Journal of Medical Virology 8:79-88 (1981)

Enzyme-Linked lmmunosorbent Assay (ELISA) for Mumps Virus Antibodies Therese Popow-Kraupp

Institute of Wrology, University of Vienna

An ELISA for the detection of mumps virus-specific IgG and IgM antibodies was developed. Three different antigen preparations were compared. Equally good results were obtained with virus concentrated by ultracentrifugation and with virus that was further purified by sucrose gradient centrifugation. Crude infected allan- toic fluid was unsuitable for use as antigen. The variability and reproducibility of the tests were satisfactory. When the ELISA was compared to conventional serolog- ical methods, a good correlation of IgG absorbance values with complement fixa- tion (CF) antibody titers was found (r = 0.574), the ELISA being more sensitive in detecting antibodies in acute-phase sera. For the determination of immunity, ELISA IgG values were compared with results obtained in hemagglutination-in- hibition (HI) and hemolysis-in-gel (HIG) tests. Again there was a good correlation with both tests (rH, = 0.528. rHl0 = 0.667). The ELISA was more sensitive than the HI and HIG test for the detection of low levels of antibodies.

Key words: ELISA, mumps, diagnosis, immunity

INTRODUCTION

of meningitis or encephalitis without evidence of parotitis. Other viral diseases may mimic the clinical picture of mumps [Zollar and Mufson, 1970; Buckley et al, 1972; Biedel, 19781. Furthermore, it may &.important to detect past exposure to mumps virus or to control success of vaccination. Conventional serological methods for the diagnosis of acute infections and for determination of immunity possess some disad- vantages: the complement fixation (CF) test is not rapid, since a fourfold rise in titer in p a i d serum samples is needed for the diagnosis of acute infections. The presence of nonspecific inhibitors makes hemagglutination-inhibition (HI) antibody assays less

The clinical diagnosis of mumps virus infection is unreliable, especially in cases

Accepted for publication June 22, 1981.

Address =print requests to Therese Popow-Kraupp, Institut f i r Virologie der Universitat Wien, Kinderspi- talgasse 15, A-1095 Wien, Austria.

0146-6615/81/0802-0079$03.00 0 1981 Alan R. Liss, Inc.

80 Popow-Kraupp

sensitive Fobbins et al, 1949; Henle, 19691. Cross-reacting antibodies to other para- myxoviruses may cause false-positive results in both tests [Lennette, 1963; Penttinen, 1967; Chanock, 19691. The neutralization test (NT) is laborious and time consuming, and depends upon tissue culture. Recently, several methods have been described that attempt to overcome these problems. However, the hemoly sis-in-gel (HIG) test [Grillner, 19761 also may show nonspecific reactions [Leinikki, 19791, and plates cannot be stored for a long time. The immunofluorescence (IF) method [Kuttner-May, 19781, which has also been camed out in this laboratory, is very sensitive and specific but depends upon tissue culture; it is technically difficult to perform and so not suited for screening large numbers of serum samples. We therefore established an enzyme immunoassay (ELISA), first described by Engvall and Perlmann [1971], for the deter- mination of class-specific antibodies to mumps virus. The objectives of this study were to find out the antigen preparation best suited for the ELISA test, to evaluate ELISA test reproducibility, and to compare ELISA results with other serological methods for detecting antibodies in acute and past mumps virus infections.

MATERIALS AND METHODS

ELISA Antigens Mumps virus-infected allantoic fluid was used as antigen source. The Enders

strain of mumps virus was propagated as described by Henle [1969]; Infected allantoic fluid (hemagglutinin titer 1:128) was diluted to lo-', and 0.2 ml of this dilution was used for egg inoculation. After 7 days at 33"C, allantoic fluid was harvested. Three preparations of mumps virus antigen were compared: 1) crude virus-containing allan- toic fluid; 2) virus pelleted by centrifugation at 60,OOOg for 1 hour at 4°C and resus- pended in approximately 1/200 of the original volume in PBS; 3) virus purified in a discontinuous sucrose gradient as described by Jensik and Silver [ 19751. Allantoic fluid from uninfected eggs was treated in the same way and served as control antigen.

