J. gen. Virol. (I978), 38, 549-559

549

Printed in Great Britain

Radioimmunoassay and Some Properties of Human Antibodies to Hepatitis B Core Antigen By A. R. N E U R A T H , W. S Z M U N E S S , C. E. S T E V E N S , N. S T R I C K AND E. J. H A R L E Y The Lindsley F. Kimball Research Institute o f The N e w York Blood Center, N e w York, N e w York 1o021, U.S.A. (Accepted IO October 1977) SUMMARY

A solid-phase radioimmunoassay for antibodies to hepatitis B core antigen (anti-HBc) is described. Polystyrene beads coated with anti-HB0, hepatitis B core antigen prepared from pooled sera of humans infected with hepatitis B virus (HBV) and 125I-labelled anti-HBo were used for the test. Distinct patterns of development and changes of anti-HBc and their immunologic properties are all related to variations of other markers specific for HBV infections. Knowledge concerning the detailed features of the immune response to hepatitis B core antigen may provide deeper insight into the pathogenesis of HBV infections. INTRODUCTION

The putative hepatitis B virus, commonly designated as the Dane particle (Dane, Cameron & Briggs, I97o) consists of a nucleocapsid 27 nm in diameter within a lipoprotein shell. Distinct antigenic determinants are associated with the outer surfaces of the lipoprotein envelope and the nucleocapsid (Alrneida, Rubenstein & Stott, I971). These determinants are referred to as hepatitis B surface antigen (HB~Ag) and hepatitis B core antigen (HBoAg) respectively. Several techniques have been developed for the assay of antibody to HBoAg (anti-HBc): complement fixation (Hoofnagle, Gerety & Barker, 1973), immune adherence haemagglutination (Tsuda et al. I975), immunofluorescence (Brzosko et al. I975), counterimmunoelectrophoresis tests (cf. Howard & Zuckerman, 1977), and radioimmunoassays using either HB~Ag containing tritiated DNA (Robinson & Greenman, I974; Moritsugu et al. I975; Greenman, Robinson & Vyas, 1975; Howard & Zuckerman, 1977) or a~sIlabelled anti-HBc (Purcell et al. I973; Vyas & Roberts, 1977). Anti-HBo develops regularly during type B hepatitis and is an indicator of either recent acute infection or of chronic infection by hepatitis B virus (Hoofnagle et al. ~973, 1974; Krugman et al. t974; Hoofnagle, Gerety & Barker, I975; Szmuness et al. I976). The titre of antiHBo in sera of HB~Ag carriers appeared to be unrelated to the concentration of Dane particles determined directly by electron microscopy or estimated indirectly by testing the sera for HB~Ag-specific DNA polymerase activity or for e antigen (HB,Ag) (Takahashi et al. I976; Imai et al. 1976), an additional soluble antigen specifically associated with hepatitis B virus (HBV) infections (Magnius & Espmark, 1972) and an apparent marker for the infectivity of HB,Ag-positive sera. We describe here a sensitive solid-phase radioimmunoassay for anti-HB, and its appli35

V~R

38

55 °

A. R. NEURATH AND OTHERS

cation to the study of anti-HB~ in individuals infected by HBV. Our results suggest that anti-HBo present in HBoAg-positive sera differs qualitatively from anti-HBo in sera containing antibodies to HB~Ag (anti-HB,), and that the magnitude of quantitative changes in the titre of anti-HBo with time may be of value in predicting the probable outcome of infections by HBV. METHODS

