Antibody to a Synthetic Oligopeptide in Subjects at Risk for

Vol. 25, No. 8

JOURNAL OF CLINICAL MICROBIOLOGY, Aug. 1987, p. 1498-1504 0095-1137/87/081498-07$02.00/0 Copyright © 1987, American Society for Microbiology

Antibody to a Synthetic Oligopeptide in Subjects at Risk for Human Immunodeficiency Virus Infection RICHARD S. SMITH,'* ROBERT B. NASO,1 JON ROSEN,' ALICE WHALLEY,' YAO-LING HOM,' KENWAY HOEY,' CAROLINE J. KENNEDY,2 J. ALLEN MCCUTCHAN,2 STEPHEN A. SPECTOR, 2 3 AND DOUGLAS D. RICHMAN2'4 Johnson & Johnson Biotechnology Center, San Diego, California 92121,1 and Departments of Medicine,2 Pediatrics,3 and Pathology,4 University of California, San Diego, and Veterans Administration Medical Center, San Diego, California 92161 Received 3 March 1987/Accepted 28 April 1987

Detection of antibodies to human immunodeficiency virus (HIV) by enzyme-linked immunosorbent assay (ELISA) is the accepted method to screen blood products at risk to transmit infection. The presence of antibodies to HIV in 565 serum specimens from 274 patients with acquired immunodeficiency syndrome (AIDS) or AIDS-related complex, symptomatic and asymptomatic subjects at risk for AIDS, and controls was determined with an ELISA that incorporates synthetic peptides (designated E32/E34) representing sequences in the envelope glycoprotein gp41. Of 105 specimens from patients with AIDS or AIDS-related complex, 3 specimens that were negative by commercially licensed ELISA and immunoblot test were similarly unreactive in the E32/E34 ELISA. For homosexual mnen with generalized lymphadenopathy, 186 specimens were positive by the E32/E34 ELISA and 63 specimens were negative. In comparison, with the licensed ELISA, 184 of these samples were positive and 65 samples were negative. The two samples that were positive in the E32/E34 ELISA but not the commercial kit were also positive by immunoblotting. Sequential sera from one individual who apparently underwent seroconversion according to the commercial assays were all positive by E32/E34 ELISA and immunoblotting. Thus, the ELISA with synthetic peptides is an extremely sensitive and specific test of antibody response to HIV and has not yet yielded a negative result with a Western blot (immunoblot)-confirmed antibody-positive serum.

The human immunodeficiency virus (HIV) (formerly called lymphadenopathy-associated virus, human T-lymphotropic virus type III, or acquired immunodeficiency syndrome [AIDS]-related virus) is the etiologic agent of AIDS and related disorders (3, 18, 22, 26). Antibodies to HIV antigens are detectable in the blood of many individuals with AIDS or at risk for infection with HIV (1, 26, 27). Propagation of this retrovirus in continuous cell lines such as H9 cells has permitted the development of widely available enzyme-linked immunosorbent assays (ELISAs) (24, 26) which are now widely available to screen blood products

The effectiveness of ELISAs now used to screen blood and thereby reduce the possibility of transfusion-related AIDS (8) is greatly impeded by an incidence of false-positive results. With sera from low-risk blood donors, as many as 90% of positive results are thought to be falsely positive (12). Therefore, confirmation of ELISA-positive samples with an immunoblotting technique (Western blot) has been recommended (17, 25). However, inmmunoblotting has not been standardized, is subjective in interpretation, and is too complex for a screening test. The frequency of false-positive results with ELISAs that use H9 cells as the source of viral antigen has stimulated attempts to develop a source of viral antigens free from human lymphocyte antigen and other antigens that apparently cause false positivity in ELISAs with H9 cells (10). Recombinant DNA technology has been used to express recombinant HIV polypeptides in Escherichia coli (5, 6). Two recombinant polypeptides representing regions of the transmembrane 41-kDa region have been shown to be immunoreactive with human sera (5, 6). The sensitivity and specificity of assays using these polypeptides remain to be established. Subsequently, synthetic peptides representing amino acid sequences of native proteins were used to define antigenic determinants that are involved in host immune responses (14). Peptide synthesis permits the manufacture of chemically defined epitopes free of any contaminating proteins. This technology has been effective in identifying immunoreactive peptides from the gp4l envelope glycoprotein of HIV (11, 30). One such peptide, designated E8, composed of 15 amino acids, was immunoreactive with all of nine sera (J. Rosen, T.-L. Hom, A. Whalley, R. S. Smith, and R. R. Naso, in P. A. Lucin, K. S. Steiner, and L. A. Overby, ed., Genetic Engineering Approaches to AIDS Diagnosis, in

(31).