Test Procedure

The tests were camed out in polystyrene U-shaped microtiterplates (Nunc, Kam- strup, Denmark). For coating, antigen and control antigen were diluted in carbonate- bicarbonate buffer, pH 9.6, to a protein content of 10 pg/ml measured by the method of Schaffner and Weissmann [1973]. Fifty p1 of antigen was placed in each well and incubated overnight at 4°C. Tests reported in this paper were camed out immediately after coating the plates. Nevertheless the plates could be stored at -80°C for several weeks without loss of activity. After coating, antigen and control antigen were re- moved and the plates were incubated with PBS, pH 7.5 (100 pl/well, containing 2% sheep serum), for 1 hour at 37°C.

In the routine diagnostic procedure, serum samples were diluted 1:100, 1:300, and 1:500 in PBS, pH 7.4, containing 2% Tween 20, and 2% sheep serum (PBS- ELISA). Plates were incubated with 50 pl/well of the serum dilutions for 2 h at 37°C. After washing the plates three times with PBS, pH 7.5, and 2% Tween 20, 50 pl/well of heavy chain specific anti-human IgG or IgM (Orion Diagnostica, Helsinki, Fin- land), diluted 150 in PBS-ELISA was added, and the plates were incubated for an- other 2 h at 37°C. Then the conjugate was washed out with PBS-Tween, and 0.1 ml of the substrate (p-nitrophenylphosphate; Sigma 104, phosphatase substrate tablets; 1 mg/ml in 10% diethanolamine buffer, pH 9.8) was added. After 30 min at mom tem-

ELISA for Mumps Virus Antibodies 81

perature the enzymatic reaction was stopped with 50 pl of 3 M NaOH. The absor- bance at 405 nm was measured by a multichannel photometer for vertical measurement (Titertec Multiscan, Flow Labs, Bonn, Federal Republic of Germany). No significant difference was found when an empty plate of the same lot or a plate coated with virus, incubated with PBS-ELISA replacing human serum, served as blank value.

Test Criteria

Serum samples were considered positive if their absorbance values exceeded the threefold value of the negative control, but absorbances had to be above a threshold level of 0.2 at a serum dilution of 1:300, or above 0.15 at a serum dilution of 1500. These threshold levels were determined by testing 13 serum samples characterized by the following criteria: 1) no past history of mumps virus infection; 2) no detectable antibodies in the HI and HIG; 3) low absorbance values in the ELISA IgG. These 13 serum samples were assayed in two different ELISA tests and threshold levels were then defined as 3 standard deviations above the mean absorbance value at a serum dilution of 1:300 and 1:500, respectively.

Other Serological Methods

Complement fixation (CF) test. Microtiter CF test with commercially avail- able mumps and parainfluenza antigens (Seromed, Munich, Federal Republic of Ger- many) was used as described by Antoniadis [1971].

Hemagglutination-inhibition (HI) test. HI antibody levels were determined using chicken erythrocytes and four units of hemagglutinin. Prior to HI testing non- specific inhibitors were removed by treatment with the receptor-destroying enzyme (RDE) of V. cholerae [Burnet, 19471.

Hemolysis-in-gel (HIG) test. Commercially available HIG plates were used (Orivir, Orion Diagnostica, Helsinki, Finland). Five microliters of undiluted inacti- vated serum was put into the wells, and the plates were sealed and incubated at 37°C for 12 h. Then the diameter of the hemolytic zone was measured.

test (RA test, Hyland Diagnostics, Illinois) in order to exclude nonspecific reactions in the IgM tests.