Preparation of the HB~Ag reagent. Sera positive for HB~Ag were centrifuged at 37ooo g for I6 h to pellet the Dane particles. In some cases, the sera were first precipitated with 6 % (w/v) polyethylene glycol, and the pellets containing Dane particles (Neurath & Strick, 1977) were resuspended in o.oi M-tris (hydroxymethyl) aminomethane-o.I 4 M-NaCl-o.o2 % NaN~ (TS) in I/5 of the original serum volume and then centrifuged as above. The pellets containing Dane particles were suspended in TS in I/IOO of the original serum volume and pronase (IOO #g/ml) was added. The mixtures were incubated for I h at 37 °C. Solid KBr was added to a buoyant density of 1.3o g/ml and the mixtures were transferred into centrifuge tubes (3 to 8 ml/tube) for either the SW 25.2 or SW 25. I rotors (Spinco Division, Beckman Instruments, Palo Alto, California). A gradient of KBr (density 1.28 to 1"15 g/ml) was layered on top of the mixtures. The tubes were centrifuged at 25o00 rev/min for 24 h. Fractions with a density of 1.2o to 1-24 g/ml which contained most of the HB~Ag and the Dane particles were pooled, dialysed against TS and then centrifuged at 370o0 g for I6 h. The pellets were resuspended in i ml of TS and treated with 5o #g of pronase for I h at 37 °C and then layered on top of a 7"5 to 20% (w/w) sucrose gradient in a centrifuge tube for the SW 65 rotor. The pellet obtained after centrifuging at 35ooo rev/min for 2 h was resuspended in 5 ml o f o . t M-NH4CI-o.o8 M-MgCI2-o'25 % mercaptoethanol-o.5 % Nonidet P4o (pH 7"6), incubated for I h at 37 °C and stored at 4 °C. No loss of HBoAg activity was observed after storage for 4 months. The properties of HBcAg will be described in a separate paper (Neurath, Huang & Strick, I978). Radioimmunoassay for anti-HBc. Polystyrene beads (diam. 6 mm) were immersed in a solution of anti-HBc [IOO/~g IgG/ml in o.I M-tris (hydroxymethyl) aminomethane, p H 8.8] prepared from an anti-HBo-positive serum as described before (Neurath & Strick, 1977). After several hours at room temperature, the solution of anti-HBo was removed, the beads were washed several times with TS and then immersed in a solution of bovine haemoglobin (IO mg/ml in TS). After standing overnight at 4 °C, the beads were washed with TS, dried and stored at 4 °C. Anti-HB~ from anti-HB, and HBoAg-positive sera were labelled with Nal~SI or with NalalI as described before (Neurath & Strick, I977)For routine determinations of anti-HBc, proper dilutions (0"4 ml) of test samples in sheep serum diluted I : 5 with TS were mixed with I #1 of the HBcAg reagent and incubated for 30 min at 37 °C- Negative (I :5 sheep serum) and positive (I:5 sheep serum+ I/~1 of the HBcAg reagent) controls were included with each set of test samples. Anti-HBo-coated beads were added and the incubation was continued for I6 h at room temperature. The beads were washed with TS and then incubated for 2 h at 37 °C with 0"3 ml of I : 5 sheep serum and with o.I ml of the labelled anti-HBo (200000 ct/min). Finally, the beads were washed with TS and counted in a y-counter. Radioactive counts corresponding to the negative control were subtracted from all results. To study the competition for determinants on HBoAg between unlabelled anti-HBo, 1~5I- and 13q-labelled anti-HBc, beads were first incubated overnight at room temperature with the HBcAg reagent (I #1 in 0"4 ml of I : 5 sheep serum). The beads were washed with

Antibodies to hepatitis B core antigen

551

TS, and a mixture of unlabelled and labelled (200000 ct/min for each label) anti-HBo in I : 5 sheep serum was added. The beads were incubated for 2 h at 37 °C, washed with TS and counted in a two-channel 7-counter. All results were corrected for spillover of xa~Icounts into the channel for counting ~25I. Other methods. Isoelectric focusing was performed as described before (Neurath & Strick, 1977). HB~Ag and antibodies to HB~Ag (anti-HB,) were determined by radioimmunoassays using commercial reagents (Abbott Laboratories, North Chicago, Illinois). Rheophoresis was used to detect HBoAg and anti-HBo (Neurath & Strick, I977). Serum glutamicpyruvic transaminase (SGPT) was tested spectrophotometrically using the Enztrate test kit (Beckman Instruments, Fullerton, California) and following the instructions of the manufacturer. Sera were collected from patients and personnel of haemodialysis units in biweekly to I month intervals. To study the class specificity of anti-HB,, 4o #1 samples of sera diluted with TS to I ml were applied to z. 5 ml columns of insolubilized antibodies to human immunoglobulins. The columns were prepared by binding the IgG fractions from antisera to human immunoglobulins (Behring Diagnostics, Somerville, New Jersey) to CNBr-Sepharose (Pharmacia, Uppsala, Sweden). The columns were incubated for 3o min at 37 °C and I h at 4 °C and then washed at 4 °C with 6 ml of TS supplemented with NaC1 (o'5 M). The adsorbed immunoglobulins were eluted at 24 °C with 6 ml of o.z M-NaHCOs--o.5 M-NaC1 (pH Io"9).