Additionally, immunoblotting and radioimmunoprecipitation assays for this antibody have helped to define the HIV-specific antigens to which infected individuals generate antibody responses (22, 23). The 17-, 15-, and 24-kilodalton (kDa) proteins and the 53 to 55-kDa precursor protein are viral core protein products of the gag gene complex (16, 19). A 160-kDa glycoprotein, a product of the env gene, appears to be the precursor for the gpl20 and gp4l envelope glycoproteins (29). The 41-kDa glycoprotein and the 24-kDa core protein appear to be major markets for seropositivity by immunoblot analyses (22, 29). Other genes, such as pol, tat, and 3' orf, also yield protein products to which individuals infected with HIV generate antibody responses (7). Although some of these gene products are probably critical for the replication of HIV, the lower concentrations of these proteins relative to the gag and env gene products make antibody responses to gag and env antigens more important for the serodiagnosis of HIV infection. *

Corresponding author. 1498

VOL. 25, 1987

IMMUNOREACTIVE PEPTIDES FROM env REGION OF HIV

1499

tat

v1~ 5

P120

g

gp41

1 1 of 1

CL

E34 A V E R Y L K D

env

Q Q

E32 I WG C S G K L I C T T A V PWN A S L L G I WG C S G K L I

FIG. 1. Sequence and approximate genetic location of synthetic oligopeptides E32 and E34. These oligopeptides were derived from highly conserved region of the gp4l region of HIV. The tat and art genes are actually produced from spliced mRNAs, the 5' end of which is synthesized upstream of env, as shown, and then spliced to the 3' end from gp4l. The tat and art sequences, however, do not overlap the peptide sequences of E32/E34 which are shown.

a

press) from patients with AIDS or AIDS-related complex (ARC). As described here, a second oligopeptide sequence, Nterminal to the E8 region, has been more recently identified which appears to increase the sensitivity for detection of antibody to HIV. In this study, two synthetic oligopeptides from the E8 region of gp4l, designated E32 and E34, were used as adsorbed antigen in an ELISA to assay sera from wellcharacterized subjects at risk for HIV infection. With this assay, sera from 274 patients were analyzed. Many of these patients had histories that indicated a clear-cut risk of HIV infection. The outcome of these studies indicated greater sensitivity than that of existing assays, with the significant advantage of freedom from false-positive responses. MATERIALS AND METHODS Study subjects and control groups. Subjects being followed in a longitudinal study of homosexual men in San Diego, Calif., were classified into seven groups based on clinical and immunological criteria. Group A met the Centers for Disease Control definition of AIDS. Group B met the Centers for Disease Control-National Institutes of Health definition of ARC. Specifically, each patient had one or more of the following clinical symptoms persisting for at least 3 months: (i) fever of more than 100°F (37.8°C), (ii) weight loss of more than 10% of body weight or 15 lb. (ca. 6.8 kg), (iii) persistent lymphadenopathy involving two or more extrainguinal chains with at least one lymph node equal to or greater than 10 mm in diameter in each chain, (iv) diarrhea, (v) fatigue sufficient to interfere with work, and (vi) night sweats. At least two abnormalities necessary for inclusion by laboratory analyses included a depressed absolute T-cell number of less than 500/mm and (i) a ratio of T4+ to T8+ lymphocyte percentages of less than 1 or at least one of four cytopenias, including leukopenia of less than 4,500/mm, (ii) thrombocytopenia of less than 100,000/mm, (iii) anemia with a hematocrit of less than 40 or absolute lymphopenia with total lymphocytes of less than 1,100/mm, or (iv) failure to respond to skin testing with six antigens (Candida sp., coccidioidin, mumps virus, purified protein derivative, tetanus toxoid, and Trichophyton sp.). Group C consisted of patients with persistent generalized lymphadenopathy as defined above but lacking the other criteria for group B. Group D was men without signs or symptoms of ARC but having at least one abnormal outcome

