- Email: [email protected]
Development and evaluation of an ELISA using secreted recombinant glycoprotein B for determination of IgG antibody to herpes simplex virus
Journal of Virological Methods, 34 (1991) 57-70 0 1991 Elsevier Science Publishers B.V. / All rights reserved / 0166-0934/91/$03.50 ADONIS 016609349100414S
57
VIRMET 01210
Development and evaluation of an ELISA using secreted recombinant glycoprotein B for determination of IgG antibody to herpes simplex virus M. Grazia Revello’, Rita Gualandri2, Roberto Manservigi2 and Giuseppe Gerna’ ’ Virus Laboratory, Institute of Infectious Diseases, University of Pavia, IRCCS Policlinico S. Matteo, Pavia, Italy and 21nstitute of Microbiology, University of Ferrara, Ferrara, Italy (Accepted 16 April 1991)
Summary
An ELISA for the determination of IgG antibody to herpes simplex virus (HSV) was developed using a secreted recombinant HSV-1 glycobrotein B (gB1s) as a solid phase. The clinical validity of the ELISA was established by testing different groups of sera containing HSV-1, HSV-2, or mixed antibody, in parallel with gB-1s ELISA and conventional HSV- 1/HSV-2 ELISA. The new gB-1s ELISA detected HSV-l/HSV-2 antibody in sera from 48 subjects with either HSV-1 or HSV-2 past infection as well as in sera from 20 patients with primary infections by either serotype, in complete agreement with the results obtained using conventional ELISA. In 7 patients with HSV-1 encephalitis the kinetics of the gB-1s serum/cerebrospinal fluid antibody-titre ratio paralleled that of conventional ELISA over a period of time of up to 4 years. Acute and convalescent-phase sera from 28 patients with acute infections by human herpesviruses other than HSV did not show a significant cross-reactivity with gB-1s. In conclusion, gB-1s ELISA is a reliable assay for determination of HSV immune status as well as for diagnosis of both primary HSV-1 and HSV-2 infections and for diagnosis of HSV-1 encephalitis. Glycoprotein B; Herpes simplex virus; Recombinant DNA technology; IgG antibody
ELISA;
Correspondence to: Prof. Giuseppe Gema, Virus Laboratory, Institute of Infectious Diseases, University of Pavia, IRCCS Policlinico S. Matteo, 27100 Pavia, Italy.
58
Introduction Humoral immune response to herpes simplex virus (HSV) type 1 and type 2 infections is partly common and partly type-specific. The relative proportion of the two types of antibody varies according to the time interval after onset of infection, type of infecting HSV strain and preexisting antibody of either HSV type. That is the basis upon which diagnostic virology laboratories determine IgG antibody to HSV by testing human sera against both HSV-1 and HSV-2 antigens. Enzyme-linked immunosorbent assay (ELISA) represents the most widely used assay system for determination of the immune status to HSV. However, standardization of different batches of antigen is difficult to achieve and a control antigen is required as a specificity control of test results. On the other hand, type-specific antibody can only be determined by experienced laboratories using microneutralization procedures, while recently proposed methods based on the use of immunoaffnity-purified HSV glycoproteins, such as gG-1 and gG-2 (Lee et al., 1985, 1986), require procedures too cumbersome for general use in a diagnostic laboratory. In order to develop an enzyme immunoassay which could be standardized for determination of both the HSV immune status and diagnosis of acute HSV1 and HSV-2 infections, we used a secreted recombinant HSV-1 glycoprotein B (gB-1s) constitutively expressed in human cells transfected with a BK virus recombinant episomal vector (Manservigi et al., 1990a). It is known that gB is an essential glycoprotein of HSV and a major target for humoral and cellmediated immune response in the infected host (Corey and Spear, 1986) and has been shown to be a strong immunogen in eliciting neutralizing antibody (Glorioso et al., 1984; Dix and Mills, 1985; Manservigi et al., 1990a,b).This glycoprotein has been suggested as a promising candidate for the development of a subunit vaccine (Pack1 et al., 1987; Manservigi et al., 1988). While the mature and completely glycosylated form of gB-1 is 903 amino acids in length ahd 120 kDa in size (Campadelli-Fiume and Serafini-Cessi, 1985), gB-ls, which is deprived of the transmembrane portion of the molecule, is 690 amino acids long and 91.6 kDa in size (Manservigi et al., 1990a). In recent studies, gB-1s has been shown to protect mice against lethal and latent HSV-1 infections and to prevent CNS disease in rabbits, while reducing the severity of herpetic keratitis (Manservigi et al., 1990a,b). In the present study, the development of an ELISA using gB-1s as a solid phase is described. The test is capable of detecting small amounts of either HSV-1 or HSV-2 antibody in human sera from patients with remote, primary or recurrent HSV-l/HSV-2 infections as well as from patients with HSV encephalitis. The assay does not require handling of infectious material and may be performed in the absence of control antigen.
59
Material and Methods
Cloning and expression of gB-Is in human cells The gB-1 gene of HSV-1, strain F, deleted of 693 nucleotides encoding the transmembrane anchor sequence and reconstructed with the extramembrane and intracytoplasmic domains, was cloned under the control of the Rous Sarcoma virus promoter in the episomal replicating vector pRP-RSV (Manservigi et al., 1990a). gB-1s was secreted into the culture medium of pRP-RSV-gB-Is-transformed 293 cells at a concentration of 0.1-0.2 pg/ml. The secreted product was identified in both radioimmunoprecipitation and ELISA by using monoclonal antibodies to HSV-1 gB (Manservigi et al., 1988) and gB1s developed in the laboratory. Using the same monoclonal antibodies, gB-1s was immunoaflinity-purified from 20 x concentrated (by ultrafiltration) and dialyzed cell culture medium. ELISA procedure Either purified or crude concentrated culture medium preparations of gB-1s were used as a solid phase in the ELISA. Following initial titration, the optimal concentration of gB-1s was found to be between 10 and 50 ng/well. Wells of 96well polystyrene microplates (M129A Dynatech, Plochingen, F.R.G.) were coated overnight at room temperature with the optimal dilution of gB-1s in 0.05 M sodium carbonate-bicarbonate buffer, pH 9.6. After washings, 100 ~1 of test sera, diluted 1:50 in phosphate-buffered saline (PBS) containing 1% foetal calf serum (FCS) and 0.1% Tween 20, were incubated for 2 h at 37°C. Plates were then washed with, PBS-Tween 20 and 100 ~1 of optimally diluted peroxidase-conjugated goat anti-human IgG (Cappel Laboratories, Cooper Biomedical, Malvern, PA, U.S.A.) were added to each well. Incubation was continued for 1 h at 37°C. After washing, enzymatic activity was detected by adding the substrate solution for 30 min at room temperature. The latter was prepared immediately before use by dissolving 3.4 mg of ortho-phenylenediamine in 10.0 ml of citrate phosphate buffer (0.05 M phosphate, 0.02 M citrate, pH 5.5) and then adding 40 ~1 of 3% H202. Absorbance values were recorded at 492 nm using a microtitre plate spectrophotometer (Titertek Multiskan, Flow Laboratories, Irvine, U.K.). The cut-off was set at 0.10 absorbance following testing of 39 HSV-negative sera. When required, quantitation of IgG antibody to gB-1s was carried out by constructing a calibration curve, which was obtained by plotting the mean absorbance value of each of five standard sera (containing different amounts of specific antibody) versus the amount of gB-ls-specific IgG expressed as antibody titre. In parallel, all sera were tested by conventional ELISA for HSV-l/HSV-2 IgG antibody determination, as previously reported (Revello and Gerna, 1988). In conventional ELISA, glycine-extracted HSV-1 and HSV-2 crude antigens as well as uninfected Vero cell control antigen were used. This assay was calibrated optimally to allow
60
typing of HSV antibody Gerna, 1988). Microneutralization
in human
sera, as reported
previously
(Revello
and
test
The microneutralization assay was carried out using growing Vero cells by testing serial twofold dilutions of heat-inactivated serum samples mixed with equal amounts of virus suspensions containing 100 50% tissue culture infectious doses of either HSV-1 or HSV-2 (Revello and Gerna, 1988). Corrected titres (tc) for each serum were calculated according to Stalder et al. (1975), and typing of HSV neutralizing antibody was achieved by calculating the tc difference between the titre to HSV-1 and that to HSV-2. Patients and sera The following groups of patients and sera were examined in parallel by gB- 1s ELISA and conventional HSV-l/HSV-2 ELISA: l 39 HSV antibody-negative sera from 39 healthy individuals; l 53 sequential sera from 14 patients with primary HSV-1 infection and 17 sequential sera from 3 patients with primary HSV-2 infection (in the absence of prior HSV-1 infection); l 48 sera from 48 patients, of which 23 contained HSV-1 antibody, 13 HSV-2 antibody and 12 dual HSV antibody, as determined previously by microneutralization; l acute- and convalescent-phase sera from 28 patients with acute infections caused by human herpesviruses other than HSV, i.e. varicella-zoster virus (VZV), human cytomegalovirus (HCMV), Epstein-Barr virus (EBV), and human herpesvirus 6 (HHV-6), and from 16 patients with primary HSV-1 infection; l 35 sequential paired serum-cerebrospinal fluid (CSF) samples from 7 patients with herpes simplex encephalitis. HSV isolation and typing Material from mucocutaneous lesions was collected by swabs, for inoculation of Vero and diploid cell cultures for virus isolation, and by scraping, for preparation of cell smears for direct virus identification and typing. HSV isolates were typed by indirect ELISA, as reported previously (Gerna et al., 1983), whereas direct typing on cell smears from clinical specimens was done by direct fluorescent antibody test, using in parallel fluorescein-labeled type-specific monoclonal antibodies both produced in our laboratory and available commercially (Syva Micro-Trak, Syva Company, Palo Alto, CA, U.S.A.).