RESULTS

Specificity and relative sensitivity of the radioimmunoassay for anti-HBo The basis of the radioimmunoassay is the inhibition by anti-HBc in serum test samples of the attachment of HBoAg to plastic beads coated with anti-HBc. The quantity of HBoAg bound to the beads is determined by measuring the amount of 125I-labelled anti-HBc adsorbed to the beads during a subsequent incubation step after they had been washed to remove all unattached HBoAg and unlabelled anti-HBo from the samples. To determine the optimum quantity of HB~Ag for the test, serial twofold dilutions of the HBcAg reagent in sheep serum diluted fivefold with TS were incubated with the beads overnight at 24 °C and the relative quantity of bound HBoAg was determined. The quantity of 125I-labelled anti-HBo bound to the beads increased non-linearly with decreasing dilution of HBoAg (Fig. 1). One #1 of the HBcAg reagent per test (corresponding to a dilution of 1:4oo) was used in subsequent experiments. Incubation of the beads with HBoAg instead of HBoAg did not cause subsequent attachment of the radioactive antibody. Addition of anti-HB, to HBoAg inhibited the binding of a2~I-labelled anti-HBc to the beads (Fig. 2). Some sera from carriers of HBsAg tested in our experiments caused a 5o % inhibition at dilutions up to 1:51ooo. The slopes of dose response curves varied for different sera, an observation which will be discussed later. To evaluate the relative sensitivity of the test for anti-HBc, a panel of sera supplied by the Bureau of Biologics, Food and Drug Administration and used for evaluation of the sensitivity of tests for HB,Ag was screened for anti-HBc. All sera were tested at a I : I6 dilution. Results in Table I show that the radioimmunoassay for anti-HB~ is less efficient in screening out HB~Ag carriers from blood donors than current tests for HB~Ag. However, the method identifies additional individuals whose serum may contain HBV although negative for HB~Ag by current test methods (Table z). To further evaluate the specificity of the radioimmunoassay for anti-HBo, sera from patients with HB~Ag-negative alcoholic liver cirrhosis (5), acute non-A non-B hepatitis 35-2

552

A. R. N E U R A T H A N D OTHERS 2800

_I

I

I

I

I

I

I

I

~

l

I

1400

2400

1200

.L-.. . . . . . . . . . . . . . .

•~ 2000 B

/ i i / ~I ~ '--- >

/o/

~1000

~1600

/P /

800

1200

.~ 600

8O0

4O0

400

20O

//I ,__o ..o °

-l

I

r

i

I

2 x2 2-~0 2-8 2-6 2-'* Antigen dilution Fig. [

0 2 -z

2 6

2-8

J

°

'

/

°

/

2-10 2-tz 2-14 2-~6 ~ 18 Antibody dilution Fig.2

Fig. i. Radioimmunoassay of HB~Ag. Relationship between 125I-labelled anti-HB~ bound to polystyrene beads and the dilution of HB~Ag. Fig. 2. Radioimmunoassay inhibition test of anti-HBo. Dose response curves for a serum containing HB~Ag ((3---(3) and anti-HB, (Q--Q), respectively. Horizontal broken line indicates the radioactivity corresponding to a positive control containing HBoAg but no anti-HB~. T a b l e i. Detection of anti-HBc in a panel of HB+Ag reference sera Relative concentration of HB+Ag in serum as indicated by the supplier of the panel

Frequency of anti-HBo detection at a serum dilution of I : I6"

A. Reactive by all methods including agar gel diffusion B. Reactive by counter-electrophoresis, reversed passive haemagglutination and radioimmunoassay C. Reactive only by reversed passive haemagglutination and radioimmunoassay (C). Only occasionally positive by third generation (C) test methods D. Non-reactive by current test methods but containing low levels of HB+Ag N. Non-reactive by current test methods

3/3 5/5 6]9

1/2 I/z

2]4

* All sera inhibited the binding of 125I-labeUedanti-HBo by more than 50 %. (Prince et al. I974) (3), a n d acute hepatitis A (5) were tested at a dilution o f I : I 6 . Only t w o o f the sera h a d anti-HB¢, a n d these two were f r o m hepatitis A virus-infected children f r o m a school o f the m e n t a l l y retarded, i.e., f r o m an e n v i r o n m e n t where hepatitis B is also prevalent. N o r m a l h u m a n sera a n d a n i m a l sera tested u n d i l u t e d inhibited the a t t a c h m e n t o f a25I-anti-HBo b y a b o u t 4o to 60 %. The n a t u r e o f this effect r e m a i n s u n k n o w n . H o w e v e r , s o m e h u m a n sera (tested u n d i l u t e d o r 1:4 diluted) h a d a higher i n h i b i t o r y effect, suggesting the presence o f low levels o f anti-HB,.

Antibodies to hepatitis B core antigen

553

Table 2. Detection of anti-HB~ in a group of HB,Ag-negative sera Characterization of serum based on previous results (Szmuness et al. 1976)* Positive for both anti-HB~ and anti-HBo Negative for anti-HB~ Positive for anti-HBo Positive for anti-HB, Negative for anti-HBo

Minimal inhibition of Frequency of anti-HB~ 12~I-anti-HBo detection by radioimmunoassay attachment observed at a serum dilution of in groupt I:I6 II/II

4° %

6/6

42 %

7/I6

20%

* Anti-HB~ was detected by complement fixation tests. 4- The highest inhibition observed with 2 out of 32 normal human sera tested at I : I6 dilutions was I3"3 ~. An inhibition equal to or higher than 2o ~ was considered specific for anti-HBo. Table 3. Prevalence of anti-HB~ in serum (negative for both HB, Ag and anti-HB~)

of patients and staff members of haemodialysis units Characterization of the test group Patients haemodialysed for more than 3 years Patients haemodialysed for less than I year Staff members employed for less than I year

