Journal of Medical Virology 3:281-289 (1979)
Solid-Phase Radioimmunoassay of IgA, IgG, and IgM Antibodies to Human Rotavirus Hannu K. Sarkkinen, Olli H. Meurman, and Pekka E. Halonen Department of Virology, University of Turku, Turku, Finland
A solid-phase radioimmunoassay (RIA) has been developed for t h e detection of human rotavirus-specific IgA, IgG, and IgM antibodies. Nebraska calf diarrhea virus grown in LLC-MK2 cell cultures in the presence of trypsin was directly adsorbed o n t o polystyrene balls, and antibodies that attached t o the virus-coated balls were detected by subsequent binding of 125 I-labeled antibodies specific t o human alpha, gamma or mu chains of human IgA, IgG, or IgM immunoglobulins, A total of 116 serum specimens from 58 adult patients were tested. Binding ratios between the positive and t h e negative serum varied between 5 and 15, occasionally being 20 o r more in the IgA and IgG assays, but rarely exceeding 3 in t h e IgM assay. T h e RIA was found to be more sensitive in detecting antibodies t o rotavirus than the complement fixation (CF) test, the RIA titers obtained being 50-100 times as high as the CF titers. The method described offers a possibility of evaluating the immune response t o human rotavirus and of detecting recent infection. Key words: radioimmunoassay, IgA, IgG, IgM, rotavirus, gastroenteritis
INTRODUCTION
Rotavirus is t h e major etiologic agent of transmissible gastroenteritis in young children [Kapikian et al, 19761 but has recently been shown to affect also older children [Hara et al, 19761 and adults [von Bonsdorff, et a1 1978, Bolivar et al, 1978, Meurman et al, 1977a, qrstavik et al, 1976bl . Human rotavirus has not been successfully propagated in continuous cell lines routinely used in diagnostic laboratories [Bryden et al, 19771. This has complicated the development of diagnostic methods. In serologic tests human rotavirus obtained from stool specimens of infected children has been used as an antigen [Flewett et al, 1974, Kapikian et al, 19741. However, because many species have their own rotaviruses [Flewett et al, 19781 which can be propagated in cell cultures, animal viruses are more often used as a n
Received October 31, 1978. Address reprint requests to Dr. H.K. Sarkkinen, Department of Virology, University of Turku, Kiinamyllynkatu 10, 20520 Turku 5 2 , Finland.
0846-6615/79/0304-0281$02.00 0 1979 Alan R. Liss, Inc.
282
Sarkkinen, Meurman, and Halonen
antigen in serology. This has been possible because the rotaviruses from different species have a common group antigen [Woode et al, 19761 which cross-reacts in serologic tests. Bovine rotavirus has been used as an antigen in complement fixation (CF) and indirect fluorescence antibody tests [Kapikian et al, 1975, qrstavik et al, 1976al and SA-11 (vervet monkey rotavirus) in radioiinmunoassay (RIA) [Watanabe et al, 19771 for human rotavirus antibody. Recently, an enzyme-linked immunosorbent assay (ELISA) using antigen obtained from the stools of calves infected with human rotavirus was developed [Yolken et al, 19781. The present report describes a solid-phase radioimmunoassay for detecting human IgA, IgG, and IgM antibodies to rotavirus employing Nebraska calf diarrhea virus (NCDV) directly adsorbed to the solid phase as an antigen. MATERIALS AND METHODS Antigen Production
The Lincoln strain of Nebraska calf diarrhea virus (kindly supplied by Dr. T. Johnson, Malmo, Sweden) was grown in cultures of LLC-MK2 cells according to Babiuk et a1 [ 19771 . The 12th passage of NCDV in LLC-MK2 cells was used to prepare the antigen. Briefly, confluent cultures of LLC-MK2 cells in Roux bottles were infected with 3 ml of NCDV virus solution in basal medium Eagle (BME) in the presence of 10 pg of trypsin (Difco Laboratories, Detroit, Michigan, 1 : 250) per ml. After a one-hour adsorption period at t37"C, BME containing 0.5% bovine serum albumin (BSA) fraction V (Armour Pharmaceutical Company Ltd., Eastbourne, England) and 5 pg of trypsin per milliliter were added. When the cells showed extensive cytopathic effect (CPE), usually in 48-72 hours, the supernatant and the cells were harvested and pooled. The pooled material was subjected to two freeze-thaw cycles followed by centrifugation at 7,OOOg in a Sorwall GSA rotor for ten minutes. The pellet was discarded and the supernatant further centrifuged at 25,000 rpm for two hours in a Spinco SW 27 rotor. The remaining pellet was resuspended in a small volume of phosphate-buffered saline (PBS), pH 7.35, and checked for virus by electron microscopy (EM). The protein content was determined by the method of Lowry et al [ 195 11 and varied from 1,000 to 1,800 pg/ml. Specimens
