc asdetermined by parallel control experiments without HBVparticles.

RESULTSDetection of HBcAg in the presence of anti-HBc. When

released from circulating HBV particles by treatment withdetergent, HBcAg is sequestered within ICs by antibodies toHBcAg present in the sera of HBV-infected individuals.Efficient purification of these HBcAg ICs was achieved by asingle-step precipitation with 2.0% PEG-1.5% Tween 20 asdescribed by Siersted et al. (22). The final pellets containedonly 0.1 to 0.2% of the protein originally present in theserum and less than 0.1% of anti-HBc as determined by1251-labeled anti-HBc antibodies added to serum samplesbefore PEG precipitation. The yield of precipitated 1251-la-beled HBcAg ICs prepared as described above was approx-imately 40% under these conditions.

J. CLIN. MICROBIOL.

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from

QUANTITATION OF HBcAg IN SERUM 595

Dissociation ofPEG-precipitated HBcAg ICs was achievedby treatment with thiocyanate, which proved to be mosteffective at concentrations of 2 and 3 M. Since proteins,including HBcAg, are adsorbed onto various matrices, evenwhen dissolved in chaotropic ion solutions of high molarity(14), the dissociated antigen and antibody components wereattached to polystyrene beads in the presence of the thio-cyanate anions. Thereby, HBcAg was linked independentlyfrom anti-HBc to the solid support and could be quantitateddirectly with the aid of '25I-labeled anti-HBc. HBcAg wasdetectable by this technique, even in the presence of anexcess of anti-HBc (Table 1). Core particles released frompurified HBV particles were mixed with increasing volumesof either NHS or serum containing anti-HBc at a titer of1:105 (HBcAg negative) and adsorbed onto polystyrene ballsin the presence of thiocyanate. Although the inhibitory effectof the anti-HBc-containing serum exceeded that of NHS, thedifference was less than a factor of 4.

Therefore, the decrease of sensitivity was primarily due tothe presence of protein contaminants competing with thecore antigen for binding sites on the polystyrene beadsrather than to the presence of anti-HBc. Furthermore, by thePEG precipitation method used, less than 0.1% of anti-HBcoriginally present in the serum was co-precipitated, corre-sponding to an amount of antibodies in 0.4 ,ul of serum when400-,li serum volumes were used for the PEG precipitation.This quantity of anti-HBc serum, however, did not interferewith the HBcAg quantitation (Table 1), even when the serumcontained anti-HBc at a titer of 1:105.With the aid of a standard serum containing a defined

number ofHBV particles, the assay may be used to estimatethe number of circulating HBV particles per milliliter ofserum. Linear dose-response curves with identical slopeswere obtained when the yield of HBcAg activity in seriallydiluted sera of four HBV-infected individuals was plottedagainst the corresponding serum dilution (Fig. 2).To assess the sensitivity of the quantitation procedure,

HBV particles from the sera of six patients with chronichepatitis B were concentrated and separated from anti-HBcas described by Hess et al. (6). After being subjected to anextensive 8-h endogenous DNA polymerase reaction, theDNA-containing HBcAg particles were separated fromempty ones by cesium chloride density gradient centrifuga-tion as previously described (6) or by ultracentrifugation inmetrizamide density gradients as reported by Takahashi etal. (24). Fractions containing the complete HBcAg particles(density in metrizamide, 1.20 to 1.23 g/cm3; density in CsCI,1.37 to 1.35 g/cm3) were then analyzed for trichloroacetic

TABLE 1. Effect of excess anti-HBc on the detection of HBcAgResults of HBcAg detec-

Material Vol Amt of tion after adsorption ontoadded to(g ) protein (kpg) polystyrene

cpm S/N %

None 10,865 24.1 100

NHS 0.31 21 11,625 25.8 1071.25 86 13,625 30.2 1255.0 345 10,305 22.9 95

20.0 1,380 6,175 13.7 57

Anti-HBc 0.31 23 11,081 24.6 102serum 1.25 91 9,945 22.1 92

5.0 365 7,315 16.3 6720.0 1,460 1,955 4.3 18

10

9'

