measurement of eleven serum proteins by microparticle

Montagne et al.: Measurement of eleven serum proteins 217

Eur. J. Clin. Chem. Clin. Biochem.Vol. 30, 1992, pp. 217-222© 1992 Walter de Gruyter & Co.

Berlin · New York

Measurement of Eleven Serum Proteins byMicroparticle-Enhanced Nephelometric Immunoassay

By P. Montagne, P. Laroche, Therese Bessou, Marie Louise Cuilliere, P. Varcin and J. Duheille

Laboratoire d'Immunologie, Faculte de Medecine, Nancy, France

(Received September 10, 1991/January 21, 1992)

Summary: Eleven proteins (immunoglobulins IgG, IgA, IgM, orosomucoid, o^-antiproteinase, haptoglobin,ceruloplasmin, C-reactive protein, transferrin, prealbumin and a2-macroglobulin) in human serum werequantitated by a new microparticle-enhanced nephelometric immunoassay. This is a one step competitiveassay, based on the nephelometric measurement of light scattered by clusters of protein-coated microparticlesspecially synthesized for that use. Statistical evaluation (precision, recovery and method comparison) showsthat the determination of serum proteins is reliable and accurate for wide ranges of concentration and thatthe method is quite adequate for strongly increased concentrations. This microparticle-enhanced nephelometricimmunoassay appears to offer an alternative method for routine measurement of a great variety of serumproteins at high and intermediate concentrations, which are usually quantified by radial immunodiffusion orconventional immunonephelometry. On account of its sensitivity, it can also be used for the determination ofrelatively low concentrations of analytes.

Introduction

The synthesis of special microparticles which serve asa solid-phase for antigen-antibody reactions has beenreported (1). These microparticles are monodisperse,hydrophilic and polyfunctional microspheres to whichproteins can be covalently bound. The light scatteredduring agglutination of microparticle-protein conju-gates with specific antibodies is measurable with aspecially designed nephelometer. Nephelometricmeasurement of the light scattered during inhibitionof this agglutination by free proteins (fig. 1) allowstheir sensitive quantitation (detection limit lower than1 g/l) (2). These studies have led to the developmentof a new microparticle-enhanced nephelometric im-munoassay (Nephelia®, Diagnostics Pasteur, Marnes,France). The present report describes this competitiveimmunoassay and its application to the quantitationof eleven serum proteins at high and intermediateconcentrations in human serum: immunoglobulinsIgG, IgA, IgM, orosomucoid, cxi-antiproteinase, hap-toglobin, ceruloplasmin, C-reactive protein, transfer-rin, prealbumin, and a2-macroglobulin.

.£

0)

0 % Inhibition AgglutinationHigh light scattering

Calibration range

100% Inhibition ILow light scattering

No agglutination

Free antigen concentration

Fig. 1 .Assay of free antigen by its inhibition of the agglutin-ation of microparticle-antigen conjugate in the presenceof specific antibody: change of the light scattering in-tensity during this inhibition as a function of the freeantigen concentration.

Microparticle-antigen conjugate

Free antigen to be assayedSpecific antibody

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 30,1992 / No. 4

218 Montagne et al.: Measurement of eleven serum proteins

Materials and MethodsReagentsAcrolein, 2-hydroxyethyl methacrylate, and methacrylic acid,were from Merck (Darmstadt, Germany). Ν,Ν'-methylene dia-crylamide was from Eastman Kodak (Rochester, NY). Sodiumdodecyl sulphate, hydroquinone, boric acid, sodium chloride,2-aminoethanol, sucrose and sodium azide were from ProlaboRhone Poulenc (Paris, France). Protein determinations werecarried out with the phosphate buffer for nephelometry, pH7.2, provided by Diagnostics Pasteur.

Purified proteins (IgG, IgA, IgM, orosomucoid, C-reactiveprotein, and prealbumin) were kindly supplied by DiagnosticsPasteur or produced (arantiproteinase, haptoglobin, cerulo-plasmin, transferrin and a2-macroglobuHn) by the Centre Na-tional de Transfusion Sanguine (Paris, France). Heavy chainsof IgG (γ) and IgM (μ) were prepared by reduction-alkylationof disulphide bridges and separated by gel filtration (SephadexG-100, Pharmacia, Uppsala, Sweden) (2). The purity of allprotein preparations was checked by polyacrylamide gel elec-trophoresis and immunoelectrophoresis.

