Analyses of functional antibody responses in HIV-1-infected individuals after vaccination with rgp160

E L S E V I E R Clinical and Diagnostic Virology 6 (1996) t15-126

Clinical and Diagnostic Virology

Analyses of functional antibody responses in HIV-l-infected individuals after vaccination with rgpl60

Kristina Broliden% Jorma Hinkula a, G6ran Bratt b, Christina Persson a, Anselmo Otero a, Anders Ekstr6m a, Eric Sandstr6m b, Per-Anders Broliden%

Britta Wahren ~'*

aMicrobiology and Tumorbiology Center, Karolinska Institute, and Swedish Institute for Infectious Disease Control, S-105 21 Stockholm, Sweden

bDepartment of Dermatovenereology, South Hospital, Karolinska Institute, S-118 83 Stockholm, Sweden

Received 14 January 1996; accepted 25 April 1996

Abstract

Background: The immune response to HIV infection has been extensively studied and the antibody response against the virus has been characterized in detail. It is, however, still unclear which immune function it is most important to stimulate when administering a vaccine against HIV.

~ i v e s : To analyze the functional antibody responses in asymptomatic HIV-l-infected individuals after vaccination with rgpl60.

Stmly deign: Forty-nine asymptomatic HIV-l-infected individuals were followed for 9 months and analyzed for changes in functional antibody responses. Forty of them received HIV-1 envelope rgpl60 injections and nine did not.

Remills: Increased levels of antibodies mediating neutralization and cellular cytotoxicity (ADCC) could be seen in subjects who also showed a better CD4 development compared with the patients without increased levels of functional antibodies. Out of nine matched HIV-infected and influenza-immunized controls, none had increased neutralizing activity and only one had an increased ADCC titer. An increased capacity to block soluble CD4 binding to gpl20 occurred in 10 immunized patients. Seroreactivity and avidity maturation were detectable to peptides representing consensus HIV-1 envelope regions, indicating an anamnestic response to the patients own virus.

Conei~io~s: The humoral immune response in HIV-l-infected individuals was moderately influenced by repeated gpl60 immunizations, while previous studies have shown that HIV-specific T-ceU reactivity was strongly increased.

Keywords: ADCC; AIDS; Avidity; HIV; Neutralization; Peptides

* Corresponding author. Tel.: + 46 87351300; fax: + 46 8272231.

0928-0197/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PII S0928-0197(96)00235-8

116 K. Broliden et al. / Clinical and Diagnostic Virology 6 (1996) 115-126

1. Introduction

There is an increasing interest in improving the human immune response against chronic virus infections by vaccinating the already infected individual. In HIV infection, several candidate AIDS vaccines based on recombinant virus vectors expressing HIV-1 envelope glycoprotein gpl60 are now being tested in asymptomatic HIV-infected volunteers to evaluate their safety and the kind of immune response they induce. Although the immune response to HIV infection has been extensively studied and the antibody response against the virus has been characterized in detail, it is still unclear which immune function is most important to stimulate by a vaccine against HIV. A conse- quence of the HIV-1 type of infection is the early loss of specific T-helper cell responses to HIV antigens (Wahren et al., 1986; Shearer and Clerici, 1991). Such losses occur even before the immune defect is manifested by decreasing CD4 ÷ cells. B-cell responses become defective too, as shown by the incompetence to mount neutralizing antibodies to new viral variants that occur during the progression of disease (Albert et al., 1990).

Studies with native and recombinant envelope proteins in naive animals suggest a role for envelope glycoprote.in in the induction of protective immune responses and justify the use of such immunogens in human trials (Berman et al., 1990; Fultz et al., 1992a; Girard et al., 1991; Bruck et al., 1994). Immunization or infection of already infected animals has shown both positive and negative results. HIV-l-infected chimpanzees be- came virus-culture-negative and resisted a further viral challenge after three HIV immunizations (Gibbs et al., 1991). Transient increases in numbers of HIV-infected cells occurred by stimulation of an infected animal, but a strong persistent T-cell-specific proliferation response was not induced until inactivated HIV-1 in adjuvant prepa- ration was given (Fultz et al., 1992b). Modification of SIV disease has taken place with a virulent SIV virus challenge following HIV-2, which is a less pathogenic lentivirus (Putkonen et al., 1990). In another SIV model, however, late immunization had no effect on immune or viral parameters (Gardner et al., 1989).

The primary and secondary antibody responses are thus important, since a potent antibody response can directly inactivate free virus by neutralization (Weiss et al., 1985; Wahren, 1994a), perform complement-aided lysis of virions (Spear et al., 1990), inhibit syncytium induction between infected and non-infected cells, inhibit cell-to-cell spread of infection (Niedrig et al., 1992) as well as mediate ADCC (Broliden et al., 1993). The antibody response following vaccination may include avidity maturation against specific antigens. We have recently shown an influence of ADCC reactivity on the prognosis of HIV-l-infected newborn children (Broliden et al., 1993) as well as an improved cell-mediated immune response in HIV- 1-infected asymptomatic individuals after immunization with recombinant gpl60 (Wahren et al., 1994b). In the present study, the same group of 40 immunized individuals were assayed for their functional antibody responses.

