Radiometric Cytolysis Inhibition Assay, a New Rapid Test forNeutralizing Antibodies to Intact and Trypsin-Cleaved Poliovirus
TAPANI HOVI* AND MERJA ROIVAINEN
Enterovirus Laboratory, Department of Virology, National Public Health Institute, SF-00300 Helsinki, Finland
Received 18 October 1988/Accepted 9 January 1989
We have developed a new rapid test, the radiometric cytolysis inhibition assay (RACINA), for the deter-mination of neutralizing poliovirus antibodies. HeLa cells prelabeled with s'Cr, [3H]leucine, or, preferentially,with [3H]uridine are used as sensitive quantitative indicators of residual infectious virus. Both suspensions andmonolayer cultures of the indicator cells can be used. Neutralization of a fraction of a high-titer viruspreparation can be scored after the first replication cycle at 8 to 10 h. By lowering the incubation temperatureto 30°C, the completion of the cytolysis due to the first replication cycle of poliovirus was delayed beyond 21 h.This makes it possible to use the RACINA, unlike the standard microneutralization assay, for measuringantibodies to trypsin-cleaved polioviruses. The RACINA was found to be as sensitive as and more reproduciblethan the standard microneutralization assay in the measurement of neutralizing poliovirus antibodies. TheRACINA is a rapid and reliable test for neutralizing antibodies and in principle it may be applicable forquantitation of neutralizing antibodies to other cytolytic agents as well.
Assays for neutralizing antibodies are widely used inserosurveys for immunity to poliovirus and to several otherpathogenic microbial agents (1). Neutralization tests are alsoused in serological diagnosis of poliomyelitis and certainother viral infections, especially in cases in which other testsare impaired by cross-reactions (4). Because of the require-ments of aseptic conditions, long incubation times, andsubjective visual scoring of results, alternative methods forthese situations would be welcomed. Straightforward solid-phase immunoassays, although relatively rapid, are oftenonly marginally less tedious and measure other types ofantibody molecules as well and, therefore, have not replacedproper neutralization assays, for instance, in the measure-ment of poliovirus antibodies.
Additional requirements for the assay are raised when seraare tested for their capacity to neutralize trypsin-cleavedpolioviruses (2, 3). This is important especially in evaluatingimmunity to type 3 poliovirus infections (6, 7). In theconventional microneutralization assay, a relatively smallamount of infectious virus is reacted with serial dilutions oftest sera, and the residual infectious virus is allowed toreplicate several cycles and to cause a cytopathic effect insuitable indicator cells (1). In this system, trypsin-cleavedpolioviruses will produce, from the first replication cycle on,a normal progeny virus with intact antigenic sites. Antibod-ies that are present in the test serum and targeted totrypsin-cleavable sites in the virus may thus confuse theresults. For unbiased results, one has to inoculate thevirus-serum mixtures on monolayer cultures of the indicatorcells and, after an adsorption period, wash off the test serum.Plaque reduction tests fulfill these requirements, and wehave successfully used a plaque assay to selectively measureserum antibodies capable of neutralizing trypsin-cleavedpolioviruses. However, the assay is rather tedious and notvery sensitive (6, 7).We have been looking for alternative methods to demon-
strate the residual infectious virus in neutralization tests. Wehave applied a principle widely used in cellular immunologyfor quantification of cell killing. The release of radioactively
* Corresponding author.
labeled cell contents turned out to be an accurate measure ofpoliovirus-induced cytopathic effect. In this article, we de-scribe a rapid and sensitive neutralization assay based onradiometric quantification of virus-induced cytolysis. Wehave named the method radiometric cytolysis inhibitionassay (RACINA).
