Journal of Virological Methods, 27 (1990) 189-202

189

Elsevier VIRMET 00975

Rapid biotin-avidin method for quantitation antiviral antibody isotypes

of

Jeffrey D. Peterson I, Stephen D. Miller and Carl Waltenbaugh Department of Microbiology-Immunology and the Graduate Program in Neurosciences, University Medical School, Chicago, IL 60611, U.S. A.

Northwestern

(Accepted 5 October 1989)

Summary A rapid and efficient method is described for isotype quantitation of antiviral antibodies in mice infected with Theiler’s murine encephalomyelitis virus (TMEV). Serum antibodies were reacted with fluorochrome-labeled TMEV in a modified fluid-phase particle concentration fluorescence immunoassay (PCFIA). Biotin and avidin were used to attach antiimmunoglobulin isotype antibodies to polystyrene particles by the separate incubation of biotinylated goat anti-mouse isotypes (IgGlIgG2a-, IgG2b-,IgG3-, or IgM-specific) with avidin coupled polystyrene particles. These anti-isotype particles captured the virus-antibody complexes. Mouse myeloma proteins were used to quantitate and standardize isotype profiles of normal mouse serum using fluorescein isothiocyanate (FITC)-labeled, goat anti-mouse isotypes and polystyrene particles coated with goat anti-mouse. These assays quantitated the affinity-purified mouse serum antiviral antibodies for the standardization of antiviral isotype assays. Immunoglobulin of all serum isotypes as well as the amount of virus-specific isotypes can be quantitated rapidly and accurately. Fluorescence immunoassay; Viral antibody; Particle concentration fluorescence immunoassay; Theiler’s murine encephalomyelitis virus; Antibody isotype quantitation

Correspondence too: Jeffrey D. Peterson, Department of Microbiology-Immunology, University Medical School, 303 East Chicago Avenue, Chicago, IL 60611, U.S.A.

0166-0934/90/$03.50 0 1990 Elsevier Science Publishers B.V. (Biomedical Division)

Northwestern

190

Introduction

Solid-phase immunoassays often utilize antigen-conjugated assay plates to capture antibodies for detection by enzyme- or fluorochrome-labeled, immunoglobulin-specific antibodies. Adaptation of this methodology for the measurement of antiviral antibodies often results in high, non-specific background levels of immunoglobulin binding (Tijssen, 1985). Solid-phase immobilization of antigen in immunoassays often traps immunoglobulins undetectable under physiological conditions, contributing to high backgrounds. Basic pH buffers are often used to bind optimal quantities of virus to polystyrene plastics, but this can cause large conformational changes in viral particles altering, destroying, or masking relevant epitopes, thus affecting antibody binding (Rueckert, 1976). In contrast, fluid-phase antibody-ligand interactions, under near physiological conditions, circumvent the problems of solid-phase induced ligand alteration, epitope masking, and high backgrounds. Jolly et al. (1984) described a modified fluorescence immunoassay (FIA), incorporating facets of both fluid and solid phase, which used a suspension of antiimmunoglobulin conjugated polystyrene particles to capture serum immunoglobulin. Specially designed 96-well plates permitted the vacuum separation of the solid and fluid phases and washing of the resulting particle pellet between assay steps. Then fluorescein isothiocyanate (FITC)-labeled anti-immunoglobulin was allowed to bind to the captured serum immunoglobulin. This modified FIA, termed PCFIA (particle concentration fluorescence immunoassay), circumvented a basic problem inherent in the use of fluorescent detection by concentrating the assay’s solid phaseassociated fluorescence into a pellet which is essentially the same diameter as the beam of light used for excitation with little or no light scatter. Peterson et al. (1989) modified PCFIA allowing antibody and FITC-labeled virus to interact in fluid phase, followed by the capture of antibody-virus complexes with anti-immunoglobulin conjugated polystyrene particles. In this modified PCFIA, phase separation and washing was only performed subsequent to the final solid-phase capture. Detection of the FITC-labeled virus showed exquisite specificity with a low background, because the specificity and signal are both contained in the antigen itself. In the present report, we expand the versatility of fluid-phase PCFIA by developing rapid, efficient immunoassays to quantitate immunoglobulin and antiviral isotypes (IgGl, IgG2a, IgG2b, IgG3, and IgM). (Throughout this paper we use immunoglobulin to mean those molecules whether or not they have antiviral activity whereas antibody indicates those molecules with known specificity.) In antiviral isotype PCFIAs a biotimavidin modification was used to produce isotype-specific polystyrene particles by incubation of biotinylated goat anti-mouse isotype antibodies (b-anti-isotypes) with small aliquots of a large stock of avidin-conjugated polystyrene particles. Antiviral isotypes were easily detected using FITC-labeled virus to bind antibody, followed by the capture of antibody-virus complexes with anti-isotype particles. Similarly, polystyrene particles coated with goat anti-mouse immunoglobulin used in conjunction with the FITC-labeled goat anti-mouse antibodies allowed the detection and quantitation of immunoglobulin isotypes following standardization against myeloma isotype proteins.