A total of 116 serum specimens were collected from 58 army trainees attending the medical ward for acute gastroenteritis. Acute rotavirus gastroenteritis was diagnosed in eight patients by electron microscopy and/or the complement-fixation test. Antisera
Porcine antisera to the heavy chains of human immunoglobulins were purchased from Orion Diagnostica (Helsinki, Finland). The heavy-chain-specific antibodies were isolated by immunoadsorbent column chromatography and iodinated with 12' I as reported in detail elsewhere [Ziola et al, 19781 . The specific activities varied from 5 to 20 pCi/pg. RIA Procedure
The antigen was adsorbed onto polystyrene balls (6.4 mm in diameter, Precision Plastic Ball Co., Chicago) by incubation of the untreated balls in an antigen solution containing 10 pg of protein per milliliter for six hours at room temperature. Carbonate buffer, pH 9.6, was used as a diluent [Voller, 19761. After adsorption, excess antigen was aspirated off and carbonate buffer, pH 9.6, containing 15 mg of bovine serum albumin fraction V per milli-
Radioimmunoassay of Rotavirus
283
liter was added. After overnight incubation at room temperature, the balls were washed twice in PBS, air-dried, and stored at t4OC until used. Fourfold serial dilutions of serum specimens in 20O-pl aliquots were pipetted into disposable plastic tubes. The serum samples were diluted in PBS containing 0.5% BSA, 0.5% Tween 20, and 0.1%NaN3 (PBS-tween) as described by Meurman et a1 [1978]. A polystyrene ball with adsorbed antigen was then added to each tube. After incubating at t37"C for one hour, the serum samples were aspirated off and the balls were washed twice with 5 ml of tapwater. A 200-pl volume of '251-labeled antibodies specific to human alpha, gamma, or mu chains of human IgA, IgG, or IgM immunoglobulins was then added to each tube. As a dilution buffer for labeled antisera, Eagle's minimum essential medium (MEM), supplemented with 0.5% lactalbumin hydrolysate, 10%heat inactivated calf serum, 1% Tween 20, and 0.1% NaN,, was used [Meurman et al, 19781. After incubating at t37"C for one hour, the radioactive solutions were aspirated off and the balls washed as described above. The balls were then placed in clean tubes and counted in an LKB 1280 gamma counter. Buffer blanks (balls incubated in PBS-tween only) and positive and negative control sera were included in each assay. The assay was standardized by diluting the iodinated anti-human alpha antibodies to a concentration which gave 5,000 cprn bound (5,000 active cpm), and the anti-human gamma and mu antibodies to a concentration which gave 2,500 cpm bound, when 200 p1 of the antisera were incubated with a ball adsorbed with 2 pg of purified human IgA, IgG, and IgM. The RIA results were evaluated by means of binding ratios which were calculated by dividing the counts per minute of the serum specimens by the counts per minute of the negative serum at the same dilution. Binding ratios higher than 2 were considered positive, with the proviso that the counts per minute of the test specimen should be at least 150. Endpoint titers were calculated from the counts per minute/dilutions curve of each serum specimen. In tables and figures, titer values are expressed as reciprocals of dilutions. OTHER TESTS
The complement fixation test was performed with the standardized microtechnique [Casey 19651. NCDV grown in primary calf kidney cell cultures was used as an antigen. For electron microscopy stool suspensions were made approximately 10%in PBS. Small aliquots of stool specimens were put on carbon-coated Formvar grids, negatively stained with 2% phosphotungstic acid, pH 6.5, and examined with a Siemens Elmiscope electron microscope. Specimens from the antigen preparations were stained and examined similarly. RESULTS
Antigen Production
In preliminary experiments human rotavirus purified from stools from infected children was used as the antigen in the solid-phase radioimmunoassay. However, it was difficult to obtain sufficient amounts of rotavirus-positive stools for antigen production. This problem was overcome by exploiting the finding of Babiuk et al [ 19771 that bovine rotavirus replicates and can be passaged in continuous cell lines when small amounts of trypsin are present throughout the whole growth period. At first BSC-1 cell cultures were
284
Sarkkinen, Meurman, and Halonen
used, but we found it more convenient to prepare the antigen in LLC-MK2 cells.