Goat or rabbit antisera, human standard and control sera wereDiagnostics Pasteur products. They were standardized accord-ing to the rules of the Societe Franchise de Biologie Cliniqueapproved by the International Federation of Clinical Chemistry.Protein concentrations of standard and control sera were de-termined by radial immunodiffusion.

Human sera used for precision, recovery and correlation studieswere collected from routinely tested patients of the UniversityHospital of Nancy (France). They were stored at 4 °C or frozenat -20 °C.

Microparticle-protein conjugates

Microparticles were synthesized as previously reported (1): amixture of distilled acrolein (1.01 mol/1), 2-hydroxyethyl meth-acrylate (0.46 mol/1), methacrylic acid (0.03 mol/1), N,N'-meth-ylenediacrylamide (0.01 mol/1) and sodium dodecyl sulphate(0.003 mol/1) was gamma-irradiated (*Co, 25 krad/cm2 · h, 3 h).The microparticle suspension thus obtained was stored forseveral years at 4 °C with hydroquinone (0.009 mol/1).

For coating, each protein (0.4 g of IgG heavy chain, 3.3 g IgA,1.2 g of IgM heavy chain, 0.5 g orosomucoid, 1.1 g αι-antipro-teinase, 1.7 g haptoglobin, 2.6 g ceruloplasmin, 3.0 g C-reactiveprotein, 0.6 g transferrin, 1.6 g prealbumin, or 3.6 g oc2-macro-globulin) was mixed with microparticles (10 g) in 0.1 mol/1borate buffer pH 8.2, 0.3 mol/1 NaCl (11) and incubated for18 h at 4°C. 2-Aminoethanol was then added to the bindingmixture (0.12 mol/1, 4h) to block unreacted groups of themicroparticles. Centrifugation (8000 g l h, 4 °C) on a discon-tinuous sucrose gradient (200/800 g/1) served to eliminate un-coupled proteins. These uncoupled proteins were measured byconventional immunonephelometry to determine the bindingyields. Microparticle-protein conjugates were stored at 4 °C in0.1 mol/1 borate buffer pH 8.2 containing 0.3 mol/1 NaCl and0.03 mol/1 NaN3.

Procedure of protein determinations

Microparticle-enhanced nephelometric immunoassay of serumproteins is a one step assay: 50 μΐ of microparticle-proteinconjugate (0.5 to 2.0 g/1), 10 to 50 μΐ of serum standard (5 or6 serial dilutions) or diluted (1/100 to 1/1000) sera (control orunknown), and 50 μΐ of prediluted antiserum (1/100 to 1/2000)were mixed with the buffer for nephelometry to give a finalvolume of 500 μΐ, in disposable microcuvettes (Nephelia cuvette,Diagnostics Pasteur). All dilutions were carried out with anautomated dilutor (Hamilton AG, Bonaduz, Switzerland).After 30 min or 1 h at room temperature, light scattering wasmeasured with a specially designed nephelometer (Nephelia

N600, Diagnostics Pasteur) (2) equipped with an helium-neonlaser (wavelength, 632.8 nm). Spline-smoothed calibrationcurves (fig. 1) and unknown sample concentrations were cal-culated with a microcomputer (MBC 885, Sanyo Electric Co.,Tokyo, Japan) fitted to the nephelometer.

Evaluat ion s tudy

The precision of the assays was assessed by measuring normaland pathological concentrations of each protein in human sera,10, 20 or 30 times during the same assay (within-run precision)and on 5 to 15 different days (between-run precision). Theprecision was expressed as the coefficient of variation (%).

Recovery studies were carried out by adding increasing amountsof each purified protein to several human sera with a low ornormal protein content. One hundred and twenty overloadingswere performed in this way, and the results subjected to linearregression analysis (recovered = b + a added). For each pro-tein, the null hypothesis H0, intercept b = 0 and slope a = 1,versus the alternative hypothesis HI, intercept b φ Ο and slopea 7* 1, were tested by F (Fisher) and t (Student), respectively.