191

Materials and Methods

Mice Female SJL/J mice were purchased from the Jackson Laboratory, Bar Harbor, ME. Mice were 2-3 months old at initiation of these experiments and were maintained on standard laboratory chow and water ad libitum. Antibodies and myeloma proteins duty-pled FITC-labeled, goat ~ti-moue IgM ~31501), IgGl (~2~1), IgG2a @X32201), IgG2b (M32401), IgG3 (M32601) and IgM + IgG (M30801) antibodies were purchased from Caltag Laboratories, South San Francisco, CA). Aft?nity-purified, biotinylated goat anti-mouse IgM (M31515), IgGl (M32015), IgG2a (M32215), IgG2b (M32415), IgG3 (M32615) and IgM + IgG (M30815) antibodies were also obtained from Caltag. Mouse myeloma proteins, TEPC 183 (IgM K, M2770), MOPC 21 (IgGl K, M9269), UPC 10 (IgG2a K, M9144), MOPC 141 (IgG2b K M8894), and FLOPC 21 (IgG3 K, M3645), purified by ion-exchange chromatography, were purchased from Sigma Chemicals, St. Louis, MO. Fluorescence measurement PCFIA assays are performed with the Fluorescence Concentration Analyzer (FCA, Pandex Division, Baxter Healthcare Corp., Mundelein, IL), a multi-wave-’ length, set-automated fluorescence reader. Socially-desired 96-well plates are used together with antigen or antibody immobilized on polystyrene particles to capture fluorochrome-labeled ligand. Fluid-phase unbound ligand is separated from matrix-bound ligand by drawing a vacuum (20 in. Hg) across a porous cellulose acetate membrane on the bottom of the assay plates. The retained ligand undergoes several wash-separation cycles (usually 3-5 cycles of 50 l.~lper well each). During the final separation the polystyrene particles bearing the bound ligand are concentrated onto the 2 mm diameter cellulose acetate well bottom. The wells are then illuminated at one or more wavelengths (365, 485, 545, or 590 nm), epifluorescence is read by a photodetector (at 450, 535, 575, 620 nm, respectively, depending upon the fluorochrome employed), and data are reported as relative fluorescent units (RFU). Virus preparation Virus was prepared as described in detail in Lipton and Friedmann (1980) and Peterson et al. (1989). Briefly, confluent BHK-21 cells were infected with Theiler’s murine encephalomyelitis virus (TMEV), BeAn 8386 strain, in tissue culture medium for 48 h. Virus in the supernatant and cellular debris was precipitated with NaCl and polyethylene glycol (PEG). The PEG precipitate, containing virus and cellular debris, was pelleted and resuspended by sonication. Virus was further pur-

192

ified by sequential discontinuous 15-30% sucrose and Cs,SO, gradients. Finally, the virus was pelleted by ultracentrifugation, resuspended, and measured for optical absorbance at A2s0 to determine concentration. FITC labeling of virus

TMEV (1 mg in 1 ml in PBS, pH 7) was incubated with 50 ~1 of fluorescein isothiocyanate (FITC) freshly prepared stock solution (4 mg/ml in 0.5 M carbonate buffer, pH 9.5) in the dark for 2 h with occasional mixing. Unconjugated FITC and TMEV-conjugated FITC (FITC-TMEV) were separated on a 2.5~40 cm Sephadex G-25 column. The respective bands were visualized by ultraviolet light. FITC-TMEV eluted in the first peak and unconjugated FITC in the second. The fluorochrome to protein (F/P) ratio of the FITC conjugate was determined by absorbance measurements at 495 and 280 nm and calculated to be 240. Assay particles