Adsorption of Antigen to the Solid Phase
In initial experiments an IgG fraction of rabbit anti-NCDV hyperimmune serum was first used to coat the balls and NCDV antigen was then immunologically bound to the specific antibodies. However, in subsequent studies NCDV antigen was directly adsorbed to the solid phase. The results of the effect of the antigen concentration on the adsorption are shown in Figure 1. The best results were obtained when the balls were incubated in an antigen solution containing 10 pg of protein per milliliter. If lower antigen concentrations were used, the binding ratios decreased, and no further increase in binding ratios could be obtained if higher antigen concentrations were used. Figure 1 also demonstrates the fact that no benefit was gained if an IgG fraction of hyperimmune serum made against NCDV was used as precoating material on the solid phase. The effect of pH on the adsorption was not critical. When the pH of the adsorption buffer was kept between 7.35 and 9.6 no significant differences were found. Lowering the pH to 5.6 or raising it to 1 2 caused a clear decrease both in specific counts and in binding ratios. Throughout this study carbonate buffer, pH 9.6, was used.
CPI
1501
loo(
501
1 5
C
Fig. 1. Effect of protein concentration on adsorption of semipurified Nebraska calf diarrhea virus (NCDV) antigen on polystyrene bails. The balls adsorbed at each NCDV antigen concentration were incubated with 1:50 dilution of rotavirus antibody-positive and -negative serum and tested for the binding of IgA antibodies to rotavirus. Values given are means of three parallel determinations. 0 - 0 ) counts per minute (cpm) of the negative serum; o---o) cpm of the positive serum; A - - b ) binding ratio (BR). *, m, A) As above but the balls were precoated with 30 pg/rnl of IgG fraction of rabbit anti-NCDV hyperimmune serum and NCDV antigen was then immunologically bound to the antibodies.
Radioimmunoassay of Rotavirus
50
200
800
28 5
3200
QlA IgG - T I T E R
Fig. 2. Intra-assay variation of radioimmunoassay IgG (RIA IgG) test using Nebraska calf diarrhea virus as antigen. Each dilution of the positive and the negative control serum was tested in ten parallel determinations. --) Mean; - - -) standard deviation.
Incubation With the Labeled Antisera
Dilutions of 12sI-labeledantihuman immunoglobulins containing 1,250,2,500,5,000, 10,000 and 20,000 active cpm (see Materials and Methods) per 200 p1 were tested. As expected, the specific counts in both the positive and the negative sera were greatly increased when higher amounts of active cpm per 200 pl were used. Although the binding ratios also increased, the change was small, and therefore for anti-human alpha 5,000 active cpm per 200 PI were used, and for anti-human gamma and mu 2,500 active cpm per 200 pld were used. Assay Results
The binding ratios in the RIA IgA and in the RIA IgG tests were consistently high, varying from 5 to 15, occasionally being 20 or more. In the RIA IgM test the binding ratios rarely exceeded 3, owing to lower binding of specific immunoglobulins to the solid phase. The backround levels for all three tests were reproducibly low, being slightly higher for the RIA IgM test than for the RIA IgA and RIA IgG tests. The intra-assay variation was studied by testing a positive and a negative control serum in ten parallel determinations (Fig. 2). The largest coefficient of variation values were k 9.9 for alpha and 9.1 for gamma. A comparison of the CF test and the RIA IgG test is shown in Figure 3 . All specimens negative by radioimmunoassay were also negative by the CF test. On the other hand 34 sera negative by CF had a titer of 50 or more in the RIA IgG test. The RIA IgG titers were 50-100 times as high as the CF titers.
*
286
Sarkkinen, Meurman, and Halonen
. . :x
3200 1600
..
1
r U
4
8 16 CF - T I T E R
$8
# .
32
64
128
Fig. 3. Comparison of antibody titers to human rotavirus obtained by complement fixation (CF) and radioimmUnoaSsay IgG (RIA IgG) tests using Nebraska calf diarrhea virus as antigen.