2. Materials and methods

2.1. Patients and CD4 counts

The study population consisted of 40 HIV-l-infected asymptomatic individuals, including 30 ho- mosexual males, 4 heterosexually infected individuals (2 females), 3 drug addicts (2 females) and 3 type A hemophiliacs. They were between l 8 and 53 years old (mean 35), and had been HIV- seropositive for a mean of 38 months. The patients were randomized for azidothymidine (AZT) treatment (200 mg x 4) or placebo for 2 weeks following each immunization. No one had been on AZT treatment before entry. Nine randomized control HIV-infected subjects were studied con- currently and matched for age, sex, CD4 ÷ cell counts and route of infection, and received influenza vaccination only. The control group did not receive AZT.

CD4 ÷ counts were performed at the Depart- ment of Immunology, Swedish Institute for Infec- tious Disease Control, Sweden. The rate of CD4 change (decline or increase) before and after the inception of the study was assessed using a linear regression analysis model, measured as the sum of

K. Broliden et al. / Clinical and Diagnostic Virology 6 (1996) 115-126 117

squares of the residuals. The monthly change in CD4 values was calculated for each individual patient for the following periods before and after the beginning of the study: between - 2 4 and 0 months; between 0 and 6 months; between 0 and 9 months; and between 0 and 12 months from the beginning of the study. Com- parisons were made between patients who had or had not increased neutralization and/or ADCC (Student's t-test, P values are two- tailed).

2.2. Vaccination

The recombinant gpl60 (rgpl60, VaxSyn, Mi- croGeneSys, Meriden, CT) was derived from HIV-1, strain LAI. The gpl60 gene was trans- fected into a baculovirus expression vector and the glycoprotein was produced in a lepidopteran cell line. The vaccine contains >95% pure rgpl60 protein which has been adsorbed on to aluminum phosphate as adjuvant. All 40 partici- pants received six doses of 160 /~g rgpl60 ad- ministered i.m. at 0, 1, 4, 8, 17 and 26 weeks. The 40 study subjects and nine matched con- trois received 0.5 ml influenza vaccine (Fluzone, Connaught/Meda) at 0 and 4 weeks. Virus isolation was performed at each vaccination from all patients from peripheral blood lymphocytes (PBL) in phytohemagglutinin (PHA)-activated allogeneic PBLs.

2.3. Antigens and peptide ELISA

Solid phase synthesized peptides based on the HIV-1 LAI and MN sequences were used as antigens in enzyme-linked immunosorbent assays (ELISAs) (Broliden et al., 1991). Sequence numbers of the Los Alamos database (Myers, 1995) were used. Two peptides, A10 (aa 304-318, RKSIRIQRGPGRAFV) and A12 (aa 309-321, RKRIHIGPGRAFY), representing the LAI and MN strains, respectively, were derived from the central part of the V3 region and A13 (aa 328- 342, GTIRQAHCNISRAKW) from the C-terminal end of the MN V3 region. The C90 peptide (aa 489-503, VKIEPLGVAPTKAKR) represents the C-terminal end of gpl20. Peptides

249 (aa 652-666, QQEKNEQELLEDKW) and 265 (732-746, GIEEEGGERDRDRSI) represent conserved neutralization-inducing regions of glMl(120). The JB-4 peptide (aa 594-613, GI- WGCSGKLICTTAVPWNAS) is derived from the immunodominant region of glM1 and was synthesized on an Applied Biosystems peptide synthesizer and chromatographed to > 99% pu- rity. The control antigens are described else- where (Wahren et al., 1994b) and were used in indirect ELISAs. The peptide ELISA was performed as previously described (Broliden et al., 1991). Briefly, micro-ELISA plates were coated with the selected synthetic peptide and stored at 4°C until use. Sera were added for 1 h at 37°C. After washing, horseradish peroxidase-conju- gated rabbit anti-human IgG were added for 30 min at 370C followed by washing and addition of the substrate o-phenylenediamine activated by H202. The reaction was stopped after 30 min, at room temperature, by adding H2SO 4.

Values three standard deviations (S.D.) above the mean optical density (OD) of negative con- trois were considered to be positive.

2.4. Recombinant soluble CD4 (rsCD4)-gp120 blocking assay

Antibodies binding to the CD4 binding site of gpl20 were tested by a rsCD4-gpl20 blocking assay described in Hinkula et al. (1994) and modified as follows. Serial dilutions of patient serum were incubated with rgpl20 (from a baculovirus expression system) bound to Galanthus nivalis agglutinin on solid phase. After washing, rsCD4 was added to the gpl20-coated wells; rsCD4 binding to gpl20 was then traced with the OKT-4 monoclonal antibody (Ortho, Rari- tan, MA) and horseradish peroxidase-labeled anti-mouse IgG conjugate. A >i 15% blocked rsCD4 binding to gpl20 was considered to be a significant difference in blocking capacity, when compared with the blocking capacity with 15 HIV-negative sera giving a mean blocking capacity of 9.6 + 2.5% (ranging from 0 to 11% of blocking capacity). Serum dilution 1:100 was used when rsCD4 blocking capacity was assayed.