Avidin-coated particles were prepared by incubating avidin (Calbiochem, 8 mg in 4 ml of 0.01 M sodium acetate buffer, pH 5.0) with 4 ml of carboxylated particles (0.83 km diameter, 5% w/v, Pandex) in the presence of 80 mg l-ethyl-3-(3dimethylaminopropyl)-carbodiimide HCl (EDC) (Calbiochem) on a rotary mixer for 2 h room temperature (22°C). The particle suspension was centrifuged at 3000 x g for 30 min to remove unbound avidin. The particles were washed 3 times by resuspending the pellet in 8 ml isotonic buffered saline (IBS) and centrifuging as above. Avidin particles were finally suspended in 80 ml of IBS. Goat anti-mouse immunoglobulin coated polystyrene particles (Fluoricon assay particles, 0.83 pm diameter, lot No. DO-l) were purchased from Pandex. PCFIA

assay

Assay optimization.

Assay quality and, hence, quantitation of antibody, depends upon reagent optimization. It is necessary that assay conditions such as FITC conjugate dilutions, particle concentrations, b-anti-isotype dilutions, and incubation times be optimized to produce binding curves with low non-specific immunoglobulin binding and maximum fluorescence lo- to 30-fold greater than background (naive sera binding in antiviral assays and buffer wells for immunoglobulin assays). We previously optimized FITC-TMEV as an indicator ligand (Peterson et al., 1989). Fig. 1 illustrates the influences of particle concentration, biotinylated anti-isotype concentration, and incubation times on the quality of the antiviral IgG2b assay. Fig. la shows that keeping b-anti-IgG2b and incubation times constant, an optimal signal:background ratio for avidin-coated particles is seen at 0.25% (w/v). Likewise, b-anti-IgG2b concentration is optimal at a dilution of l/100, providing no significant increase at greater concentrations (Fig. lb). Anti-isotype capture by anti-mouse Ig particles is very efficient, requiring as little as 5 min to show significant signal over background (Fig. lc). Subsequent experiments showed

193

Fig. 1. Assay optimization. The antiviral IgG2b assay was optimized by varying individual assay conditions while keeping other parameters constant. Full immune serum (*-a) and normal serum (o---o) titer curves from l/50 to 1151200 were assayed, and results were expressed as RFU at serum dilutions of l/200, providing half-maximal fluorescence that would be subject to either an increase or decrease as parameters were varied. (a) Avidin particle concentration was assayed from 0% to 1% (w/v) while maintaining b-anti-IgG2b at a dilution of 11100and FITC-TMEV at 156 @well. (b) Biotin-antiIgG2b dilution ranged from 1150to 1/800, using an avidin particle concentration of 0.25% w/v and FITCTMEV at 156 @well. (c) Incubation time for the solid phase capture was examined by incubating FITCTMEV (156 @well) with serum dilutions for 30 min. Subsequently, an avidin particle concentration of 0.25% w/v, conjugated with l/100 b-anti-IgGZb, was added to assay plates at five minute intervals, from 5 to 55 min.

incubation times of 30 min would assure sufficient signal in serum samples from low responder mice. Incubations greater than 30 min provided no added benefit.

194

Standard procedure. Goat anti-mouse isotype particles were prepared by incubating avidin particles (0.25% w/v) with a l/200 dilution of biotinylated goat antimouse isotype antibodies (b-anti-IgGl, b-anti-IgG2a, b-anti-IgG2b, b-anti-IgG3, or b-anti-IgM purchased from Caltag Laboratories, South San Francisco, CA) at room temperature with occasional mixing. Mouse sera (20 l.~lin log, serial dilutions) were incubated with 40 ng of FITC-labeled TMEV in 20 ~1 PBS in 96-well Fluoricon Assay Plates (Pandex) for 30 min at room temperature. Twenty (*l of goat anti-mouse isotype-coated particles (0.25% [w/v] suspension) were added to each well and incubated for an additional 30 min at room temperature. Automated instrumentation (FCA, Pandex) performed phase separation, washing, and determination of relative fluorescence units (RFU) (at 535 nm for emission of FITC). Data are expressed as RFUs. Antibody concentrations determined by logarithmic linear regression analysis of the RFU values are expressed as pg/ml based upon a standardized curve obtained by PCFIA analysis. immunoglobulin