61
Development of the indirect ELISA for determination of IgG antibody to gB-1s
Initially, the optimal concentration of gB- 1s to be used in the indirect ELISA for IgG antibody detection was determined. For this purpose, human sera with past and primary HSV-1 infections, containing only HSV-1 antibodies, were used. Low levels of HSV-1 antibody could be clearly detected when using a concentration of gB-1s of 10-15 ng/well. However, when early convalescentphase sera from primary HSV-2 infections in subjects with no prior experience of HSV-1 infection were tested, the amount of gB-is/well had to be increased to 40-50 ng in order to allow detection of low levels of early HSV-2 antibody. Afterwards, the cut-off was determined on 39 HSV-negative sera previously characterized by neutralization and ELISA. The mean Ad92 was 0.03 + 0.03. Thus, a cut-off of 0.10 was selected for further testing (0.03 + 2SD). Antibody response to gB-1s in primary HSV-1 and HSV-2 infections
All subsequent tests were run in parallel using the gB- 1s ELISA and the HSVl/HSV-2 ELISA. Initially, two monoclonal antibodies to gB and four to gB-1s a.0
_
1s
.
1.0
.
1.7
.
1.0
.
000 0
0 1s.
1.4
.
1.5
.
11
.
1.1
.
1.0
.
0.0
.
0
0
0
.
0 :
ao. f
a7
0 .
-
0
.
0.
.
0
Dae.
0 Eos. 5
00
.b
I--
O
0
0
0
C‘OJ. az.
Fig. 1. Absorbance values (A& in the gB-1s vs. HSV-I/HSV-2 ELISA of: 53 sera collected from 14 patients during the first month (0) of a primary HSV-1 infection or afterwards (0); 17 sera collected from 3 patients during the first month (m) of a primary HSV-2 infection or afterwards (0).
62
(shown in competitive binding assays to recognize different gB epitopes) were tested in both ELISAs. All reacted in the gB-1s ELISA with titres of 1O-5-1O-7, whereas none reacted at a dilution of 10e3 in the HSV-l/HSV-2 ELISA. Measurement of IgG antibody to gB-1s during primary HSV-1 infections was achieved by examining 53 sequential sera from 14 patients undergoing primary HSV- 1 infection. Among 34 sera collected within 30 days after onset of symptoms, 9 collected within the first 5 days were negative by both ELISA systems, 23 were concordant positive and 2 (collected 3 and 5 days after onset, respectively) were weak positive by gB-1s ELISA only (Fig. 1). Ad92 values measured using gB-1s ELISA and the HSV-l/HSV-2 ELISA were also grossly comparable, with the exception of 2 sera collected within 30 days and 2 sera collected afterwards, which gave significantly higher (greater than a fourfold difference in titre) reactivity in the gB-1s ELISA. Comparable results were observed when 17 sera from 3 patients undergoing primary HSV-2 infection (in the absence of prior HSV- 1 infection) were examined. Of 6 sera collected within 30 days, 5 were detected by both ELISAs and one only by the HSV-2 ELISA
.
5.8
.
5.7
.
33
.
3.5
.
a1
.
0 . .
. . . . . .
. . .
2.9
.
2.7
.
2.5
.
23
.
2.1
.
1.0
.
1.7
.
1.s
.
1.3
.
1.1
_
‘;.
0.2
_
f
0.7
.
z
03
.
0.3
.
0.1
.
5 s :
8.0
oe
.
0
.
0
0 . .
.
e
0
8
. .
.
0 0 0 0 0 0
. 8
0
. .
. .‘.
Am
WIT”
ttONOLOOW8
HW
ANTKILN
Fig. 2. Comparison of ELISA reactivities with gB-1s vs. HSV-l/HSV-2 crude antigens in 48 sera previously typed for HSV antibody by neutralization. Sera were divided into three groups containing: 23, HSV-1 (0); 13, HSV-2 (O), and 12, mixed (0) antibody, respectively. For sera containing mixed antibody, the higher Adp2 value in the HSV-l/HSV-2 ELISA was plotted.
63
with an A492 value close to cut-off. In addition, all 11 sera collected after 30 days were found positive by both systems. However, most late sera showed a tendency to give higher HSV-2 than gB-1s A492values (Fig. 1). Thus, although HSV-1 and HSV-2 gBs share a high degree of homology, at least some sera from primary HSV-1 infections seem to recognize gB-1s somewhat better than sera from HSV-2 infections. Determination of HSV immune status in patients with HSV-I, HSV-2 or mixed antibody The HSV immune status was determined by gB-1s ELISA and HSV-l/HSV2 ELISA in sera from 48 patients, in which HSV antibody type was previously determined by microneutralization. As shown in Fig. 2, all test sera were detected as positive by both ELISA systems. However, HSV-1 and mixed antibody sera gave A 492 values higher in the gB-1s ELISA as compared to HSV-l/HSV-2 ELISA, whereas most HSV-2 sera showed an equivalent reactivity between the two ELISAs. These data again suggest that HSV-1 antibody reacts more strongly with gB-1s than HSV-2 antibody. Study of the heterologous antibody response to gB-Is in patients with acute infections by HHVs other than HSV Since there are high levels of homology among gB genes of different human herpesviruses, we investigated the level of heterologous antibody response to 22 24
(MllAnv)
(CIIUAW)
(rnlllill)
( CIImAmv)
Fig. 3. Heterologous antibody response to gB-1s in acute and convalescent-phase sera from 28 patients with acute HHV infections other than HSV (6 with zoster, 5 with chickenpox, 10 with primary HCMV, 5 with primary EBV and 2 with primary HHV-6 infection). Homologous antibody response in acute and convalescent sera from 16 patients with primary HSV-1 infection is reported for comparison. Numbers inside columns indicate gB-Is-negative (shaded areas) or -positive (blank areas) sera. A, acute; C, convalescent serum.
64
gB-1s in acute and convalescent sera from 28 patients with acute infections by HHVs other than HSV (6 with zoster, 5 with chickenpox, 10 with primary HCMV infection, 5 with EBV infectious mononucleosis and 2 with primary HHV-6 infection). For comparison, acute and convalescent sera from 16 patients with primary HSV-1 infection were examined. Mean Ad92 values obtained in the gB-1s ELISA with different groups of sera are reported in Fig. 3. No significant difference (greater than a twofold dilution) in mean Ad92 between acute and convalescent sera was observed in any of the 5 groups of HSV-positive patients (where high gB-1s antibody titres reflected past HSV-1 infections), whereas seroconversion from seronegative to seropositive was found in all 16 TABLE 1 Heterologous Patient No.