MATERIALS AND METHODSCell cultures. Vero cells, a continuous cell line of African
green monkey kidney origin, were used for virus propaga-tion. HeLa-S cells originally derived from human epithelialcancer were used as indicator cells in the neutralizationassays. We initially labeled the cells in suspension; mono-layer cultures were trypsinized, were suspended at 2 x 107to 3 x 107 cells per ml in minimal Eagle medium (MEM) forsuspension cultures (S medium) supplemented with 10%fetal calf serum (FCS) (S-10 medium), and were labeled in aroller tube (for details, see Results). Later, we found that themost practical way of labeling the cells was as follows. Aconfluent culture in an 800-ml plastic flask (7 x 107 to 12 x107 cells) was rinsed with phosphate-buffered saline and wasfed with 3 ml of the regular MEM supplemented with 10%FCS thoroughly dialyzed against phosphate-buffered saline(MEMdF) and 250 ,uCi of [3H]uridine. After 60 min at 36°C,the cells were rinsed three times with phosphate-bufferedsaline and were supplemented with 30 ml of MEMdF.
Virus strains. Poliovirus strain type 3 Saukett NPHI, asupposed derivative of type 3 poliovirus/USA/Saukett/50maintained at this institute, was propagated in Vero cells.Crude virus preparations were cleared from cell debris bylow-speed centrifugation and were used as such or afterbeing treated with trypsin (6).
Neutralization assays. Standard microneutralization (1)was carried out as described earlier (8). In the early phase ofthe development of the RACINA, test sera and the virusstocks were both diluted in Hanks balanced salt solutionsupplemented with 0.1% bovine serum albumin (H-BSA). Inthe suspension culture assay, 10 ,u.l of serum was mixed with30 ,Il of virus in Eppendorf tubes. After 1 h at 36°C andovernight at 4°C, the indicator cell suspension was added(105 cells in 100 Ftl of S medium supplemented with 1% FCS
[S-1 medium], unless otherwise indicated). After indicatedtimes at 36°C in a horizontal rotatory shaker, sets of tubeswere centrifuged for 2 min at 10,000 x g to pellet the cells.Samples of supernatant (150 pil) were taken for determina-tion of extracellular radioactivity.
In the monolayer culture assay, sera were diluted on96-well flat-bottom microdilution plates (Nunc AG,Roskilde, Denmark) at a ratio of 1:4, leaving a final volumeof 75 pul per well. Half of the volume was transferred to aparallel plate, using a multichannel Finnpipette. Thereafter,35 ,ul of virus dilution was added per well and the mixtureswere incubated as in the suspension culture assay. Afterneutralization, 5 x 104 labeled cells in 150 pul of regular MEMplus 1% FCS were added per well, and incubation wascontinued overnight at the indicated temperature. The plateswere centrifuged for 10 min at 500 x g, and 100-pul samplesof supernatant were harvested for determination of extracel-lular radioactivity. In the final mode of the assay, test sera,virus, and [3H]uridine-labeled HeLa cells were all diluted inMEM supplemented with 2% FCS.
Radioisotopes. All radioactive isotopes were from Amer-sham International plc, Amersham, United Kingdom. Spe-cific activities were as follows: sodium [51Cr]chromate (codeCJS.4), 10 to 35 mCi/mmol; [4,5-3H]leucine (code TRK 683),161 Ci/mmol; and [3H]uridine (code TRK 178), 30 Ci/mmol.51Cr was counted in an LKB-Wallac gamma counter, and thespecimens (100 to 150 pli) containing 3H isotopes were mixedwith 5 ml of Aqueous Counting Scintillant (Amersham) andcounted in an LKB-Wallac beta counter.