E~~~0~~~

c-m 0~~~

Cut>oo

1:.2 1:8 13 1128Dilution ofseruC

FIG. 2. Detection of HBcAg in serum serially diluted with NHS.Samples of sera (400 plI each) from four patients with chronichepatitis B (O, *, A, *) and from two healthy donors (*, A) wereserially diluted with pooled NHS and assayed for HBcAg induplicate according to the standard quantitation procedure. Thecutoff level represents an S/N ratio of 2.1.

acid-precipitable 3H radioactivity (6, 24) and HBcAg. As-suming that the viral DNA polymerase completes, on theaverage, single-stranded gaps constituting 25% of the circu-lar HBV genome (see reference 20), the acid-insolubleradioactivity provides a measure for the quantity of isolatedcomplete HBcAg particles. By testing various amounts ofthese purified HBcAg particles in the presence of PEGprecipitates obtained from 400-,ul NHS samples, the sensi-tivity of the HBcAg assay proved to be in the range of 1 x107 to 4 x 107 HBV particles.

Effect of serum protein concentration on HBcAg detection.Since the precipitation of ICs at low PEG concentrations isstrongly influenced by pH, ionic strength, and protein con-centration (9), we analyzed the yield of precipitated HBcAgICs from samples containing a constant amount of purifiedHBV particles but various volumes of anti-HBc serum. Theprecipitation of HBcAg ICs was decreased when the proteinconcentration was diminished (Table 2). However, despitean eightfold lower protein concentration, the loss of HBcAgdetection sensitivity did not exceed 20%, pointing to variouseffects compensating each other. In the presence of lowprotein concentrations, a decreased yield of precipitated

VOL. 21, 1985

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from

596 BREDEHORST, VON WULFFEN, AND GRANATO

TABLE 2. Effect of anti-HBc serum protein concentration onPEG precipitation and subsequent detection of sequestered

HBcAgaResults of core antigen

Addition of detection in PEG-precipitatedVol (,ul) Arnt of Dane sample

protein (mg) particlescpm (103) S/N %

400 28.8 - 0.4400 28.8 + 18.5 46.3 100200 14.4 + 17.2 43.0 93100 7.2 + 16.2 40.5 8850 3.6 + 15.2 38.0 82

a Ten-microliter volumes of resuspended HBV particle concentrates, pre-pared as described in the DNAP assay, were added to various volumes ofserum containing anti-HBc at a titer of 1:1Q0 (HBcAg negative), made up to afinal volume of 4.01 mi with PBS (pH 7.2), and processed according to thestandard quantitation procedure.

HBcAg ICs appeared to be accompanied by an increasedadsorption of HBcAg to polystyrene due to reduced compe-tition of protein contaminants with the core for binding sites.Therefore, the reproducibility of the quantitation procedureis assured even when protein concentrations in serum varyconsiderably.

Stability of HBcAg in serum and thiocyanate. Sera from sixpatients with chronic type B hepatitis were incubated atroom temperature and assayed for HBcAg after variousperiods of incubation. HBcAg proved to be very stable (Fig3A). In none of the six different sera could any degradationof the core antigen be noticed during an incubation period of10 days at room temperature.To estimate the loss of core antigen activity by treatment

with thiocyanate as observed by Ohori et al. (16), PEG-pre-cipitated HBcAg ICs from two sera of patients with chronictype B hepatitis were dissolved in 3 M thiocyanate andassayed for HBcAg after various periods of incubation at4TC. HBcAg levels showed only moderate degradation dur-ing the first 2 days of treatment, ranging between 5 and 10%(Fig. 3B). Even after 2 weeks of treatment at 4°C, approxi-mately 50% of the original core antigen activity cold bedetected. Thus, the loss of immunogenic activity of HBcAg

by 3 M thiocyanate during the standard quantitation proce-dure appears to be negligible.