Zoster RS A C R22 A C R25 : R26 2 R30 A C R31 A C Varicella 6806A A C 6163A A C 6251A f? 12198 A C 11968
antibody
response to gB-1s in 6 patients with zoster and 5 patients with varicella A 492
vzv
gB-1s (titre)
HSV-1 (titre)
1.12 2.28
2.00 (3,100) 1.99 (3,100)
1.28 (6,000) 1.12(3,500)
1.29 2.47
2.23 (4,000) 2.60 (7,000)
1.18(4,500)
0.75 1.97
2.52 (5,600) 2.64 (7,000)
0.94 (2,000) 1.04(2,500)
1.66 2.34
2.60 (7,000) 2.60 (7,000)
1.36 (8,000) 1.44 (10,000)
1.49 2.30
2.70 (7,000) 2.64 (7,000)
1.40 (9,500) 1.22 (7,000)
0.34 2.17
0.06 (< 50) 0.04 (< 50)
0.01 (<50) 0.07 (< 50)
0.15 0.71
2.76 (7,000) 3.28 (12,600)
1.40 (9,500) 1.59 (18,000)
0.03 0.35
2.68 (7,000) 2.16 (3,500)
1.40 (9,500) 1.24 (5,500)
0.13 1.12
1.63 (2,000) 2.37 (4,000)
1.27 (6,000) 1.31 (6,500)
0.03 0.57
1.65 (2,000) 2.41 (4,200)
1.18 (4,500) 1.18(4,500)
0.12 0.67
0.05 (< 50) 0.04 (< 50)
0.01 (<50) 0.03 (< 50)
A = acute serum; C = convalescent
serum.
1.19 (4,500)
65
patients with primary HSV-1 infection. When paired sera from individual patients of each group were examined, it was found that only one patient with primary HCMV infection showed a fourfold rise in antibody titre to gB-1s. However, this patient was a heart transplant recipient possibly undergoing an HSV recurrence in the post-transplant period. With respect to VZV acute infections, in 3 of 4 HSV-1 seropositive patients with chickenpox a fair rise in Ad92 with gB-1s was observed, but it corresponded only to a twofold rise in antibody titre and was thus considered not clinically significant (Table 1). In only one of 5 patients with zoster infection a slight increase in AdP2with gB-1s was observed, again within a twofold dilution difference in titre. Comparable results with paired sera from patients with varicella or zoster were obtained with conventional ELISA. In addition, 6 patients (1 with zoster, 1 with chickenpox, 1 with primary HCMV, 2 with primary EBV and 1 with primary HHV-6 infection) negative for gB-1s (and HSV) antibody in the acute serum remained seronegative in the convalescent phase of the relevant infection. Antibody to gB-Is in patients with herpes simplex encephalitis Thirty-five sequential paired serum-CSF samples from 7 patients with HSV encephalitis (4 by a primary and 3 by a recurrent HSV-1 infection) were examined for determination and ouantitation of IgG antibody response by gB1s and HSV-1 ELISAs. The kinetics of the IgG antibody levels in two representative patients (1 with primary and 1 with recurrent HSV-1 infection) are reported in Fig. 4. The curves of IgG antibody to gB-1 s and HSV-1 in
0.0
.
4.1
.
4.4
.
4.1
.
a.0
.
0.0
.
LE
.
E.0
.
l.0
.
*.a
Fig. 4. Kinetics of virus-specific IgG antibody as determined by ELISA using gB-1s (-) and HSV-1 crude antigen (- - -) in sequential paired serum (0) and CSF (A) samples from two patients with HSV encephalitis due to a primary (A) and to a recurrent (B) HSV-1 infection, respectively. (a- - -a), HSV-1 serum IgG; (A- - -A), HSV-I CSF IgG; (a----a), gB-1s serum IgG; (A---&, gB-Is CSF IgG antibody.
66
. 6.0
2
r
0.87
’ 14 IS) DAYS
APIER
(No PAIRED
vi*
I21
28 12)
ONSET OF NEUROLOQICAL SANHES
JO-180 6%
m-ass (3)
an (3)
2 i
4vn (4)
SIYPTOYS
AVAILABLE 1
Fig. 5. Comparison of the kinetics of gB-1s and HSV-1 mean serum/CSF antibody titre ratios vs. Link index in 7 patients with HSV encephalitis.
serum and CSF are almost parallel in both patients. When data available on sequential paired samples from the 7 patients were considered together, the mean serum/CSF antibody titre ratio obtained with the gB-1s ELISA gave a curve generally parallel to that provided by the mean ratio with the HSV-1 ELISA (Fig. 5). The lowest levels of both ratios were reached concomitantly with the highest value of the Link index (Tibbling et al., 1977) at the end of the first month after onset of symptoms (the Link index represents, when > 0.67, a marker of intrathecal IgG synthesis); then ratios rose slowly and progressively until they reached the cut-off value of 20 after 24 years. Contrariwise, the Link index decreased, approximating the cut-off value of 0.67 after 4 years. Again, the antibody response measured by the gB-1s ELISA appeared to overlap that measured by the HSV ELISA.
Discussion The results of the present study clearly demonstrate that gB-1s is suitable for use in an indirect ELISA for determination of IgG antibody to either HSV-1 or HSV-2 and that the gB-1s ELISA can promptly replace the conventional HSVl/HSV-2 ELISA for either determination of the immune status to HSV or for the diagnosis of acute HSV infection. When developing a new immunoassay using a recombinant DNA product, the first question to be addressed is whether detection of the immune response
to a single viral protein may be sufficiently representative of the whole antibody response elicited in the patient by viral infection. It was shown previously that antibody to gB is among the first to appear during the early convalescent phase of HSV infections involving either mucocutaneous sites or the CNS (Eberle and Mou, 1983; Kahlon et al., 1986; Grimaldi et al., 1988). The comparable kinetics of the IgG antibody response observed in our study in acute HSV infections when using gB-1s and conventional ELISAs suggest that the same type of antibody was detected by the two assays. However, this was not the case, since monoclonal antibodies to gB and gB-ls, while highly reactive in the gB-1s ELISA, were practically nonreactive in the conventional ELISA, where gB was poorly represented in the glycine-extracted antigens. Thus, although the profile of the antibody response did not appear to display significant variation among the two assays, the type of IgG antibody detected by the two ELISA assays was totally different as regards antigen specificity. Although the amino acid sequences of gB-1 and gB-2 are 86% conserved (Bzik et al., 1986; Stuve et al., 1987), type-specific epitopes identified in the gB molecule (Kousoulas et al., 1988; Eberle and Courtney, 1989) may cause sera from patients with HSV-1 antibody to react in gB-1s ELISA more markedly than sera from patients with HSV-2 antibody. This might have been the reason why, in the initial development of gB-1s ELISA, the antigen concentration had to be increased fourfold, in order to enable the test to detect low amounts of HSV-2 antibody in sera from patients with HSV-2 infection. Recently, human sera were tested by radioimmunoprecipitation using different forms of truncated gB, showing that the greater the deletion the lower was the percentage of reactive sera (Ali and Forghani, 1990). In our study, gB-Is, i.e. a form of gB deprived only of the transmembrane region and of two adjacent short intracytoplasmic and extramembrane sequences, was shown to retain all immunoreactivity versus human sera in an indirect ELISA. The gB-1s used in the ELISA could be readily purified from the culture medium of cells expressing gB- 1s. However, it should be noted that purification of gB-1s was not an essential step for the development of the ELISA, since coating of the solid phase by crude cell culture medium provided comparable results in terms of sensitivity and specificity (unpublished results). Nevertheless, because of a fair degree of variability in the gB-1s content among different batches of culture medium, standardization of the ELISA assay was achieved only by using purified gB- 1s. A high degree of homology is shared by gB of HSV and gBs of the other HHVs, including VZV (Davidson and Scott, 1986; Keller et al., 1986), HCMV (Cranage et al., 1986) and EBV (Pellet et al., 1985). In particular, a crossreactive serological response between HSV and VZV has been reported repeatedly (Kapsenberg, 1964; Schmidt et al., 1969; Harbour and Caunt, 1979; Bernstein et al., 1990). In addition, antigenic relationships have been detected between some VZV glycoproteins and some HSV glycoproteins (with special reference to gB) by immunoprecipitation using hyperimmune sera to HSV and VZV (Shiraki et al., 1982) as well as monospecific antibody to VZV gp3 or
68
monoclonal antibody to gB of HSV (Kitamura et al., 1986). In the present study, by testing in the gB-Is ELISA several paired sera from patients with acute infections caused by HHVs other than HSV, a significant rise in heterotypic antibody response to gB-1s was not found. This suggests that, despite the high homology among gBs of different herpesviruses, crossreactivity does not seem to occur at the level of serological response to HHV infections, at least in the gB-1s ELISA. This is in agreement with previous observations showing that cross-reactions observed with some tests, such as complement fixation, were not detectable by other assays, such as gel precipitation (Schmidt et al., 1969). In conclusion, the gB-1s ELISA represents a reliable, sensitive and specific immunoassay for determination of IgG antibody to HSV which can be standardized. It can be utilized for both determination of the immune status HSV or for serological diagnosis of primary HSV-1 or HSV-2 infections, as well as for early retrospective diagnosis of HSV encephalitis. For determination of HSV type-specific antibodies, we have recently been able to obtain expression of gG of HSV-2 in the 293 human cell line both in a membranebound (gG-2) and secreted form (gG-2s) (Sabbioni et al., unpublished results). Both gG-2 and gG-2s were shown to be suitable for determination of HSV-2specific antibodies.