RESULTS
Spontaneous and poliovirus-induced release of radioactivityfrom prelabeled HeLa celis. Since release of 51Cr label fromtarget cells within a given period is successfully used as anindicator of cytotoxicity of activated T lymphocytes andnatural killer cells, we first wanted to see if this methodcould be used as a quantitative measure of the poliovirus-induced cytopathic effect. HeLa cells were incubated in S-10medium overnight at 36°C in a roller bottle to allow completeregeneration of poliovirus receptors. Serial dilutions of type3 poliovirus Saukett in 100 pul of S-1 medium were mixedwith standard amounts of 51Cr-labeled cells, and the suspen-sions were incubated at 36°C in a horizontal rotatory shaker.The virus-induced release of label was proportional to thecalculated multiplicity of infection and took place accordingto a time schedule reminiscent of the replication cycle ofpoliovirus (Fig. 1). The highest multiplicity of infection usedresulted in an extra release of 51Cr label to a level two- tothreefold that of the spontaneous release. At 21 h or after anincubation time theoretically allowing three cycles of virusreplication, the virus-dependent release of 51Cr label wasstill proportional to the input multiplicity (Fig. 1A).
Since the use of 51Cr labeling is considered somewhatinconvenient in most laboratories, we wanted to see if cellslabeled with other types of radioisotopes could be usedinstead. We first tested several labeling times and chaseperiods, using [3H]leucine. As with the 51Cr label, therelease of [3H]leucine label was very regular, and paralleltests showed little variation. However, the rate of thespontaneous release was always at least half of that seenwith the virus-infected cells (Fig. 1B).The next and final radioisotope tested was [3H]uridine.
Kinetics of spontaneous and poliovirus-induced release werecompared in two sets of HeLa cells labeled either for 1 h orovernight with [3H]uridine (10 ,uCi/ml). As with the two
*6%- /O-
E
1.0
20.
W 4
5 10 15 20Hours after tntection
FlG. 1. Spontaneous (O, W1, and A\) and poliovirus-inducedrelease of radioactivity from prelabeled HeLa cells. (A) Cells (150 x106) were labeled with 0.5 mCi of sodium [5'Cr]chromate for 1 h at36°C, were suspended after being washed at 2 x 10'/ml in S-1medium, and were distributed as 0.5-ml aliquots in Eppendorf tubescontaining various concentrations of type 3 Saukett poliovirus.After indicated incubation times, tubes were centrifuged for 1 min at10,000 x g and 400 ,ul of the supernatant was counted for extracel-lular radioactivity. Multiplicities of infection were 10 (0), 1 (à)* 0.1(O), and 0.01 (V) TCID.5. per cell. (B) Cells were labeled with[3H]fleucine for 1.5 h (O and *) or for 14 h (Qî and *) and weredivided into 105-cell aliquots. Part of the cultures was exposed totype 3 Saukett poliovirus (multiplicity of infection, 100 TCID5, percell) at time 0 (0) and another part was exposed after 10.5 h (arrow)of incubation (3) or was left uninfected (O and [j]). (C) Cells werelabeled with [3H]uridine for 1 h (O and *b) or for 12 h (Eand A),were distributed in 105-cell aliquots, and were infected at a multi-plicity of infection of 100 TCID50 per cell (- and A) or were leftuninfected (O and A). In all cases the range of two to four paralleldeterminations was minimal and within the area covered by thesymbols.
previous isotopes, the spontaneous release of cellular[3H]uridine label from uninfected HeLa cells was veryregular and after an initial rapid phase assumed a linearpattern. Infection of the cells with poliovirus type 3 Saukettresulted in a threefold increase in the release of uridine label(Fig. 1C). Labeling time of the cells did not influence theresults. Therefore, in subsequent experiments we regularlyused 1-h labeling times. In other experiments (not shown), it
FIG. 2. Inhibition by serum antibodies of poliovirus-induced
release of radioactivity from HeLa cells prelabeled with [51Cr]
chromate (A and B), [3H]leucine (C and D), or [3H]uridine (E and F).
Type 3 Saukett virus (A, C, and E, 106 TCID50; B, D, and F, i05TCID50 per tube) was incubated with the indicated serum dilutions
for h at 36°C; i05 prelabeled cells were added and the incubation
was continued for 10 h at 36°C. Serum from a rabbit immunized with
type 3 Sabin virus (E1), two human sera (* and *), a virus control
(V), and an uninfected cell control (O) were used. Duplicate tests
are shown by separate symbols if results were distinguishable from
each other.
was found that longer than 24-h incubation of infected celîs
decreased the relative proportion of the virus-specific label
released.