in laboratory tests. Group E was homosexual men without physical signs, symptoms, or abnormal laboratory test results. Subjects were classified independently of antibody status to HIV. Control subjects, healthy heterosexual medical and laboratory personnel, were designated as group F. Group G consisted of patients from other risk groups, including spouses of HIV-infected patients. Serum and plasma samples were obtained by venipuncture, coded, and stored at -70°C. In addition, 102 control sera from patients with autoimmune diseases and other infections including systemic lupus erythematosus, rheumatoid arthritis, and cytomegalovirus, Epstein-Barr virus, and herpesvirus infection were assayed as controls in the E32/E34 ELISA. Definition of E32/E34 env peptides. The HIV genome and the sequence of the E32 and E34 peptides are diagrammed in Fig. 1. The E8 env peptide originally shown to be reactive with most, if not all, sera from AIDS-ARC patients was selected for synthesis based upon its conserved sequence among various HIV isolates sequenced. This peptide also had two cysteines in close proximity, which potentially could allow for intramolecular, intermolecular, or both types of disulfide bond formation, resulting in a peptide with a native peptide-like conformation. The sequence of E8 is IWGCSGKLICTTAVP. The E32 and E34 peptides were overlapping extensions of E8 that retained specificity for antibodies to HIV gp4l while increasing the sensitivity of the assay. The E32/E34 peptides were derived from a sequence of amino acids contained in a conserved region of the gp4l viral envelope protein (19). These peptides were selected on the basis of the hydrophilicity of the sequence and the presence of cysteine residues in the peptides. E32/E34 ELISA. The E32 and E34 peptides were solubilized by dissolving the peptides at 200 ,ug/ml in 0.1 M sodium bicarbonate buffer, pH 9.6. Under these solubilization conditions, the E32 peptide exists as a cyclic monomer and E34 is dimerized as determined by high-pressure liquid chromatographic analysis (unpublished data). The ELISA was performed on Removawell 12-well microtiter strips (Immulon Il; Dynatech Industries, Inc., McLean, Va.). Each well was coated with 200 ,ul of a 5-,ug/ml concentration of each peptide in the previously described buffer for a minimum of 12 h at 4°C. The peptide solution was discarded, and excess binding sites were quenched with 300 ,ul of 10% bovine serum albumin per well in phosphate-buffered saline (PBS; 0.15 M NaCI, 0.05 M NaH2PO4, 0.05 M Na2HPO4, pH 7.3). This solution was incubated for 90 min in a 37°C humidified

1500

J. CLIN. MICROBIOL.

SMITH ET AL.

chamber. The wells were emptied by inversion and dried for 60 min at 37°C. A volume of 200 il containing 1% pooled bovine gamma globulin, 5% bovine serum albumin, and 0.05% Tween 20 in PBS was then added to the wells before addition of 20 tl of test serum, which was mixed by pipetting at least four times. This mixture was incubated for 30 min at 37°C and then washed five times with 350 pl of 0.05% Tween 20 in PBS. A solution of 200 pl of horseradish peroxidase-conjugated mouse monoclonal antibody to human immunoglobulin G (Ortho Diagnostic Systems, Inc., Raritan, N.J.) was diluted (1:3,500) in 50% heat-inactivated fetal bovine serum-1% heat-inactivated horse serum-0.05% Tween 20 in PBS and added to the wells. After 30 min of incubation at 37°C, the wells were washed with 350 pl of 0.05% Tween 20 in PBS five times and incubated with 200 pl of o-phenylenediamine solution (one 10-mg o-phenylenediamine tablet in 15 ml of water and 6.25 pi of 30% H202; Sigma Chemical Co., St. Louis, Mo.) for 30 min at room temperature. The reaction was stopped by addition of 50 pl of 4 N H2SO4. The optical density or absorbance of the solution was then determined at 490 nm. Reference HIV antibody procedures. Antibody to HIV in serum and plasma was measured with two commercially available ELISA kits: HTLV-III EIA (Abbott Laboratories, North Chicago, Ill.), henceforth termed ELISA 1, arnd Virgo HTLV-III ELISA (Electro-Nucleonics, Inc., Columbia, Md.), termed ELISA 2. These assay procedures were performed according to manufacturer instructions. ELISA 2 was used on all specimens. ELISA was used as an additional ireference procedure on specimens from subjects who serocccnverted. Immunolblot assay. Selected samples were analyzed by 200 186