118 K. Broliden et al . / Clinical and Diagnostic Virology 6 (1996) 1t5-126

2.5. Avidity assay

The assay was performed according to Steward et al. (1991) with a few modifications. Serial 3- fold dilutions of peptides were selected from previously prepared inhibition curves for each peptide. Peptide concentrations were in the range of 50-2000 nM, Controls in triplicate were performed in the absence of inhibitor. 50 pl of each peptide dilution were added together with 50 p l of the selected antibody dilutions (the dilution of each serum giving an OD value of 0.5-1.0). After incubation for 1 h at 37°C and washing, the assay was continued as described for peptide ELISA. The molar inhibitory concentration required for 50% inhibition (IC50) was calculated by interpola- tion in linear regression analysis of log OD and log peptide concentration. Interpolated values were used when the r 2 curve fitting was over 0.90. The reciprocal values of these represent a directly proportional measure of average anti-peptide antibody avidity (Rath et al., 1988). All serum samples were tested three times on separate occasions. Estimated antibody avidity ranged between 104xM - 1 (peptides C90, A10) to 106xM-i (peptides JB-4, A12). An increase in avidity of a factor of I> 1.5 was considered significant.

2.6. Immunocomplex formation

Immunocomplex formation (Mathiesen et al., 1989) was assayed in serum by ability to bind Clq in ELISA (Immunomedics, Warren, N J). Values above 20 ~tg/ml were measurable, values above 400 pg/ml were considered high.

2.7. ADCC assay

ceeded 10%. HIV-specific ADCC was determined as follows: specific 5~Cr-release with HIV-poSitive serum minus specifc 5~Cr-release with HIV-negative sera. Serum with a specific 5~Cr-release greater than 2 S.D. above the specific 5~Cr-release obtained by the mean value of five HIV-antibody- negative sera was considered positive. HIV-antibody-positive serum samples with known ADCC titers were included in each test. The reciprocal of the last dilution step with a positive ADCC value was taken as the ADCC titer.

2.8. Neutralization assay

A 100-/d aliquot of the HIV-1 LAI or SF2 virus supernatant (50 tissue culture infectious doses) was preincubated for 1 h at 37°C with serial dilutions of heat-inactivated serum. The serum-virus mixture was added to 5 x 104 Jurkat/ tat cells for another 1 h at 37°C. After centrifuga- tion (200 g, 10 min) and washing twice, the cells were cultured in 96-well plates (Nunc, Roskilde, Denmark) for 7 days in RPMI-1640 medium (GIBCO, Grand Island, NY) supplemented by 10% heat-inactivated fetal calf serum, 2 mM L- glutamine, and antibiotics. On day 2, the plates were centrifuged and fresh medium was added. Supernatants were collected after 7 days and analyzed for p24 content. Neutralization was defined as 80% reduction of the p24 antigen in the supernatant when compared with the mean value of five HIV-negative serum controls. The reciprocal of the last dilution step showing neutralization was taken as the neutralizing titer. HIV-l-antibody-positive sera and an anti-gpl20-directed monoclonal antibody with known neutralizing titer were included in each test.

The monocytoid cell line U937, clone 2, contin- uously infected with the HIV-1 strains LAI or SF2, provided target cells. PBMC obtained from HIV-antibody-negative blood donors were used as effector cells. 51Cr-labeled target cells (1 x 10 4) and lymphocytes as effector cells (2 x l05) were mixed with serum dilutions (six steps, 3-fold serial dilutions, starting with 1 : 30). Supernatants were harvested after 3 h and released radioactivity was calculated. The spontaneous release never ex-

3. Results

3.1. Presence of heterologous and autologous neutralization

The neutralizing activity was analyzed at 0, 6 and 9 months in sera from the 40 patients immunized with rgpl60 and influenza (Table 1). HIV-1- infected control subjects immunized with

K. Broliden et al. / Clinical and Diagnostic Virology 6 (1996) 115-126

Table 1 Comparison of neutralizing activity before and after vaccination with rgpl60

119

Virus Months of follow-up Titer increase ~ Stably positive Titer decrease b Stably negative

Vaccinated subjects (n = 40) LAI 6 7 (4) 29 3 1 SF2 6 14 (4) 23 0 4 LAI 9 10 (5) 28 2 (1) 0 SF2 9 16 (6) 22 0 2

Control subjects LAI 6 (n - 8) 0 8 0 0 SF2 6 (n = 8) 0 8 0 0 LAI 9 (n - 9) 0 7 2 0 SF2 9 (n = 9) 0 8 1 0

a No. of patients with ~4-fold increase Of neutralizing titer (no. of patients in this group with neutralization serooonversion). b No. of patients with >I 4,fold decrease of neutralizing titer (no. of patients in this group that had neutralizing titers /> 20 before vaccination but lacked neutralizing activity at the time of follow-up). All sera were tested twice with reproducible titers.

influenza vaccine were studied at the same time intervals. Prior to immunization, 35 of the 40 patients had neutralizing antibodies against LAI and 32 against SF2. All controls could neutralize both strains. After 9 months, 39 patients had LAI-specific neutralization (patient no. 29 lacked detectable neutralizing antibodies) and all but two (patients nos. 31 and 33) had SF2-specific neutralization.

