799

Naturally Occurring Soluble CD4 in Patients with Human Immunodeficiency Virus Infection Mark Peakman, Giorgio Senaldi, Nigel Foote, Thomas J. McManus, and Diego Vergani

Departments of Immunology and Genito-Urinary Medicine. King's College School of Medicine and Dentistry. United Kingdom; WHO Immunology Research and Training Centre. Department of Pathology. University ofGeneva. Switzerland

Depletion ofhelper T lymphocytes bearing the CD4 accessory molecule is the hallmark ofinfection with human immunodeficiency virus (HIV) [ I], and the measurement of circulating levels ofCD4lymphocytes is an established laboratory index ofthe severity ofHI V-related disease [2]. In asymptomatic patients, a low level of CD4 lymphocytes is the best parameter for predicting progression to AIDS, and it is also used to guide decisions regarding the initiation and monitoring of treatment in advanced HIV-related disease [3]. The CD4 molecule on the surface ofT lymphocytes is the receptor for HIV [4], and viral entry follows high-affinity binding between the CD4 a I-domain and the HIV envelope glycoprotein 120 (gpI20) [5]. Several lymphocyte surface receptors are secreted into the circulation and their measurement is of proven value in monitoring disease activity in some immune-mediated diseases. In rheumatoid arthritis, autoimmune chronic active hepatitis, and renal transplant rejection, disorders in which interleukin- 2 receptor (IL- 2R)-bearing T cells are believed to provoke immune-mediated damage, serum soluble IL-2R levels correlate closely with clinicolaboratory indices of disease activity [6-8].

Received 26 August 1991; revised 9 December 1991. Presented in part: British Society for Immunology spring meeting. London. April 1991 (abstract 75). All patients gave informed consent. Grant support: M.P. is a Wellcome Trust Research Training Fellow. Reprints or correspondence: Dr. D. Vergani, Department of Immunology. King's College School of Medicine and Dentistry. Bessemer Road. London SE5 9PJ. UK. The Journal of Infectious Diseases 1992;165:799-804 © 1992 by The University of Chicago. All rights reserved. 0022-1899/92/6505-0002$01.00

Recombinant sCD4 has been proposed as therapy in HIV infection [9], but nothing is known about the spontaneous levels of this receptor in the disease. In this study we measured sCD4 and investigated its potential use as a marker of HIV-related disease by comparing it with the conventional markers CD4 lymphocyte count and serum {j2-microglobulin. In addition, we studied the presence of sCD4-gp 120 complexes in patients with HIV infection and the ability of naturally occurring sCD4 to bind to gp 120 in vitro.

Materials and Methods Subjects. We studied 108 patients (98 male; mean age, 33.6 years; range, 22-56) with HIV infection in whom the diagnosis was made on the basis of positive testing for the presence of anti-p24 antibody (Wellcome Diagnostics, Dartford, UK) and p24 antigen (Abbott Diagnostics, Maidenhead, UK). Using the Centers for Disease Control (CDC) classification, 40 (35 male) were asymptomatic, CDC II; 35 (32 male) had persistent generalized lymphadenopathy, CDC III; 7 (6 male) had AIDS-related complex (ARC), CDC IV-A; and 26 (25 male) had AIDS, CDC IVB-D. At the time of the study, 51 patients were receiving zidovudine (250 mg twice daily) and 5 were receiving treatment as part of the Concorde trial. Twenty-five healthy laboratory staff (22 male; mean age, 32.5 years; range, 18-46) without known risk factors for HIV infection were recruited as normal control subjects. Blood (5 ml) was taken into 10 mM EDTA and immediately processed for lymphocyte subset analysis; 5 ml was allowed to clot at room temperature for 2 h and was centrifuged at 1000 g at 4°C for 15 min and the serum separated and stored at -70°C. Soluble CD4 (sCD4) measurement. Levels of sCD4 were measured in serum using a commercially available sandwich ELISA (T Cell Sciences, Cambridge, MA). In the capture phase,

Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 30, 2015

To investigate its use as a marker of disease severity, serum soluble CD4 (sCD4) was measured by ELISA in patients with human immunodeficiency virus (HIV) infection. Levels of sCD4 were higher in patients than in controls (P < .001) but did not increase with disease severity. sCD4 release per CD4 lymphocyte showed a linear increase with disease severity and performed as well as p2-microglobulin, a widely used marker. To study the role of sCD4 in the pathogenesis of HI V infection, an ELISA to detect sCD4 complexed with glycoprotein 120 (gpI20) HIV envelope protein was developed. Preformed sCD4-gp120 complexes were not detectable in patient serum, but addition of recombinant gp120 showed that circulating sCD4 is capable of binding HIV envelope proteins. This study indicates that the sCD4-to-CD4 lymphocyte ratio increases linearly with disease severity and may be a useful marker of CD4 lymphocyte damage. In addition, serum sCD4 can bind viral particles, which may have implications for the use of recombinant sCD4 as a therapy in HIV infection.

800

Peakman et al.

To investigate the ability of sCD4 present in test sera to bind to HIV envelope proteins, a similar ELISA was used with the single modification that undiluted test serum was preincubated with a saturating quantity of recombinant gp 120 (ABT; final concentration, l ug/ml) for 2 h at 37°C. Subsequent steps in the ELISA were identical to those described above. A standard curve was derived using a fixed concentration of recombinant gp 120 (10 lLg/ml)and reducing concentrations of sCD4 from I to 100 pg/ ml. The concentration of complexes detected was read from the standard curve and expressed as the concentration of sCD4 (in picograms per milliliter) associated with the complex. CD4 lymphocyte count. Using Simultest and Leucogate monoclonal antibodies, SimulSET software, and a FACScan flow cytofluorimeter (Becton-Dickinson, Mountain View, CA), the absolute number of CD4 lymphocytes was counted using a lysed whole blood technique according to the manufacturer's instructions, with results expressed in cells per microliter, using lymphocyte counts derived from an automated cell counter (Coulter Stacker; Coulter Electronics, Hialeah, FL). To express sCD4 release per CD4 lymphocyte, the ratio of sCD4 to CD4 lymphocytes (sCD4:CD4 in picograms per 1000 cells) was derived. {jrmicroglobulin. {j2-microglobulin was measured in serum by RIA using a commercially available kit (Pharmacia, Milton Keynes, UK), and results were expressed in milligrams per liter. Statistical analysis. All variables had the normality of their distribution tested in all subject groups by the KolmogorovSmimov goodness of fit test. Only the percentage of CD4 lymphocytes had a distribution consistent with the hypothesis of normality. To lessen the problem of unequal variances in the other variables, therefore, logarithmic transformations of sCD4, CD4 lymphocytes, sCD4:CD4, and {j2-microglobulin were made, and all had a distribution consistent with the hypothesis of normality. The mean values of these parameters and the percentage of CD4 lymphocytes in the four CDC stage groups (II, III, IV A, IVB-D) and in controls were compared using an analysis of variance (ANOVA) method with tests for linearity and multiple Student's t tests. Correlations among log sCD4, log CD4 lymphocytes, log sCD4:CD4, log {j2-microglobulin, and percentage ofCD41ymphocytes were evaluated using Pearson's linear regression analysis. Untransformed data are also given for sCD4, CD4 lymphocytes, sCD4:CD4, and {j2-microglobulin. These were compared using the Kruskal-Wallis one-way ANOVA and multiple MannWhitney U tests. Statistical computations were done using the Statistical Package for the Social Sciences of the University of London Computer Centre.