isotype assays

Log, dilutions of serum samples, beginning at l/500, were made in IBS containing 0.05% Tween-20 (IBST). Duplicate 50 p,l serum dilutions together with 20 ~1 FITC-labeled goat anti-mouse isotype antibodies (Caltag) were added at l/200 dilution to Fluoricon assay plates and incubated for 30 min at room temperature. Twenty l.rJ goat anti-mouse Ig particles (0.25% w/v) were added to each well and incubated for 30 min at room temperature. Plates were processed and read by the FCA as described above. Serum immunoglobulin of all isotypes could be correlated to relative fluorescence units using FITC-labeled goat anti-mouse isotype. AfJinity purification of anti- TMEV antibodies

An immunoadsorbent gel was prepared by coupling TMEV to CNBr-activated Sepharose 4B (Pharmacia) at 1 mg virus/ml gel as previously described for protein antigens (Lei et al., 1982). One ml gel was packed into a 3 ml syringe fitted with a plastic fritted disc and washed with 100 ml IBS. A 2 ml pool of serum, diluted l/2 in IBS, from TMEV-infected SJL/J mice was applied to the column as a slow drip. The column was washed with 20 ml IBS to remove unbound protein and bound anti-TMEV antibody was eluted with 0.1 M glycine-HCl buffer, pH 2.6. The eluate was immediately neutralized to pH 7.0 with NaOH to prevent denaturation of anti-TMEV antibodies. The column was regenerated by sequentially washing with 0.1 M glycine-HCl, pH 2.6, and IBS buffers until an effluent pH of 7.0 was attained. The neutralized eluate was passaged over the TMEV-Sepharose column a second time, the column washed with IBS, bound material eluted with the glytine-HCl buffer, and the eluate neutralized to pH 7.0 with NaOH. Isotype concentrations of this twice affinity-purified anti-TMEV antibody were determined as described in the Results.

195

35 ‘\

‘1

30

2.00

1.

2.50 LoolO

0

RFWIOOO

=

-16.7x

+

75.1

.

PFUl1000

=

-21.1x

+

60.4

3.00 Rec~pocal

3.50 Seru-n

4.m

4.50

Dilution

Fig. 2. Data analysis. An example of data analysis is shown using a pooled serum standard from mice infected 80 days with TMEV (m-m) to quantitate antiviral IgG2b in serum from a TMEV-hyperimmunized mouse (0-O). The linear portion of each titration curve, as described by three to five sequential points, was analyzed by logarithmic linear regression as described in Materials and Methods. The calculated equations for these lines were used to determine intersections with an RFU value which was lo-fold over background (i.e., RFU in the absence of added serum indicated by the dotted line). These intersections established points which were used to compare the standard with the unquantitated hyperimmune sample.

Data analysis The linear portion of each titration curve, as described by three to five sequential points, was used for data analysis. Logarithmic linear regression utilizing Slidewrite Plus (Version 3.10, Advanced Graphics Software Inc., Sunnyvale, CA) was performed to calculate the equation describing this portion of the titration curve. In all cases this method provided a correlation coefficient (r) between -0.95 and - 1 .O, where - 1 .O represents complete correlation. Calculated intercepts with lofold background RFU (i.e., RFU in the absence of added serum) were multiplied by the serum standard antiviral isotype concentration (1 g/ml) and divided by the serum standard lo-fold background intercept value to calculate isotype concentrations. For example, Fig. 2 illustrates a standardization curve for IgG2b antiviral isotype. The linear portion of the curve for the standard is described by the three data points (solid boxes) indicated by the solid arrows. Regression analysis yields the equation Y = -21.2X + 68.4 with a correlation coefficient (r) of -0.99. Solving for logi reciprocal serum dilution using an RFU value lo-fold over background (10 x 1000 RFU as indicated by the dotted line) yields a value of 2.77 log,, and an antilog of 586 (i.e. a l/586 dilution). Performing a similar procedure on an unquantitated serum sample (open boxes and open arrows) yields a value of 3.9 log,, and an antilog of 7910 (i.e. 117910 dilution). Multiplying the antilog of the unknown serum (7910) by the known antiviral IgG2b concentration of the serum standard (235 kg/ml) and dividing this quantity by the antilog of the serum standard (586) yields a concentration of 3172 &ml of antiviral IgG2b in the unknown.