The majority of the patients and controls had stable activity during the follow-up. A 4-fold titer increase or neutralization seroconversion were defined as significant changes. The number of subjects showing this increase was higher in the rgpl60-immunized group compared with the control patients (at 6 months P < 0.03 when tested against SF2, at 9 months P <0.08 against LAI and P < 0.02 against SF2, Fischer's exact test). A few patients lost their serum neutralizing capacity, defined as a 4-fold decrease or loss of detectable neutralizing antibodies, against LAI and none against SF2.

Furthermore, sera from 13 of the patients were tested for the presence of neutralizing antibodies against their own virus isolate. Autologous neutralization was seen in 11 of these at month 0 with no significant titer change at month 9 (data not shown).

3.2. Presence o f A D C C

Many patients already had high ADCC titers before vaccination (Table 2). The titers against LAI and SF2 were ) 2 4 3 0 in 26 and 23 patients, respectively, and t> 7290 in 20 and 17 patients, respectively (data not shown). Improvement of ADCC was defined as an increase of at least two titration steps ( = 9-fold) or ADCC seroeonver- sion. In spite of the large number of patients already having high ADCC titers prior to immunization, improvement could be seen in five patients at 6 months against both virus strains and in eight and seven patients at 9 months against LAI and SF2, respectively. An ADCC decrease was seen in three and one patients at 6 months, and in four and three patients at 9 months, against LAI and SF2, respectively. The majority of patients and controls showed stable titers.

3.3. Presence o f both neutralization and A D C C

ADCC and neutralization reactivities were not always present in the same sera. This lack of correlation was seen both pre- and postvaccination. High titers of both neutralization (/> 500) and ADCC (>f 7290) were seen in eight patients against LAI and in five patients against SF2 in prevaccination sera (data not shown). Improve-

120 K. Broliden et al. / Clinical and Diagnostic" Virology 6 (1996) 115-126

Table 2 Comparison of ADCC activity before and after vaccination with rgpl60

Virus Months of follow-up Titer increase ~ Stably positive Titer decrease b Stably negative

Vaccinated subjects (n = 40) LAI 6 5 (3) 30 3 (2) 2 SF2 6 5 (4) 31 1 (1) 3 LAI 9 8 (4) 27 4 (2) 1 SF2 9 7 (4) 27 3 (2) 3

Control subjects LAI 6 (n = 7) 0 7 0 0 SF2 6 (n = 7) 0 7 0 0 LAI 9 ( n = 9 ) 1 (1) 7 0 1 SF2 9 (n = 9) 1 7 0 1

a No. of patients with ~>9-fold increase of ADCC titer (no. of patients with ADCC seroconversion). b NO. of patients with /> 9-fold decrease of ADCC titer (no. of patients in this group that had ADCC titers 1> 30 before vaccination but lacked ADCC at the time of follow-up). No ADCC activity could be detected in any serum sample against uninfected target cells or in any of 20 HIV-antibody-negative control sera.

ment of both functions against LAI and/or SF2 was seen in eight patients at 9 months (Table 3). None of the patients showed a decrease of both functions against LAI or SF2 at 9 months.

3.4. Correlation of functional antibodies with CD4 counts, virus isolation and A Z T treatment at 9 months of follow-up

Out of the eight patients who demonstrated both neutralizing and ADCC improvements, seven had improved CD4 trends when comparing the time period prior to ( - 24-0 months) and after (0-9 months) the start of the study (Table 3). These eight patients had worse mean monthly CD4 development before the beginning of the study compared with those who did not have neutralizing and ADCC increases. When the difference in mean monthly CD4 change before the start of the study ( - 24-0 months) was compared with different periods postimmunization (0-6 and 0-9 months, respectively), a significantly higher increase (P = 0.07 and P = 0.03 for the respective time periods) was seen in the patients with improved neutralization and ADCC (n = 8) when compared with the patients without improvement

of both functions (n = 32). The former group (n = 8) also had a significantly improved development in mean monthly CD4 change between the period 0-12 months and the period - 2 4 - 0 months as compared to the matched HIV-positive, influenza-vaccinated controls (n= 9) ( + 26.3 x 10 6 cells/l/month and + 9.5 x 106 cells/I/month, respectively; P--0.037).

Changes of CD4 levels were not significantly related to any other serological titer or affinity development in our asymptomatic patients.

Virus could be isolated from 32 patients at 9 months of follow-up. No correlation was seen between presence of the functional antibodies and ability to isolate virus from patient lymphocytes. Virus could be isolated from all patients with decreasing titers of neutralizing and ADCC antibodies (data not shown).

Half of the patients were treated with AZT for 14 days following each immunization. This treatment did not significantly influence any serological parameter (titer increase, avidity increase, ADCC or neutralizing values) or viral isolation from PBL. AZT was, however, related to plasma viral decrease as measured by polymerase chain reaction (PCR) (data not presented).

