0095-1137/82/050752-05$02.00/0
Indirect Hemagglutination
Test
That Uses
Glutaraldehyde-Fixed Sheep Erythrocytes Sensitized with
Extract
Antigens
for Detection of Pasteurella
Antibody
TAKUO SAWADA,* RICHARD B. RIMLER,ANDKEITH R. RHOADES
National Animal Disease Center, Agricultural Research, Science and EducationAdministration, U.S.
Department
of Agriculture,
Ames,Iowa50010Received 23 July1981/Accepted 29 December 1981
Glutaraldehyde-fixed sheep erythrocytes (GA-SRBC)wereused in the indirect
hemagglutination testfor the detection of Pasteurella antibody. GA-SRBCwere
stable for at least 6 months. Heat extract or potassium thiocyanate extract
antigens of Pasteurella strains could be adsorbed ontoGA-SRBCortanned
GA-SRBC, respectively. The indirect hemagglutination test reaction was capsular
group specific with heat extractantigen-sensitized GA-SRBC but not potassium
thiocyanate extract antigen-sensitized tanned GA-SRBC.
Indirect
hemagglutination
(IHA) tests havebeen employed
successfully for
theidentifica-tion of
serological
typesof Pasteurella
multo-cida
(4) and P.haemolytica
(1). Inaddition,
Carter and Rappay
(5)
havedescribed
amodified
IHA test
that
usesformalinized
group 0 humanerythrocytes (RBC).
Although
group 0 humanRBC
havebeen
found to be the mostsatisfac-tory
erythrocytes
for
theserotyping of
P. multo-cida(4)strains, they
aredifficult for
somelabo-ratories to obtain.
Also,
formalinized,
sensitizedRBC have a tendency to clump and thus become
unsuitable
for
the assayof
antibody
(6).Na-mioka
and Murata (10) havedemonstrated
thatsheep RBC (SRBC)
canbesubstituted for
groupO human
RBC. Mukker
andNilakantan
(8, 9)
have
described
an IHA test that uses tannicacid-treated SRBC
to detectantibody
in cattleimmunized with
P.multocida.
Noattemptto useSRBC fixed
with agents other thanFormalin
inthe IHA test to
detect Pasteurella
antibody
hasbeen
reported.
The purpose
of
this report is to describe amodified IHA test that uses
glutaraldehyde-fixed SRBC
(GA-SRBC)
for the detection ofPasteurella
antibody.
MATERIALSANDMETHODS
Bacteria. Fourstrains of P. multocida andonestrain
ofP. haemolytica were used. Strains 1059 and P-3827ofP. multocida,isolatedfromaturkeythatdied offowlcholera andfromabovinethat hadpneumonia, respectively, areclassified inCarter'scapsulargroup A. Strains P-1256 and P-1235 of P.multocida,
respec-tively isolated fromabuffalo andabovinethatdied of hemorrhagicsepticemia, areincapsulargroups B and
E,respectively. Strain L-101 ofP. haemolytica
sero-type 1wasisolated from thelungofacalf that diedof
acuterespiratorydisease. Colonies of all strains grown
on dextrosestarch agar (DSA) plateswereiridescent in obliquetransmittedlight, indicating that the bacteria
wereencapsulated. P.multocidaP-3827alsoproduced
mucoid colonies.
Rabbit antisera. We prepared anti-P. multocida
group A serumby immunizingyoungadult maleNew
Zealand whiterabbits with Formalin-killed P.
multo-cidaP-1059cells. The rabbitswereinoculated subcu-taneouslywith 0.1 mlofacellsuspension (equivalent indensityto aMcFarlandnephelometer standardof1)
preparedfroma4-h culturegrown onDSA.
Thereaf-ter,therabbitswereinoculated intravenouslyat
week-ly intervals for5weeks:weadministered0.2, 0.3, 0.4,
0.5, and0.6mlatweeks1,2,3, 4,and 5,respectively. Serawerecollected from the rabbits1 week after the lastinjection and preserved with 0.01% thimerosaland
0.06%phenol at4°C. Anti-P. mlultocida groupB and
groupE sera, described previously(12), were used.
Anti-P. haemolytica serotype 1 and serotype 2 sera were kindly providedby G. H. Frank, National Ani-mal Disease Center. The sera were prepared as de-scribedby Frank and Wessman(7).
Heat extractantigen. The cells of P. multocida
P-3827, P-1256 and P-1235 and P. haemolytica L-101,
grown onheavily seeded DSA in standard petri plates
for 18 h at 37°C, were harvested in 1 ml of 0.02 M
phosphate-bufferedsaline(PBS; pH 7.2)perplate.The cellsuspensionofP.multocida P-3827 was mixed with
anequalvolumeof PBScontaining200 Uof
hyaluroni-dase (bovine testis; Calbiochem) per ml, and this mixture was incubatedat37°C for2h withoccasional
shaking to remove hyaluronic acid, which interferes with the IHA test. All of the cell suspensions were heated at 100°C for 1 h. The cells were pelleted by centrifugation at 10,000 x g for 20 min, and the supernatantfluidwasdesignatedheatextractantigen.
The extractantigenswerepreservedwith0.1%sodium
azideandkeptat4°C.