Acknowledgements This work was partially supported by the Minister0 della Sanita, Istituto Superiore di Sanita, Progetto Nazionale AIDS 1990, contract No. 4204 18, and by Consiglio Nazionale delle Ricerche, Progetto Finalizzato Biotecnologie e Biostrumentazione, contract Nos. 89.00164.PF70, 90.00046.PF70 and 90.00 164.PF70. We thank Linda d’Arrigo for her help with the English revision.
References Ali, M.A. and Forghani, B. (1990) Serologic specificity of antibodies to herpes simplex virus glycoprotein B present in human sera: analysis by transient expression of glycoprotein Bderivatives. Arch. Virol. 112, 129-138. Bernstein, D.I., Frenkel, L.M., Bryson, Y.J. and Myers, M.G. (1990) Antibody response to herpes simplex virus glycoproteins gB and gD. J. Med. Virol. 30, 4549. Bzik, D.J., Debroy, C., Fox, B.A., Pederson, N.E. and Person, S. (1986) The nucleotide sequence of the gB glycoprotein gene of HSV-2 and comparison with the corresponding gene of HSV-I. Virology 155, 322-333. Campadelli-Fiume, G. and Seralini-Cessi, F. (1985) Processing of the oligosaccharide chains of herpes simplex virus type 1 glycoproteins. In: B. Roizman (Ed), The Herpesviruses, Vol. 3, Plenum Press, New York, pp. 357-382. Corey, L. and Spear, P.G. (1986) Infection with herpes simplex viruses. N. Engl. J. Med. 314, 686 691 and 749-757.
69 Cranage, M.P., Kousarides, T., Bankier, A.T., Satchwell, S., Weston, K., Tomlinson, P., Barrell, B., Hart, H., Bell, S.E., Minson, AC. and Smith, G.L. (1986) Identification of the human cytomegalovirus glycoprotein B gene and induction of neutralizing antibodies via its expression in recombinant vaccinia virus. EMBO J. 5, 3057-3063. Davidson, A.J. and Scott, J.E. (1986) The complete DNA sequence of varicella-zoster virus. J. Gen. Virol. 67, 1759-1816. Eberle, R. and Courtney, R.J. (1989) Topological distribution of virus-specific and cross-reactive antigenic determinants on the gB glycoprotein of the herpes simplex viruses. J. Med. Virol. 27, 309-3 16. Eberle, R. and Mou, S.W. (1983) Relative titres of antibodies to individual polypeptide antigens of herpes simplex virus type 1 in human sera. J. Infect. Dis. 148, 436444. Gerna, G., Battaglia, M., Revello, M.G. and Torsellini Gema, M. (1983) Typing of herpes simplex virus isolates by enzyme-linked immunosorbent assay: comparison between indirect and doubleantibody sandwich techniques. J. Clin. Microbial. 17, 942-944. Grimaldi, L.M.E., Roos, R.P., Manservigi, R., Spear, P.G., Lakeman, F.D. and Whitley, R.J. (1988) An isoelectric focusing study in herpes simplex virus encephalitis. Ann. Neurol. 24, 227232. Harbour, D.A. and Caunt, A.E. (1979) The serologic relationship of varicella-zoster virus to other primate herpes viruses. J. Gen. Virol. 45, 469477. Kahlon, J., Lakeman, F.D., Ackermann, M. and Whitley, R.J. (1986) Human antibody response to herpes simplex virus specified polypeptides following primary and recurrent infection. J. Clin. Microbial. 23, 725-730. Kapsenberg, J.G. (1964) Possible antigenic relationship between varicella zoster virus and herpes simplex virus. Arch. Ges. Virusforsch 15, 6673. Keller, P.M., Davidson, A.J., Lowe, R.S., Bennet, C.D. and Ellis, R.W. (1986) Identification and structure of the gene encoding gpI1, a major glycoprotein of varicella-zoster virus. Virology 152, 181-191. Kitamura, K., Namazue, J., Campo-Vera, H., Ogino, T. and Yamanishi, K. (1986) Induction of neutralizing antibody against varicella-zoster virus (VZV) by VZV gp3 and cross-reactivity between VZV gp3 and herpes simplex viruses gB. Virology 149, 74-82. Kousoulas, K.G., Huo, B. and Pereira, L. (1988) Antibody-resistant mutations in cross-reactive and type-specific epitopes of herpes simplex virus 1 glycoprotein B map in separate domains. Virology 166, 423343 1. Lee, F.K., Coleman, R.M., Pereira, L., Bailey, P.D., Tatsuno, M. and Nahmias, A.J. (1985) Detection of herpes simplex virus type 2-specific antibody with glycoprotein G. J. Clin. Microbial. 22, 641-644. Lee, F.K., Pereira, L., Griffin, C., Reid, E. and Nahmias, A.J. (1986) A novel glycoprotein for detection of herpes simplex virus type l-specific antibodies. J. Virol. Methods 14, 11 l-l 18. Manservigi, R., Gualandri, R., Negrini, M., Albonici, L., Milanesi, G., Cassai, E. and BarbantiBrodano, G. (1988) Constitutive expression in human cells of herpes simplex virus type 1 glycoprotein B gene cloned in an episomal eukaryotic vector. Virology 167, 284-288. Manservigi, R., Grossi, M.P., Gualandri, R., Balboni, P.G., Marchini, A., Rotola, A., Rimessi, P., Di Luca, D., Cassai, E. and Barbanti-Brodano, G. (1990a) Protection from herpes simplex virus type 1 lethal and latent infections by secreted recombinant glycoprotein B constitutively expressed in human cells with a BK virus episomal vector. J. Virol. 64, 431436. Manservigi, R., Incorvaia, C., Di Luca, D., Rotola, A., Balboni, P.G., Sebastiani, A., Rossi, A. and Cassai, E. (199Ob) Experimental keratitis in rabbits by human HSV-1 variants: prevention and treatment. J. Med. Virol. 32, 148-154. Packl, C., Burke, R.L., Stuve, L., Sanchez-Pescador, L., Nest, G.V., Masdiarz, F. and Dina, D. (1987) Expression of cell-associated and secreted forms of herpes simplex virus type 1 glycoprotein B in mammalian cells. J. Virol. 61, 315-325. Pellet, P.E., Biggin, M.D., Barrell, B. and Roizman, B. (1985) Epstein-Barr virus genome may encode a protein showing significant amino acid and predicted secondary structure homology with glycoprotein B of herpes simplex virus 1. J. Virol. 56, 807-813.
70 Revello, M.G. and Gema, G. (1988) Herpes simplex virus type-specific antibody determination by enzyme-linked immunosorbent assay in human sera. Serodiagn. Immunother. Infect. Dis. 2, 415422. Schmidt, N.J., Lennette, E.H. and Magofftn, R.L. (1969) Immunological relationship between herpes simplex and varicella-zoster viruses demonstrated by complement-fixation, neutralization and fluorescent antibody test. J. Gen. Virol. 4, 321-328. Shiraki, K., Okuno, T., Yamanishi, K. and Takahashi, M. (1982) Polypeptides of varicella-zoster virus (VZV) and immunological relationship of VZV and herpes simplex virus (HSV). J. Gen. Virol. 61, 255-269. Stalder, H., Oxman, M.N. and Hermann, K.L. (1975) Herpes simplex virus neutralization: a simplification of the test. J. Infect. Dis. 131, 430-432. Stuve, L.L., Brown-Shimer, S., Packl, C., Najarian, R., Dina, D. and Burke, R.L. (1987) Structure and expression of the herpes simplex virus type 2 glycoprotein gB gene. J. Virol. 61, 326335. Tibbling, G., Link, H. and dhman, S. (1977) Principles of albumin, and IgG analyses in neurological disorders. I. Establishment of reference value. Stand. J. Clin. Lab. Invest. 37, 385392.