For comparison of methods, proteins were measured in humansera (1084 sera with widely ranging concentrations for all ofthe eleven proteins), using radial immunodiffusion (NOR-Par-tigen, Behring, Marburg, Germany) for IgG, C-reactive protein,and prealbumin, and conventional immunonephelometry(Behring Laser Nephelometry) for IgG, IgA, IgM, orosomu-coid, cxi-antiproteinase, haptoglobin, ceruloplasmin, C-reactiveprotein, transferrin, prealbumin, and a2-macroglobulin. TheArray System Beckman (Fullerton, CA) was used for IgG, IgA,and transferrin and the Technicon RA-1000 System (Tarrytown,NY) for IgG, IgA, C-reactive protein and transferrin. In ad-dition, C-reactive protein was assayed by fluorescence polaris-ation (TDx Abbott, Chicago, IL). These assays were carriedout strictly following the manufacturer's recommendations.Linear regression analysis was performed for each comparisonand a 95% confidence interval was calculated for each corre-lation coefficient which was used as a measure of the closenessof the association.Rheumatoid factors were determined in human sera with alatex-slide agglutination test (latex-RF reagent, Behring).

Results

Microparticle-protein conjugates

Copolymerization of acrylic monomers produced mi-crospheres of 235 nm (standard deviation = 3 nm,n = 36) diameter with a mean yield to 63%. Proteinswere covalently bound to the microparticles by for-mation of imine bonds between aldehyde groups ofthe microsphere and primary amino groups of theproteins (binding yield from 23 to 100%). Micropar-ticle-protein conjugate suspensions remained immu-noreactive for six months at least when they werestored at 4 °C with NaN3 (0.03 mol/1).

Calibration ranges

Decreasing sigmoidal inhibition curves (fig. 1) wereobserved for all serum protein determinations. Cali-bration ranges are given table 1. The limits of eachcalibration range corresponded to protein concentra-tions giving light scattering intensity significantly dif-



ferent (3 SD upper and 3 SD lower at least) from thelight scattered by dispersed microparticle-antigen con-jugate alone (0% agglutination) and by aggregatedmicroparticle-antigen conjugate in the presence ofantiserum (100% agglutination). Calibration rangesallowed the measurement of normal as well as higherand lower pathological concentrations. Under-esti-mation of the highest pathological concentrations wasavoided by the used inhibition mode.

Tab. 1. Calibration ranges for eleven serum proteins

Protein Calibration range (g/1)

IgGIgAIgMOrosomucoidαϊ - AntiproteinaseHaptoglobinCeruloplasminC-reactive proteinTransferrinPrealbuminoc2-Macroglobulin

Tab. 2. Within-run

Protein

IgG

IgA

IgM

Orosomucoid

αϊ -Antiproteinase

Haptoglobin

Ceruloplasmin

C-reactive protein

Transferrin

Prealbumin

a2-Macroglobulin

1.350.250.040.23

-43.70- 8.00-10.46- 3.65

0.44 - 7.100.45 - 7.250.065- 2.0800.003- 0.2000.30 - 9.690.038- 1.200

precision

Num-ber ofdata

302030

302020

1010

1010

1010

1010

1010

20203010

302020

202020

1010

0.26

Mean

(g/1)

1.588.55

18.40

0.272.785.72

1.743.58

0.951.95

2.306.94

1.504.35

0.421.24

0.01520.05150.10870.1520

0.664.197.79

0.05530.11370.2231

2.698.39

- 8.40

Standarddeviation

(g/1)

0.100.401.09

0.030.080.24

0.130.31

0.020.09

0.040.19

0.050.13

0.020.06

0.00140.00120.00320.0103

0.040.120.20

0.00370.00500.0157

0.050.06

Coeffi-cient ofvariation(%)

6.34.75.9

9.52.84.2

7.58.7

2.14.8

1.72.7

3.33.0

5.05.2

9.32.42.96.8

5.62.82.6

6.64.47.0

1.90.7

Evaluation study

Tables 2 and 3 present the within-run and between-run precision results, respectively. The coefficients ofvariation calculated for large distributions of analyteconcentration ranged between 0.7% and 9.5% for thewithin-run and between 0.7% and 9.6% for the be-tween-run study.

All the analytical recovery results are reported in table4. Percentage recovery was very close to 100% for alleleven proteins. Statistical studies of the parametersof the linear regression (recovered = b + a added)performed for each protein over a large concentrationrange showed that the intercepts (b) were never sig-nificantly different from 0 (P < 0.05), and the slopes(a) were not significantly different from 1 (P < 0.05),except for IgG (1.03) and C-reactive protein (1.05).