184

Positive

I

Negative

cn

E

7670

m

~~~~~~~21

26 o

A

B

C

E

D

F

G

Subject Study Group FIG. 3. Results from commercial ELISA 2

the 565 serum

on

samples specified in Fig. 2. ELISA 2 was performed as directed in the package insert, and the cutoff used as 0.10 absorbance or optical density unit. The number of subjects in each group is indicated

above the bar.

immunoblotting. These samples included those from subjects who seroconverted or provided setum that yielded discordant results. The assay was performed as described by Symington et al. (28) except when sera were diluted 1:20 in a diluent that contained 0.05 M Tris, 0.150 M NaCl, 2.5% fetal bovine serum, 0.2% bovine serum albumin, and 0.05% Tween 20, pH 7.6. Binding of human antibody to HIV-

I

Positive Negative

encoded

proteins

on

the

blots

was

detected

by using goat

anti-human

immunoglobulin conjugated with horseradish peroxidase (Cooper Biomedical, Inc., West Chester, Pa.). A specimen was considered to be reactive by the immunoblot-

ting technique when antibody directed to both the 24- and 41-kDa bands was present (22). Statistical analyses. Mean values are expressed with their standard deviations. Outlying normal values were removed

ca

cn, o

from the E z

F

76

75

76

75

normal-range studies by using by Barnett (2).

the statistical

theory

described

Peptides. The human T-lymphotropic type III env peptides

70

631

E32

and

E34

solid-phase quences

Yi22 E5

5 3 ma w 2> 2 1 Ét = _ _ % _ 3 rx A B C D E

used

method

were

in

this

of

study

were

Merrifield

characterized

by

synthesized by

(15).

amino

The

acid

peptide

the

se-

analysis (4150

acid analyzer, LKB Instruments, Inc., Rockville, Md.) and anlino acid sequence (Applied Biosystems 470 H protein sequences). Analytical high-pressure liquid chromatography of the pooled proteins showed the peptide to be greater than 70% pure. The peptide was lyophilized to remove acetic acid.

amino

~~~~13 _i F G

Subject Study Grbup FIG. 2. Results from the E32/E34 ELISA of 565 sera from the study population classified as follows: group A, AlDS; group B, ARC; group C, subjects with generalized lymphadenopathy arid immune abnormalities; group D, homosexual men with an immunological abnormality; group E, heterosexual male and female controls; group G, subjects from other risk groups, including spouses of HIV-infected patients; group F, heterosexual controls. The number of subjects in each group is indicated above the bar.

RESULTS Determination of normal ranges. Serum or plasma was obtained from 380 healthy volunteers and assayed in the E32/E34 ELISA. Three samples were rejected from the normal-range determination as statistically significant outliers according to the method of Barnett (2). The normal range for the E32/E34 ELISA was then calculated from the assay

IMMUNOREACTIVE PEPTIDES FROM

VOL. 25, 1987

TABLE 1. Comparison of results from E32/E34 ELISA and ELISA 2 for 565 sera from the study population No. of results that were:

Method

E32/E34 ELISA ELISA 2

Truly Falsely Truly Falsely

Predictive %

% Sensi-

Speci100

positive

positive

negative

negativea

322

0

240

3

99.1

321

lb

238

5C

98.5

tivity

in ~~value this popu-

tion (%) f'ylto

ficity

100

99.6

99.7

Includes three sera from group A (AIDS) that were negative for antibody by ELISAs 1 and 2 and immunoblot. b Positive by ELISAs 1 and 2, negative by E32/E34 ELISA and immunoblot. c Includes two sera that were positive by E32/E34 ELISA and immunoblot and came from subjects who later seroconverted by ELISA 2. a

results of testing 377 samples. According to Barnett, the mean value plus 2 standard deviations determines the usual cutoff level for samples that are considered positive. Because of the critical nature of this disease, we chose to use 6 standard deviations to locate the positive cutoff value. The mean optical density value or absorbance was 0.024 ± 0.018. The mean optical density value plus 6 standard deviations was 0.15. E32/E34 ELISA results on unknown sera above an optical density of 0.15 were considered to be positive. If the assay results from these three outliers were to be included, the mean optical density and standard deviation would be 0.034 + 0.040. The sera from patients with autoimmune diseases and other infections were negative in the E32/E34 ELISA. Comparison of the E32/E34 ELISA with ELISA kits 1 and 2. A total of 565 specimens from 274 study subjects (including controls) that had been classified by clinical and immunological status into the groups described earlier were examined by E32/E34 ELISA, and the results are shown in Fig. 2. The results from ELISA 2 on the same samples appear in Fig. 3. Of 79 samples in group A (AIDS), 3 were shown to be