Spççificity of the assay for core determinants. Nonlabeledanti-HBs, anti-HBe, and anti-HBc antibodies were tested fortheir ability to block the detection of core epitopes by theapplied '25I-labeled anti-HBc. After preincubation with an-ti-HBs, no decrease in the level of detectable HBcAg couldbe noticed (Table 3). After preincubation with anti-HBc,however, binding of '25I-labeled anti-HBc was totally inhib-ited, indicating specific quantitation of core determinants bythe detection system.Comparison with other HBcAg assays. The technique de-

scribed here was compared with procedures requiring priorremoval of anti-HBc. For this purpose, serum samples froma patient with chronic hepatitis B (anti-HBc titer by RIA, 1:5X 104) were centrifuged as described by Sagnelli et al. (21) toobtain serum pellets devoid of anti-HJ3c. The subsequentquantitation of core antigen in the serum pellets was per-formed by the HBcAg RIA described previously (21) and bythe solid-phase RIA described by Purçell et al. (17). Thehighest sensitivity was obtained by the PEG precipitationmethod without prior removal of anti-HB3c (Fig. 4). In theserum pellets obtained by the centrifugation procedure de-scribed by Sagnelli et al. (21), reasonable amounts of HBcAgcould only be detected in the solid-phase RIA described byPurcell et al. (17). The application of HBcAg-coated beads inthe competitive RIA described by Sagnelli et al. (21) resultedin determinations of HBcAg activities not exceeding thecutoff level (N/S ratio of 2.1), even at the highest HBcAgconcentrations tested in these experiments.

Application of the liBcAg assay in HBeAg-positive sera andcomparison with assays of HBV DNA, DNAP, and HBeAg.The assay was applied for the determination of HBcAg in185 HBeAg-positive sera. When correlated to HBeAg con-

centrations (131 of the 185 sera), HBc antigenemia was

found in 99% of the sera with HBeAg titers of -1:256, butonly in 25 of the 58 sera with lower levels of HBeAg (Table4). In 6 of the 34 HBcAg-negative sera, HBV DNA could bedetected by the blot hybridization technique, pointing to a

lower sensitivity of the HBcAg assay than the HBV DNAassay for the detection of HBV particles.

104 -

E

m-

1031

A: Serum 1051 B: Thlocyanate

104 -

1 3

O 4 8 12Incubation time [days]

O 4 8 12Incubation time [days]

FIG. 3. Stability of HBcAg in serum and in thiocyanate. (A) Serafrom six patients with chronic type B hepatitis were stored at roomtemperature and analyzed for HBcAg according to the standardquantitation procedure on various days as indicated in the figure. (B)PEG-precipitated HBcAg ICs from 2 ml of serum from two patientswith chronic hepatitis B were dissolved each in 1.3 ml of 3 MKSCN; after different incubation periods at 4°C, 200-pl sampleswere assayed for HBcAg as described in the text.

TABLE 3. Specificity of the RIA for HBcAg determinants"Results of binding of

Preincubation of coated "251-labeled anti-HBcpolystyrene beads with:

cpm ± SD S/N

PBS 3,621 ± 126 9.05

NHS 3,502 ± 192 8.76

Anti-HBs serum (anti-HBs +, 3,997 ± 143 10.00anti-HBc -, anti-HBe -)

Anti-HBc serum (anti-HBs -, 421 ± 40 1.05anti-HBc +, anti-H3e -)

Anti-HBe serum (anti-HBs -, 398 ± 20 0.99anti-HBc +, anti-HBe +)a After the attachment of PEG precipitates (obtained from an HBeAg-

positive serurn) to polystyrene beads by the standard quantitation procedure,the beads were washed four times with distilled water and then incubated for 4h at 37°C with 200 .LI of PBS (pH 7.2) pooled NHS, anti-HBs serum (708 mU/ml) from a patient immunized with HBsAg, anti-HBc serum (titer by RIA,1:104; negative for anti-HBe and anti-HBs), or serum containing anti-HBc(titer, 1:104) and anti-HBe (titer, 1:5 x 102). The sera were diluted 1:25 in PBS(pH 7.2). After six washes with distilled water, the balls were processed asdescribed in the standard quantitation procedure.

J. CLIN. MICROBIOL.

:=

a--- - m

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from

QUANTITATION OF HBcAg IN SERUM 597

Comparison of the HBcAg assay with the DNAP activitytest (139 of the 185 HBeAg-positive sera) showed HBcantigenemia in all DNAP-positive sera and in 39% of the serawithout detectable DNAP activity (Table 5). In these DNAP-negative sera, significant amounts of HBV DNA were de-tected only in 3 of the 27 HBcAg-negative sera, but in 16 ofthe 17 sera with HBc antigenemia. Therefore, the HBcAgassay appears to be more sensitive than the viral DNAPassay for the detection of HBV particles with DNA-contain-ing cores.