Linear regression parameters computed for each ofthe 22 comparisons between methods are given intable 5. Correlation coefficients ranged from 0.92 to0.99, except for 2 comparisons (IgG and transferrin)

Tab. 3. Between-run precision

Protein Num-ber ofdata

Mean

(g/1)

Standarddeviation

(g/D

Coeffi-cient ofvariation

IgG

IgA

101010

10

4.419.08

17.34

0.65

0.230.451.13

0.05

5.14.96.5

7.2

IgM

Orosomucoid

ar Antiproteinase

Haptoglobin

Ceruloplasmin

C-reactive protein

Transfemn

Prealbumin

a2-Macroglobulin

1210

55

55

55

55

55

10101010

101015

101010

55

5.3410.99

1.653.43

0.951.85

2.276.81

1.554.47

0.421.24

0.01230.05410.10950.1400

2.214.518.62

0.05960.11630.4754

2.668.42

0.160.84

0.140.33

0.020.11

0.110.29

0.120.20

0.020.07

0.00090.00170.00370.0130

0.130.160.45

0.00450.00420.0219

0.040.06

3.07.6

8.59.6

2.35.9

4.84.3

7.54.5

5.35.3

7.63.13.49.3

6.03.55.3

7.53.64.6

1.50.7



Tab. 4. Analytical recovery

Protein

IgGIgAIgMOrosomucoidcti -AntiproteinaseHaptoglobinCeruloplasminC-reactive proteinTransferrina2-Macroglobulin

Rangeinvestigated(g/D

4.90.931.120.872.551.650.95

-41.0- 5.80- 3.60- 2.48- 5.22- 6.60- 1.90

0.025- 0.1251.282.12

- 7.34- 4.80

Number οdata

18886

108

1023209

- RecoveryΓ (°/ ^

Mean

100.6101.295.4

101.8101.4101.995.4

104.199.2

102.0

SD

4.94.55.03.62.25.34.77.53.42.4

Linear regression Significance Sign(recovered = b + a added) F test for t tes

b (g/1)

-0.290.09

-0.100.060.110.020.08

-0.001-0.14-0.12

a

1.031.020.980.990.931.020.971.051.051.04

r

0.9980.9720.9680.9830.9810.9930.9830.9970.9920.992

0.75 NS0.32 NS0.38 NS0.88 NS0.66 NS0.07 NS0.92 NS0.71 NS1.01 NS0.81 NS

1.830.380.260.341.230.400.542.731.570.71

ificancet for1 vs a φ \

NSNSNSNSNSNS

NSNS

NS: not significant at P < 0.05

Tab. 5. Comparison with other methods

Comparedmethod

Radial immuno-diffusion

Behringlasernephelometry

BeckmanArray System

TechniconRA 1000

Abbott TDx

Protein

IgGC-reactive proteinPrealbumin

IgGIgAIgMOrosomucoidoti -AntiproteinaseHaptoglobinCeruloplasminC-reactive proteinTransferrinPrealbumina2-Macroglobulin

IgGIgATransferrin

IgGIgAC-reactive proteinTransferrin

C-reactive protein

Range investigated Numbx(compared method) of(g/1) sample

3.6 -37.00.010- 0.1210.10 - 0.44

3.3 -38.21.16 -10.200.65 - 9.940.66 - 2.901.76 - 4.980.23 - 7.500.24 - 1.690.010- 0.1331.77 - 9.820.08 - 0.401.28 - 9.50

4.3 -32.11.09 - 9.161.43 - 5.12

3.9 -32.80.92 - 9.500.010- 0.1881.10 - 4.74

0.003- 0.090

531860

8145404544457045446037

525660

70742834

23

sr Linear regression 95% confidence(Y = b + a compared method) interval for r

!Sb (g/1)