57
VIRMET 01210
Development and evaluation of an ELISA using secreted recombinant glycoprotein B for determination of IgG antibody to herpes simplex virus M. Grazia Revello’, Rita Gualandri2, Roberto Manservigi2 and Giuseppe Gerna’ ’ Virus Laboratory, Institute of Infectious Diseases, University of Pavia, IRCCS Policlinico S. Matteo, Pavia, Italy and 21nstitute of Microbiology, University of Ferrara, Ferrara, Italy (Accepted 16 April 1991)
Summary
An ELISA for the determination of IgG antibody to herpes simplex virus (HSV) was developed using a secreted recombinant HSV-1 glycobrotein B (gB1s) as a solid phase. The clinical validity of the ELISA was established by testing different groups of sera containing HSV-1, HSV-2, or mixed antibody, in parallel with gB-1s ELISA and conventional HSV- 1/HSV-2 ELISA. The new gB-1s ELISA detected HSV-l/HSV-2 antibody in sera from 48 subjects with either HSV-1 or HSV-2 past infection as well as in sera from 20 patients with primary infections by either serotype, in complete agreement with the results obtained using conventional ELISA. In 7 patients with HSV-1 encephalitis the kinetics of the gB-1s serum/cerebrospinal fluid antibody-titre ratio paralleled that of conventional ELISA over a period of time of up to 4 years. Acute and convalescent-phase sera from 28 patients with acute infections by human herpesviruses other than HSV did not show a significant cross-reactivity with gB-1s. In conclusion, gB-1s ELISA is a reliable assay for determination of HSV immune status as well as for diagnosis of both primary HSV-1 and HSV-2 infections and for diagnosis of HSV-1 encephalitis. Glycoprotein B; Herpes simplex virus; Recombinant DNA technology; IgG antibody
ELISA;
Correspondence to: Prof. Giuseppe Gema, Virus Laboratory, Institute of Infectious Diseases, University of Pavia, IRCCS Policlinico S. Matteo, 27100 Pavia, Italy.
58
Introduction Humoral immune response to herpes simplex virus (HSV) type 1 and type 2 infections is partly common and partly type-specific. The relative proportion of the two types of antibody varies according to the time interval after onset of infection, type of infecting HSV strain and preexisting antibody of either HSV type. That is the basis upon which diagnostic virology laboratories determine IgG antibody to HSV by testing human sera against both HSV-1 and HSV-2 antigens. Enzyme-linked immunosorbent assay (ELISA) represents the most widely used assay system for determination of the immune status to HSV. However, standardization of different batches of antigen is difficult to achieve and a control antigen is required as a specificity control of test results. On the other hand, type-specific antibody can only be determined by experienced laboratories using microneutralization procedures, while recently proposed methods based on the use of immunoaffnity-purified HSV glycoproteins, such as gG-1 and gG-2 (Lee et al., 1985, 1986), require procedures too cumbersome for general use in a diagnostic laboratory. In order to develop an enzyme immunoassay which could be standardized for determination of both the HSV immune status and diagnosis of acute HSV1 and HSV-2 infections, we used a secreted recombinant HSV-1 glycoprotein B (gB-1s) constitutively expressed in human cells transfected with a BK virus recombinant episomal vector (Manservigi et al., 1990a). It is known that gB is an essential glycoprotein of HSV and a major target for humoral and cellmediated immune response in the infected host (Corey and Spear, 1986) and has been shown to be a strong immunogen in eliciting neutralizing antibody (Glorioso et al., 1984; Dix and Mills, 1985; Manservigi et al., 1990a,b).This glycoprotein has been suggested as a promising candidate for the development of a subunit vaccine (Pack1 et al., 1987; Manservigi et al., 1988). While the mature and completely glycosylated form of gB-1 is 903 amino acids in length ahd 120 kDa in size (Campadelli-Fiume and Serafini-Cessi, 1985), gB-ls, which is deprived of the transmembrane portion of the molecule, is 690 amino acids long and 91.6 kDa in size (Manservigi et al., 1990a). In recent studies, gB-1s has been shown to protect mice against lethal and latent HSV-1 infections and to prevent CNS disease in rabbits, while reducing the severity of herpetic keratitis (Manservigi et al., 1990a,b). In the present study, the development of an ELISA using gB-1s as a solid phase is described. The test is capable of detecting small amounts of either HSV-1 or HSV-2 antibody in human sera from patients with remote, primary or recurrent HSV-l/HSV-2 infections as well as from patients with HSV encephalitis. The assay does not require handling of infectious material and may be performed in the absence of control antigen.
59
Material and Methods
Cloning and expression of gB-Is in human cells The gB-1 gene of HSV-1, strain F, deleted of 693 nucleotides encoding the transmembrane anchor sequence and reconstructed with the extramembrane and intracytoplasmic domains, was cloned under the control of the Rous Sarcoma virus promoter in the episomal replicating vector pRP-RSV (Manservigi et al., 1990a). gB-1s was secreted into the culture medium of pRP-RSV-gB-Is-transformed 293 cells at a concentration of 0.1-0.2 pg/ml. The secreted product was identified in both radioimmunoprecipitation and ELISA by using monoclonal antibodies to HSV-1 gB (Manservigi et al., 1988) and gB1s developed in the laboratory. Using the same monoclonal antibodies, gB-1s was immunoaflinity-purified from 20 x concentrated (by ultrafiltration) and dialyzed cell culture medium. ELISA procedure Either purified or crude concentrated culture medium preparations of gB-1s were used as a solid phase in the ELISA. Following initial titration, the optimal concentration of gB-1s was found to be between 10 and 50 ng/well. Wells of 96well polystyrene microplates (M129A Dynatech, Plochingen, F.R.G.) were coated overnight at room temperature with the optimal dilution of gB-1s in 0.05 M sodium carbonate-bicarbonate buffer, pH 9.6. After washings, 100 ~1 of test sera, diluted 1:50 in phosphate-buffered saline (PBS) containing 1% foetal calf serum (FCS) and 0.1% Tween 20, were incubated for 2 h at 37°C. Plates were then washed with, PBS-Tween 20 and 100 ~1 of optimally diluted peroxidase-conjugated goat anti-human IgG (Cappel Laboratories, Cooper Biomedical, Malvern, PA, U.S.A.) were added to each well. Incubation was continued for 1 h at 37°C. After washing, enzymatic activity was detected by adding the substrate solution for 30 min at room temperature. The latter was prepared immediately before use by dissolving 3.4 mg of ortho-phenylenediamine in 10.0 ml of citrate phosphate buffer (0.05 M phosphate, 0.02 M citrate, pH 5.5) and then adding 40 ~1 of 3% H202. Absorbance values were recorded at 492 nm using a microtitre plate spectrophotometer (Titertek Multiskan, Flow Laboratories, Irvine, U.K.). The cut-off was set at 0.10 absorbance following testing of 39 HSV-negative sera. When required, quantitation of IgG antibody to gB-1s was carried out by constructing a calibration curve, which was obtained by plotting the mean absorbance value of each of five standard sera (containing different amounts of specific antibody) versus the amount of gB-ls-specific IgG expressed as antibody titre. In parallel, all sera were tested by conventional ELISA for HSV-l/HSV-2 IgG antibody determination, as previously reported (Revello and Gerna, 1988). In conventional ELISA, glycine-extracted HSV-1 and HSV-2 crude antigens as well as uninfected Vero cell control antigen were used. This assay was calibrated optimally to allow
60
typing of HSV antibody Gerna, 1988). Microneutralization
in human
sera, as reported
previously
(Revello
and
test
The microneutralization assay was carried out using growing Vero cells by testing serial twofold dilutions of heat-inactivated serum samples mixed with equal amounts of virus suspensions containing 100 50% tissue culture infectious doses of either HSV-1 or HSV-2 (Revello and Gerna, 1988). Corrected titres (tc) for each serum were calculated according to Stalder et al. (1975), and typing of HSV neutralizing antibody was achieved by calculating the tc difference between the titre to HSV-1 and that to HSV-2. Patients and sera The following groups of patients and sera were examined in parallel by gB- 1s ELISA and conventional HSV-l/HSV-2 ELISA: l 39 HSV antibody-negative sera from 39 healthy individuals; l 53 sequential sera from 14 patients with primary HSV-1 infection and 17 sequential sera from 3 patients with primary HSV-2 infection (in the absence of prior HSV-1 infection); l 48 sera from 48 patients, of which 23 contained HSV-1 antibody, 13 HSV-2 antibody and 12 dual HSV antibody, as determined previously by microneutralization; l acute- and convalescent-phase sera from 28 patients with acute infections caused by human herpesviruses other than HSV, i.e. varicella-zoster virus (VZV), human cytomegalovirus (HCMV), Epstein-Barr virus (EBV), and human herpesvirus 6 (HHV-6), and from 16 patients with primary HSV-1 infection; l 35 sequential paired serum-cerebrospinal fluid (CSF) samples from 7 patients with herpes simplex encephalitis. HSV isolation and typing Material from mucocutaneous lesions was collected by swabs, for inoculation of Vero and diploid cell cultures for virus isolation, and by scraping, for preparation of cell smears for direct virus identification and typing. HSV isolates were typed by indirect ELISA, as reported previously (Gerna et al., 1983), whereas direct typing on cell smears from clinical specimens was done by direct fluorescent antibody test, using in parallel fluorescein-labeled type-specific monoclonal antibodies both produced in our laboratory and available commercially (Syva Micro-Trak, Syva Company, Palo Alto, CA, U.S.A.).