DISCUSSIONThe serum HBcAg assay described here includes the

following four steps (as schematically outlined in Fig. 1): (i)the release of the viral core from circulating HBV particlesby treatment with Nonidet P-40 detergent; (ii) isolation ofsubsequently formed HBcAg ICs by PEG precipitation; (iii)dissociation of the PEG-precipitated HBcAg ICs and attach-ment to polystyrene beads under dissociating conditions;and (iv) detection of adsorbed HBcAg with 125I-labeledanti-HBc.The most important advantage of this method is the

minimal interference by even high levels of anti-HBc,

3-

o

.i_

Cn

s 2-cmC-E?

c=

1-

O-

1

Cut off \

A

A

1:4 1:16 1:64Dilution of serum

FIG. 4. Comparison of the standard quantitation procedure withother serum HBcAg assays. All assays were performed with serumsamples from the same patient (anti-HBc titer, 1:5 x 104) and with

1-labeled anti-HBc antibodies of identical specific radioactivity(4.5 x 104 cpm of immunoglobulin G per ,ug). For the standardquantitation procedure (H), 400-,ul samples of serum were seriallydiluted with pooled NHS and assayed for HBcAg as described in thetext. In the HBcAg assays described by Purcell et al. (17) (A) andSagnelli et al. (21) (A), comparable volumes of HBcAg-rich serum

pellets prepared as described by Sagnelli et al. (21) were analyzed.The results are expressed as the natural logarithm of S/N ratios or,with respect to the HBcAg RIA (A), the natural logarithm of N/Sratios. The cutoff level represents a ratio of 2.1.

TABLE 4. H4Bc antigenemia in 131 sera with various HBeAgconcentrations

HBeAg HBcAgconcn Positive sera Negative sera(1/titer) No. % No. %

2256 72 99 1 116-128 23 74 8 261-8 2 7 25 93

whereas other assays require the removal of even traceamounts of anti-HBc. Due to the formation of HBcAg ICs,standard immunoassays could not be applied for the detec-tion of HBcAg; therefore, sequestered HBcAg was sepa-rated first from other serum components by PEG precipita-tion as described by Siersted et al. (22). The relatively lowyield of precipitated HBcAg ICs, not exceeding 40% underthese conditions, is less important for the sensitivity of theassay than the highly efficient purification of the HBcAg ICsby this technique. Improved precipitation of sequesteredHBcAg by an increase in the PEG concentration from 2.0 to3.5% would be accompanied by co-precipitation of immuno-globulin G and immunoglobulin M in the range of 4 and 8%,respectively (2), leading to an increase of protein contami-nants in the PEG precipitates and thereby to reduced attach-ment of HBcAg to the polystyrene beads. After PEG pre-cipitation, the HBcAg ICs were dissociated by treatmentwith 3 M KSCN. Although successful attempts to identifyantigens within ICs have been reported (3, 8), these investi-gations were performed with ICs containing antigen excess,whereas HBcAg ICs can be expected to contain antibodyexcess. Without prior removal of the thiocyanate ions, thedissociated ICs were then adsorbed onto polystyrene, sinceunder dissociating conditions HBcAg and anti-HBc antibod-ies originating from the ICs are linked independently of eachother to the solid support (14), even when anti-HBc ispresent in excess. Thus, adsorbed HBcAg could easily bequantitated with the aid of commercially available 12'I-la-beled anti-HBc, but any other sensitive detection system,including enzyme immunoassay techniques, may be em-ployed.The HBcAg assay proved to be specific and reproducible.

Nearly identical serum levels of HBcAg were determined ondifferent days in samples of the same sera (Fig. 3A). Changesin immunospecific core antigen activity by treatment withthiocyanate as observed by Ohori et al. (16) were negligibleunder the conditions of the assay. Furthermore, differingprotein concentrations in the sera did not influence HBcAgdetection significantly, presumably as a result of compensat-

TABLE 5. Comparison of the HBcAg assay to the DNAP assayin 139 HBeAg-positive sera

Amt of HBV- HBcAg

DNAP activity Positive sera Negative sera(pmOl/ml [102]) No. % No. %

in:

Positive sera10.1-100 57 100 0 01.1-10 32 100 0 00.1-1.0 6 100 0 0

Negative sera 17 39 27 61

VOL. 21, 1985

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from

598 BREDEHORST, VON WULFFEN, AND GRANATO

ing effects: a decreased precipitation of HBcAg-ICs wasfollowed by an increased adsorption onto polystyrene beadsdue to the presence of less contaminating proteins in thePEG precipitates, and vice versa. Finally, the reproducibil-ity of the HBcAg assay was supported by the remarkablestability of HBcAg in particles in serum, allowing reliabledeterminations of HBcAg in sera which had been stored atroom temperature for 2 weeks. Even partial disruption oftheHBV particles, which may result from the conditions ofstorage (see reference 17), would not affect the quantitationof HBcAg by the technique described here, since the assayis based on the detection of sequestered HBcAg.Comparison of this method to HBcAg assays requiring

prior removal of anti-HBc (17, 21) revealed that, in thepresence of high anti-HBc concentrations (titer by RIA, 1:5x 104), the highest sensitivity was achieved by the techniquedescribed here. Presumably, residual amounts of anti-HBcin the HBcAg-rich serum pellets prepared as described bySagnelli et al. (21) were responsible for the lower sensitivityof the other assays. Rizzetto et al. (19) demonstrated resid-ual anti-HBc activities up to titers of 1:103 in serum pelletsobtained from sera with high anti-HBc levels (titer by RIA,>1:5 x 104) by the same centrifugation procedure. Theyapplied gel filtration, with subsequent ultracentrifugation ofthe eluate, for the efficient removal of anti-HBc from serumsamples.To assess the value of the described HBcAg assay for

clinical practice, 185 HBeAg-positive sera were screened forHBc antigenemia, and the results were compared withdeterminations of HBV DNA (78 sera), DNAP activity (139sera), and HBeAg titers (131 sera). The HBcAg assayproved to be less sensitive than the HBV DNA hybridizationtechnique, but more sensitive than the DNAP assay, for thedetection of circulating HBV particles with DNA-containingcore. Furthermore, the routinely applied HBeAg RIA is lessreliable as a substitute for the HBV DNA hybridization testthan the described HBcAg assay. In more than 50% of the 44DNAP-negative sera with HBeAg, no HBV DNA could bedetected, whereas HBc antigenemia in this group of sera wasassociated with a positive HBV DNA hybridization assay in16 of 17 sera. In summary, we believe that in clinicalpractice, the HBcAg assay will help to identify infectioussera of HBV-infected individuals and assess the synthesis ofHBcAg in the absence of a liver biopsy in HBsAg carriers,which has been shown to prevail in patients with chronichepatitis of the agressive and persistent type (19).

ACKNOWLEDGMENTSWe thank R. Laufs for financial support to this work and B. Weiss

for technical assistance. We also thank W. H. Gerlich and E. Zyzikfor being so kind as to perform the HBV DNA hybridization assayon our serum samples.

LITERATURE CITED1. Arnold, W., K. H. Meyer zum Bueschenfelde, G. Hess, and J.

Knolle. 1975. The diagnostic significance of intrahepatocellularhepatitis-B surface antigen (HBsAg), hepatitis-B core antigen(HBcAg) and IgG for the classification of inflammatory liverdiseases. Klin. Wochenschr. 53:1069-1074.

2. Creighton, W. D., P. H. Lambert, and P. A. Miescher. 1973.Detection of antibodies and soluble antigen-antibody complexesby precipitation with polyethylene glycol. J. Immunol.111:1219-1227.

3. Cunningham-Rundles, C. 1981. The identification of specificantigens in circulating immune complexes by an enzyme-linked

immunosorbent assay: detection of bovine casein IgG com-plexes in human sera. Eur. J. Immunol. 11:504-509.

4. Gerlich, W., and R. Thomssen. 1975. Standardized detection ofhepatitis B surface antigen: determination of its serum concen-tration in weight units per volume. Dev. Biol. Stand. 30:78-7.

5. Gudat, F., L. Bianchi, W. Sonnabend, G. Thiel, W. Aenishaen-slin, and G. Stalder. 1975. Pattern of specific immune responsein hepatitis B. Lab. Invest. 32:1-9.