Rheumatoid factor. This was determined by a commerically available latex

Serum Specimens

Samples were taken from sera sent to this laboratory in 1979. The sera were stored at -20°C. There were three groups of patients: 1) Thirty-three patients had the clinical diagnosis of mumps or mumps-meningitis showing a fourfold increase of CF antibodies to mumps virus in paired serum samples. These sera were also screened for mumps-specific IgG and IgM antibodies by the ELISA technique and for CF anti- bodies to paminfluenza viruses. 2) Sixty-three patients had the clinical diagnosis of mumps, parotitis, mumps-meningitis, or meningitis serosa but without diagnostic in- creases of CF antibodies to mumps virus in paired serum samples. These sera were also tested for the presence of mumps-specific antibodies in the HI test and ELISA IgG and IgM. 3) One hundred eighteen sera, sent to this laboratory for rubella anti- body screening, were also tested for antibodies against mumps virus in the HI and HIG tests and ELISA IgG.

82 Popow-Kraupp

Statistical Analysis

the mean, coefficients of variation, t-tests, linear correlations, and two-way analysis of variance were calculated by the methods of Winer [ 197 11.

The statistical parameters, mean values, standard deviations, standard emrs of

RESULTS

Comparison of Different Antigens for the ELISA Test and Reproducibility of the Test System

Crude virus-containing allantoic fluid could not be used as it was not possible to distinguish between positive and negative sera. Besides, no difference in the reactivity of the sera with antigen and control antigen was detectable. In contrast, specific anti- gen-antibody reaction was observed with pelleted virus antigen and with discontinous gradient-purified virus antigen. In order to find out whether there is a difference in results when the sera are tested with these two different antigen preparations and to investigate the reproducibility of our test system, 13 sera were assayed for mumps- specific IgG antibodies in a dilution of 1:300: a) on different days with one and the same antigen preparation (pelleted virus antigen = PVA); b) on the same day with pelleted virus and discontinuous gradient-purified virus antigen (= GPVA); c) on dif- ferent days with pelleted and discontinuous gradient-purified virus antigen.

Results are shown in Table I. There was no significant difference between the absorbance values of the tests using PVA and GPVA as antigen (Ib). This was also confirmed when GPVA was compared to PVA in tests cmied out on different days (Ic) *

Our results also exhibited the same degree of reproducibility when sera were tested with the same antigen preparation on different days (Ia) or with two different antigen preparations on the same and on different days (Ib, Ic). Since ELISA results were similar with PVA and GPVA, we decided to use the PVA in our further testing.

Variability of Absorbance Values in One Test

Three sera, one from the second day, one from the 11th day of illness, and one negative serum, were tested 24 times at a dilution of 1:300 for mumps virus-specific IgG and IgM antibodies in one and the same test. Results are shown in Table 11. It will be seen that ranges of the absorbance values of the negative, intermediate, and strongly positive sera were distinctly different. Coefficients of variation of absorbance values were fairly low for repeated tests of positive semm samples.

Comparison of ELISA and CF for Serodiagnosis of Mumps Virus Infection

Thirty-three paired sera with the clinical diagnosis of mumps and a fourfold rise of titer in the CF test were screened for mumps virus-specific IgG and IgM anti- bodies. Figure 1 shows the correlation between CF titers and ELISA IgG absorbance values at a serum dilution of 1:300. In the first serum sample only two patients who were negative in the CF test were also negative for ELISA IgG antibodies. In 14 of the 16 CF-negative sera IgG antibodies were found by the ELISA test. Thirty of the 33 patients possessed IgM antibodies (see Table HI). In the second serum sample all

TAB

LE 1

. C

ompa

riso

n of

Tw

o D

iffer

ent A

ntig

en P

repa

ratio

ns a

nd R

epro

duci

bilit

y of

ELI

SA T

est W

ith t

he S

ame

and

With

Tw

o D

iffe

rent

Ant

igen

Pre

pa-

ratio

ns (

abso

rban

ce v

alue

s of

13

sera

test

ed f

or p

rese

nce

of I

gG a

ntib

odie

s to

mum

ps v

irus

in a

dilu

tion

of 1

:300

)

Stan

dard

R

ange

of

emr o

f C

oeff

icie

nt

abso

rban

ce

Coe

ffic

ient

M

ean

the

mea

n of

var

iatio

n va

lues

1'

Sign

ifica

nce

of c

orre

latio

n Si

gnifi

canc

e

077

P<O

.O1

0.094-0.675

PVA

, day

2

0.373

0.083

0.80

2 0.130-1.134

n.s.

a)

PV

A, d

ay 1

0.323

0.04

8 0.533

0.873

1.81

1 n.

s. 0.82

P<O

.O1

b)

PVA

, day

2

0.373

0.083

0.80

2 0.130-1. I34

GPV

A, d

ay 2

0.

