Biomnjugate Chem. 1992, 3, 187-193
187
Microparticle-Enhanced Nephelometric Immunoassay with Microsphere-Antigen Conjugates Paul Montagne,' Pierre Varcin, Marie Louise Cuillihre, and Jean Duheille Immunology Laboratory, Faculty of Medicine, BP 184,F-54505 Vandoeuvre lea Nancy, France. Received September 17,1991 y-Irradiation of acrolein and other acrylic monomers allowed the synthesis of spherical polyfunctional hydrophilic microparticles in the size range of 50 to 300 nm, on which antigens (immunoglobulins G, chorionic gonadotropin hormone, prealbumin) could be covalently bound. Microsphere-antigen conjugates clustered together in the presence of specific antiserum or monoclonal antibodies and their agglutination was quantified by light-scattering measurement performed with a specially designed nephelometer. Essential factors concerning the conjugate agglutination and ita quantitation (size of microsphere, amount of antigen bound on microsphere, concentration of conjugate, concentration of agglutinating reagent, angle of light-scattering observation) were successivelystudied. A microparticleenhanced nephelometric immunoassay for prealbumin was finally developed as an example of application. It was based on the inhibition of the immunoagglutination of microspheres-prealbumin conjugate by free prealbumin. This prealbumin immunoassay was easy to perform (one-step assay without washing or phase separation), fast (30min), reliable (variation coefficients ranged from 3.6% to 7.5% for withinand between-assay determination), and sensitive (1 pg/L detected). It was correlated with conventional immunonephelometry and radial immunodiffusion (correlation coefficients, 0.98). Microparticleenhanced nephelometric immunoassay offered many advantages over the last two methods. Its better sensitivity allowed alower reagent consumption and a larger sample dilution (contrary to the conventional immunonephelometry, sample pretreatment and sample blank measurement were unnecessary). Ita inhibition mode induced a total accuracy for sample with high analyte concentration (a risk of underevaluation in antigen excess conditions existed in all method based on a noncompetitive antigenantibody reaction) and provided the possibility to quantify haptens. Fully automated and fast, it was better adapted to large series of measurement and produced results more rapidly than radial immunodiffusion.
INTRODUCTION
For several years, conventional immunonephelometry (CIN)' has been used to quantify various proteins in biological fluids: serum, plasma, urine, and cerebrospinal fluid (Sieber & Gross, 1976;Schliep & Felgenhauer, 1978; Schmitz-Huebner et al., 1980). CIN is based on the nephelometric measurement of antigen-antibody complexes. This method is easy t o perform, but ita detection limit (about 1 mg/L) remains poor, compared with other immunoassays such as radioimmunology or immunoenzymology. Several authors have attempted to improve the sensibility of the CIN by using agglutination of latexes coated with antigens or antibodies (Grange et al., 1977; Von Schulthess et al., 1976,1980;Ripoll et al., 1980).Cell counting (Cambiaso et al., 1977;Cambiaso & Limet, 1989), turbidimetry (Kimura, 1980), and laser Doppler spectrometry (Uzgiris, 1976)have also been used to quantify latex agglutination. These latexes, usually made of polystyrene, were hydrophobic and often unstable. Stable hydrophilic immunolatexes have been synthetized as markers for scanning electron microscopy (Molday et al., 1975;Rembaum et al., 1979,Bene et al., 1982) and later on as the solid phase in radioimmunoassay (Pines & Margel, 1986). They were also used for the nephelometric detection of Clq-binding immune complexes (Montagne et al., 1980)and as immunization carrier (Jambon et al., 19811,but they were unsuitable for general use as
* Author to whom correspondence should be addressed. Abbreviations used CIN, conventional immunonephelometry; Ms,microsphere; PA, prealbumin; IgG, immunoglobulins G; hCG, chorionic gonadotropin hormone; PB, phosphate buffer; HSA, human serum albumin; RID, radial immunodiffusion. 1043-1802/92/2903-0187$03.00/0
a reagent in nephelometric immunoassay (extra steps of activation for protein binding, inadequate size, overlylong reaction time). This study describes the synthesis of stable hydrophillic microspheres (Ms) that are physicochemically welldefined and optimally devised for immunological reaction support. It reports the preparation of Ms-antigen conjugates and assesses conditions of their use as reagent in an improved microparticle-enhanced nephelometric immunoassay. The performance of this assay is demonstrated by an example of application to the prealbumin (PA) quantitation in human serum. EXPERIMENTAL PROCEDURES