Inhibition of poliovirus-induced radiolabel release by neu-
tralizing antibodies. In initial neutralization experiments,
samples of two relatively high concentrations of the Saukett
virus were mixed with serial dilutions of a high-titer rabbit
antiserum to type 3 poliovirus and of a human serum
specimen. After 1 h at 36°C and overnight incubation at 4°C,a standard amount of prelabeled HeLa cells was added to
each tube. After 8 to 10 h in a rotatory shaker at 36°C, the
tubes were briefly centrifuged and extracellular radioactivity
was measured. Neutralization of the poliovirus-induced cy-
tolysis was dose dependent (Fig. 2). As expected, the
sensitivity of the test was better with the lower multiplicity
of infection. Practically identical results were obtained with
51Cr-, [3H]leucine-, and [3H]uridine-labeled HeLa cells.Normal rabbit serum, FCS, and mouse serum did not inhibit
the virus-induced release of cell contents, but both FCS andnormal rabbit serum decreased the spontaneous release by10 to 20% (not shown). Accordingly, levels of label lowerthan background were sometimes observed with the higherconcentrations of the test sera (Fig. 2).
Overnight assay in microdilution plates. For practical pur-poses, the scoring of virus neutralization after the firstreplication cycle at 8 to 10 h is cumbersome and not possiblewithin normal working hours. We therefore worked on asystem in which HeLa cells labeled in suspension with[3H]uridine were allowed to adhere on microdilution platesimmediately after being mixed with the virus. Extracellularradioactivity was measured after an overnight incubation at36°C. In pilot experiments (results not shown), it was foundthat the dose dependence of virus-induced release of[3H]uridine label was evident also in this system and that thereproducibility of the test was slightly improved by a 10-mincentrifugation at 500 x g before the extracellular fluidspecimen was taken for counting.
Serial fourfold dilutions of test sera and monoclonal anti-bodies were made on microdilution plates and were mixedwith standard virus preparations. [3H]uridine-labeled HeLacells were added, and extracellular radioactivity wascounted after 24 h. Because of the differences in the shape ofthe serum dilution curve (Fig. 3), a 50% reduction of thevirus-induced release of [3H]uridine label was adopted as theendpoint titer of a test serum for further experiments. Notinfrequently it was found that the highest dilutions of somesera, when mixed with the virus, brought about slightly morecytolysis than the virus alone and vice versa, i.e., that thewells with the highest serum concentrations often showedlower counts than the uninfected cell controls. The reasonfor these unexpected observations is not known. The phe-nomenon rarely influenced the calculation of the 50% inhi-bition endpoint titers. The small intra-assay variability al-lowed very accurate expression of the results.
Sensitivity and reproducibility of RACINA. Sixty-two hu-man serum specimens were tested with the RACINA, andthe results were compared with those of a standard micro-neutralization assay. Both tests were carried out only onceand without duplicate wells. Correlation of the results wasfairly good (Fig. 4). The titer values obtained were, however,slightly higher in the microneutralization assay than in theRACINA and consequently, some sera showing low titersin the microneutralization test appeared negative in theRACINA. On the other hand, the RACINA appeared to bea more accurate method than the microneutralization assay,as shown by the difficulty of determining the endpoint titer inseveral sera when using the latter assay.To test the reproducibility of the results, five selected
serum specimens were tested several times in both theRACINA and the microneutralization assay. Two dilutionseries were made from each serum specimen, one starting ata fivefold-greater dilution than the other. The results ob-tained with the RACINA were highly reproducible, with the1:5-prediluted specimen always yielding a titer about four-fold lower (Table 1). More variation was seen in the standardmicroneutralization assays of the same sera. Furthermore,the 1:5-prediluted specimens sometimes yielded titers ashigh as did the undiluted specimens. For two sets of dilutionsthe RACINA incubation was also carried out at 30°C, forreasons described below. RACINA titers obtained at 30°Cwere about fourfold higher than those scored at 36°C. Thus,the RACINA carried out at the lower temperature seems tobe as sensitive as the standard microneutralization test.RACINA for measuring antibodies to trypsin-cleaved polio-