280

0.047

0.600

0.091-0.593

n.s.

0.

77

PcO

.01

0.094-0.675

GPV

A, d

ay 2

0.280

0.047

0.600

0.091-0.593

1.398

c)

PVA

, day

1

0.325

0.04

8 0.532

Two-

way

ana

lysi

s of

varia

nce

betw

een

all t

ests

: n.s

. PV

A =

pel

lete

d v

im an

tigen

, GPV

A =

dis

cont

inou

s gra

dien

t pur

ified

vim

ant

igen

. n.

s. =

not

sig

nific

ant (

P 0.

05).

tx =

Stu

dent

's t-t

est.

TAB

LE II.

Var

iabi

lity

of A

bsor

banc

e V

alue

s in

One

and

the

Sam

e EL

ISA

Tes

t (ea

ch o

f th

ree

diff

eren

t ser

a te

sted

24

times

for

IgG

and

IgM

an

tibod

ies a

gain

st M

umps

vir

us a

t a d

ilutio

n of

1:3

00)

Stan

dard

St

anda

rd

erro

r of

Coe

ffic

ient

R

ange

of

abso

rban

ce

Mea

n M

edia

n de

viat

ion

the

mea

n of

var

iatio

n va

lues

Empt

y m

icro

titer

pla

tes

0.005

0.005

0.004

0.001

0.776

0.000-0.014

Neg

ativ

e se

mm

0.

056

0.057

0.02

8 0.

006

0.500

0.003-0.105

Inte

rmed

iate

pos

itive

IgG

Stro

ngly

pos

itive

IgG

Inte

rmed

iate

pos

itive

IgM

Stro

ngly

pos

itive

IgM

(sec

ond

day

of i

llnes

s)

0.753

0.760

0.087

0.01

8 0.116

0.455-0.863

(1 1 t

h da

y of

illn

ess)

1.

491

1 SO

8 0.097

0.019

0.06

5 1.276- 1.679

0.04

4 0.650-0.775

(sec

ond

day of i

llnes

s)

0.703

0.694

0.03

1 0.006

(1 It

h da

y of

illn

ess)

1 .

504

1.513

0.061

0.013

0.04

1 1.401-1.604

2 P 4

00

w

84 Popow-Kraupp

1.5-

1.0 -

E E

a 0 d

w

0.5-

t POS

. - - . a

I . : "

I - 4 ' .

a .

: I = 0.574

( p <0.01)

--- I :

4 8 16 32 64 128 256 512

CF -Ti ter

Fig. 1 . Correlation of ELISA IgG absorbance values at a serum dilution of 1:3W with CF titers.

patients were positive for IgG antibodies, but the number of patients positive for IgM antibodies remained the same. In the three IgM-negative patients a fourfold rise in titer of CF antibodies to parainfluenza virus antigen could be demonstrated, indicating the presence of cross-reacting IgG antibodies. Cf titers correlated well with the absor- bance values of the ELISA IgG (r = 0.574, PcO.01) but showed no definite correla- tion with the ELISA IgM values (r = 0.24, P<0.05).

Sixty-three paired serum samples from patients with the clinical diagnosis of mumps that remained negative in the CF test (titers5 1:8) were assayed for the pres- ence of IgG and IgM antibodies in the ELISA. Of these, 41 had IgG antibodies and four also had IgM. One of the latter was from a patient showing a fourfold rise in HI titer.