196

Results

Antibody is important in the immune response to either bacterial or viral infection. Different antibody classes or isotypes often play diverse roles in immune responses to infection (Lafrenz et al., 1981; Farkas et al., 1982). Until the present, no quantitative method has been available for determining antiviral antibody isotypes in virally infected mice. Here we demonstrate quantitative isotype assays for immunoglobulin and antibody to TMEV, a picornavirus which induces demyelinating disease in susceptible strains of mice (Lipton, 1975; Dal Canto and Lipton, 1976; Lipton and Friedmann, 1980). Immunoglobulin

and antiviral isotype assays

Serum pools, obtained from SJL/J mice 80 days following either sham or TMEV infection, were used as standards for the development of the isotype assays. Immunoglobulin isotype assays used FITC-labeled anti-isotype (anti-IgGl-, antiIgG2a-, anti-IgG2b-, anti-IgG3-, or anti-IgM-FITC) to bind serum immunoglobulin in fluid phase. Immune complexes were captured by goat anti-mouse Ig-conjugated particles. Parallel assays for the antiviral component of each isotype were performed by a different methodology. FITC-TMEV and serial dilutions of serum samples were incubated in assay plates for 30 min to allow antibody-antigen binding in fluid phase. The resulting antibody-FITC-TMEV complexes were captured by 30 min incubation with anti-isotype particles, produced by incubating avidinconjugated particles with biotinylated anti-isotype for 30 min. The wells were washed twice with IBS and the fluid phase removed by vacuum filtration, illuminated at 485 nm, read at 535 nm, and reported as relative fluorescence units (RFU). Fig. 3 shows Ig (solid circles) and anti-TMEV antibody (open squares) binding curves for (a) total, (b) IgM, (c) IgGl, (d) IgG2a, (e) IgG2b, and (f) IgG3. These data contain log-linear or log-sigmoidal portions suitable for use in quantitative analysis. The background fluorescence in negative assay wells for immunoglobulin isotype assays was always less than 500 RFU (data not shown), and binding of normal mouse sera in antiviral assays was insignificant (solid triangles). The sensitivity of each of the immunoglobulin isotype assays is comparable, showing significant detection over background at comparable ranges of dilutions. Standardization of isotype assays

For quantitation of immunoglobulin isotypes, myeloma protein standards were used for each isotype. The curve for each myeloma standard was compared to the serum curves from TMEV-infected and sham-infected animals. Logarithmic-linear regression analysis of the linear portions of RFU (535 nm) curves determined the intersection with RFU cut-off values as detailed in the data analysis section of the Materials and Methods. FITC-labeled goat anti-isotype antibodies were used to quantitate an affinity-purified (on a TMEV-Sepharose column) anti-TMEV serum antibody pool. No anti-TMEV activity was seen in the effluent of the TMEV-Se-

e.

IgG2b

Fig. 3. Immunoglobulin and antiviral isotype assays. Serum from TMEV-infected SJUJ mice were assayed for total (0-a) and mti-TMEV (-) isotypes by incubating serial dilutions with either FITG anti-Ig or FITC-TMEV, reqectiveIy. Serum from shaminfected (A.....A) SW.F mioe was assayed only for antiviraf isotype concentrations. For illustrative purposes, immunoglobulin and antiviral isotype curves are placed in the same panels. These assays measured total immunoglobulin and total antiviral antibody (a); IgM and antiviral IgM (h); I@1 and antiviral IgGl (c}: IgGZa and’aativirai IgG2a (d); IgGZb and antiviral IgG2b {e); and Ig3 and antivixal Ig3 (f)_ Data are expressed as relative ihtorescent units