DilutionFIG. 3. Determination of 50% inhibition endpoint titer of various antibody preparations. Human sera from three individuals (0, FI, and
*) (A), sera from three mice (S, E, and M) immunized with type 3 Saukett poliovirus (B), and MAbs 175 (H) and 882 (0) (C) were dilutedon microdilution plates as indicated and were mixed with 5 x 10' TCID(,) of type 3 Saukett virus. After 1 h at 36°C, followed by an overnightincubation at 4°C, 5 x 104 [3H]uridine-labeled HeLa cells were added per well. Extracellular radioactivity was determined after anotherovernight incubation at 36°C. Virus control (Y) and uninfected cell control (O) are shown. Dashed horizontal lines indicate the 50% inhibitionlevels of cytolysis. Serum titers were scored as shown by the closed arrowhead (for specimen E) in panel A. Variation between the duplicatespecimens (vertical bars) did not much impair the accuracy of titer recording (open arrowheads).
x
-
._
co
N
C
O
o
1024
512
256
128
64
32
16
8
4
<4
<4 8 32 128 5124 16 64 256 1024
Racina titer (1/x)FIG. 4. Correlation between neutralization titers obtained with
standard microneutralization assay and RACINA. Sera from inac-tivated poliovirus vaccine-immunized 7- to 24-month-old childrenwere tested against type 3 Saukett poliovirus at 36°C. Completeneutralization of 100 TCID50 of the virus in Vero cells was recordedas the endpoint titer in the microneutralization assay while theRACINA titer represents a reciprocal of serum dilution causing 50%1cinhibition of cytolysis caused by 5 x 10' TCID,(< of virus in 5 x 104[3H]uridine-labeled HeLa cells.
virus. Since the main goal of this work was to develop a testfor measuring antibodies in human sera, capable of neutral-izing trypsin-cleaved polioviruses (6, 7), we wanted to testwhether the RACINA would be a suitable method. Trypsin-cleaved poliovirus type 3 Saukett induced the release of[3H]uridine-labeled HeLa cell contents in a manner similarto that caused by its intact counterpart (Fig. 5). Antibodiesneutralizing trypsin-cleaved poliovirus type 3 Saukett couldbe readily measured if the cytolysis caused by the residualinfectious virus was scored by the end of the first replicationcycle at 8 to 12 h (data not shown). Because of the inconve-nience of the harvesting time, we wanted to see if anovernight incubation could be used instead. Pilot experi-ments had indicated that the counts released were propor-tional to the multiplicity of infection as late as at 24 h afterinfection. A small set of human sera, previously studied in aplaque reduction test (6), was tested for neutralizing anti-bodies to trypsin-cleaved poliovirus type 3 Saukett, usingthe RACINA. A fairly good correlation was found betweenthe tests (Fig. 6).
In some subsequent tests, however, unexpectedly highlevels of antibodies to the cleaved virus were measured incertain sera. Therefore, we wanted be sure that the cytolysismeasured was really caused by the input trypsin-cleavedvirus and not by a secondary replication cycle due to theintact progeny virus generated during the overnight incuba-tion. Trypsin-cleaved Saukett virus was incubated withserial dilutions of monoclonal antibodies (MAbs) 175 and 882(5) and serial dilutions of MAb 882 supplemented with a 1:10dilution of MAb 175. As expected (3), the site 1-specificMAb 175 was not able to neutralize the trypsin-cleavedvirus. However, when combined with serial dilutions ofMAb 882 it brought about a 16-fold increase in the titer of the
"Results are reciprocal of endpoint dilution. Sera were diluted on plates and dilutions were divided in half, one half for RACINA at 36°C and the other halffor microneutralization. RACINA at 30'C was made of parallel dilutions. For B specimens. the corresponding A specimen was prediluted 1:5 in a test tube beforebeing pipetted onto the plate. Two figures separated by a shill represent results with the smaller serum dilution inhibiting definitely more than 50% of label releaseand the greater dilution inhibiting definitely less than 50%.
latter antibody. This result indicated that part of the cytoly-sis scored at 24 h was indeed due to the intact progeny virus.Hence, this system could not be used for measuring antibod-ies to the trypsin-cleaved virus in human sera, since anti-genic site 1-specific antibodies that are always present mightcompromise the result.