Chemical Reagents. Acrolein, 2-hydroxyethyl methacrylate, and methacrylic acid were obtained from Merck (Darmstad, Germany);NP-methylenebisacrylamidewas from Eastman Kodak Co. (Rochester, NY); Sodium dodecy1sulfate,hydroquinone, 2-aminoethanol,sucrose,sodium chloride, sodium dihydrogenophosphate, disodium hydrogenophosphate, and sodium azide, of analytical-reagent grade, were purchased from Prolabo-Rhbne Poulenc (Paris, France). Most immunonephelometric assays were carried out with the buffer for nephelometry supplied by Diagnostics Pasteur (Marnes, France). Biochemical Reagents. Human immunoglobulins G (IgG) were isolated from a pool of normal human sera by DEAE-Trisacryl (IBF, Paris, France) chromatography using an elution gradient of increasing ionic strength (Cuilliere et al., 1991). The purity of the IgG preparation (about 98% ) was checked by immunoelectrophoresis and polyacrylamide gel electrophoresis. Human chorionic gonadotropin (hCG) from urine was a Sigma (St. Louis, MO) 0 1992 American Chemical Society
188 BbconJtigate Chem., Vol. 3, No. 2, 1992
product (3326 IU/mg accordingto the second international standard). Eight anti-hCG monoclonal antibodies (mAb) were kindly supplied by Sanofi (Paris, France): six of them were hCG-8 specific with affinity about lo8M-l, the other two reacted against the whole hCG molecule (a+ 0) with higher affinity (109and 10'0 M-1). Purified PA (95% electrophoretically pure), anti-PA goat antiserum, PA standard (0.35 f 0.03 g/L), and anti-IgG ( 7 ) goat antiserum were Diagnostics Pasteur products. Behring (Marburg, Germany) protein standards were used as control sera in the PA immunoassay; their PA concentrations, determined by radial immunodiffusion and compared with international WHO standard preparations, were 0.12 f 0.02 g/L and 0.40 f 0.05 g/L. Human sera were randomly chosen from the patients of the University Hospital of Nancy (France). Preparationand Characterizationof Ma. Acrolein, 2-hydroxyethyl methacrylate, and methacrylic acid were freshly distilled under Ar and mixed with N,N'-methylenebisacrylamide in respective percentages: 47 % ,49.7 % , 2 % , and 1.3 % (v/v) of total monomers. Various concentrations of total monomers (50,80,100, and 120 g/L) were used. Sodium dodecyl sulfate was added as a surfactant at concentrations of 0.6, 0.8, 0.9, and 1 g/L. Deaerated and Ar-saturated monomer mixtures were y-irradiated using a 'Wo source (ORIS, Nucleart, CENG, Grenoble, France) for 3 h under vacuum at a flux of 23 krad.cm-2.h-1. After irradiation, the polymerized Ms were stored in aminefree suspension, at 4 "C under Ar, with hydroquinone (1 g/L) (Duheille et al., 1982). Ms suspension concentration was calculated by dry weight determination at 110 "C. Ms size measurement was carried out by transmission electron microscopy (JEOL 200 CX; Tokyo, Japan) after drying on a formvar filmed grid (200 mesh) and with a granulometer (Autosizer 11; Malvern Instruments, Worcs, England) in aqueous suspension. Shape and dispersion were evaluated by scanning electron microscopy (Cambridge S250) after gold shadowing. Binding of Antigens to MS.Antigen (IgG, PA, and hCG) solution and Ms suspension were mixed in 0.1 M phosphate buffer (PB) (pH 7.2) and 0.3 M NaCl for IgG, 0.01 M PB (pH 7.2) and 0.14 M NaCl for PA, or 0.005 M PB (pH 7.2) for hCG. The final concentrations in the binding mixtures were 10 g/L for Ms and 1.3 g/L for IgG, 1.6 g/L for PA, and 0.2,0.4,0.8,1.6, and 3.2 g/L for hCG. After gentle stirring for 18 h at 4 "C, unreacted aldehyde groups of the Ms were blocked by 2-aminoethanol, 0.12 M in pH 8 buffered solution, for 4 h at room temperature. The uncoupled proteins were then eliminated by centrifugation on a discontinuous sucrose gradient (200/800 g/ L) buffered at pH 7.2 (Ultracentrifuge Spinco L, Rotor Beckmann SW50,3000-12000g according to the Ms size, 1 h, 4 "C). The Ms-antigen conjugates were finally collected at the interface of sucrose solutions and stored at 4 "C in 0.1 M PB (pH 7.2) containing 2 g/L of sodium azide. The yields of antigen binding were calculated by measuring the uncoupled protein by CIN for IgG and PA and by determining the amount of labeled hCG in MshCG conjugates, using a lZ5I-hCGprobe prepared by the chloramine-T method (Greenwood & Hunter, 1963). Amounts of immunoreactive hCG on the Ms-hCG conjugates were evaluated with anti-hCG mAb by a radioimmunoassay using 1251-hCG as competitor and free hCG as calibrator. Nephelometric and Immunophelometric Assays. The influence of several factors (Ms size, amount of antigen
Montagne et al.