K. Broliden et al. / Clinical and Diagnostic Virology 6 (1996) 115-126 121

Table 3 Neutralizing and A D C C activity in relation to CD4 changes

Patient code Neutralizing titer against A D C C titer against Monthly CD4 change × 106 cells/1 a

LAI SF2 LAI SF2

Virus isolation at 9 mon ths

AZT

Omo. 9mo. Omo. 9mo. Omo. 9mo. Omo. 9mo. -24-0mo. 0-9mo.

rgp 160-vaccinated subjects 3 < 20 50 400 500 30 270 90 810 7.1 0.8 + - 10 20 40 50 500 0 270 0 0 - 5 3 . 2 - 4 . 6 + + 11 600 800 50 400 7290 7290 30 7290 - 7.4 18.6 + + 12 <20 40 200 500 90 2430 2430 7290 - 13.1 2.7 - - 15 40 40 20 200 2430 2430 270 2430 -21 .1 - 1 . 8 + + 16 <20 40 <20 100 0 810 0 270 - 19.1 4.5 - - 30 <20 40 20 40 90 7290 7290 7290 - 9 . 2 34.8 - - 40 40 200 <20 50 270 7290 7290 7290 - 2 0 . 2 19.3 + +

a First column denotes CD4 changes from - 2 4 mon ths to baseline, second column denotes baseline to 9 months after first vaccination. Only subjects improving both their neutralizing activity and A D C C are shown for the rgpl60 immunized group. Bold figures denote either improvement o f neutralization and/or A D C C (at least a 4-fold titer increase or conversion from a lack of neutralizing activity to a titer of 1> 20; at least a 9-fold titer increase or conversion from a lack of A D C C activity to a titer of /> 30) or decrease of neutralization and/or A D C C (at least a 4-fold titer decrease or reversion from a titer of /> 20 to a lack of neutralizing activity; at least a 9-fold titer decrease or reversion from a titer of >/30 to a lack of A D C C activity). Seven of nine controls had improved CD4 levels, in this case not related to improved neutralization or ADCC.

3.5. Antibody seroreactivity and avidity measured by peptide inhibition ELISA

All patients showed seroreactivity to rgpl60, both preimmunization and during the follow-up period. A 4-fold titer increase, as compared with the preimmunization values, was seen in 13/40 patients (Table 4) and the titers were generally high (104-107). Seroreactivity was also analyzed against peptides representing antigenic and neutralization-inducing regions of the HIV-1 envelope, as well as against control antigens, for the general antiviral antibody response influenced by the rgpl60 immunizations. The initial response to the immunodominant JB4/LAI peptide before immunization was high, and significant increases of these antibody levels were seen in only four patients. The majority of patients showed seroreactivity to the peptides (except to peptide A13/MN) and a titer increase was seen in many patients against both HIV, influenza and CMV and HIV/ p24 antigens. An interesting trend to higher influenza A vaccine responses was seen in the gpl60- and influenza-immunized recipients compared

with those who received influenza vaccine only (53% versus 22%, P = 0.09).

The avidity of antibodies against several of the HIV-l-derived peptides was measured. The avidity maturation calculated as a ratio is shown in Fig. 1 for three of the peptides. Avidity measure- ments were not performed with multivalent antigens (gpl60, p24, CMV or influenza) or with the CD4 binding region. An avidity increase of >i 1.5 times to the V3 region could be identified in more patients against the A12/MN peptide than against the A10/LAI peptide (12 and 4 of 40 patients, respectively). Only a few of the patients had decreased avidity values after immunization compared with preimmunization values. Avidity was not evaluated in control patients.

High neutralizing activity (titer >I 500) to SF2 at 6 or 9 months postvaccination correlated to increased avidity to the A12/MN peptide ( P = 0.05, Fischer's exact test), and also to increased seroreactivity to the A10/LAI peptide (P = 0.04, Fischer's exact test). It was not, however, possible to correlate the 4-fold increase in the titer of neutralizing capacity to any increase in antibody titer or avidity against the peptides tested.

122 K. Broliden et al. / Clinieal and Diagnostic Virology 6 (1996) 115-126

Table 4 Seroreactivity in rgpl60-immunized individuals and control patients 9 months after immunization

Antigen No. of responding individuals

(a.a. no.) rgpl60-immunized subjects (n = 40) Controls (n = 9)

Seropositive Seroconversion Titer increase Titer increase

rgp160/LAI 40 0 13 0 A10/LAI (304-318) 34 2 8 0 AI2/MN (309-321) 39 2 8 0 A13/MN (328-342) 16 0 2 0 JB4/LAI (594-613) 40 0 4 0 C90/LAI (489-503) 32 2 8 0 249/LAI (652-666) 32 0 3 0 265/LAI (732-746) 40 1 7 0

Control antigens p24/LAI 40 0 4 ND CMV 36 1 l 1 0 Influenza A 39 0 21 2 Influenza B 40 0 14 2

Titer increase was defined as a 4-fold increase of mean preimmunization titers.

3.6. Gp I20-CD4 blocking capacity and immunocomplex formation

Gpl20-CD4 blocking capacity in serum increased by ~ 15% (as described in Section 2) in 10 patients, three of whom were seroconverters (data not shown). In three patients, by contrast, the blocking capacity decreased by > 15%. Of the 27 patients who had stable values, the majority (19) had high levels of blocking capacity (in the range 25-88%). After further titration of sera, 10 addi- tional patients developed increasing levels of rsCD4 blocking capacity. Among the control patients, the blocking capacity increased in one patient, decreased in two patients, one patient had low capacity and five patients had stable levels (data not shown). No, or weak, correlation was seen when increased rsCD4 blocking capacity was compared with increased neutralization capacity.

All 40 patients and the nine controls were assayed for occurrence and development of im- munocomplexes and all were found to have low values ( < 150 pg/ml). Below this value, five patients increased their values 2-fold and two patients decreased 2-fold during this study.