.-I
co,o
E
._
o
>1,
._
a0
lu
ce
w
3
2
3
2
1
5 10 15 20
Hours after InfectionFIG. 5. Delay of poliovirus-induced cytolysis at lowered temper-
ature. [3H]uridine-labeled HeLa cells were exposed to standard (-)or trypsin-cleaved (El) poliovirus type 3 Saukett or were leftuninfected (O). Parallel cultures were incubated at 36 and 30'C, andextracellular radioactivity was determined as indicated. The numberof cells per well in the cultures at 36°C was about half of that at 30°C.
We then tested whether lowering the incubation tempera-ture to 30°C would delay the first replication cycle of thevirus and allow scoring of the results after an overnightincubation. Kinetics of cytolysîs (Fig. 5) indicated a signifi-cant delay in label release at 30°C. In other experiments, itwas found that virus-induced cytolysis at 30°C continued toincrease up to 31 h but the ratio between virus-induced andspontaneous release of radioactivity was not improved fromthat seen at 21 h. A test with MAbs, similar to that describedabove, indicated that cytolysis scored after 21 h at 30°Cprobably was all due to residual input trypsin-cleaved virusthat was escaping neutralization (Fig. 7).
DISCUSSIONWe have described in this article a new type of neutrali-
zation assay, based on radiometric quantification of residual
51
256x
128
. 64
*» 32-
sc 8
4 .
1 2 3
Neutralization indexFIG. 6. Correlation between levels of neutralizing antibodies to
trypsin-cleaved poliovirus type 3 Saukett measured by RACINAand plaque reduction. In the RACINA, [3H]uridine-labeled HeLacells were incubated with virus-serum dilution mixtures overnight at36°C. Results of the plaque reduction assay were expressed as thelogarithm of the reduction of stock virus plaque titer brought aboutby 1-h incubation of the virus with 1:4-diluted serum (6).
Antibody dilutionFIG. 7. Antigenic site-specific neutralization of intact (0) and trypsin-cleaved (El) poliovirus type 3 Saukett. (A) The virus was incubated
with serial dilutions of MAb 175, which is specific for the trypsin-cleavable antigenic site 1. (B) Corresponding experiment with antigenic site2-specific MAb 882. Trypsin-cleaved virus incubated with dilutions of MAb 882 each supplemented with 1/10 dilution of MAb 175 (A) isshown. Cytolysis of [3H]uridine-labeled HeLa cells was measured after 21-h incubation at 30°C. Intact (V) and trypsin-cleaved (O) viruscontrols and an uninfected cell control (O) were used.
cytopathogenic virus. We have studied several parametersinfluencing the sensitivity and reproducibility of the test,which was found to be suitable for measuring separatelyneutralizing antibodies to intact and trypsin-cleaved poliovi-ruses.
Quantification of the poliovirus-induced cytopathic effectby radiometric assay of cytolysis turned out to be a sensitiveand reproducible method. The kinetics of the virus-inducedrelease of cell contents were unaffected by the isotope usedfor labeling the cells. The 51Cr-release method widely used incytotoxicity assays was adapted and resulted in an accept-able rate of spontaneous release and very small variationbetween duplicate specimens. 51Cr-labeled cells would be a
suitable indicator system for residual infectious virus inneutralization tests in laboratories with facilities for workwith gamma-radiating isotopes. We went on searching foralternative labeling methods since low-energy beta-radiatingisotopes such as tritium are more widely accepted for normallaboratory work. While [3H]leucine-labeled cells might alsoperform satisfactorily as the indicator system, we preferred[3H]uridine as the labeling isotope because of the relativelylower rate of spontaneous release of the label. By modifyingthe labeling procedure and perhaps the subsequent incuba-tion of the cells before use in the test, it may be possible tofind conditions for obtaining cell preparations with even
lower background release, but we did not pursue the matterfurther.