bound on Ms, stability of Ms-antigen conjugate, concentration of Ms-antigen conjugate in reaction mixture, concentration of agglutinating reagent, angle of observation of the light scattering) were defined by nephelometric assays of native Ms and immunonephelometric assays of Ms-antigen conjugates. These studies were performed by measurement of the light scattered by dispersed native Ms, Ms-antigen conjugates, and clusters of Ms-antigen conjugate, performed with the Diagnostics Pasteur nephelometer (Nephelia N600), perfected during this work. Its light source was an He-Nelaser (power,2 mW; wavelength, 632.8 nm). The reagents were dispensed in disposable microcuvettes (light path, 1 cm) and scattered light was collimated at an angle of 10" on a light-sensitive silicium diode. To measure the light scattered at various angles, an experimental prototype (laboratory made) photogoniometer was used. Its physical characteristics are the same as those of the Diagnostics Pasteur nephelometer, but the light scattering can be measured between 7.4" and 40.2' every 0.8O with a scan period of 4 s. Microparticle-Enhanced Nephelometric Immunoassay of PA. The microparticle-enhanced nephelometric immunoassay of serum P A was a one-step immunoassay: 0.03 mL of 200-fold diluted unknown human serum was mixed with anti-PA antiserum, Ms-PA conjugate (125 nm in diameter), and buffer for nephelometry in a total volume of 0.3 mL, to obtain a sample dilution in the reaction mixture of l/zm with 2100-fold diluted antiserum and 100 mg/L of Ms-PA conjugate. All predilutions and dispensations were performed with an automated dilutor (Hamilton, Bonaduz, Switzerland). Light scattering was measured with the Diagnostics Pasteur nephelometer after 30-min incubation at room temperature. For the standard curves, six serial dilutions of the PA standard were used in the place of unknown serum. The precision of the assay was assessed by measuring high, middle, and low PA concentration in 20 assays repeated within 1day (within-assay precision) and assays repeated on each of 10 days (between-assay precision) with fresh reagents taken each time from the same batches. Other Methods. For interference and comparison studies, rheumatoid factor was determined with a latex slide-agglutination test (Latex-RF reagent, Behring), PA concentrations were measured in human sera by CIN (Behring laser nephelometer and Behring reagents) and radial immunodiffusion (RID) (Behring M-Partigen) following the manufacturer's recommendations. RESULTS
Ms-Antigen Conjugates. Above-described copolymerizations produced spherical, monodispersed, polyfunctional, and hydrophilic microparticles. Effects of polymerization time, irradiation flux, agitation, and relative concentration of monomers and surfactant on yield, size, and chemical characteristics of Ms were previously reported (Duheille et al., 19821, Sizes of the Ms synthesized in this study are given in Table I. Mean dry diameters were between 50 and 295 nm with an acceptable dispersion (2% to 17% ) for each sample. IgG, PA, and hCG were covalently bound to the native Ms in physiological conditions by a one-step reaction with formation of imine bonds between aldehyde functions on the Ms (acrolein residues) and primary amino groups of the ligand. Previous reaction of the Ms with concentrated HSA solution (higher than 12 g/L) inhibited the ligand binding. Ms-antigen conjugates were stable several months at 4 "C with sodium azide as preservative and could be frozen and lyophilized. On account of this good
Sioconjugite Chem., Vol. 3, No. 2, 1992
Micropa~lcle-EnhancedNephelometric Immunoassay
Table I. Influence of Relative Concentrations of Total Monomers (TM) and Sodium Dodecyl Sulfate (SDS) on the Size of Microspheres (Ms) dry diameter of Ms concentrationof reagenta, g/L
TM 50 50
SDS 1.0
80 100 120
0.8 0.6 0.9
0.8
mean,nm 50 100 125 200 295
SD,IInm 5 17 16 6 4
189
1 0 0 0
nb 39 30 43 27 7
SD, standard deviation. b n, number of data. Table 11. Influence of the Ms Size on the Light Scattered at an Angle of 10' by Dispersed Ms and Ms Clusters Ms concentration,mg/L amplificationof range with light scattered linearity between Ms during Ms diameter, concentrationand significantly detectedb agglutinationc light scatteredn nm 20 32 300-4800 50 3 21 40-1700 100 2 17 10-300 125 0.2 12 6-230 200 0.01 4 2.5-150 295 P < 0.001. b Expressed as the Ms concentrationgiving an intensity of light scattering 3 SD higher than the intensity of light scattered by the empty nephelometermicrocuvettes. c Expressed as the ratio of the maximum light scattered by Ms clusters (pH 4) to the light scattered by the same dispersed Ms (pH 7).