env

REGION OF HIV

1501

antibody negative by the E32/E34 ELISA, by ELISA 2, and by immunoblotting. Retrospective analysis of the records of these three patients indicated features atypical of AIDS, raising questions about the accuracy of the clinical diagnosis. In group B (ARC) all 26 samples were positive in both ELISA procedures. In group C (persistent lymphadenopathy syndrome), 186 specimens were scored positive by the E32/E34 ELISA and 63 specimens were negative; ELISA 2 scored 184 samples positive and 65 samples negative in group C. The results from our E32/E34 ELISA and ELISA 2 were in accord with all 100 samples in group D. In group E, both the E32/E34 ELISA and ELISA 2 scored the same 3 of 73 specimens positive. In group F (normal controls), all 22 samples were negative by the E32/E34 ELISA; however, ELISA 2 scored one sample as positive. Concordance between methods was observed in group G when all 16 samples were assayed. Immunoblots. When discordant results were checked by immunoblot, no false-positive results were observed for the 565 sera tested with the E32/E34 ELISA; however, falsenegative specimens were found for 3 of 79 individuals with AIDS (Table 1). These were observed out of a total of 322 specimens positive by E32/E34 ELISA and produced a 1% false-negative rate. These three samples were also negative by ELISA 2 and by immunoblotting, suggesting that these patients with advanced disease truly had no measurable antibody. Assay results with ELISA 2 resulted in one false-positive (F group) and five false-negative results, including the three antibody-negative samples from AIDS patients (Table 1). The false-negative rate for ELISA 2 was established as 1.5%. Of a total of 565 specimens, 63 were analyzed for the presence of HIV-specific antibody by the immunoblot technique. Of these 63 specimens, 2, both from study group C, appeared to have antibodies directed to the 24-kDa band and were scored as negative both in the E32/E34 ELISA and in the two commercial HIV ELISAs. All other results with samples assayed by immunoblotting concurred with the results of the E32/E34 ELISA. In addi-

TABLE 2. Detection of HIV seroconversion by HIV ELISA and immunoblotting E32/E34 ELISA

Patient no. and study group

ELISA 1

ELISA 2

Draw Drawreu satudy nro. dand date Score Score Score ODa OD OD

E C B

7-17-84 5-7-85 11-19-85

0.019 2.0 2.0

+ +

0.035 1.3 2.0

C C C

8-21-84 6-17-85 2-10-85

0.015 1.8 2.0

+ +

0.07 0.5

D C C

8-27-84 6-3-85 10-29-85

0.03 1.45 0.39

+ +

C C C

10-23-85 10-6-85 11-19-85

2.0 2.0 2.0

+ + +

+

0.005 0.58 0.22

+ +

+

+ +

2

-

0.009

-

1.14

+ +

0.37 0.62

+ +

0.048 1.7

-

0.002

-

+

0.73

NDC

ND

0.79

+ +

+ +

0.084 0.18 0.19

-

0.096 0.25 0.19

-

+ +

+ + +

+ + 24

3 -

4

a

b

c

OD, Optical density. Questionable p24 band seen. ND, Not determined.

+ +

J. CLIN. MICROBIOL.

SMITH ET AL.

1502

310-

i11n.

-

c E E

EI 1E32/E34 ElISA

C cn o-

-

.0

-

a>

w

E32/E34 ELISA ELISA method il

-

Co c,

cn

-

a>

E2-

j 0

F% v% PSr n e 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

f~~~~~~~lrxl»nS

FI|M ]rlSnrM 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

Absorbance

Absorbance

Cutoff

o

FIG. 4. Frequency distribution of ELISA results from all sera in the study population. The absorbance values for the E32/E34 ELISA and ELISA 2 are shown on the abscissa. The results from the E32/E34 ELISA are shown in panel A. Test results from the assay of these same sera by ELISA 2 are shown in panel B. A composite of the E32/E34 ELISA and ELISA 2 results is shown in panel C.

,à in

B E: ELISA method /l

DISCUSSION a..

o

.0

E

z

0

u-_

0.

0l.

0

1

n1

n 1.

1. 1

112

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0

t

Cutoff

Absorbance

tion, two samples that were positive by both immunoblot and E32/E34 ELISA were negative by ELISA 2 (Table 1). Seroconversion of study subjects detected by the ELISAs and immunoblots. From our study population of 274 individuals tested, 4 seroconverted for antibody to HIV according to ELISA 2 (Table 2). Overall, test results agreed among the four assays used, with one important exception. HIVspecific antibodies were detected on all sample dates from patient no. 4 (group C) by the E32/E34 ELISA and the immunoblotting technique. However, on the 23 October 1985 sampling date, ELISAs 1 and 2 yielded negative results. ELISAs 1 and 2 yielded negative results on a serum that was clearly from an infected individual and was positive by Western blot and E32/E34 ELISA. With the other seroconversion cases, there was agreement among the assays.