Comparison of ELISA, HI, and HIG for Determination of Immunity to Mumps Virus Infection

We compared HI titers of 118 sera with ELISA IgG absorbance values at a se- rum dilution of 1:300 (Fig. 2A). In addition, 70 of these sera were also tested in the HIG test (Fig. 2B). Correlations of ELISA with both tests were significant, with a somewhat higher coefficient of correlation between HIG and ELISA (r = 0.667) than HI test and ELISA (r = 0.528). The ELISA seemed to be more sensitive than HI (Table IV), since 13 out of 21 sera (62%) without detectable HI antibodies were posi-

ELISA for Mumps Virus Antibodies 85

1.5

1 .o

E v) 0 0 W

0.5

t P O S

- 0.2.

‘i‘ A

” ..

1.5

1.0

..

. .. .. ... .. . ”

r = 0.528 0.5

(p <0.01)

t DOS

. . . . . . 10 20 40 80 160 320

HI-Ti ter

. .

r ~ 0 . 6 6 7 .. (p< 0.01 )

4 5 6 7 8 9 1 0 1 1 1 2 1 3 - neg I pos - HHIG m m B

Fig. 2. the diameters of the HIG test. A) ELISA IgG - HI titers. B) ELISA IgG - diameters of HIG test.

Correlation of ELISA IgG absorbance values at a semm dilution of 1:300 with HI titers and with

tive in the ELISA IgG test. One serum with an HI-titer of 1:20 (Fig. 2A) was nega- tive in the ELISA, but this serum was also negative in the HIG test, and 5 out of 45 sera with HI titers of 1: 10 (1 1 %) were negative in the ELISA and HIG tests. There- fore it cannot be excluded that an HI titer lower than or equal to 1:20 may represent a false-positive result.

Antibodies to mumps virus were detected in 25/3 1 (8 1 %) of the HIG-negative sera (Table IV), indicating a higher sensitivity of the ELISA test compared to HIG.

DISCUSSION

These investigations demonsmte that the ELISA test for mumps virus antibodies is a sensitive and specific method for both the rapid diagnosis of acute infections and for determination of immunity. Its sensitivity allows confirmation of the clinical diag- nosis of mumps infection from a single serum sample as soon as the second to fourth day of illness. The test is relatively simple to perform, needs only small amounts of serum, and can be used for screening of large numbers of specimens. There is no need for pretreatment of the sera, and sensitized microtiter plates can be stored at -80°C for several weeks. The reproducibility of our test when performed with different anti-

86 Popow-Kraupp

TABLE 111. Results of ELISA IgG and IgM Determinations Compared to Results of CF Test in Paired Serum Samples (numbers of sera)

ELISA-IgG-positive ELISA-lgM-positive n

CF antibodies 4-fold titer rise

CF antibodies 2 X negative (5 1:8)

33

41

30 33

4 63

TABLE IV. Results of Immunity Screening (numbers of positive i- negative sera) With ELISA IgG Compared to HIG and HI Test

HIG HI - + - +

( 2 5 mm) (cS mm) (21:lO) (< 1:lO)

+ Abs. 20 .2

ELISA 1gG

Abs. <0.2 n =

39 25 91 13

6 6 8 -

70 118

gens and on different days (Table I) was good, as differences between the tests mea- sured by t-test and analysis of variance were not significant and linear correlation studies were adequate.

tions and on different test days. The plates were sensitized separately for each test in order to get a similar variation of test conditions as in routine testing. Linear comla- tion studies revealed no significant difference between tests performed with different antigen preparations on the same day and with the same antigen on different days. This indicates that the pelleted virus antigen was equivalent to the discontinuous gradi- ent purified virus antigen, and suggests that differences between tests are possibly due to factors of test processing and adsorption of antigen.

individual wells shows marked variation [Kricka, 19801. The amount of antigen ad- sorbed per well is further influenced by the various steps of washing and incubation procedures [Salonen, 1979; Lehtonen, 19801. Therefore the coefficient of variation of our positive samples in one test that ranged from 4.1 % to 11.6% (Table II) must be considered as satisfactory. The high coefficient of variation in the measurements of empty plates and of the negative replicates is related to the low absolute values of the negative sera [Heck, 19801 and to photometric inaccuracy at low absohance values [Ruitenberg, 19801. There was a marked difference between ranges of negative, inter- mediate, and strongly positive sera.