4. Discussion

Previous experience indicates that immunization against viral proteins may take place while infection with the same virus is active. A reduced morbidity compared to non-immunized individuals has been demonstrated with cytomegalovirus, herpes simplex and hepatitis B viral antigens (Plotkin et al., 1990; Stanberry, 1991; Pol et al., 1993). During primary HIV infection, an apparent failure occurs in the CD4 response which is likely to be followed by incapacity to produce efficient neutralizing antibodies. Considering the very large viral burden and the immense HIV replication capacity even early in HIV disease (Embretson et al., 1993; He et al., 1995; Wei et al., 1995), it is likely that only combined regimens, i.e. a combi- nation of chemotherapies, or combined immune- and chemotherapy, may reduce the viral burden for any sustained period.

The humeral immunity of 40 asymptomatic HIV-l-infected subjects who received repeated immunizations with recombinant HIV-gpl60 of a LAI strain was monitored. Increased seroreactivity postimmunization was seen in 30 of the 40 HIV-immunized patients against at least one of

K. Broliden et al. / Clinical and Diagnostic Virology 6 (1996) 115-126

four immunogenic regions represented by peptides from gpl20 and gp41. No significantly decreased responses were seen in the same group. These findings confirm data by Redfield et al. (1991) who discovered new peptide reactivities in sera of rgp160-immunized individuals.

High-avidity antibodies are more efficient than

a I A12/MN (V3 region) 7 -

6 -

i' 1 4 ¸

8 ¸

2 ¸

0 ' _!

Ratio

b 7 CgO/LAI (Op120) e

| ,

2

o t

1 o °

, . . . i . . , , ,

Ratio ~ ;

4

" . . ,

7 7 7 7 7 7 7 ' 7 " . . . .

R a t i o

Fig. 1. Avidity changes against peptides before and after immunization, calculated as a ratio of reciprocal molar peptide inhibitory concentrations: (a) A12/MN (n = 32), the peptide derives from the HIV-1-MN V3 region; (b) C90/LAI (n = 18), the peptide derives fi'om the HIV-1-LAI C-terminal region; (c) JB4/LAI (n = 39), the peptide derives from the HIV-I-LAI gp4l immunodominant region.

123

antibodies with low avidity in a number of biological reactions, including virus neutralization, complement fixation and elimination of antigen (Steward and Steensgaard, 1983). The avidity of IgG is initially low after primary infection or immunization and matures slowly. In various infectious diseases, the measurement of I5(3 avidity has been shown to be a specific and sensitive diagnostic tool. The principle in the inhibition enzyme immunoassay, which we have used to measure avidity in the present study, is to inhibit an antigen-antibody reaction occurring on the solid phase by soluble peptides (Rath et al., 1988), In this way we could demonstrate increased avidity to antigenic peptides representing both conserved and variable regions. Decreased avidity against the same peptides was also seen in a few subjects. The avidity increase seen to a V3 peptide representing the MN strain correlated with high neutralizing titers, which confirms the V3 region as the principal neutralizing domain, it is of interest to note that avidity was increased to the V3 MN peptide, while the vaccine contained V3 of the IIIB LAI strain. This indicates an anamnestic response.

Antibodies reacting with the CD4 binding site of gpl20 have been implicated, together with anti- V3 antibodies, as important for inhibition of HIV infection. Between 25% and 50% of the immunized patients in the present study also improved their serological response with respect to the capacity to block soluble CD4 binding to gpl20. Since it is quite unlikely that the denatured rgpl60 induces primary antibodies reacting to disconti- nous sites, again this may be an anamnestic response.

Sequential serum samples were analyzed for the presence and titers of neutralizing antibodies which can bind cell-free virus, and of ADCC, which may react with virally infected cells. These functional antibodies were tested against two strains of HIV-1, LAI and SF2. Although ADCC titers were already generally high before vaccination, some patients improved their response. An improvement of neutralization was observed in about 25% and 40% of the patients against LAI and SF2, respectively, whereas a decrease in neutralizing activity was seen in 5% and 0% of the


patients, respectively. More patients thus reacti- vated their response to SF2, which is geneti- cally more homologous to the consensus European and US strains than the vaccine represented as LAI. Since better affinity increases were also seen to peptides of the consensus strain MN compared with the vaccine strain LAI peptide, we cautiously interpret these findings as indicating anamnestic responses to the patients own viruses. Sequencing of sequential virus isolates has indeed shown that all have V3 regions characteristic of the consensus B clade (Myers, 1995).

Nine asymptomatic, HIV-l-infected control subjects received influenza vaccine only instead of rgpl60. They did not show the same level of HIV-specific titer increase. It is of particular interest to note that influenza vaccination gave a higher frequency of titer increases in patients who also received gpl60 vaccine (53%) than in controls receiving influenza vaccine only (22%).