TABLE 1. Complement Fixation (CF), Radioimmunoassay (RIA) IgA, RIA IgG, and RIA IgM Titers in Serum and Electron Microscopy (EM) From Stools in Eight Patients With Rotavirus Gastroenteritis
Patient
Days after onset of disease
Titer EM
CF
RIA IgA
RIA IgG
3 13
+
Solid-Phase Radioimmunoassay of IgA, IgG, and IgM Antibodies to Human Rotavirus Hannu K. Sarkkinen, Olli H. Meurman, and Pekka E. Halonen Department of Virology, University of Turku, Turku, Finland
A solid-phase radioimmunoassay (RIA) has been developed for t h e detection of human rotavirus-specific IgA, IgG, and IgM antibodies. Nebraska calf diarrhea virus grown in LLC-MK2 cell cultures in the presence of trypsin was directly adsorbed o n t o polystyrene balls, and antibodies that attached t o the virus-coated balls were detected by subsequent binding of 125 I-labeled antibodies specific t o human alpha, gamma or mu chains of human IgA, IgG, or IgM immunoglobulins, A total of 116 serum specimens from 58 adult patients were tested. Binding ratios between the positive and t h e negative serum varied between 5 and 15, occasionally being 20 o r more in the IgA and IgG assays, but rarely exceeding 3 in t h e IgM assay. T h e RIA was found to be more sensitive in detecting antibodies t o rotavirus than the complement fixation (CF) test, the RIA titers obtained being 50-100 times as high as the CF titers. The method described offers a possibility of evaluating the immune response t o human rotavirus and of detecting recent infection. Key words: radioimmunoassay, IgA, IgG, IgM, rotavirus, gastroenteritis
INTRODUCTION
Rotavirus is t h e major etiologic agent of transmissible gastroenteritis in young children [Kapikian et al, 19761 but has recently been shown to affect also older children [Hara et al, 19761 and adults [von Bonsdorff, et a1 1978, Bolivar et al, 1978, Meurman et al, 1977a, qrstavik et al, 1976bl . Human rotavirus has not been successfully propagated in continuous cell lines routinely used in diagnostic laboratories [Bryden et al, 19771. This has complicated the development of diagnostic methods. In serologic tests human rotavirus obtained from stool specimens of infected children has been used as an antigen [Flewett et al, 1974, Kapikian et al, 19741. However, because many species have their own rotaviruses [Flewett et al, 19781 which can be propagated in cell cultures, animal viruses are more often used as a n
Received October 31, 1978. Address reprint requests to Dr. H.K. Sarkkinen, Department of Virology, University of Turku, Kiinamyllynkatu 10, 20520 Turku 5 2 , Finland.
0846-6615/79/0304-0281$02.00 0 1979 Alan R. Liss, Inc.
282
Sarkkinen, Meurman, and Halonen
antigen in serology. This has been possible because the rotaviruses from different species have a common group antigen [Woode et al, 19761 which cross-reacts in serologic tests. Bovine rotavirus has been used as an antigen in complement fixation (CF) and indirect fluorescence antibody tests [Kapikian et al, 1975, qrstavik et al, 1976al and SA-11 (vervet monkey rotavirus) in radioiinmunoassay (RIA) [Watanabe et al, 19771 for human rotavirus antibody. Recently, an enzyme-linked immunosorbent assay (ELISA) using antigen obtained from the stools of calves infected with human rotavirus was developed [Yolken et al, 19781. The present report describes a solid-phase radioimmunoassay for detecting human IgA, IgG, and IgM antibodies to rotavirus employing Nebraska calf diarrhea virus (NCDV) directly adsorbed to the solid phase as an antigen. MATERIALS AND METHODS Antigen Production
The Lincoln strain of Nebraska calf diarrhea virus (kindly supplied by Dr. T. Johnson, Malmo, Sweden) was grown in cultures of LLC-MK2 cells according to Babiuk et a1 [ 19771 . The 12th passage of NCDV in LLC-MK2 cells was used to prepare the antigen. Briefly, confluent cultures of LLC-MK2 cells in Roux bottles were infected with 3 ml of NCDV virus solution in basal medium Eagle (BME) in the presence of 10 pg of trypsin (Difco Laboratories, Detroit, Michigan, 1 : 250) per ml. After a one-hour adsorption period at t37"C, BME containing 0.5% bovine serum albumin (BSA) fraction V (Armour Pharmaceutical Company Ltd., Eastbourne, England) and 5 pg of trypsin per milliliter were added. When the cells showed extensive cytopathic effect (CPE), usually in 48-72 hours, the supernatant and the cells were harvested and pooled. The pooled material was subjected to two freeze-thaw cycles followed by centrifugation at 7,OOOg in a Sorwall GSA rotor for ten minutes. The pellet was discarded and the supernatant further centrifuged at 25,000 rpm for two hours in a Spinco SW 27 rotor. The remaining pellet was resuspended in a small volume of phosphate-buffered saline (PBS), pH 7.35, and checked for virus by electron microscopy (EM). The protein content was determined by the method of Lowry et al [ 195 11 and varied from 1,000 to 1,800 pg/ml. Specimens