Frequency of anti-HB~ detection at a serum dilution of I : I6 12/25 I/I9 1/22

Development and changes in anti-HB~ in the serum of selected individuals infected with HBV Hepatitis B is prevalent among patients and staff members of haemodialysis units. The radioimmunoassay test for anti-HBo identified additional individuals who were infected with HBV but whose serum was negative for both HB~Ag and anti-HB~ and did not show elevated levels of SGPT (Table 3). Considering the frequency of positive tests for each HB~Ag, antiHB~ and anti-HBo, at least 75 % of patients treated for more than 3 years were infected with HBV. The high rate of HBV infections in haemodialysis units and the long-term surveillance of patients and staff members allowed us to select individuals recently infected with HBV who showed distinct patterns of anti-HBc development in relation to other markers for HBV infection. The results summarized in Fig. 3 indicate that: (I) unless present at the initiation of surveillance (panels 5, 6, 7), anti-HBc appeared abruptly at titres /> ~ :27; (2) the initiation of synthesis of anti-HBo coincided approximately with the first signs of liver damage as indicated by elevated levels of serum SGPT (panels I, 2, 3, 4, 8, 9, io); (3) clearance of HBoAg from serum and/or detection of anti-HB~ usually occurred during a decline of anti-HBc levels (panels ~ to 7); (4) a low rate of clearance of anti-HBo (panel 5) and a prolonged period in which levels of anti-HBc increased (panels 8 to Io) seemed to be associated with sporadic and prolonged liver damage, respectively; (5) HBoAg appeared in serum during a time interval in which titres of anti-HBc increased (panels 8 to ~o); (6) infection with HBV is not necessarily associated with even transient presence of HB~Ag (detectable by radioimmunoassay) in serum (panel 4); (7) the fast clearance rate of antiHBo from serum (panels 2, 3, 4, 6, 7) corresponded to an anti-HBo half-life of I6 to 53 days; (8) after clearance of HBsAg from serum and before anti-HB~ becomes detectable, anti-HBc represents the only marker for HBV infection (panels r, 2, 3, 5, 6, 7, 8). In order to determine the class of anti-HBo appearing early in the course of infection, samples of the first specimen positive for anti-HB, (panel 9, Fig. 3) were chromatographed

8

I

I

i

I

I

50

I

0

llllIIlllll[]lllllll

I

(1)

I

I

I

i

I

~

I

I

I

HIIIIIIIIIIIIIIIIIIIIIII!

V//////////////~.

I

I

100 150 200 250

I

I

I[llll[lllllll[ll[ I

I

I

I

I

I

I

I

I

I

I

I

I

I flllIllllllfllllllllllll

I

I

I

I

50

I

0

(7) I I

l

100 150 200 250

v

I ,IIIIIII[III[IIIIIIIIIIEIIIHIIIIIIHIII[II,

(2)

t

0

I

I

I

I

I

I

1

I

I

I

I

i

I

I

I

I

50 100 150 200 250 Days of surveillance

(8)

Illllilll/ll[llllllllllllllL]lLllllllIJll

kO000~

(3)

I

|llllllllll[lllllll]lllll[l

I

I

0

I

I

I

I

/

I

I

I

I

I

I

I

\

[ 50 100 150 200 250

1

I

I

I

(9)

(4)

I

I

I

I

V//////////////////////////~ I

I

I

50

l

I

I

0

I

I

(10)

(s)

I

I

/

I

I

I

I

I

100 150 200 250

I

'HHIHH]HHH'HqIHII

I

I

EIIIll[lllIIIIllllilHIIIIllilllt[ilJ[llllJlll

. ~

I

Fig. 3. Development of anti-HBo in relation to other markers of infection by HBV in patients (panels 5 to Io) and staff members (panels i to 4) of haemodialysis units. All sera were tested for anti-HB~ at a 1:52 dilution. Sera negative for anti-HB~ were retested at a i :8 dilution and were again found negative in all cases. The 50 ~ inhibition end-points were determined from dose response curves (see Fig. 2) and plotted on Fig. 3 ( 0 - - 0 ) . Short perpendicular lines at the ends of the bars representing different HBV markers indicate positivity at the start or termination of surveillance. Position of bars along vertical axis has no meaning. [m, Anti-HBc; V~, HB~Ag; [], anti-HB~; i , elevated SGPT; ~ , HB~Ag.