The ELISA IgM test proved to be useful in the diagnosis of acute mumps infec- tion: In the 33 paired serum samples with a fourfold increase in CF antibody titers to mumps virus (Table 111), three sera were negative for ELISA IgM. These sera also

Values in Table I are derived from tests performed with different antigen prepara-

As in other microtiter plate assays, adsorption of virus protein onto the surface of

ELISA for Mumps Virus Antibodies 87

showed a fourfold rise in titer of parainfluenza antibodies in the CF test. We therefore assume that these patients had cross-reacting IgG antibodies to parainfluenza viruses, which is a common finding [Lennette, 1963; Penttinen, 1967; Chanock, 19691. Fur- thermore, these results indicate that ELISA IgM determinations may be more specific than CF testing. In 63 CF antibody-negative serum samples we detected four IgM-pos- itive sera by the ELISA. These four patients had the clinical diagnosis of serous meningitis or serous meningitis with parotitis. The ELISA IgM results were reproduc- ible, and false-positive IgM antibody reactions due to rheumatoid factor were excluded by negative latex agglutination test. Unfortunately we did not receive any specimens for virus isolation from these patients. One of the four sera also showed a fourfold titer rise in the HI test; the remaining three had mumps HI titers of more than or equal to 20. No significant CF titers or titer rises against enteroviruses, herpes simplex virus, and herpes zoster virus could be detected. No IgM antibodies against the Ep- stein-Barr virus were detectable by the immunofluorescence method in these patients. We therefore assume that these patients had acute mumps infection despite negative CF tests. ELISA IgG and IgM tests seem to be more sensitive than CF tests. This is indicated by the ability to demonstrate relatively high ELISA absorbance values in the early phase sera, which have only low levels of CF antibodies (Fig. 1) . Significant increases in the ELISA absorbance values between early phase and convalescent sera may therefore not be detectable. Correlation between CF titers and ELISA IgG absor- bances was significant (r = 0.574, P<O.Ol). This has also been shown by Daugharty [1973] in a radioimmunoassay for mumps antibodies. Ukkonen et a1 [1980] found a better correlation (r = 0.9) between the two tests. This might be due to the fact that these authors used a different antigen in their CF test. Furthermore, low ELISA IgG values in early phase sera suggest that sampling of the first serum was done only in a very early phase of the disease, whereas we found relatively high specific ELISA IgG absorbance values related to low CF titers. This indicates that at least some of our first serum samples were not drawn on the first days of illness, and this may have influ- enced our correlation coefficient.

There was only a weak correlation (r = 0.24, PcO.05) between CF titers and ELISA IgM absorbance values at a serum dilution of 1:300, which was also found by Daugharty [1973]. This may be explained by the fact that IgG antibodies have a greater affinity to virus antigens and thus partly block the antigenic sites for IgM anti- bodies, especially in convalescent phase sera. The ELISA IgG test again proved to be a sensitive tool for the evaluation of past infection with mumps virus. Comparison with HIG and HI test showed good correlation with both tests, and it seems, shown by the results in Figure 2, that the ELISA has a higher capacity to detect low levels of antibodies. The sensitivity of the ELISA test also proved to be very valuable to detect seroconversion after mumps vaccination (in preparation). However, the extent of cross- reactions with antibodies to parainfluenza viruses in determinations of immunity status against mumps virus with the ELISA IgG method remains to be investigated.

ACKNOWLEDGMENTS

The author wishes to thank Mrs. E. Haugeneder and Mrs. H. Dippe for excellent technical assistance, and Dr, G. Kundi for helpful advise in statistical analysis.