We have developed an ELISA (called here E32/E34) that detects antibody to HIV by its immunoreactivity to chemically synthesized oligopeptides used as the antigen. The amino acid sequences selected were deduced from the published nucleotide sequence of a highly conserved region of the gp4l envelope glycoprotein of HIV (19). When 543 serum samples from a study population of 274 high-risk individuals were examined, most test results were bimodally distributed into those displaying extremely high or extremely low reactivity manifested in optical density analysis. These results contrasted with those produced by two licensed ELISA kits, the results of which yielded a spectrum of optical density or absorbance values spanning the arbitrary cutoff point (Fig. 4). The combination of this unclear cutoff point for positive results and failure to detect all truly positive specimens in test blood has resulted in a situation in which >90% of positive results in populations of low-risk blood donors are falsely positive (17). Of 79 sera from patients with AIDS (group A), 3 were negative for antibody by all ELISA procedures and immunoblotting. Others have described antibody-negative patients with AIDS or ARC despite virus-positive cultures (9, 21). The HIV-specific antibody may be undetectable (i) because it has bound to circulating antigen or (ii) as a consequence of immunosuppression by either HIV or antigenically distinct viruses, including members of the herpesvirus family. Although most results obtained with the E32/E34 ELISA, the commercial ELISA kits, and immunoblotting agreed, the few discordant results were of particular interest. In group C (asymptomatic lymphadenopathy), two samples were positive by the E32/E34 ELISA but yielded what we interpret as

VOL. 25, 1987

IMMUNOREACTIVE PEPTIDES FROM env REGION OF HIV

falsely negative results with ELISA 2. One of these samples was the first specimen from a patient (Table 1), which was also positive by E32/E34 ELISA and immunoblot but negative by ELISA 2. Three subsequent specimens from this patient were positive by all assays. The second specimen from a different patient was also positive by E32/E34 ELISA and immunoblot but negative by ELISA 2. In group F, one sample, a female heterosexual control, was positive by ELISA 2 but negative by both E32/E34 ELISA and immunoblotting, suggesting a false-positive result by ELISA 2. Of particular interest are the three serum specimens that had only p24 bands on immunoblotting. It is clear that many sera (including bona fide normal sera) yielding such bands on immunoblots react nonspecifically to HIV p24 or react to human cellular antigens unrelated to HIV (20); however, it has been reported that antibody to the p24 core protein may be the first to appear during seroconversion (13, 25). One such serum sample, from a patient with asymptomatic lymphadenopathy, was positive by both E32/E34 and immunoblot but negative by ELISAs 1 and 2. Since no extraneous antigen is present in the E32/E34 peptide assay, these results suggest that the licensed ELISAs were falsely negative and that the immunoblot is relatively insensitive for env antigens. Because of artifacts in antigen preparation or antigen exposure on filters, assay for antibody by immunoblot may be relatively less sensitive for env gene products than for gag gene products. Thus, conclusions regarding the biological significance of the relative antibody responses to env and gag gene products that are based upon immunoblots may reflect laboratory artifacts rather than actual immune responses to HIV infection. Two test sera reacted with p24 only in immunoblots but were negative by all ELISAs. The significance of these p24 bands in the absence of other HIV-specific bands and in the absence of corroboration by ELISA remains an unresolved practical problem. The possibility of false-positive scores in immunoblots showing core-only reactivity has been raised previously (4). The use of synthetic peptides as antigens in the E32/E34 ELISA, by virtue of their specificity, should preclude false-positive results of these kinds of cases because of the specificity of the E32/E34 ELISA. The specificity and accuracy of the E32/E34 ELISA reduces the false-positive results observed in commercially available ELISAs. Therefore, this assay, because of its inherent specificity, could serve as a better method for screening blood products than are commercially available ELISAs. The fact that the E32/E34 ELISA has detected all confirmed, antibody-positive sera that have been examined to date demonstrates that a chemically defined, readily synthesized antigen is sufficient to formulate an immunoassay as sensitive as any yet described. In addition, these results indicate that virtually all individuals who possess antibody to HIV react with this single, well-conserved epitope of HIV gp4l. ACKNOWLEDGMENTS We thank D. Elliot Parks of Johnson & Johnson Biotechnology Center for advice and discussions. We are indebted to Rickey Richardson for manuscript preparation and Clifton Trammel, Mark Bergseid, Mark Feinstein, and Martin Campbell for technical assistance. This research was supported in part by Johnson & Johnson, by Public Health Service grants RO1 HL 32471, AI 52578, and HB 67019 from the National Institutes of Health, by the State of California Universitywide Task Force on AIDS, and by the Veterans Administration.