198 TABLE 1 Isotype concentrations

in the sera of TMEV-infected

Sample”

and sham-infected

Concentration IgM

SJLiJ mice

(ug/ml)b

IgGl

IgG2a

0.10

Anti-TMEV serum pool

Anti-TMEV antibody

1

39

186

235

9

540

3250

2190

1110

1820

330

8700

0

0

0

0

0

0

1800

920

1130

2300

300

6500

Anti-TMEV antibody Total Ig

0.14

Total

Anti-TMEV antibody

Sham-infected serum pool

0.26

IgG3

Affinity purified anti-TMEV

Total Ig

0.13

IgG2b

0.03

0.67

“Affinity-purified anti-TMEV antibodies or serum pools from sham-infected and TMEV-infected mice 80 days post infection. “Antibody concentrations were determined by logarithmic linear regression analysis in comparison to a serum standard as described in Materials and Methods. All isotype determinations showed coefficients of variation (CV) less than lo%, except for zero values which indicate that calculations could not be performed since anti-TMEV antibodies were undetectable.

pharose column, indicating that all detectable anti-TMEV was absorbed by the column (data not shown). Once quantitated, the affinity-purified antibody was used in the anti-TMEV isotype assays and compared to serum from TMEV-infected and sham-infected mice. In this way the antiviral isotype assays were standardized against the immunoglobulin isotype assays. The quantitative sum of isotype concentrations of the affinity-purified antibody is in close agreement with the concenTABLE 2 Serum immunoglobulin

isotype profiles in normal and TMEV-infected

SJL/J mice

Percent

Sample

IaM

IaGl

IaG2a

IaG2b

InG3

Sham-infected serum pool

Ig isotype% of total Ig”

28

14

18

36

5

Anti-TMEV serum pool

Ig isotype% of total Ig”

37

25

13

21

4

Total

Anti-TMEV% of total anti-TMEVb

0.2

8.3

39.5

50.0

1.9

-

Anti-TMEV% of total Ig isotypeC

0.03

1.8

16.8

12.9

2.7

6.0

“Individual immunoglobulin isotype concentrations from Table 1 are expressed as percentages of the sum of IgM, IgGl, IgGZa, IgG2b, and IgG3. Due to rounding, sum totals may not equal 100%. bIndividual antiviral isotype concentrations are expressed as percentages of the sum of antiviral IgM, IgGl, IgG2a, IgG2b, and IgG3. ‘Antiviral isotype concentration is expressed as percentage of the same immunoglobulin isotype.

199

tration of total antiviral antibody suggesting that the affinity purified TMEV-specific antibody contained little or no non-specific immunoglobulin. Tables 1 and 2 show immunoglobulin and antiviral isotype profiles in serum from both.normal and TMEV-infected SJL mice expressed as both kg/ml and percentages respectively. Table 1 shows infected animals contain approximately 30% more immunoglobulin (8700 l&ml vs 6500 l&ml) than do sham-infected controls. We have observed this to occur consistently and it may indicate an adjuvant effect of viral infection since only 540 &/ml is TMEV-specific. Isotype profiles of TMEVinfected mice show increased IgM (37% vs 28%) and IgGl (25% vs 17%) and decreased IgG2b (21% vs 35%) compared to the control (sham-infected) animals (Table 2). Overall, these profiles agree with isotype profiles seen in other strains of mice (Fahey et al., 1965; Natsuume-Sakai et al., 1977; and Popp, 1979). Table 2 shows that approximately 6% (540 Fg of 8700 ug/ml) of the total immunoglobulin of a pooled immune serum had anti-TMEV activity, with IgG2b (50.0%) and IgG2a (39.5%) as the major antiviral isotypes, followed by IgGl (8.3%), IgG3 (1.9%), and IgM (0.2%). Antiviral activity of the IgG2a and IgG2b, in mice 80 days post infection, account for a large proportion (16.8% and 12.9%, respectively) of each of these immunoglobulin isotypes. In contrast, IgGl anti-TMEV activity accounts for only 2% of the IgGl immunoglobulin isotype even though total IgGl levels are more than twice that seen in serum from sham-infected mice. The low IgM and IgG3 concentrations of anti-TMEV suggest that these isotypes play a minor role in the antibody response approximately three months post-infection.