A better CD4 development was previously shown to be correlated to improved specific and recall T-cell reactions in vitro and in vivo (Wahren et al., 1994b; Redfield et al., 1991; Blatt et al., 1993). We identified specific HIV titer and affinity increases in the present group of gpl60-immunized patients, and found a cer- tain correlation between increasing levels of neutralizing and ADCC antibody titers, and CD4 improvement. Individuals who responded with increased ADCC and neutralizing antibodies to LAI and SF2 had a more rapid decline in CD4 values prior to immunization. After immunization and acquisition of these antibody reactivities, they improved their CD4 development to a significantly greater extent than those without increased ADCC and neutralization. However, titer rises were seen both with and without CD4 improvement and did not mirror the recall B-cell reactivities to CMV or p24.

Concern has been expressed regarding the possible immunosuppressive effect of HIV envelope proteins and thus the possible negative effect on the host immune system after vaccination with the rgpl60. Some parts of the gpl60 have therefore been depleted in the present vac-

cine in order to avoid cross-reactivity with cellular proteins. No signs of side effects were noted in the patients following immunization. In fact, many of the patients decreased their autoantibody reactivities to HIV-homologous HLA peptides (Lundholm et al., 1994). In order to inhibit viral replication which might the- oretically occur upon activation of CD4 ÷ T cells in HIV-infected individuals, AZT was ad- ministered for 14 days following each vaccination to half of the patients (Wahren et al., 1994b).

Evidence as to which protective immune function is most important in vaccinated hu- mans and animals is still lacking. Interestingly, both humoral and cellular immunities were en- hanced when seronegative, vaccinia-naive individuals were immunized with a live vaccinia recombinant expressing HIV envelope followed by boosting with rgpl60 (Cooney et al., 1993; Graham et al., 1993). Vaccination of seronegative individuals against HIV may need a combined vaccine regimen, whereas vaccination with viral recombinant proteins appears suffi- cient to induce humoral and cellular responses in subjects who can respond anamnestically.

We have shown increased levels of functional antibodies in some of the rgpl60-immunized individuals. In concordance with the T-cell prolif- erative responses (Wahren et al., 1994b), the functional antibody responses could for some patients be related to improved CD4 development. The regained capacity to form humoral and cellular antiviral responses during HIV infection indicates that attempts to modulate the host immune response are of interest.

Acknowledgements

Financial support was obtained from the Swedish Medical Research Council, the Swedish Board for Technical Development (NUTEK) and Physicians against AIDS Research Founda- tion. Informed consent was obtained from all patients. Approval was obtained from the ethi- cal committee at the South Hospital and Karolinska Institute, Stockholm, Sweden.

K. Broliden et aL / Clinical and Diagnostic Virology 6 (1996) 115-126 125

References

Albert, J., Abrahamsson, B., Nagy, K. et al. (1990) Rapid development of isolate specific neutralizing antibodies after primary HIV-1 infection and consequent emergence of virus variants which resist neutralisation by autologous sera. AIDS 4, 107-112.

Berman, P.W., Gregory, T.J., Riddle, L. et al. (1990) Protec- tion of chimpanzees from infection with recombinant glycoprotein gpl20 but not gpl60. Nature 345, 622-625.

Blatt, S.P., Hendrix, C.W., Butzin, C.A. et al. (1993) Delayed- type hypersensitivity skin testing predicts progression to AIDS in HIV-infected patients. Ann. Intern. Med. 119, 177-184.

Broliden, P.A., M~ikitalo, B., /kkerblom, L. et al. (1991) Identification of aminoacids in the V3 region of gpl20 critical for virus neutralization by human HIV-1 specific antibodies. Immunology 73, 371-376.

Broliden, K., Sievers, E., Tovo, P.A. et al. (1993) Antibody-de- pendent cellular cytotoxicity and neutralization in vertical transmission of HIV-infection. Clin. Exp. Immunol. 93, 56-64.

Bruck, C., Thiriart, C., Fabry, L. et al. (1994) HIV-1 envelope- elicited neutalizing antibody titres correlate with protection and virus load in chimpanzees. Vaccine 12, 1141-1148.

Cooney, E.L., McElrath, M.J., Corey, L. et al. (1993) En- hanced immunity to human immunodeficiency virus (HIV) envelope elicited by a combined vaccine regimen consisting of priming with a vaccinia recombinant expressing HIV envelope and boosting with gpl60 protein. Proc. Natl. Acad. Sci. USA 90, 1882-1886.

Embretson, J., Zupancic, M., Ribas, J.L. et al. (1993) Massive covert infection of helper T lymphocytes and macrophages by HIV during the incubation period of AIDS. Nature 362, 359-362.

Fultz, P.N., Nara, P., Barre-Sinoussi, F. et al. (1992a) Vaccine protection of chimpanzees against challenge with HIV-1- infected peripheral blood mononuclear cells. Science 256, 1687-1690.

Fultz, P.N., Gtuckman, J.C., Muchmore, E. and Girard, M. (1992b) Transient increases in numbers of infectious cells in an HIV-infected chimpanzee following immune stimulation. AIDS Res. Hum. Retroviruses 8, 313-317.

Gardner, M.B., Jennings, M., Carlson, J.R. et al. (1989) Postexposure immunotberapy of simian immunodeficiency virus (SIV) infected rhesus with an SIV immunogen. J. Med. Primatol. 18, 321-328.

Gibbs, C.J., Peters, R., Gravell, M. et al. (1991) Observations after human immunodeficiency virus immunization and challenge of human immunodeficiency virus seropositive and seronegative chimpanzees. Proc. Natl. Acad. Sci. USA 88, 3348-3352.