61
Development of the indirect ELISA for determination of IgG antibody to gB-1s
Initially, the optimal concentration of gB- 1s to be used in the indirect ELISA for IgG antibody detection was determined. For this purpose, human sera with past and primary HSV-1 infections, containing only HSV-1 antibodies, were used. Low levels of HSV-1 antibody could be clearly detected when using a concentration of gB-1s of 10-15 ng/well. However, when early convalescentphase sera from primary HSV-2 infections in subjects with no prior experience of HSV-1 infection were tested, the amount of gB-is/well had to be increased to 40-50 ng in order to allow detection of low levels of early HSV-2 antibody. Afterwards, the cut-off was determined on 39 HSV-negative sera previously characterized by neutralization and ELISA. The mean Ad92 was 0.03 + 0.03. Thus, a cut-off of 0.10 was selected for further testing (0.03 + 2SD). Antibody response to gB-1s in primary HSV-1 and HSV-2 infections
All subsequent tests were run in parallel using the gB- 1s ELISA and the HSVl/HSV-2 ELISA. Initially, two monoclonal antibodies to gB and four to gB-1s a.0
_
1s
.
1.0
.
1.7
.
1.0
.
000 0
0 1s.
1.4
.
1.5
.
11
.
1.1
.
1.0
.
0.0
.
0
0
0
.
0 :
ao. f
a7
0 .
-
0
.
0.
.
0
Dae.
0 Eos. 5
00
.b
I--
O
0
0
0
C‘OJ. az.
Fig. 1. Absorbance values (A& in the gB-1s vs. HSV-I/HSV-2 ELISA of: 53 sera collected from 14 patients during the first month (0) of a primary HSV-1 infection or afterwards (0); 17 sera collected from 3 patients during the first month (m) of a primary HSV-2 infection or afterwards (0).
62
(shown in competitive binding assays to recognize different gB epitopes) were tested in both ELISAs. All reacted in the gB-1s ELISA with titres of 1O-5-1O-7, whereas none reacted at a dilution of 10e3 in the HSV-l/HSV-2 ELISA. Measurement of IgG antibody to gB-1s during primary HSV-1 infections was achieved by examining 53 sequential sera from 14 patients undergoing primary HSV- 1 infection. Among 34 sera collected within 30 days after onset of symptoms, 9 collected within the first 5 days were negative by both ELISA systems, 23 were concordant positive and 2 (collected 3 and 5 days after onset, respectively) were weak positive by gB-1s ELISA only (Fig. 1). Ad92 values measured using gB-1s ELISA and the HSV-l/HSV-2 ELISA were also grossly comparable, with the exception of 2 sera collected within 30 days and 2 sera collected afterwards, which gave significantly higher (greater than a fourfold difference in titre) reactivity in the gB-1s ELISA. Comparable results were observed when 17 sera from 3 patients undergoing primary HSV-2 infection (in the absence of prior HSV- 1 infection) were examined. Of 6 sera collected within 30 days, 5 were detected by both ELISAs and one only by the HSV-2 ELISA
.
5.8
.
5.7
.
33
.
3.5
.
a1
.
0 . .
. . . . . .
. . .
2.9
.
2.7
.
2.5
.
23
.
2.1
.
1.0
.
1.7
.
1.s
.
1.3
.
1.1
_
‘;.
0.2
_
f
0.7
.
z
03
.
0.3
.
0.1
.
5 s :
8.0
oe
.
0
.
0
0 . .
.
e
0
8
. .
.
0 0 0 0 0 0
. 8
0
. .
. .‘.
Am
WIT”
ttONOLOOW8
HW
ANTKILN
Fig. 2. Comparison of ELISA reactivities with gB-1s vs. HSV-l/HSV-2 crude antigens in 48 sera previously typed for HSV antibody by neutralization. Sera were divided into three groups containing: 23, HSV-1 (0); 13, HSV-2 (O), and 12, mixed (0) antibody, respectively. For sera containing mixed antibody, the higher Adp2 value in the HSV-l/HSV-2 ELISA was plotted.
63
with an A492 value close to cut-off. In addition, all 11 sera collected after 30 days were found positive by both systems. However, most late sera showed a tendency to give higher HSV-2 than gB-1s A492values (Fig. 1). Thus, although HSV-1 and HSV-2 gBs share a high degree of homology, at least some sera from primary HSV-1 infections seem to recognize gB-1s somewhat better than sera from HSV-2 infections. Determination of HSV immune status in patients with HSV-I, HSV-2 or mixed antibody The HSV immune status was determined by gB-1s ELISA and HSV-l/HSV2 ELISA in sera from 48 patients, in which HSV antibody type was previously determined by microneutralization. As shown in Fig. 2, all test sera were detected as positive by both ELISA systems. However, HSV-1 and mixed antibody sera gave A 492 values higher in the gB-1s ELISA as compared to HSV-l/HSV-2 ELISA, whereas most HSV-2 sera showed an equivalent reactivity between the two ELISAs. These data again suggest that HSV-1 antibody reacts more strongly with gB-1s than HSV-2 antibody. Study of the heterologous antibody response to gB-Is in patients with acute infections by HHVs other than HSV Since there are high levels of homology among gB genes of different human herpesviruses, we investigated the level of heterologous antibody response to 22 24
(MllAnv)
(CIIUAW)
(rnlllill)
( CIImAmv)
Fig. 3. Heterologous antibody response to gB-1s in acute and convalescent-phase sera from 28 patients with acute HHV infections other than HSV (6 with zoster, 5 with chickenpox, 10 with primary HCMV, 5 with primary EBV and 2 with primary HHV-6 infection). Homologous antibody response in acute and convalescent sera from 16 patients with primary HSV-1 infection is reported for comparison. Numbers inside columns indicate gB-Is-negative (shaded areas) or -positive (blank areas) sera. A, acute; C, convalescent serum.
64
gB-1s in acute and convalescent sera from 28 patients with acute infections by HHVs other than HSV (6 with zoster, 5 with chickenpox, 10 with primary HCMV infection, 5 with EBV infectious mononucleosis and 2 with primary HHV-6 infection). For comparison, acute and convalescent sera from 16 patients with primary HSV-1 infection were examined. Mean Ad92 values obtained in the gB-1s ELISA with different groups of sera are reported in Fig. 3. No significant difference (greater than a twofold dilution) in mean Ad92 between acute and convalescent sera was observed in any of the 5 groups of HSV-positive patients (where high gB-1s antibody titres reflected past HSV-1 infections), whereas seroconversion from seronegative to seropositive was found in all 16 TABLE 1 Heterologous Patient No.