I0

12

14

"~ 16 o

"2

10

12

14

16 I

0 ~Z

>

Z

>

4~

Antibodies to hepatitis B core antigen

555

Table 4- Association of early anti-HB~ with distinct immunoglobulin classes Anti-HB~ titret of c-

Type of immunosorbent column* Anti-IgG Anti-IgA Anti-IgM

Unadsorbed immunoglobulins 1:2 4 i :23 I :3

Adsorbed immunoglobulins subsequently eluted at pH Io'9 I : Io

1:9 t : 21

* Immunodiffusion tests with the unadsorbed, and the adsorbed and subsequently eluted proteins revealed that each immunosorbent column specifically and completely adsorbed the immunoglobulins of the proper class. Anti-lgD and anti-IgE columns adsorbed very little anti-HB~, probably non-specifically. Corresponding to 50 70 inhibition end.points determined from anti-HB~ dose response curves (Fig. 2). Table 5. Comparison of slopes of anti-HB, dose response curves

from HB,Ag- and anti-HB,-positive sera*

Characterization of anti-HBc HBeAg-positive serum Anti-HBe-positive serum F(ab)~ fragment from: HB~Ag-positiveserum Anti-HB~-positive serum

Magnitude of maximum slope ct/min ] (logs (anti-HBc dilution/ I 5 0 ; 2 2 0 ; 2 3 0 ; 2 5 0 ; (meant: 2 I 3 ~ 2 2 ) 700; 450; 370; 300; (mean~ : 455 +-87)

I5O (I73)~ 350 (4Io)~

* The slopes reflect the avidity of antibodies (Celada, Schmidt & Strom, I969). t The difference between the two means is statistically significant (t test, P < o'oI). :~ Values obtained with two different batches of anti-HBo-coated beads. on columns of insolubilized antibodies to the distinct classes of human immunoglobulins. The unadsorbed, and the adsorbed and subsequently eluted immunoglobulins were tested for anti-HBc. Results summarized in Table 4 show that anti-HBo activity was associated predominantly with lgM. In a specimen taken from the same patient Io6 days later, antiHBo was detected mostly in I g G as determined by the same technique.

Qualitative differences between anti-HB, apparently related to the presence of either HB,Ag or anti-HB, in individual sera Differences in anti-HBo dose response curves were observed between sera positive for HBoAg and anti-HBe, respectively (Fig. 2). Similar results were obtained with three additional randomly selected pairs of HBeAg- and anti-HBe-positive sera and with F(ab)~ fragments prepared from I g G isolated from two of the respective sera (Table 5). In order to explain the observed differences in the slope of the dose response curves, anti-HBo I g G prepared from HBoAg- and anti-HBo-positive sera were labelled with lzaI and a25I, respectively. Experiments using beads coated with HBcAg revealed that the x25I- and 13XI-labelled anti-HB0 competed for the same antigenic sites on HB,Ag and that the l~5I-labelled antibodies were bound preferentially when an excess of both radioactive antibodies was added to the beads. This suggested that anti-HBo from anti-HBo-positive sera may have a higher average affinity for antigenic sites on HBcAg than corresponding antibodies from HBoAgpositive sera. Therefore, it was expected that addition of unlabelled anti-HBo to HBoAgcoated beads suspended in a mixture of a25I- and ~3q-labelled anti-HBo (added in constant