It is advisable to standardize the labeling procedure as
much as possible, including the growth phase and concen-
tration of the cells, to obtain a regular level of specificactivity for the indicator cells. We found that the reliabilityof the test somewhat decreased with low-specific-activitycells. HeLa-S cells were originally selected for the indicatorsystem because we wanted to carry out both the labeling ofthe cells and the viral infection in suspension in Eppendorftubes to simplify the sampling of extracellular radioactivity.
In later experiments, we found that it was more practical tocarry out both virus neutralization and subsequent incuba-tion with the indicator cells in microdilution plates. More-over, whereas most of the experiments described in thisreport were carried out by using cells labeled in suspension,we found later that HeLa-S cells could be readily labeled asmonolayers. This observation suggests that the RACINA isapplicable to other cell types as well and, consequently, formeasuring neutralizing antibodies to cytolytic viruses notgrowing in HeLa-S cells. It is clear that the kinetics of thespontaneous and virus-induced label release have to bestudied separately for each cell type and virus strain. Thekinetics of label release from poliovirus-infected HeLa cellsseem, however, to be similar for wild-type and attenuatédstrains of all serotypes (unpublished observations).Another aspect that has to be considered is the dynamic
nature of the cytolysis. Although it is possible to score theresults as early as 8 h or as late as 24 h after infection, it isimportant to note that the neutralization titers observed fortest sera may be different at different time points. Further-more, when analyzing large numbers of sera, the timerequired for harvesting of the specimens may be long enoughto cause a significant increase in the released radioactivity inthe cells harvested last.
If the neutralization results are scored at 8 to 12 h, it ispossible to quantitate the neutralization capacity of a stan-dard dilution of the test serum by determining the amount ofresidual infectious virus with the aid of a standard curve.Unfortunately, this time schedule does not fit in normalworking hours. Another alternative is to make a series ofdilutions of the test serum and to determine the endpointdilution which significantly reduces the virus-induced cytol-ysis.
In the overnight assays we found it practical to take the50% inhibition as the endpoint titer. The dilution curves ofmost sera had a sigmoidal shape with the steepest part of the
curve in the middle. The use of higher or lower (e.g., 75 and25%) extents of inhibition as the breaking points was notattempted because of the fact that the upper and lowerplateau-phase values in some of the curves did not coincidewith virus and cell controls, respectively. On the other hand,determination of the 50% inhibition titer seemed relativelyaccurate with the fourfold dilution system used. As forpositive sera, at least one of the dilutions used gave a valueclearly differing from that of either the virus or the cellcontrol.The sensitivity of the RACINA at the regular temperature
(36°C) was found to be slightly lower than that of thestandard microneutralization assay. Interestingly, loweringthe incubation temperature to 30°C, which resulted in post-poning the virus-induced cytolysis, increased the sensitivityabout fourfold. Another factor influencing the sensitivity ofthe test was the amount of virus used in the test. A titrationcurve of the challenge virus was found to be necessary ineach test to guarantee the use of a standard amount of virusand to enable comparison of antibody titers against differentvirus preparations. While the antibody titer scores could beincreased by lowering the virus concentration from the -10750% tissue culture infective doses (TCID50) per well used inour test, this would require a longer incubation time andwould also bring about greater variation. One reason for theslightly better sensitivity but greater variability of the micro-neutralization test, compared with those of the RACINA at36°C, is most likely the smaller amount of virus used perwell.Under the conditions used in our experiments, the repro-
ducibility of the RACINA was excellent both within andbetween assays. Maximally twofold differences in the titerswere found to be due to technical variation of the test. TheRACINA, carried out at 30°C overnight, was found to besuitable for measuring antibodies to trypsin-cleaved poliovi-rus and capable of replacing the cumbersome plaque reduc-tion assay previously used for this purpose (6). This makes itpossible to study on a larger scale the effect of phenotypicchanges in virion structure on virus neutralization. Wewould like to emphasize that by now we have used theRACINA to study much larger numbers of sera than thosedescribed above and the results support the above conclu-sions about the sensitivity and reproducibility of the test.