Zoster RS A C R22 A C R25 : R26 2 R30 A C R31 A C Varicella 6806A A C 6163A A C 6251A f? 12198 A C 11968
antibody
response to gB-1s in 6 patients with zoster and 5 patients with varicella A 492
vzv
gB-1s (titre)
HSV-1 (titre)
1.12 2.28
2.00 (3,100) 1.99 (3,100)
1.28 (6,000) 1.12(3,500)
1.29 2.47
2.23 (4,000) 2.60 (7,000)
1.18(4,500)
0.75 1.97
2.52 (5,600) 2.64 (7,000)
0.94 (2,000) 1.04(2,500)
1.66 2.34
2.60 (7,000) 2.60 (7,000)
1.36 (8,000) 1.44 (10,000)
1.49 2.30
2.70 (7,000) 2.64 (7,000)
1.40 (9,500) 1.22 (7,000)
0.34 2.17
0.06 (< 50) 0.04 (< 50)
0.01 (<50) 0.07 (< 50)
0.15 0.71
2.76 (7,000) 3.28 (12,600)
1.40 (9,500) 1.59 (18,000)
0.03 0.35
2.68 (7,000) 2.16 (3,500)
1.40 (9,500) 1.24 (5,500)
0.13 1.12
1.63 (2,000) 2.37 (4,000)
1.27 (6,000) 1.31 (6,500)
0.03 0.57
1.65 (2,000) 2.41 (4,200)
1.18 (4,500) 1.18(4,500)
0.12 0.67
0.05 (< 50) 0.04 (< 50)
0.01 (<50) 0.03 (< 50)
A = acute serum; C = convalescent
serum.
1.19 (4,500)
65
patients with primary HSV-1 infection. When paired sera from individual patients of each group were examined, it was found that only one patient with primary HCMV infection showed a fourfold rise in antibody titre to gB-1s. However, this patient was a heart transplant recipient possibly undergoing an HSV recurrence in the post-transplant period. With respect to VZV acute infections, in 3 of 4 HSV-1 seropositive patients with chickenpox a fair rise in Ad92 with gB-1s was observed, but it corresponded only to a twofold rise in antibody titre and was thus considered not clinically significant (Table 1). In only one of 5 patients with zoster infection a slight increase in AdP2with gB-1s was observed, again within a twofold dilution difference in titre. Comparable results with paired sera from patients with varicella or zoster were obtained with conventional ELISA. In addition, 6 patients (1 with zoster, 1 with chickenpox, 1 with primary HCMV, 2 with primary EBV and 1 with primary HHV-6 infection) negative for gB-1s (and HSV) antibody in the acute serum remained seronegative in the convalescent phase of the relevant infection. Antibody to gB-Is in patients with herpes simplex encephalitis Thirty-five sequential paired serum-CSF samples from 7 patients with HSV encephalitis (4 by a primary and 3 by a recurrent HSV-1 infection) were examined for determination and ouantitation of IgG antibody response by gB1s and HSV-1 ELISAs. The kinetics of the IgG antibody levels in two representative patients (1 with primary and 1 with recurrent HSV-1 infection) are reported in Fig. 4. The curves of IgG antibody to gB-1 s and HSV-1 in
0.0
.
4.1
.
4.4
.
4.1
.
a.0
.
0.0
.
LE
.
E.0
.
l.0
.
*.a
Fig. 4. Kinetics of virus-specific IgG antibody as determined by ELISA using gB-1s (-) and HSV-1 crude antigen (- - -) in sequential paired serum (0) and CSF (A) samples from two patients with HSV encephalitis due to a primary (A) and to a recurrent (B) HSV-1 infection, respectively. (a- - -a), HSV-1 serum IgG; (A- - -A), HSV-I CSF IgG; (a----a), gB-1s serum IgG; (A---&, gB-Is CSF IgG antibody.
66
. 6.0
2
r
0.87
’ 14 IS) DAYS
APIER
(No PAIRED
vi*
I21
28 12)
ONSET OF NEUROLOQICAL SANHES
JO-180 6%
m-ass (3)
an (3)
2 i
4vn (4)
SIYPTOYS
AVAILABLE 1
Fig. 5. Comparison of the kinetics of gB-1s and HSV-1 mean serum/CSF antibody titre ratios vs. Link index in 7 patients with HSV encephalitis.
serum and CSF are almost parallel in both patients. When data available on sequential paired samples from the 7 patients were considered together, the mean serum/CSF antibody titre ratio obtained with the gB-1s ELISA gave a curve generally parallel to that provided by the mean ratio with the HSV-1 ELISA (Fig. 5). The lowest levels of both ratios were reached concomitantly with the highest value of the Link index (Tibbling et al., 1977) at the end of the first month after onset of symptoms (the Link index represents, when > 0.67, a marker of intrathecal IgG synthesis); then ratios rose slowly and progressively until they reached the cut-off value of 20 after 24 years. Contrariwise, the Link index decreased, approximating the cut-off value of 0.67 after 4 years. Again, the antibody response measured by the gB-1s ELISA appeared to overlap that measured by the HSV ELISA.
Discussion The results of the present study clearly demonstrate that gB-1s is suitable for use in an indirect ELISA for determination of IgG antibody to either HSV-1 or HSV-2 and that the gB-1s ELISA can promptly replace the conventional HSVl/HSV-2 ELISA for either determination of the immune status to HSV or for the diagnosis of acute HSV infection. When developing a new immunoassay using a recombinant DNA product, the first question to be addressed is whether detection of the immune response
to a single viral protein may be sufficiently representative of the whole antibody response elicited in the patient by viral infection. It was shown previously that antibody to gB is among the first to appear during the early convalescent phase of HSV infections involving either mucocutaneous sites or the CNS (Eberle and Mou, 1983; Kahlon et al., 1986; Grimaldi et al., 1988). The comparable kinetics of the IgG antibody response observed in our study in acute HSV infections when using gB-1s and conventional ELISAs suggest that the same type of antibody was detected by the two assays. However, this was not the case, since monoclonal antibodies to gB and gB-ls, while highly reactive in the gB-1s ELISA, were practically nonreactive in the conventional ELISA, where gB was poorly represented in the glycine-extracted antigens. Thus, although the profile of the antibody response did not appear to display significant variation among the two assays, the type of IgG antibody detected by the two ELISA assays was totally different as regards antigen specificity. Although the amino acid sequences of gB-1 and gB-2 are 86% conserved (Bzik et al., 1986; Stuve et al., 1987), type-specific epitopes identified in the gB molecule (Kousoulas et al., 1988; Eberle and Courtney, 1989) may cause sera from patients with HSV-1 antibody to react in gB-1s ELISA more markedly than sera from patients with HSV-2 antibody. This might have been the reason why, in the initial development of gB-1s ELISA, the antigen concentration had to be increased fourfold, in order to enable the test to detect low amounts of HSV-2 antibody in sera from patients with HSV-2 infection. Recently, human sera were tested by radioimmunoprecipitation using different forms of truncated gB, showing that the greater the deletion the lower was the percentage of reactive sera (Ali and Forghani, 1990). In our study, gB-Is, i.e. a form of gB deprived only of the transmembrane region and of two adjacent short intracytoplasmic and extramembrane sequences, was shown to retain all immunoreactivity versus human sera in an indirect ELISA. The gB-1s used in the ELISA could be readily purified from the culture medium of cells expressing gB- 1s. However, it should be noted that purification of gB-1s was not an essential step for the development of the ELISA, since coating of the solid phase by crude cell culture medium provided comparable results in terms of sensitivity and specificity (unpublished results). Nevertheless, because of a fair degree of variability in the gB-1s content among different batches of culture medium, standardization of the ELISA assay was achieved only by using purified gB- 1s. A high degree of homology is shared by gB of HSV and gBs of the other HHVs, including VZV (Davidson and Scott, 1986; Keller et al., 1986), HCMV (Cranage et al., 1986) and EBV (Pellet et al., 1985). In particular, a crossreactive serological response between HSV and VZV has been reported repeatedly (Kapsenberg, 1964; Schmidt et al., 1969; Harbour and Caunt, 1979; Bernstein et al., 1990). In addition, antigenic relationships have been detected between some VZV glycoproteins and some HSV glycoproteins (with special reference to gB) by immunoprecipitation using hyperimmune sera to HSV and VZV (Shiraki et al., 1982) as well as monospecific antibody to VZV gp3 or
68
monoclonal antibody to gB of HSV (Kitamura et al., 1986). In the present study, by testing in the gB-Is ELISA several paired sera from patients with acute infections caused by HHVs other than HSV, a significant rise in heterotypic antibody response to gB-1s was not found. This suggests that, despite the high homology among gBs of different herpesviruses, crossreactivity does not seem to occur at the level of serological response to HHV infections, at least in the gB-1s ELISA. This is in agreement with previous observations showing that cross-reactions observed with some tests, such as complement fixation, were not detectable by other assays, such as gel precipitation (Schmidt et al., 1969). In conclusion, the gB-1s ELISA represents a reliable, sensitive and specific immunoassay for determination of IgG antibody to HSV which can be standardized. It can be utilized for both determination of the immune status HSV or for serological diagnosis of primary HSV-1 or HSV-2 infections, as well as for early retrospective diagnosis of HSV encephalitis. For determination of HSV type-specific antibodies, we have recently been able to obtain expression of gG of HSV-2 in the 293 human cell line both in a membranebound (gG-2) and secreted form (gG-2s) (Sabbioni et al., unpublished results). Both gG-2 and gG-2s were shown to be suitable for determination of HSV-2specific antibodies.