556

A. R. NEURATH AND OTHERS 2'5

2'0

b

O O

ooo

oo

O

1"5-

o o0

0 0

°°

0

ooo

ooo oo

OooO.ooo

000000

o

O

1"0-

O00O00 0o% oOgoo° 0 0

oo-oo t 0 000

oOoo .Oo 0 ~0

000

0 O 00

O~D O

oo

a¢ 0'5-

o

00000000 000 °

°-,, 8 °~°8

o

0

oO::oo

00e ° 0 OeO0 ~

t

o:: •

01:27

1 : 270 Serum dilution

Fig. 4. Results of three-way competition tests between anti-HB, present in individual sera (tested at dilutions of 1:27 and 1:27o) positive for either HBeAg(©) or anti-HB~ (O), ]~SI-labelledantiHB~ (isolated from an anti-HB~-positiveserum) and ~3]I-labelledanti-HBo(isolated from an HB~Agpositive serum). Both labelled anti-HB¢ were added at fixed ratios in identical amounts for each test. Circles with bars on the right side of individual plotted experimental results correspond to the mean values. The length of the bars indicates the magnitude of the standard error of the mean. quantities) would suppress the binding of each of the labelled antibodies unequally, depending on the origin of the cold anti-HBc. This would result in a change of the ratio of 1251/131I anti-HBc bound to the beads as compared with the proportion of each labelled anti-HBo adsorbed to the beads when cold anti-HBc was absent. Indeed, additions of 2to 128-fold dilutions of an HB~Ag- and an anti-HBo-positive serum (each positive for anti-HBc) to identical mixtures of labelled anti-HBo, altered the 12~IylI ratio of bound radioactivity r'o4- to 2.2-fold and 0"35- to o'93-fold, respectively. The lower numbers in each case corresponded to lower serum dilution and vice versa. Based on these results, additional anti-HBo-positive (2i) and HBoAg-positive (69) sera were tested in the same way at two distinct dilutions. The results (Fig. 4) indicate a statistically significant difference (P < o.oi for a t test of significance between two sample means) between anti-HBo in HBoAg- and anti-HBo-positive sera. The preferential binding of 125I-labelled antibodies as compared to ~3~I-labelled immuneglobulins was not due to the distinct isotopes used. When each antibody was labelled with 125I, the tagged immunoglobulin from the anti-HBe-positive serum was less efficiently displaced by unlabelled anti-HB0 from HBcAg-coated beads than was the l~SI-anti-HBo from HBoAg-positive serum. To determine whether the observed differences in the behaviour of anti-HBo depending on their origin were reflected in their physical properties, HB~Ag- and anti-HBe-positive sera from long-term HB,Ag carriers were submitted to isoelectric focusing. Results in Fig. 5 indicate that regardless of its origin, anti-HBc was heterogeneous with respect to charge

557

A n t i b o d i e s to h e p a t i t i s B c o r e a n t i g e n I

I

I

I

I

10

100 90

9

80

8

7O 60

7

I

•= 50 r--

0 °

6

" 40 o o

30

zz

5

0 o

o 00

20

o D

000

4

10 I

I

I

12

16

20

~_ 24 28 32 36 Fraction number

3

I

40

44

48

52

Fig. 5. Isoelectric focusing of anti-HB~ f r o m HB,Ag-positive sera containing either HBeAg ( ~ , fractions tested at a 1:20 dilution) or anti-HB~ ([~, fractions tested at a 1:8o dilution); 0 , pH.

and therefore probably polyclonal. The distribution of anti-HBo in the pH gradient was within the pH range of 7 to 9"75- The largest portion of anti-HBo was recovered in fractions with a pH between 8"25 and 9.25. DISCUSSION

Brzosko et al. (I975) detected anti-HBo predominantly in IgG by an indirect immunofluorescent test when they examined sera collected during the first week of overt clinical illness from patients acutely infected with HBV. These authors did not notice a shift in the class specificity of anti-HBo from IgM to IgG during the course of illness and concluded that HBV infections may differ from other viral and non-viral infection with respect to development of the immune response to the particular antigens. Our limited results obtained by methods more amenable to quantitative evaluation than immunofluorescence, indicate a switch in anti-HBc production from IgM to IgG, as would be expected (Svehag, I966). As noticed before (Hoofnagle et al. I974; Krugman et al. 1974; Hoofnagle et al. I975), anti-HB, becomes detectable in serum at or close to the time during which clinical symptoms and/or abnormal levels of liver enzymes in serum are evident, i.e., much before appearance of a humoral or cellular immune response to HB~Ag (Vyas et al. 1975). The SGPT activity returned to normal levels in serum usually during a decline of anti-HBc titres (Fig. 3). In other individuals, a prolonged period during which titres of anti-HB~ increased was associated with prolonged liver damage. These results suggest a correlation, not necessarily causal, between liver injury, the synthesis of HBcAg and the immune response to it. In some individuals with probably good prognosis for recovery, anti-HB~ was cleared rapidly from serum in agreement with previous results (Krugman et al. I974; Hoofnagle et al. I975). The relatively short half-life of anti-HBo (16 to 53 days) as compared with the average half-life of IgG of 2I days (Stevenson, I974) suggests elimination of the immunologic stimulus, i.e., cessation of the synthesis of HBoAg. In other cases, anti-HB, continued to rise for a relatively long period, indicating continued synthesis of HB~Ag. This is supported by the detection of HBoAg, a marker for infectivity and for continued synthesis of