1503

LITERATURE CITED 1. Barin, F., M. F. McLane, J. S. Allan, T. H. Lee, J. E. Groopman, and M. Essex. 1985. Virus envelope protein of HTLV-III represents major target antigen for antibodies in AIDS patients. Science 228:1094-1096. 2. Barnett, R. M. 1979. Clinical laboratory statistics, 2nd ed, p. 124. Little, Brown, & Co., Boston. 3. Barre-Sinoussi, F., J. C. Chermann, F. Rey, M. T. Nugeyre, S. Charmaret, J. Gruest, C. Dauguet, C. Axler-Blin, F. VezinerBrun, C. Rouzioux, W. Rozenbaum, and L. Montagnier. 1985. Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immune deficiency syndrome (AIDS). Science

220:868-871. 4. Biberfield, G., U. Bradberg-Raden, B. Bottiger, P. O. Putkonen, J. Blomberg, and G. Wadell. 1986. Blood donor sera with false-positive Western blot reactions to human immunodeficiency virus. Lancet ii:289-290. 5. Cabradilla, C. D., J. E. Groopman, J. Lanigan, M. Renz, L. A. Lasky, and D. J. Capon 1986. Serodiagnosis of antibodies to the human AIDS retrovirus with a bacterially synthesized env polypeptide. Bio/Technology 4:128-133. 6. Chang, T. W., I. Keto, S. McKinney, P. Chanda, A. D. Barone, F. Wong-Staal, R. C. Gallo, and N. T. Chang. 1985. Detection of antibodies to human T-cell lymphotropic virus-TII (HTLV-III) with an immunoassay employing a recombinant Escherichia coli-derived viral antigenic peptide. Bio/Technology 3:905-909. 7. Franchini, G., M. Guroff-Robert, and F. Wong-Staal. 1986. Expression of the protein encoded by the 3' open reading frame of human T-cell lymphotropic virus type III in bacteria: demonstration of its immunoreactivity with human sera. Proc. Natl. Acad. Sci. USA 83:5282-5285. 8. Goldsmith, M. F. 1985. HTLV-III testing of donors imminent, complex issues remain. J. Am. Med. Assoc. 253:173-181. 9. Groopman, J. E., F. W. Chen, J. A. Hope, J. M. Andrews, R. L. Swift, C. V. Benton, J. L. Sullivan, P. A. Volberding, D. P. Sites, S. Landesman, J. Gold, L. Baker, D. Craven, and F. S. Boches. 1986. Serological characterization of HTLV-III infection in AIDS and related disorders. J. Infect. Dis. 153:736-742. 10. Hunter, J., and J. Menitore. 1986. HLA antibodies detected by ELISA HTLV-III antibody kits. Lancet ii:397. 11. Kennedy, R. C., R. D. Henkel, D. Pauletti, J. S. Allan, T. H. Lee, M. Essex, and G. R. Dressman. 1986. Antiserum to a synthetic peptide recognizes the HTLV-III envelope glycoprotein. Science 231:1556-1559. 12. Landesman, S. H., H. M. Ginzburg, and S. H. Weiss. 1985. The AIDS epidemic. N. Engl. J. Med. 312:521-525. 13. Lange, J. M. H., H. Bergvanden, L. J. Dooren, J. M. J. J. Vossen, W. Kuis, and J. Goudsmit. 1986. HTLV-III/LAV infection in nine children infected by a single plasma donor: clinical outcome and recognition patterns of viral proteins. J. Infect. Dis. 154:171-174. 14. Lerner, R. A. 1984. Antibodies of predetermined specificity in biology and medicine. Adv. Immunol. 36:1-44. 15. Merrifield, R. B. 1963. Solid phase peptide I. The synthesis of a tetrapeptide. J. Am. Chem. Soc. 85:2149-2154. 16. Muesing, M. A., D. H. Smith, C. D. Cabradilla, C. V. Benton, L. A. Lasky, and D. J. Capon 1985. Nucleic acid structure and expression of human AIDS lymphadenopathy retrovirus. Nature (London) 313:450-458. 17. Osterholm, M. T., R. J. Bowman, M. W. Chopek, J. J. McCuliough, J. H. Korlath, and H. F. Polesky. 1985. Screening donated blood and plasma for HTLV-III antibody. N. Engl. J. Med. 312:1185-1188. 18. Popovic, M., M. G. Sarngadharan, E. Read, and R. C. Gallo. 1984. Detection, isolation and continuous production of cytopathic retrovirus (HTLV-III) from patients with AIDS and pre-AIDS. Science 224:497-500. 19. Ratner, L., W. Haseltine, R. Patarcia, K. J. Livak, R. Starcich, S. F. Josephs, E. R. Doran, J. A. Rafalski, E. A. Whitehorn, K. Baumeister, L. Ivanoff, S. R. Petteway, Jr., M. L. Pearson, J. A. Lautenberger, T. S. Rapas, J. Ghrayeb, N. T. Chang, R. C. Gallo, and F. Wong-Staal. 1985. Complete nucleotide sequence of the AIDS virus HTLV-lII. Nature (London) 313:277-284.