Acknowledgements This work was partially supported by the Minister0 della Sanita, Istituto Superiore di Sanita, Progetto Nazionale AIDS 1990, contract No. 4204 18, and by Consiglio Nazionale delle Ricerche, Progetto Finalizzato Biotecnologie e Biostrumentazione, contract Nos. 89.00164.PF70, 90.00046.PF70 and 90.00 164.PF70. We thank Linda d’Arrigo for her help with the English revision.
References Ali, M.A. and Forghani, B. (1990) Serologic specificity of antibodies to herpes simplex virus glycoprotein B present in human sera: analysis by transient expression of glycoprotein Bderivatives. Arch. Virol. 112, 129-138. Bernstein, D.I., Frenkel, L.M., Bryson, Y.J. and Myers, M.G. (1990) Antibody response to herpes simplex virus glycoproteins gB and gD. J. Med. Virol. 30, 4549. Bzik, D.J., Debroy, C., Fox, B.A., Pederson, N.E. and Person, S. (1986) The nucleotide sequence of the gB glycoprotein gene of HSV-2 and comparison with the corresponding gene of HSV-I. Virology 155, 322-333. Campadelli-Fiume, G. and Seralini-Cessi, F. (1985) Processing of the oligosaccharide chains of herpes simplex virus type 1 glycoproteins. In: B. Roizman (Ed), The Herpesviruses, Vol. 3, Plenum Press, New York, pp. 357-382. Corey, L. and Spear, P.G. (1986) Infection with herpes simplex viruses. N. Engl. J. Med. 314, 686 691 and 749-757.
69 Cranage, M.P., Kousarides, T., Bankier, A.T., Satchwell, S., Weston, K., Tomlinson, P., Barrell, B., Hart, H., Bell, S.E., Minson, AC. and Smith, G.L. (1986) Identification of the human cytomegalovirus glycoprotein B gene and induction of neutralizing antibodies via its expression in recombinant vaccinia virus. EMBO J. 5, 3057-3063. Davidson, A.J. and Scott, J.E. (1986) The complete DNA sequence of varicella-zoster virus. J. Gen. Virol. 67, 1759-1816. Eberle, R. and Courtney, R.J. (1989) Topological distribution of virus-specific and cross-reactive antigenic determinants on the gB glycoprotein of the herpes simplex viruses. J. Med. Virol. 27, 309-3 16. Eberle, R. and Mou, S.W. (1983) Relative titres of antibodies to individual polypeptide antigens of herpes simplex virus type 1 in human sera. J. Infect. Dis. 148, 436444. Gerna, G., Battaglia, M., Revello, M.G. and Torsellini Gema, M. (1983) Typing of herpes simplex virus isolates by enzyme-linked immunosorbent assay: comparison between indirect and doubleantibody sandwich techniques. J. Clin. Microbial. 17, 942-944. Grimaldi, L.M.E., Roos, R.P., Manservigi, R., Spear, P.G., Lakeman, F.D. and Whitley, R.J. (1988) An isoelectric focusing study in herpes simplex virus encephalitis. Ann. Neurol. 24, 227232. Harbour, D.A. and Caunt, A.E. (1979) The serologic relationship of varicella-zoster virus to other primate herpes viruses. J. Gen. Virol. 45, 469477. Kahlon, J., Lakeman, F.D., Ackermann, M. and Whitley, R.J. (1986) Human antibody response to herpes simplex virus specified polypeptides following primary and recurrent infection. J. Clin. Microbial. 23, 725-730. Kapsenberg, J.G. (1964) Possible antigenic relationship between varicella zoster virus and herpes simplex virus. Arch. Ges. Virusforsch 15, 6673. Keller, P.M., Davidson, A.J., Lowe, R.S., Bennet, C.D. and Ellis, R.W. (1986) Identification and structure of the gene encoding gpI1, a major glycoprotein of varicella-zoster virus. Virology 152, 181-191. Kitamura, K., Namazue, J., Campo-Vera, H., Ogino, T. and Yamanishi, K. (1986) Induction of neutralizing antibody against varicella-zoster virus (VZV) by VZV gp3 and cross-reactivity between VZV gp3 and herpes simplex viruses gB. Virology 149, 74-82. Kousoulas, K.G., Huo, B. and Pereira, L. (1988) Antibody-resistant mutations in cross-reactive and type-specific epitopes of herpes simplex virus 1 glycoprotein B map in separate domains. Virology 166, 423343 1. Lee, F.K., Coleman, R.M., Pereira, L., Bailey, P.D., Tatsuno, M. and Nahmias, A.J. (1985) Detection of herpes simplex virus type 2-specific antibody with glycoprotein G. J. Clin. Microbial. 22, 641-644. Lee, F.K., Pereira, L., Griffin, C., Reid, E. and Nahmias, A.J. (1986) A novel glycoprotein for detection of herpes simplex virus type l-specific antibodies. J. Virol. Methods 14, 11 l-l 18. Manservigi, R., Gualandri, R., Negrini, M., Albonici, L., Milanesi, G., Cassai, E. and BarbantiBrodano, G. (1988) Constitutive expression in human cells of herpes simplex virus type 1 glycoprotein B gene cloned in an episomal eukaryotic vector. Virology 167, 284-288. Manservigi, R., Grossi, M.P., Gualandri, R., Balboni, P.G., Marchini, A., Rotola, A., Rimessi, P., Di Luca, D., Cassai, E. and Barbanti-Brodano, G. (1990a) Protection from herpes simplex virus type 1 lethal and latent infections by secreted recombinant glycoprotein B constitutively expressed in human cells with a BK virus episomal vector. J. Virol. 64, 431436. Manservigi, R., Incorvaia, C., Di Luca, D., Rotola, A., Balboni, P.G., Sebastiani, A., Rossi, A. and Cassai, E. (199Ob) Experimental keratitis in rabbits by human HSV-1 variants: prevention and treatment. J. Med. Virol. 32, 148-154. Packl, C., Burke, R.L., Stuve, L., Sanchez-Pescador, L., Nest, G.V., Masdiarz, F. and Dina, D. (1987) Expression of cell-associated and secreted forms of herpes simplex virus type 1 glycoprotein B in mammalian cells. J. Virol. 61, 315-325. Pellet, P.E., Biggin, M.D., Barrell, B. and Roizman, B. (1985) Epstein-Barr virus genome may encode a protein showing significant amino acid and predicted secondary structure homology with glycoprotein B of herpes simplex virus 1. J. Virol. 56, 807-813.
70 Revello, M.G. and Gema, G. (1988) Herpes simplex virus type-specific antibody determination by enzyme-linked immunosorbent assay in human sera. Serodiagn. Immunother. Infect. Dis. 2, 415422. Schmidt, N.J., Lennette, E.H. and Magofftn, R.L. (1969) Immunological relationship between herpes simplex and varicella-zoster viruses demonstrated by complement-fixation, neutralization and fluorescent antibody test. J. Gen. Virol. 4, 321-328. Shiraki, K., Okuno, T., Yamanishi, K. and Takahashi, M. (1982) Polypeptides of varicella-zoster virus (VZV) and immunological relationship of VZV and herpes simplex virus (HSV). J. Gen. Virol. 61, 255-269. Stalder, H., Oxman, M.N. and Hermann, K.L. (1975) Herpes simplex virus neutralization: a simplification of the test. J. Infect. Dis. 131, 430-432. Stuve, L.L., Brown-Shimer, S., Packl, C., Najarian, R., Dina, D. and Burke, R.L. (1987) Structure and expression of the herpes simplex virus type 2 glycoprotein gB gene. J. Virol. 61, 326335. Tibbling, G., Link, H. and dhman, S. (1977) Principles of albumin, and IgG analyses in neurological disorders. I. Establishment of reference value. Stand. J. Clin. Lab. Invest. 37, 385392.