A total of 116 serum specimens were collected from 58 army trainees attending the medical ward for acute gastroenteritis. Acute rotavirus gastroenteritis was diagnosed in eight patients by electron microscopy and/or the complement-fixation test. Antisera
Porcine antisera to the heavy chains of human immunoglobulins were purchased from Orion Diagnostica (Helsinki, Finland). The heavy-chain-specific antibodies were isolated by immunoadsorbent column chromatography and iodinated with 12' I as reported in detail elsewhere [Ziola et al, 19781 . The specific activities varied from 5 to 20 pCi/pg. RIA Procedure
The antigen was adsorbed onto polystyrene balls (6.4 mm in diameter, Precision Plastic Ball Co., Chicago) by incubation of the untreated balls in an antigen solution containing 10 pg of protein per milliliter for six hours at room temperature. Carbonate buffer, pH 9.6, was used as a diluent [Voller, 19761. After adsorption, excess antigen was aspirated off and carbonate buffer, pH 9.6, containing 15 mg of bovine serum albumin fraction V per milli-
Radioimmunoassay of Rotavirus
283
liter was added. After overnight incubation at room temperature, the balls were washed twice in PBS, air-dried, and stored at t4OC until used. Fourfold serial dilutions of serum specimens in 20O-pl aliquots were pipetted into disposable plastic tubes. The serum samples were diluted in PBS containing 0.5% BSA, 0.5% Tween 20, and 0.1%NaN3 (PBS-tween) as described by Meurman et a1 [1978]. A polystyrene ball with adsorbed antigen was then added to each tube. After incubating at t37"C for one hour, the serum samples were aspirated off and the balls were washed twice with 5 ml of tapwater. A 200-pl volume of '251-labeled antibodies specific to human alpha, gamma, or mu chains of human IgA, IgG, or IgM immunoglobulins was then added to each tube. As a dilution buffer for labeled antisera, Eagle's minimum essential medium (MEM), supplemented with 0.5% lactalbumin hydrolysate, 10%heat inactivated calf serum, 1% Tween 20, and 0.1% NaN,, was used [Meurman et al, 19781. After incubating at t37"C for one hour, the radioactive solutions were aspirated off and the balls washed as described above. The balls were then placed in clean tubes and counted in an LKB 1280 gamma counter. Buffer blanks (balls incubated in PBS-tween only) and positive and negative control sera were included in each assay. The assay was standardized by diluting the iodinated anti-human alpha antibodies to a concentration which gave 5,000 cprn bound (5,000 active cpm), and the anti-human gamma and mu antibodies to a concentration which gave 2,500 cpm bound, when 200 p1 of the antisera were incubated with a ball adsorbed with 2 pg of purified human IgA, IgG, and IgM. The RIA results were evaluated by means of binding ratios which were calculated by dividing the counts per minute of the serum specimens by the counts per minute of the negative serum at the same dilution. Binding ratios higher than 2 were considered positive, with the proviso that the counts per minute of the test specimen should be at least 150. Endpoint titers were calculated from the counts per minute/dilutions curve of each serum specimen. In tables and figures, titer values are expressed as reciprocals of dilutions. OTHER TESTS
The complement fixation test was performed with the standardized microtechnique [Casey 19651. NCDV grown in primary calf kidney cell cultures was used as an antigen. For electron microscopy stool suspensions were made approximately 10%in PBS. Small aliquots of stool specimens were put on carbon-coated Formvar grids, negatively stained with 2% phosphotungstic acid, pH 6.5, and examined with a Siemens Elmiscope electron microscope. Specimens from the antigen preparations were stained and examined similarly. RESULTS
Antigen Production
In preliminary experiments human rotavirus purified from stools from infected children was used as the antigen in the solid-phase radioimmunoassay. However, it was difficult to obtain sufficient amounts of rotavirus-positive stools for antigen production. This problem was overcome by exploiting the finding of Babiuk et al [ 19771 that bovine rotavirus replicates and can be passaged in continuous cell lines when small amounts of trypsin are present throughout the whole growth period. At first BSC-1 cell cultures were
284
Sarkkinen, Meurman, and Halonen
used, but we found it more convenient to prepare the antigen in LLC-MK2 cells.