Discussion We describe a procedure to quantitate anti-viral isotypes rapidly and efficiently. This methodology has the major advantage of directly quantitating antiviral antibodies and isotypes instead of presenting the data as titers (Coutelier et al., 1987). Our technique has several technical advantages compared to commonly employed ELISA methods for detection of antiviral activity. Virus-antibody reactions occur in fluid phase, more closely mimicking in vivo immune reactions. This is followed by the solid-phase capture of the resulting virus-antibody complexes with a suspension of anti-isotype antibody-coupled particles. The obvious benefit to our technical approach is that by allowing FITC-labeled virus to bind antibody, specificity and signal are contained within one element. This results in lowered background interference allowing the measurement of very small amounts of anti-viral antibody subclasses. In fact, our assays routinely measure as little as 50 ng (1 &ml) antiviral antibody per well. PCFIA differs from most other immunoassays in that it uses a particle suspension as the solid phase. This avoids edge effects and gradient problems found with antigen-coated assay plates since particles are made and stored in large batches. Assay variance is dependent on pipetting accuracy. In comparison, polystyrene plate-based solid-phase immunoassay variance results from gradients of tempera-

200

ture, static charge, binding, and humidity across the plate (Tijssen, 1985). In the experiments reported here, intra-assay coefficient of variation (CV) for immunoglobulin and antiviral isotype PCFIAs ranges from 1.5 to 5.9% (sample size n = 15), and a 9% CV between assays (n = 8). An important feature of PCFIA is its greatly increased surface area compared to polystyrene plates. A 50 p,l sample incubated with 20 ~1 of 0.25% 0.83 p_rnparticles is exposed to 4.5 times greater surface area compared to a similar volume in an ELISA well. Further, fluid-phase and Brownian motion of the submicron particles in suspension contributes to reduced assay incubation times (Fig. lc). The immunoglobulin isotype assay curves as well as the antiviral isotype assay curves (Fig. 3) show comparable sensitivities between different isotypes. For illustrative purposes, immunoglobulin and antiviral isotype curves are placed in the same panel. Assignment of endpoint titers to these curves would be misleading, since each curve was derived from reagents with different relative affinities (antimouse isotype antibodies, antibody- or avidin-coupled assay particles) and different fluorochrome-protein ratios (FITC-TMEV and FITC-labeled anti-isotypes). Precise quantitation of separate antiviral isotypes avoids this problem. Comparisons between different immunoglobulins, between immunoglobulin and antiviral isotypes, and between different antiviral isotypes can only be determined accurately after standardization of reagents used for detection. For example, examination of the titer curves for IgM (Fig. 3b) and IgG2a (Fig. 3d) would suggest that the anti-TMEV content of these two isotypes were roughly equivalent. However, when antiviral isotypes are quantitated in pg/ml using a pool of affinity-purified anti-TMEV antibodies (derived from mice 80 days post-infection) as a standard, antiviral IgM is found to be only 0.03% of the total immunoglobulin, while IgG2a is 16.8% (Table 2). Serum immunoglobulin and isotypes concentrations vary greatly between mouse strains. Total IgG ranges from 3.22 to 12.72 mg/ml (Popp, 1979). Immunoglobulin concentrations increase as much as lo-fold in the first eight months of life in some strains (Natsuume-Sakai et al., 1977). IgM can range from 0.041 to 0.44 mg/ml, IgGl from 0.35 to 3.69 mg/ml, IgG2a from 0.25 to 3.46 mg/ml, and IgG2b from 0.25 to 1.27 mg/ml (Natsuume-Sakai et al., 1977; Popp, 1979). The immunoglobulin isotype profiles we find in the present study fall well within these ranges. The only exception is the IgM levels which may reflect differences in mouse strains used, the age of the mice, or possibly due to environmental conditions (i.e., vivarium flora, etc.). Total IgM and IgGl are increased in sera from TMEV-infected mice compared to sera from sham-infected mice (3250 vs. 1800 p,g/ml, and 2190 vs. 920 kg/ml, respectively). These elevated immunoglobulin levels, however, do not reflect increased anti-TMEV IgM and IgGl antibody concentrations. Only 0.03% of the IgM (1 of 3250 pg/ml) and 1.78% of the IgGl (39 of 2190 pg/ml) have anti-TMEV activity. Large increases in total IgM and IgGl in TMEV-infected mice may reflect responses associated with increased susceptibility to other opportunistic infections. The vast majority of the antiviral antibody response consists of IgG2a and IgG2b (Table 2)) agreeing with previous studies (Coutelier et al., 1987) which have found