0.89-0.003

0.01

-0.680.02

-0.340.280.500.0020.210.0060.040.030.14

0.200.340.59

-0.66-0.05

0.0130.21

0.003

a

1.041.1760.87

0.981.030.800.840.840.9641.160.9640.910.950.86

1.000.801.18

0.950.960.8021.03

1.108

r

0.960.990.98

0.960.950.940.970.970.990.930.980.970.980.98

0.860.980.86

0.920.970.970.94

0.97

0.93-0.970.97-0.990.96-0.99

0.94-0.980.91-0.970.89-0.960.95-0.980.94-0.980.98-0.990.89-0.960.96-0.990.94-0.980.96-0.990.96-0.990.77-0.920.96-0.990.77-0.910.87-0.950.95-0.980.93-0.990.88-0.96

0.93-0.99

with the Array System Beckman (0.86). For somesera, Ceruloplasmin, prealbumin and particularly C-reactive protein concentrations were below the limitof detection of the reference methods. These wererejected from our study, although they were measur-able by microparticle-enhanced nephelometric im-munoassay. No effect (such as a under- or over-evaluation of results) of rheumatoid factors, whichwere found in some sera, was detected in micropar-ticle-enhanced nephelometric immunoassay or in thecomparison methods.

Discussion

Radial immunodiffusion (3) and conventional im-munonephelometry (4) are commonly used to meas-ure high and intermediate concentrations of serumproteins, but these methods have certain drawbacks.A prolonged diffusion time (from 24 to 48 h) is nec-essary to obtain reliable results by radial immuno-diffusion, and it is ill-adapted for large numbers ofsamples. Although it is an easy-to-perform method,conventional immunonephelometry has poor sensitiv-



ity (1 mg/1), requiring a blank measurement of dilutedsamples and clarification of turbid samples by pre-treatment (5). Furthermore, in conventional immu-nonephelometry, as in all immunoassays based on anuncompetitive reaction, there is a risk of error whenthe antigen is in excess.

The microparticle-enhanced nephelometric immu-noassay described in this work and used to measureeleven serum proteins, preserves the advantages ofconventional immunonephelometry and avoids someof its drawbacks.

It is a highly sensitive assay (detection limit lowerthan 1 g/l) allowing a great dilution of samples(1/1000 to 1/50000 in the reaction mixture) even forproteins that are present in very low concentrations.Such dilutions removed any interference due to tur-bidity, sample blank measurement, or sample pre-treatment. Although this high sensitivity was not re-quired for the determination of the most serum pro-teins tested in this work, it permitted the assay ofproteins such as ceruloplasmin, prealbumin, and par-ticularly C-reactive protein, whose concentration insome sera used in the comparison study was belowthe limit of detection of radial imrmmodiffusion andconventional immunonephelometry.

Microparticle-enhanced nephelometric immunoassayis a rapid (30 min to 1 h), fully automatable one stepassay, which is suitable for the measurement of agreat number of samples (360 per hour), but also easyto use and therefore suitable for emergencies.

The inhibition mode, chosen for the determination ofserum proteins, enables analyte underevaluation in

antigen excess conditions to be eliminated and pro-vides the possibility of assaying haptens (6, 7).

Precision studies showed that adequate assay rangesfor the quantitation of the eleven tested proteins wereeasily covered with an accuracy that was comparableto and sometimes better than that of conventionalimmunonephelometry and radial immunodiffusion(8 — 11), in spite of a larger dilution of samples. Re-covery studies showed a close agreement between thetheoretical and measured concentrations except forIgG and C-reactive protein (slight over-recovery).Method comparison indicated a generally strong cor-relation between the concentrations measured by mi-croparticle-enhanced nephelometric immunoassayand by the reference methods.

Microparticle-enhanced nephelometric immunoassayusing microparticles and a nephelometer specificallyconceived for that use, thus appears to offer an alter-native method for routine determination of a greatvariety of serum proteins at high and intermediateconcentrations. It has been also used to quantifyproteins in other biological fluids such as cerebrospi-nal fluid (2) and milk (12 — 15) and other biologicalmolecules with lower concentration (1, 6, 7). In ad-dition to these above-mentioned reports, this workshows the general applicability of the method.

AcknowledgementsThis study was supported in part by grants 81 MO835 and 84Ml 121 from the Ministere Frangais de la Recherche et de laTechnologie. P. Montagne is a research engineer, Institut Na-tional de la Sante et de la Recherche medicale. M. L. Cuilliereis a study engineer, Centre National de la Recherche Scienti-flque.