55 8

A. R. N E U R A T H

AND

OTHERS

Dane particles (Magnius & Espmark, 1972; Takahashi et aL 1976; Imai et aL ~976) in the sera during the same time interval. The course of hepatitis B infections may be modified by administration of either interferon or of interferon inducers (Greenberg et al. I976; Desmyter et al. I976; Purcell et al. 1976). However, in most cases, anti-HB~ titres in serum did not decrease after such treatment indicating sufficient synthesis of HBcAg to stimulate a continued immune response. Therefore, the probability that the interferon-treated subjects will eliminate HB,Ag may be less than for individuals clearing anti-HBo rapidly, as suggested by our results. Anti-HBo consists of antibodies with relatively high isoelectric points. This is in agreement with the relatively low isoelectric point of HB~Ag (pH 4"4; Howard & Zuckerman, 1977) and the general rule of an inverse relationship between the net electric charge of antigens and their corresponding antibodies (Sela, 1966). The dose response curves of anti-HB, originating from HBoAg- and anti-HBo-positive sera differed significantly not only for the intact immunoglobulins but also for the corresponding F(ab)~ fragments. Therefore, this dissimilarity cannot be explained by the possible presence in anti-HBo positive sera of rheumatoid factors (Kacaki, Siem & Brouwer, 1977; Tedder & Briggs, 1977) reacting with the Fc portion of IgG and possibly enhancing the reaction between HBoAg and anti-HBo (Ashe & Notkins, 1966). The distinct dose response curves may rather be attributed to a relatively low affinity of anti-HBo from HBoAgpositive sera for antigenic determinants on HBoAg. As mentioned before, HB,Ag positivity is associated with serum infectivity and with continued synthesis of Dane particles and their nucleocapsid component. On the other hand, anti-HBo positivity is accompanied by greatly diminished production of HBcAg. The corollary of our findings is the probable association between the biosynthesis of HBoAg and the avidity of antibodies to HB~Ag, suggesting that they may either be involved in the modulation of HBV infections, or that their properties may merely reflect the immune competence of the host. It also seems plausible that a portion of anti-HBo is used up in the formation of antigen-antibody complexes in individuals continuously producing HBoAg. In such circumstances, the proportion in serum of unbound anti-HBc with low avidity could increase. This study was supported by grant HL o9o11-14 from The National Heart, Lung and Blood Institute and by contract AI-62512 from The National Institute of Allergy and Infectious Diseases. We thank Dr Richard D. Aach for HB~Ag-negative sera from patients with liver cirrhosis, Dr P. Beasley for HBeAg-positive sera, and Dr A. M. Prince for re-viewing this paper. REFERENCES ALMEIDA, J. D.~ RUBENSTEIN, D. & STOTT, E. J. (t971). New antigen-antibody system in Australia-antigen-positive hepatitis. Lancet ii, Izz5-Izz7. ASHE, W. K. & NOTKINS, A. L. (I966). Neutralization of an infectious herpes simplex virus-antibody complex by anti-v-globulin. Proceedings of the National Academy of Sciences of the United States of America 56, 447-45 I. BRZOSKO, W. J., MIKULSKA,B., CIANCIARA, J. & BAB1UCH, L. 0975)' Immunoglobulin classes o f antibody to hepatitis B core antigen. Journal of Infectious Diseases x32, I-5. CELADA,r., SCHMIDT, D. & STROM, R. (I 969). Determination o f avidity o f anti-albumin antibodies in the mouse. Influence of the number o f cells transferred on the secondary adoptive response. Immunology I7, 189198. DANE, D. S., CAMERON, C. H. & BRIGGS, M. (I970). Virus-like particles in serum of patients with Australia antigen-associated hepatitis. Lancet i, 695-698. DESMYTER, J.) DEGROOTE~J., DESMET~V. J., BILLIAU) A.~ RAY, M. B.~ BRADBURNE,A. F.~ EDY, V. G.) DESOMER,P. & MORTELMANS,J. (1976). Administration of h u m a n fibroblast interferon in chronic hepatitis B infections. Lancet ii, 645-647.

Antibodies to hepatitis B core antigen

559

GREENBERG, H. G., POLLARD, R. B., LUTWICK, L. I., GREGORY, P. B., ROBINSON, W. S. & MERIG/~.N.T.C. (1976).