6. Hess, G., W. Arnold, and K. H. Meyer zum Bueschenfelde. 1981.Demonstration and partial characterization of an intermediateHBcAg (Dane particle) population. J. Med. Virol. 7:241-250.

7. Hess, G., W. Arnold, B. Moller, G. M. Gahl, and K. H. Meyerzum Bueschenfelde. 1980. Inhibition of hepatitis B virus specificDNA polymerase by intercalating agents. Med. Microbiol.Immunol. 168:25-34.

8. Husby, S., S.-E. Svehag, H. Nielsen, N. H0iby, and P. O. Schi0tz.1981. Methodological approaches to antigen identification insoluble immune complexes: a model study. Acta Pathol. Mi-crobiol. Scand. Sect. C 89:155-160.

9. Ingham, K. C. 1978. Precipitation of proteins with polyethyleneglycol: characterization of albumin. Arch. Biochem. Biophys.186:106-113.

10. Kaplan, P. M., R. L. Greenman, J. L. Gerin, R. H. Purcell, andW. S. Robinson. 1973. DNA polymerase associated with humanhepatitis B antigen. J. Virol. 12:995-1005.

11. Laurell, C. B. 1972. Electroimmunoassay. Scand. J. Clin. Lab.Invest. 29(Suppl. 124):21-37.

12. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall.1951. Protein measurement with the Folin phenol reagent. J.Biol. Chem. 193:265-275.

13. Mollison, P. L. 1975. Blood transfusion in clinical medicine.Blackwell Scientific Publications, Oxford.

14. Neurath, A. R., N. Strick, L. Baker, and S. Krugman. 1982.Radioimmunoassay of hidden viral antigens. Proc. Natl. Sci.U.S.A. 79:4415-4419.

15. Ohori, H., S. Onodera, and N. Ishida. 1979. Demonstration ofhepatitis B e antigen (HBeAg) in association with intact Daneparticles. J. Gen. Virol. 43:423-427.

16. Ohori, H., M. Yamaki, S. Onodera, E. Yamada, and N. Ishida.1980. Conversion from HBcAg to HBeAg by degradation ofhepatitis B core particles. Intervirology 13:74-82.

17. Purcell, R. H., J. D. Gerin, J. B. Almeida, and P. V. C. Holland.1974. Radioimmunoassay for the detection of the core of theDane particles and antibody to it. Intervirology 2:231-243.

18. Ray, M. B., V. J. Desmet, A. F. Bradburne, J. Desmyter, J.Fevery, and J. DeGrote. 1976. Differential distribution of hepa-titis B surface antigen and hepatitis B core antigen in the liver ofhepatitis B patients. Gastroenterology 71:462-469.

19. Rizetto, M., J. W. K. Shih, G. Verme, and J. L. Gerin. 1981. Aradioimmunoassay for HBcAg in the sera of HBsAg carriers:serum HBcAg, serum DNA polymerase activity and liverHBcAg immunofluorescence as markers of chronic liver dis-ease. Gastroenterology 80:1420-1427.

20. Robinson, W. S., D. A. Clayton, and R. L. Greenman. 1974.DNA of human hepatitis B virus candidate. J. Virol. 14:384-391.

21. Sagnelli, E., C. Pereira, G. Triolo, S. Vernace, and F. Paronetto.1982. Radioimmunoassay for hepatitis B core antigen. Clin.Chem. 28:351-353.

22. Siersted, H. C., J. Brandslund, S.-E. Svehag, and J. C. Jensen-ius. 1981. Quantitation of circulating immune complexes bycombined PEG precipitation and immunoglobulinspecific radi-oimmunoassay (PICRIA). Methods Enzymol. 74:538-550.

23. Takahashi, K., Y. Akahane, T. Gotanda, T. Mishiro, M. Imai,Y. Miyakawa, and M. Mayumi. 1979. Demonstration of hepatitisB e antigen in the core of Dane particles. J. Immunol.122:275-279.

24. Takahashi, T., K. Kaga, Y. Akahane, T. Yamashita, Y.Miyakawa, and M. Mayumi. 1980. Isolation of Dane particlescontaining a DNA strand by metrizamide density gradient. J.Med. Microbiol. 13:163-166.

J. CLIN. MICROBIOL.

on April 10, 2019 by guest

http://jcm.asm

.org/D

ownloaded from