References

1. Duheille, J., Pau, B. & Gros, P. (1982) Reactif permettantun dosage de tres haute sensibilite de Pantigene caracter-istique du virus de l'hepatite B dans les liquides biologiqueshumains, European Patent EPO 0104101B1, Sanofi(France).

2. Cuilliere, M. L., Montagne, P., Bessou, Th., El Oman, R.,Riochet, D., Varcin, P, Laroche, P., Prud'homme, Ph.,Marchand, J., Flecheux, O., Pau, B. & Duheille, J. (1991)Microparticle-Enhanced Nephelometric Immunoassay (Ne-phelia®) for Immunoglobulins G, A, and M. Clin. Chem.37, 20-25.

3. Mancini, G., Carbonara, A. C. & Heremans, J. F. (1965)Immunochemical quantitation of antigens by single radialimmunodifTusion. Immunochemistry 2, 235 — 254.

4. Sieber, A. & Gross, J. (1976) Determination of proteins bylaser nephelometry. Laboratoriumsblätter 26, 117 — 123.

5. Voigt, H. W. (1977) Klärung lipämischer Seren durch neuesVerfahren. Laboratoriumsblätter 27, 168 — 172.

6. Duheille, J., Montagne, P, Riochet, D., Bessou, Th., Var-cin, P, Cuilliere, M. L., Marchand, J. & Pau, B. (1984)Immunonephelometric assay of serum thyroxine and triio-dothyronine, in Abstracts from the Twelfth InternationalCongress of Clinical Chemistry, Rio de Janeiro, p. 1410.

7. Gartner, A., Carles, C., Montagne, P., Cuilliere, M. L. &Duheille, J. (1991) A Microparticle Enhanced Nephelo-metric Immunoassay (Nephelia®) Applied to ThymulinMeasurement. J. Immunoassay 72, 521 — 542.

8. Virella, G. & Funderberg, H. (1977) Comparison of im-munoglobulin determination in pathological sera by radialimmunodiffusion and laser nephelometry. Clin. Chem. 23,1925-1928.

9. Guiguet, M., Padieu, P. & Mack, G. (1983) Laser nephe-lometric measurement of seven serum proteins comparedwith radial immunodiffusion. J. Clin. Chem. Clin. Biochem.27,217-221.



10. Schmitz-Huebner, U, Nachbar, J. & Asbeck, F. (1980) The 14. Marchal, E., Collard-Bovy, C., Humbert, G., Linden, G.,determination of antithrombin III, alpha 2-macroglobulin Montagne, P., Duheille, J. & Varcin, P. (1991) Microparti-and alpha 2-antiplasmin in plasma by laser nephelometry. cle-Enhanced Nephelometric Immunoassay: 2-Measure-J. Clin. Chem. Clin. Biochem. 18, 221—225. ment of Alpha-Lactalbumin and Beta-Lactoglobulin. J.

11. Daigneault, R. & Lemieux, D. (1978) Evaluation of Behring Dairy Sei. 74, 3702-3708.laser-nephelometer prototype in the measurement of IgG, 15. Humbert, G., Collard-Bovy, C., Marchal, E., Linden, G.,IgA and IgM. Clin. Biochem. 11, 28-31. Montagne, P., Duheille, J. & Varcin, P. (1991) Microparti-

12. Montagne, P., Gavriloff, C., Humbert, G., Cuilliere, M. L., cle-Enhanced Nephelometric Immunoassay: 3-ApplicationDuheille, J. & Linden, G. (1991) Microparticle-Enhanced to Milk and Dairy Products. J. Dairy Sei. 74, 3709-3715.Nephelometric Immunoassay for Immunoglobulins G inCow Milk. Lait 71, 493-499.

13. Collard-Bovy, C., Marchal, E., Humbert, G., Linden, G., Paul MontagneMontagne, P., El Bari, N., Duheille, J. & Varcin, P. (1991) Laboratoire d'ImmunologieMicroparticle-Enhanced Nephelometric Immunoassay: l- Faculte de MedecineMeasurement of Alpha- and Kappa-Caseins. J. Dairy Sei. BP 18474, 3695-3701. F-54505 Vandoeuvre les Nancy Cedex

Eur. J. Clin. Chem. Clin. Biochem. / Vol. 30, 1992 / No. 4

measurement of eleven serum proteins by microparticle

Documents