Effect of h u m a n leucocyte interferon on hepatitis B virus infection in patients with chronic active hepatitis. New England Journal of Medicine 295, 517-522. GREENMAN, R. L., ROBINSON, W. S. & VYAS, G. N. (1975). A sensitive test for antibody against hepatitis B core antigen (anti-HBc). Vox Sanguinis 29, 77-80. HOOFNAGLE, J. H., GERETY, R. J. & BARKER, L. F. (1973). Antibody to hepatitis B virus core in man. Lancet ii, 869-873. HOOFNAGLE, 1. H., GERETY, R. J. & BARKER, L. F. (I975). I n Transmissible Disease andBlood Transfusion, c h a p t e r 4, PP. 43-55. Edited by T. J. Greenwalt and G. A. Jamieson. New York: Grune & Stratton. HOOENAGLE, J. H., GERETV, R. S., NI, L. V. & BARKER, L. r. (I974)- Antibody to hepatitis B core antigen. A sensitive indicator of hepatitis B virus replication. New England Journal of Medicine 29o, I336-I34O. HOWARD, C. R. & ZUCKERMAN,A. J. (I977). Core antigen and circulating anti-core antibody in hepatitis B infection. Journal of Immunological Methods 14, 291-30 I. IMAI, M., TACHIBANA, F. C., MOKITSUGU, Y., MIYAKAWA, Y. & MAWMI, M. (I976). Hepatitis B antigen-associated deoxyribonucleic acid polymerase activity and e antigen/anti-e system. Infection and Immunity 14, 631635. KACAKI, J. N., SIEM, T. H. & BROUWER, R. (I977). Anti-e and rheumatoid factor activity in hepatitis B. Lancet i, IOO8-IOO9. KRUGMAN, S., HOOFNAGLE, J. t{., GERETY, R. J., KAPLAN, P. M. & GERIN, J. L (I974). D N A polymerase a c t i v i t y and antibody to hepatitis B core antigen. New England Journal of Medicine 29 o, I33I-I335. MAGNIUS, L. O. & ESPMARK, J. A. (1972)- New specificities in Australia antigen positive sera distinct from LeBouvier determinants. Journal of Immunology lO9, 1oi7-IO21. MORITSUGU, Y., GOLD, J. W. M., WAGNER, J., DODD, R. Y. & PURCELL, R. H. (I975). Hepatitis B core antigen. Detection of antibody by radioimmunoprecipitation. Journal of Immunology i i 4 , I792-I798. NEUa~,ATH,A. R., HUANG, S.-N. & STRICK, N. (I978). Some properties of hepatitis B core antigen isolated from serum of infected humans. Journal of General Virology 39, (in the press). NEURATH,A. R. & STRICK,N. (1977)- Host specificity of a serum marker for hepatitis B: evidence that ' e antigen' has the properties of an immunoglobulin. Proceedings of the National Academy of Sciences of the United States of America 74, I7O2-I7O6. PRINCE, A. M., BROTMAN, B., GRADY, G. F., KUHNS, W. J., HAZZI, C., LEVINE, R. W. & MILIAN, S. J. 0974). Longincubation post-transfusion hepatitis without serological evidence of exposure to hepatitis B virus. Lancet ii, 241-246. PURCELL, R. H., GERIN, J. L., ALMEIDA, J. B. & HOLLAND, P. V. (1973). Radioimmunoassay for the detection of the core of the Dane particle and. antibody to it. Intervirology 2, 231-243. PURCELL, R. H.~ LONDON, W. T.~ MCAULIFFE, V. J., PALMER, A. E., KAPLAN, P. M., GERIN, J. L., WAGNER, J.,

POPVER, H., LVOVSKV,~., WONG, O. C. & LEVY, H. a. (1976). Modification of chronic hepatitis B virus infection in chimpanzees by administration of an interferon inducer. Lancet ii, 757-761. ROBINSON, W. S. & GREENMAN,R. L. (I974). D N A polymerase in the core of the h u m a n hepatitis B virus candidate. Journal of Virology I3, 1231-1236. SELA, M. (1966). Chemical studies of the combining sites of antibodies. Proceedings of the Royal Society, London x66, 188-2o6. STEVENSON, G. T. (I974). In Structure and Function of Plasma Proteins, vol. I, chapter 8, pp. 223-263. Edited by A. C. Allison. London: Plenum Publishing Company, Ltd. SVEHAG,S. E. (1966). Diversity of antibodies formed against viruses. In Antibodies to Biologically Active Molecules, vol. I, pp. 3Ol-348. Proceedings of the 2nd Meeting of the Federation of European Biochemical Societies, Vienna, 1965. Oxford: Pergamon Press. SZMUNESS, W., HOOFNAGLE, J. H., STEVENS, C. E. & PRINCE, A. M. (1976). A n t i b o d y a g a i n s t t h e h e p a t i t i s t y p e B

core antigen. A new tool for epidemiologic studies. American Journal of Epidemiology Io4, 256-262. TAKAHASHI, K., IMAI, M., TSUDA, F., TAKAHASHI, T., MIYAKAWA, Y. & MAYUMI, M. (1976). A s s o c i a t i o n o f D a n e

particles with e antigen in the serum of asymptomatic carriers of hepatitis B surface antigen. Journal o f

Immunology XXT, IO2-IO5. TEDDER, R. S. & BRIGGS, M. (1977). Anti-e and rheumatoid factor activity in hepatitis B. Lancet i, 1262-1263. TSUDA, F., TAKAHASHI, T., TAKAHASHI, K., MIYAKAWA, Y. & MAYUMI, M. (1975). Determination of antibody to

hepatitis B core antigen by means of immune adherence haemagglutination. Journal of Immunology I I 5 , 834-838. VYAS, G. N. & ROBERTS, I. M. (1977). Radioimmunoassay of hepatitis B core antigen and antibody with autologous reagents. Vox Sanguinis (in the press). VYAS, G. N., ROBERTS, I., MACKAY, I. R. & GUST, I. D. (1975). I m m u n o l o g i c m e c h a n i s m s i n h e p a t i t i s 13 a s s a y e d

by antigen-binding lymphocytes. American Journal of Medical Sciences 270, 241-246.

(Received 17 August I 9 7 7 )