stability, imine bonds of the Ms-antigen conjugates were not reduced by treatment with a metallic hydride. Binding yield varied with binding conditions, particularly with relative concentrations of antigen and Ms in the binding mixture. Binding yield of hCG ranged from 19% to 94 % when hCG concentration decreased from 3.2 to 0.2 g/L. The best immunoreactivity of bound hCG mol) molecules (76%)was obtained when 1.6 g (4 X of hCG was present in the binding mixture (1L)with 10 g of Ms (200 nm in diameter). Under these conditions, each Ms was coated by about 1000 hCG molecules, but only a fraction of these bound hCG molecules was thus recognized by anti-hCG mAb. The ratio of immunoreactive versus total hCG molecules bound to Ms changed (18% to 76%) according to the mAb used, and there was probably a close connection between the large disparity of the immunoreactivity and a poor accessibility of some epitopes on the hCG molecules bound on the Ms surface. Nephelometric Assays of Native and Aggregated Ms. According to the physical laws (Rayleigh, 1899; Mie, 1908; Debye, 1944), the intensity of light scattered by a suspension of Ms depended on the Ms size and the concentration of the suspension. Light scattering increased with increasing Ms concentration, but reached a maximum and decreased by absorption for the highest concentrations (above 300 mg/L) of the largest Ms (295 nm in diameter). As can be seen in Table 11,the intensity of light scattered by dispersed Ms and the concentration of the suspension were in linear relation (P< 0.001) for large ranges of concentration, variable with the Ms size. When the Ms diameter increased from 50 to 295 nm, the Ms concentration significantly detected by the nephelometer, as giving an intensity of light scattering 3 SD higher than the intensity of light scattered by the empty nephelometer microcuvettes, decreased from 20 to 0.01 mg/L (Table 11). In solutions of neutral pH and physiological ionic strength, Ms were stabilized by their hydrophilic and charged surface (hydroxyl and carboxyl groups of 2-hy-
6
.'
* * * * * * * *
*
V
V
A
A
4 ' 2 .
'd
' 1 0 '20 '30 '40 '50 '60 '70 '80 R e a c t i o n Time
(inin)
Figure 1. Agglutination kinetics of Ms-IgG conjugate (100 nm in diameter, 200 mg/L) with anti-IgGantiserum (120-folddiluted) for various angles of measure of the light-scattering intensity: 9.8O (O), 12.2O (n),20.2: (*), 31.4' (v),and 40.2' (A). Lightscattering amplification is the ratio of the light scattered by MsIgG conjugate clusters to the light scattered by dispersed MsIgG conjugates (0.01M PB, pH 7.2,0.14 M NaCI; experimental photogoniometer).
droxyethyl methacrylate and methacrylic acid residues). They were not spontaneously autoagglutinated and their sedimentation was slow: intensity of light scattered by the largest Ms (295 nm) slightly decreased (1%to 6 % according to their concentration) when Ms remained at rest for 24 h. But the intensity of light scattered increased (Table 11)when Ms were aggregated by increasing ionic strength or decreasing pH. Amplification of the light scattering during the Ms agglutination was a maximum for the smallest Ms (32-fold for 50 nm diameter Ms) and decreased when their size increased (4-fold for the 295 nm diameter Ms only). Immunonephelometric Assays of Ms-Antigen Conjugates. Figure 1shows the importance of the angle of light-scattering observation. At the beginning of the immunological agglutination of Ms-IgG conjugate (100nm diameter) by anti-IgG antiserum, the intensity of light scattered by small clusters of Ms-IgG conjugate was slight and not very different, whatever the angle of measurement between 9.8O and 40.2' (Rayleigh's diffusion). Then, with the formation of big aggregatesof Ms-IgG conjugate (Mie's diffusion), light-scattering amplification increased and was higher (up to 17-fold the light scattered by the suspended conjugate alone) at small angles (9.8O and 1 2 . 2 O ) than at the largest angles (between 2- and 5-foldthe light scattered by unagglutinated conjugate only). In addition to the detection of a limited amplification of the light, the measure at a large angle was more disturbed by the light scattered by small scattering centers (Rayleigh's diffusion) present in the reaction mixture (components of antiserum, sample, buffer). Level of light-scattering amplification during agglutination of Ms-antigen conjugate and kinetics of this agglutination were functions of the Ms-antigen conjugate concentration and antiserum dilution. The curves of Figure 2 were obtained by variation of anti-IgG antibody concentration while Ms-IgG conjugate concentration was kept constant. Increasing dilution of antiserum reduced the light scattering amplification and lengthening of the reaction time was required to obtain the same lightscattering levels. However, Figure 2 shows that under
Montagne et el.
190 BloconjoBete Chem., Vol. 3, No. 2, 1992
97
400
.
350
'
c
0
8 '
Ll
0
u "
7
>
5 300
L1-
2 6 a
07
m c
'Z
5
01
0"
4
c,
cn
0
cl
c , 3
u
c,
m
200
'
150
."
2
-I
L
.?
*'
c, c,
0)
m
A
II
'z 250
E