Girard, M., Kiney, M., Pinter, A. et al. (1991) Immunization of chimpanzees confers protection against challenge with human immunodeficiency virus. Proc. Natl. Acad. Sci. USA 88, 542-546.

Graham, B.S., Matthews, T.J., Belshe, R.B. et al. (1993) Augmentation of human immunodeficiency virus type 1 neutralizing antibody by priming with gpl60 recombinant vaccinia and boosting with rgpl60 in vaccinia-naive adults. J. Infect. Dis. 167, 533-537.

Hinkula, J., Gidlund, M., Persson, C., Osterhaus, A. and Wahren, B. (1994) Enzyme immuno assay (ELISA) for the evaluation of antibodies directed to the CIM-receptor- binding site of the HIV gpl20 molecule. J. Immunol. Methods 175, 37-46.

Ho, D.D., Neumann, A.U., Perelson, A.S., Chert, W., Leon- ard, J.M. and Markowitz, M. (1995) Rapid turnover of plasma virions and CD4 lymphocytes in HIV-I infection. Nature 373, 123-126.

Lundholm, P., Wahren, M., Sandstr6m, E., Volvovitz, F. and Wahren, B. (1994) Autoreactivity in HIV-infected individuals does not increase during vaccination with envelope rgpl60. Immunol. Lett. 41, 147-153.

Mathiesen, T., S6nnerborg, A. and Wahren, B. (1989) Detec- tion of antibodies against myelin basic protein and increased levels of HIV-IgG antibodies and HIV antigen after solubilization of immune complexes in sera and CSF of HIV infected patients. Viral Immunol. 2, 1 - 11.

Myers, G. (1995) Database Human Retroviruses and AIDS. Los Alamos National Laboratory, Los Alamos, NM.

Niedrig, M., Harthus, H.P. and Hinkula, J. (1992) Inhibition of viral replication by monoclonal antibodies against human immunodeficiency virus gpl20. J. Gen. Virol. 73, 2451-2455.

Plotkin, S., Starr, S., Friedman, H., G6nzc61, E. and Bray- man, K. (1990) Vaccines for the prevention of human cytomegalovirus infection. Rev. Infect. Dis. 12, 827.

Pol, S., Driss, F., Carnot, F., Michel, M., Berthelot, P. and Brechot, C. (1993) Vaccination against hepatitis B virus: an efficient immunotherapy against hepatitis B multiplication. C. R. Acad. Sci. 316, 688-691.

Putkonen, P., Thorstensson, R., Albert, J. et al. (1990) Infec- tion of cynomolgus monkeys with HIV-2 protects against pathogenic consequences of a subsequent simian immunodeficiency virus-infection. AIDS 4, 783-789.

Rath, S., Stanley, C.M. and Steward, M.W. (1988) An inhibition enzyme immunoassay for estimating relative antibody affinity and affinity heterogeneity. J. Immunol. Methods 106, 245-249.

Redfield, R.R., Birx, D., Ketter, W. et al. (1991) A phase I evaluation of the safety and immunogenicity of vaccination with recombinant gpl60 in patients with early human immunodeficiency. N. Engl. J. Med, 324, 1677-1684.

Shearer, G.M. and Clerici, M. (1991) Early T-helper cell defects in HIV infection. AIDS 5, 245-253.

Spear, G.T., Sullivan, B.L., Landay, A.L. and Lint, T.F. (1990) Neutralization of human immunodeficiency virus type 1 by complement occurs by viral lysis. J. Virol. 64, 5869-5873.

Stanberry, L. (1991) Immune parameters in HSV infection. Rev. Infect. Dis. 13, $920-$923.


Steward, M.W. and Steensgaard, J. (1983) Antibody Affinity: Thermodynamic Aspects and Biological Significance. CRC Press, Boca Raton, FL.

Steward, M.W., Stanley, C.M., Dimarchi, R., Mulcahy, G. and Doel, T.R. (1991) High-affinity antibody induced by immunization with a synthetic peptide is associated with protection of cattle against foot-and-mouth disease. Im- munology 72, 99-103.

Wahren, B. (1994a) New observations on functional and neutralizing humoral immune response to HIV. AIDS 8 (Suppl. 1), S147 S153.

Wahren, B., Morfeldt-Mgmsson, L., Biberfeld, G. et al. (1986) Impaired specific cellular response to HTLV-IlI before

other immune defects in patients with HTLV-III infection. N. Engl. J. Med. 315, 393 394.

Wahren, B., Bratt, G., Persson, C. et al. (1994b) Improved cell-mediated immune responses in HIV-1 infected asymptomatic individuals after immunization with envelope glycoprotein gpl60. J. AIDS 7, 220-229.

Wei, X.P., Ghosh, S.K., Taylor, M.E. et al. (1995) Viral dynamics in human immunodeficiency virus type 1 infection. Nature 373, 117-122.

Weiss, R.A., Clapham, P.R., Cheingsong-Popov, R. et al. (1985) Neutralization of human T-lymphotropic virus type III by sera of AIDS and AIDS-risk patients. Nature (Lon- don) 316, 69 72.

Documents

Analyses of functional antibody responses in HIV-1-infected individuals after vaccination with rgp160