1504

J. CLIN. MICROBIOL.

SMITH ET AL.

20. Saag, M. S., and J. Britz. 1986. Asymptomatic blood donor with a false positive HTLV-III Western blot. N. Engl. J. Med. 314:118-119. 21. Salahuddin, S. Z., P. D. Markham, R. R. Redfield, M. Essex, J. E. Groopman, M. G. Sarngadharan, M. F. McLane, A. Sliski, and R. C. Gallo. 1984. HTLV-III in symptom-free seronegative persons. Lancet ii:1418-1420. 22. Sarngadharan, M. G., M. Popovic, L. Bruch, J. Schupbach, and R. C. Gallo. 1984. Antibodies reactive with human Tlymphotropic retrovirus (HTLV-III) in the serum of patients with AIDS. Science 224:506-508. 23. Sarngadharan, M. G., F. Veronese diMarzo, S. Lee, and R. C. Gallo. 1985. Immunological properties of HTLV-III antigens recognized by sera of patients with AIDS and AIDS-related complex and of asymptomatic carriers of HTLV-III infection. Cancer Res. 45(Suppl.):4574-4577. 24. Saxinger, W. C., and R. C. Gallo. 1983. Application of the indirect enzyme-linked immunosorbent assay microtiter to the detection and surveillance of human T-cell leukemia-lymphoma virus (HTLV). Lab. Invest. 49:371-377. 25. Schupbach, J., O. Haller, and M. Vogt. 1985. Antibodies to HTLV-III in Swiss patients with AIDS and pre-AIDS and in groups at risk for AIDS. N. Engl. J. Med. 312:265-270. 26. Schupbach, J., M. Popovic, R. V. Gilden, M. A. Gonda, M. G.

27.

28.

29.

30.

31.

Sarngadharan, and R. C. Gallo. 1984. Serological analysis of a subgroup of human T-lymphotropic retroviruses (HTLV-III) associated with AIDS. Science 224:503-505. Shaw, G. M., S. Broder, M. Essex, and R. C. Gallo. 1984. Human T-cell leukemia virus; its discovery and role in leukemogenesis and immunosuppression. Adv. Intern. Med. 30:127. Symington, J., M. Green, and K. Brackmann. 1981. Immunoautoradiographic detection of proteins after electrophoretic transfer from gels to diazo-paper. Analysis of adenovirus encoded proteins. Proc. Natl. Acad. Sci. USA 78:177-181. Veronese Di Marzo, F., A. L. DeVico, T. D. Copeland, S. Oroszlan, R. C. Gallo, and M. G. Sarngadharan. 1985. Characterization of gp4l as the transmembrane protein coded by the HTLV-III/LAV envelope gene. Science 229:1402-1405. Wang, J. J. G., S. Steel, R. Wisniewolski, and C. Y. Wang. 1986. Detection of antibodies to human T-lymphotropic virus type III by using a synthetic peptide of 21 amino acid residues corresponding by a highly antigenic segment of gp4l envelope protein. Proc. Natl. Acad. Sci. USA 83:6159-6163. Weiss, S. H., J. J. Goedert, M. G. Sarngadharan, A. J. Bodner, R. C. Gallo, and W. A. Blattner. 1985. Screening test for HTLV-III (AIDS agent) antibodies. J. Am. Med. Assoc. 253: 221-225.