201

predominant IgG2a and IgG2b response to TMEV-infection in C57BW6 mice. The sum of the separate isotype concentrations agrees well with the independently determined total immunoglobulin concentration. For instance, the individual antiviral isotypes add to 470 &ml compared to total antiviral antibody of 540 kg/ml (Table I). The summed totals for immunoglobulin isotypes are 8700 l&ml for serum from TMEV-infected mice and 6450 l.rg/rnl for normal mouse serum pools, and the measured immunoglobulin concentrations were determined as 8700 l.&nl and 6500 l&ml, respectively. In conclusion, we have developed rapid, quantitative assays for antiviral antibody isotypes. Although pooled sera were used in the present studies, our modification of PCFIA allows for the determination of antiviral isotypes of individual mice. As little as 50 ~1 of serum can be used to determine complete antiviral isotype profiles. Future experiments will address the possible role of antibody isotypes in histopathology and susceptibility in TMEV-induced demyelinating disease. The use of fluid-phase PCFIA will simplify and expedite the quantitation of total isotype and antibody isotype profiles in large numbers of infected mice. a

References Comelier, J.-P., Van der Logt, J.T.M., Heessen, F.W.A., Warnier, G. and Van Snick, J. (1987) IgG2a restriction of murine antibodies elicited by viral infections. J. Exp. Med. 16.5, 64-69. Dal Canto, M.C. and Lipton, H.L. (1976) Primary demyelination in Theiler’s virus infection. An ultrastructural study. Lab. Invest. 33, 626-637. Fahey, J.L., Wunderlich, J. and Mishell, R. (1965) The immunoglobulins of mice. I. Four major classes of immunoglobulins: 7S,,., 7S,,., ylA(8ra)- and 18S,,,-globulins. J. Exp. Med. 120, 223-242. Farkas, AI., Medgyesi, G.A., Fust, G., Miklos, K. and Gergely, J. (1982) Immunogenicity of antigen complexed with antibody. I. Role of different isotypes. Immunology 45, 483-492. Jolley, M.E., Wang, C-H.J., Ekenberg, S.J., Zuelke, M.S. and Kelso, D.M. (1984) Particle concentration fluorescence immunoassay (PCFIA): a new, rapid immunoassay technique with high sensitivity. J. Immunol. Methods 67, 21-35. Lafrenz, D., Strober, S. and Vitetta, E. (1981) The relationship between surface isotype and the immune function of murine B lymphocytes. J. Immunol. 127, 867-872. Lei, H.-Y., Dorf, M.E. and Waltenbaugh, C. (1982) Regulation of immune responses by I-J gene products. II. Presence of both I-Jb and I-Jk suppressor factors in (nonsuppressor x nonsuppressor) F, mice. J. Exp. Med. 155, 955-967. Lipton, H.L. (1975) Theiler’s virus infection in mice: an unusual biphasic disease process leading to demyelination. Infect. Immun. 11, 1147-1155. Lipton, H.L. and Friedmann, A. (1980) Purification of Theiler’s murine encephalomyelitis virus and analysis of the structural virion polypeptides: correlation of the polypeptide profile with virulence. J. Virol. 33, 1165-1172. Natsuume-Sakai, S., Motonishi, K. and Migita, S. (1977) Quantitative estimations of five classes of immunoglobulin in inbred mouse strains. Immunology 32, 861-866. Peterson, J.D., Kim, J.Y., Melvold, R.W., Miller, SD. and Waltenbaugh, C. (1989) A rapid method for quantitation of antiviral antibodies. J. Immunol. Methods 119, 83-94. Popp, D.M. (1979) Basal serum immunoglobulin levels. I. Evidence for genetic control in BlO and BIO.F Mice. Immunogenetics 9, 125-135. Rueckert, R.R. (1976) On the Structure and Morphogenesis of Picornaviruses. In: H. Fraenkel-Conrat and R.R. Wagner (Eds), Comprehensive Virology, Vol. 6, pp. 131-213. Plenum Press, New York and London. Tijssen, P. (1985) The immobilization of immunoreactants on solid phases. In: R.H. Burdon and P.H. Knippenberg (Eds), Practice and Theory of Enzyme Immunoassays. Laboratory techniques in biochemistry and molecular biology, Vol. 15, pp. 297-328. Elsevier, Amsterdam.