Results Levels of sCD4, CD4 lymphocyte counts, sCD4:CD4, and levels of{j2-microglobulin. Levels ofsCD4 were significantly higher in all HIV -infected patients than in controls (figures I and 2, tables I and 2), but there were no significant differences in sCD4 between successive CDC groups. Levels in controls were similar to those previously described [10]. Absolute numbers ofCD4 lymphocytes were significantly lower in all HIV -infected patients than in controls and were

Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on May 30, 2015

100 ILl of monoclonal antibody to CD4 was coated onto 96-well microtiter plates for 48 h at 4°C. After the coating solution was discarded, nonspecific binding sites were blocked with 100 ILl of PBS supplemented with 0.2% Tween 20 and I% bovine serum albumin (BSA) and incubated for 2 h at 37°C. After wells were washed four times with PBS supplemented with 0.2%Tween 20, 50 ILl of sample diluent (PBS supplemented with BSA) was added to all wells, followed by 50 ILl of standard or test serum (undiluted or diluted 1:2). After 2 h of incubation at 37°C, plates were washed four times and bound sCD4 was revealed using horseradish peroxidase-conjugated monoclonal antibody to CD4 interacting with an epitope distinct to that recognized by the coating antibody. After a further 2 h at 37°C and four washes, the enzymatic reaction was developed at room temperature using o-phenylenediamine (0.4 mg/ml in citrate-phosphate buffer, pH 5.0) supplemented with 4 ILI/ml 3% hydrogen peroxide. After termination of the reaction with 50 ILl of 4 N H 2S04 , optical density was read in a multichannel plate reader (Dynatech, Billingshurst, UK) at 490 nm. Levels of sCD4 in units per milliliter were calculated from the recombinant sCD4 standards provided and expressed in picograms per milliliter (I unit/ml is equivalent to 10 pg/ml). Detection of preformed sCD4-HIV envelope protein complexes. Preformed complexes of viral envelope proteins and sCD4 in the serum of patients and controls were detected using a sandwich ELISA. Wells ofmicrotiter plates (Immulon II; Dynatech) were coated overnight at room temperature with purified IgG fraction of human antiserum to the gp 120 and gp 160 envelope proteins of HIV (donated by U. Beckford, Wellcome Research Laboratories, Beckenham, UK) diluted to a final concentration of 2 lLg/ml in carbonate/bicarbonate buffer (pH 9.0). Wells were washed three times in PBS with 0.05% Tween 20 (used in all subsequent washing steps). Nonspecific binding sites were blocked by incubation with PBS supplemented with 0.5% BSA (Sigma Chemical, Poole, UK) for 2 h at 37°C (the same incubation conditions were used in subsequent steps). Wells were then washed and incubated with undiluted test serum. After washing, envelope proteinsCD4 complexes were detected by the addition of monoclonal antibody to CD4 conjugated with fluorescein isothiocyanate (FITC-OKT4; Ortho Diagnostics, High Wycombe, UK) diluted I:200 in PBS. After being washed, wells were incubated with horseradish peroxidase-conjugated rabbit anti-fluorescein antiserum (Dakopatts, Glostrup, Denmark) diluted 1:2000 in PBS and the reaction developed as above. A standard preparation comprising preformed complexes of gp 120 and recombinant sCD4 was obtained by incubation of recombinant gp 120 with recombinant sCD4 (both at 10 lLg/ml; American Bio-Technologies [ABT], Cambridge, MA) for 2 hat 37°C. A checkerboard titration using two sources of recombinant gpl20 (5-20 lLg/ml; ABT and Celltech, Slough, UK) and one of recombinant sCD4 (5-20 lLg/ml; ABT) was used in preliminary experiments to establish the optimum reagents and conditions for the assay; those giving the highest optical density were chosen in preference. A reference curve was obtained by dilution of the standard preparation in PBS to between 1:2 and 1:104 • Assessment of binding of serum sCD4 to recombinant gp120.

lID 1992; 165 (May)

Soluble CD4 in HIV Infection

lID 1992; 165 (May)

A

palmi 4

• •• • ••

• J

1000

~

~

.. • •

••

G .... .0 ....0='

2

•

100

III

....0

-+-

••

-c: ••••••• ••• "iT • ••

2

10

II I) ~

~

0

0

~ Q.

• .!.

1.0

...a

·••••.• . .. .... . .. . .·..•

~

• ••• •••••

+

~

~

u