Adsorption of Antigen to the Solid Phase
In initial experiments an IgG fraction of rabbit anti-NCDV hyperimmune serum was first used to coat the balls and NCDV antigen was then immunologically bound to the specific antibodies. However, in subsequent studies NCDV antigen was directly adsorbed to the solid phase. The results of the effect of the antigen concentration on the adsorption are shown in Figure 1. The best results were obtained when the balls were incubated in an antigen solution containing 10 pg of protein per milliliter. If lower antigen concentrations were used, the binding ratios decreased, and no further increase in binding ratios could be obtained if higher antigen concentrations were used. Figure 1 also demonstrates the fact that no benefit was gained if an IgG fraction of hyperimmune serum made against NCDV was used as precoating material on the solid phase. The effect of pH on the adsorption was not critical. When the pH of the adsorption buffer was kept between 7.35 and 9.6 no significant differences were found. Lowering the pH to 5.6 or raising it to 1 2 caused a clear decrease both in specific counts and in binding ratios. Throughout this study carbonate buffer, pH 9.6, was used.
CPI
1501
loo(
501
1 5
C
Fig. 1. Effect of protein concentration on adsorption of semipurified Nebraska calf diarrhea virus (NCDV) antigen on polystyrene bails. The balls adsorbed at each NCDV antigen concentration were incubated with 1:50 dilution of rotavirus antibody-positive and -negative serum and tested for the binding of IgA antibodies to rotavirus. Values given are means of three parallel determinations. 0 - 0 ) counts per minute (cpm) of the negative serum; o---o) cpm of the positive serum; A - - b ) binding ratio (BR). *, m, A) As above but the balls were precoated with 30 pg/rnl of IgG fraction of rabbit anti-NCDV hyperimmune serum and NCDV antigen was then immunologically bound to the antibodies.
Radioimmunoassay of Rotavirus
50
200
800
28 5
3200
QlA IgG - T I T E R
Fig. 2. Intra-assay variation of radioimmunoassay IgG (RIA IgG) test using Nebraska calf diarrhea virus as antigen. Each dilution of the positive and the negative control serum was tested in ten parallel determinations. --) Mean; - - -) standard deviation.
Incubation With the Labeled Antisera
Dilutions of 12sI-labeledantihuman immunoglobulins containing 1,250,2,500,5,000, 10,000 and 20,000 active cpm (see Materials and Methods) per 200 p1 were tested. As expected, the specific counts in both the positive and the negative sera were greatly increased when higher amounts of active cpm per 200 pl were used. Although the binding ratios also increased, the change was small, and therefore for anti-human alpha 5,000 active cpm per 200 PI were used, and for anti-human gamma and mu 2,500 active cpm per 200 pld were used. Assay Results
The binding ratios in the RIA IgA and in the RIA IgG tests were consistently high, varying from 5 to 15, occasionally being 20 or more. In the RIA IgM test the binding ratios rarely exceeded 3, owing to lower binding of specific immunoglobulins to the solid phase. The backround levels for all three tests were reproducibly low, being slightly higher for the RIA IgM test than for the RIA IgA and RIA IgG tests. The intra-assay variation was studied by testing a positive and a negative control serum in ten parallel determinations (Fig. 2). The largest coefficient of variation values were k 9.9 for alpha and 9.1 for gamma. A comparison of the CF test and the RIA IgG test is shown in Figure 3 . All specimens negative by radioimmunoassay were also negative by the CF test. On the other hand 34 sera negative by CF had a titer of 50 or more in the RIA IgG test. The RIA IgG titers were 50-100 times as high as the CF titers.
*
286
Sarkkinen, Meurman, and Halonen
. . :x
3200 1600
..
1
r U
4
8 16 CF - T I T E R
$8
# .
32
64
128
Fig. 3. Comparison of antibody titers to human rotavirus obtained by complement fixation (CF) and radioimmUnoaSsay IgG (RIA IgG) tests using Nebraska calf diarrhea virus as antigen.
TABLE 1. Complement Fixation (CF), Radioimmunoassay (RIA) IgA, RIA IgG, and RIA IgM Titers in Serum and Electron Microscopy (EM) From Stools in Eight Patients With Rotavirus Gastroenteritis
Patient
Days after onset of disease
Titer EM
CF
RIA IgA
RIA IgG
3 13
+