88 Popow-Kraupp

REFERENCES Antoniadis G (1971): Beschreibung einer Mikromethode zur Durchfiihrung der Komplementbindungsrktion

fiir die Virusserologie. Zentralblatt fiir Bakteriologie 216:544-561. Biedel CW (1978): Recurrent mumps parotitis following natural infection and immunization. Amencan

Journal of Diseases of Children 132:678-680. Buckley JM, Poche P, McIntosh K (1972): Parotitis and parainfluenza 3 virus. American Journal of Dis-

eases of Children 124:789. Burnet FM, Stone JD (1947): The receptor-destroying enzyme of V. cholerae. Australian Journal of Experi-

mental Biology and Medical Sciences 25:227-233. Chanock RM (1969): Parainfluenza viruses. In Lennette EH, Schmidt NJ (4s): “Diagnostic Procedures for

Viral and Rickettsial Diseases.” New York: American Public Health Association, pp 434455. Daughany H, Warfield DT, Hemingway WD, Casey HL (1973): Mumps class-specific immunoglobulins in

radioimmunoassay and conventional serology. Infection and Immunity 7:380-385. Engvall E, Perlmann P (1971): Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of im-

munoglobulin G. Immunochemistry 8:871-874. Grillner L, Blomberg J (1976): Hemolysis-in-gel and neutralization tests for deterrmnation of antibodies to

mumps virus. Journal of Clinical Microbiology 4: 11-15. Heck FC, Williams JD, Pruett J (1980): Interpretation of spectrophotometric absorbance values to define

results of enzyme-linked immunosotbent assays. Journal of Clinical Microbiology 11 :398-401. Henie W (1969): Mumps virus. In Lennette EH, Schmidt NJ (eds): “Diagnostic Procedures for Viral and

Rickettsial Diseases.” New York: American Public Health Association, pp 457480. Jensik SC, Silver S (1976): Polypeptides of mumps virus. Journal of Virology 17:363-373. Kricka LJ, Carter TJN, Burt SM, Kennedy JH, Holder RL, Halliday MI, Telford ME, Wisdom GB (1980):

Variability in the adsorption properties of rmcrotitre plates used as solid supports in enzyme immu- noassay. Clinical Chemistry 26:741-744.

lipps-University Marburg, Federal Republic of Germany. Kuttner-May S (1978): Die Serodiagnose und Seroepidenuologie der Mumpsinfekion. Dissertation, Phi-

Lehtonen OP, Viljanen MK (1980): Antigen attachment in ELISA. Journal of Immunological Methods

Leinikki PO, Shekarchi I, Tzan N, Madden DL, Sever JL (1979): Evaluation of enzyme-linked immunosor- 34:6 1-70.

bent assay (ELISA) for mumps virus antibodies. Proceedings of the Society for Experimental Biol- ogy and Medicine 160:363-367.

in patients with mumps virus infection. Journal of Laboratory and Clinical Medicine 61:780-788.

cinae Experimentalis Fennicae 45:202-205. ’

diagnosis of infections by the mumps virus. Journal of Immunology 61235-242.

for through-the-plate reading of microplates to be used in enzyme-linked immunosorbent assay. Journal of Clinical Microbiology 11: 132-134.

Salonen EM, Vaheri A (1979): Immobilization of viral and mycoplasma antigens and of immunoglobulins on polystyrene surface for immunoassays. Journal of Immunological Methods 30:2O9-2 18.

Schaffner W. Weissmann C (1973): A rapid, sensitive, and specific method for the determination of protein in dilute solution. Analytical Biochemistry 56502-514.

Ukkonen P, Vaianen D, Penttinen K (1980): Enzyme-linked immunosorbent assay for mumps and pa- rainfluenza type l immunoglobulin G and immunoglobulin M antibodies. Journal of Clinical Micro- biology 11:319-323.

Lennette EH, Jensen FW, Guenther RW, Magoffin RL (1963): Serologic responses to para-influenza viruses

Penttinen K, Cantell K(1967): Parainfluenza 1 (Sendai) antibodies and mumps vaccination. Annales Medi-

Robbins FC, Kilham L, Levens JH, Enders J F (1949): An evaluation of the test for antihemagglutinin in the

Ruitenberg EJ, Sekhuis VM, Brosi BJM (1980): Some characteristics of a new multiple-channel photometer

Winer BJ (1971): “Statistical Principles in Experimental Design.” New York: McGraw-Hill, pp 1-907. Zollar ML, Mufson MA (1970): Acute parotitis associated with parainfluenza 3 virus infection. American

Journal of Diseases of Children 119: 147-148.