In conclusion, we have developed a new type of virusneutralization test, the RACINA, which is rapid, reproduc-ible, and as sensitive as the standard microneutralizationassay. We used poliovirus and HeLa cells as the experimen-tal model, but in principle the test can be used to measureneutralizing antibodies to any other cytolytic virus or otheragent as well.
ACKNOWLEDGMENTS
This work was supported by grants from the Finnish Academy,the Finnish Cultural Foundation, and Sigrid Jusélius Foundation,Helsinki, and by Institut Mérieux, Lyon, France.We are grateful to Morag Ferguson for the MAbs, to Mira
Stenvik for part of the microneutralization results, to HelenaKayhty for some of the human serum specimens, and to MarkettaPiilo and Marita Stenvik for technical assistance.
LITERATURE CITED1. Albrecht, P., G. van Steenis, and A. L. van Wezel. 1984. Stan-
dardization of poliovirus neutralizing antibody test. Rev. Infect.Dis. 6(Suppl. 2):S540-S544.
2. Fricks, C. E., J. P. Icenogle, and J. M. Hogle. 1985. Trypsinsensitivity of the Sabin strain of type 1 poliovirus: cleavage sitesin virions and related particles. J. Virol. 54:856-859.
3. Icenogle, J. P., P. D. Minor, M. Ferguson, and J. M. Hogle. 1986.Modulation of humoral response to a 12-amino-acid site on thepoliovirus virion. J. Virol. 60:297-301.
4. Melnick, J. L., H. A. Wenner, and C. A. Phillips. 1979. Entero-viruses, p. 471-534. In E. H. Lennette and N. J. Schmidt (ed.),Diagnostic procedures for viral, rickettsial and chlamydial infec-tions. American Public Health Association, Washington, D.C.
5. Minor, P. D., M. Ferguson, A. Phillips, D. I. Magrath, A.Huovilainen, and T. Hovi. 1987. Conservation in vivo of proteasecleavage sites in antigenic sites of polioviruses. J. Gen. Virol.68:1857-1865.
6. Roivainen, M., and T. Hovi. 1987. Intestinal trypsin can signifi-cantly modify antigenic properties of polioviruses: implicationsfor the use of inactivated poliovirus vaccine. J. Virol. 61:3749-3753.
7. Roivainen, M., and T. Hovi. 1988. Cleavage of VP1 and modifi-cation of antigenic site 1 of type 2 polioviruses by intestinaltrypsin. J. Virol. 62:3536-3539.
8. Roivainen, M., C.-J. Thodén, M. Stenvik, T. Poyry, and T. Hovi.1987. Virus excretion and strain-specific antibody responses afteroral poliovirus vaccination in previously immunized children. J.Med. Virol. 23:249-256.
![Page 1: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/1.jpg)
![Page 2: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/2.jpg)
![Page 3: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/3.jpg)
![Page 4: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/4.jpg)
![Page 5: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/5.jpg)
![Page 6: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/6.jpg)
![Page 7: Radiometric Cytolysis Inhibition Assay, Neutralizing ...710 HOVI ANDROIVAINEN [S-1 medium], unless otherwise indicated). After indicated times at 36°C in a horizontal rotatory shaker,](https://reader034.documents.pub/reader034/viewer/2022042114/5e90cf563596f64bcf00e0c8/html5/thumbnails/7.jpg)