KSCNextractantigen.Cells harvested in PBS from
DSA culturesasdescribed abovewerewashed twice with PBSbycentrifugation.The washedcells,except
for P. multocida P-3827cells, weresuspended in 0.5
752
on February 7, 2020 by guest
http://jcm.asm.org/
ml of PBS per plate and mixed withanequal volume of PBS containing 2.0 M potassium thiocyanate (KSCN) (pH 7.0). BecauseP.multocida P-3827 cells couldnot
beeasily suspended inonly 0.5mlofPBS, theywere
suspended in 1.5 ml and then treated in the same
manner as the other cell suspensions. Each mixture was stirred at4°C for 2 h. The cellswerepelleted by centrifugationat10,000x gfor20min (orat14,500 x
g for 40 min forP.multocidaP-3827), and the superna-tant fluid was removed by aspiration. The pelleted cellsofeachplate were suspendedin 1.0 ml of PBS containing1.0 MKSCN(pH 7.0) and stirredat4°C for
2 h. The cells were pelleted by centrifugation as
described above, and the supernatant fluids from the first and second extractions were combined. The
ex-tractedantigenswerefilteredthrougha0.45-,um
mem-branefilter(MilliporeCorp.) and dialyzed exhaustive-ly against PBS. The antigens were preserved with 0.1% sodium azide andkeptat4°C.
Fixation of SRBC. SRBC were treated with glutar-aldehyde byamodification of the method ofBingetal. (3). A 100-ml suspension of fresh SRBC in Alsever solutionwaswashedby centrifugation (650x gfor20
min) six times with fiveorsix volumes of saline(0.85%
NaCl). After the last wash, the packed cells were
suspended in PBStoyielda10%suspension (vol/vol)
and chilledto4°C in an ice bath. A 25% solution of glutaraldehyde (Eastman Kodak Co.)was diluted to
1% (vol/vol) with PBS and chilled to 4°C. A 10% suspension of washed SRBCwasmixed withanequal
volume of the1% solution ofglutaraldehyde, and the mixture was incubatedat 4°Cfor 30 min withgentle
stirring.The mixturewasthencentrifugedat650 x g
for 10 min at 25°C. The pelleted, fixed cells were
suspended inPBS, washed three times with PBS by
centrifugation,andsuspendedinPBScontaining0.1%
sodium azide to yield a 10% suspension. The GA-SRBCwerestoredat4°C.
T-GA-SRBC. The GA-SRBC tobe sensitizedwith
protein antigens of the KSCN extract were treated withtannic acid.A10% suspension of GA-SRBC was mixed withanequal volume of PBS containing 0.005% tannic acid(wt/vol), and the mixture was incubated at 37°C for30min with occasional shaking. The tanned GA-SRBC(T-GA-SRBC)werepelleted by centrifuga-tion at650 x gfor10 minat 25°C and washed three times with PBS. After the last wash, the T-GA-SRBC
weresuspended in PBS to yielda10% suspension. SensitizationofGA-SRBCwith heat extractantigen.
A 10% suspension of GA-SRBC was mixed with an equal volume of a serial twofold dilution of heat
extractantigen. The mixturewasincubatedat37°C for
1hwith occasionalshaking. The sensitized cells were washed three times with PBS by centrifugation and suspended in PBS containing 0.25% bovine serum
albumin(BSA-PBS)and0.1%sodium azide to yield a 0.5% suspension (vol/vol). The optimal dilution of antigen forthe sensitization of GA-SRBC was deter-mined by box titration against homologous rabbit
antiserum. Thedilution of antigen used in the IHA test was the fourfold-higher (4 hemagglutination units)
concentration thatshowed thehighest specific titer in
theantiserum (1 hemagglutination unit). Fresh SRBC washed three times with PBS were sensitized with heatextractantigen ofP.multocida P-1256 or P-1235,
and abox titration was done as described above for comparisonwithsensitizedGA-SRBC.
SensitizationofT-GA-SRBCwithKSCNextract anti-gen. A 10% suspension of T-GA-SRBC was mixed with an equal volume ofa serial twofold dilution of KSCN extractantigen. The mixturewasincubatedat
37°C for 30 min withoccasional shaking. The sensi-tized cellswerewashed threetimeswithBSA-PBSby centrifugation andsuspended in BSA-PBS to yielda
0.5% suspension (vol/vol). The optimal dilution of antigen for the sensitization of T-GA-SRBCwas deter-minedby box titration.
IHA test. The IHA test was performed with a
microtiter system (DynatechLaboratories, Inc.). Seri-altwofold dilutions of antiserumwere made in BSA-PBS, and 0.025 mlof the sensitized SRBCwasadded
to 0.025 ml of the antiserum dilution in U-bottom plates. The plates were shaken and allowed to stand for1 to 2hat25°C before SRBC settling patternswere
read. TheIHAtiterwasexpressedasthereciprocal of the highestdilution of serumshowing adefinite posi-tive pattern (flat sediment), as compared with the pattern of the negative control (smooth dot in the centerof the well). Controls consisted of unsensitized SRBC plus test serum and sensitized SRBC plus diluent. Whenheterophile antibodiesweredetected in sera,they were removed by absorption with unsensi-tized GA-SRBCat25°C for2hbeforebeing used inthe IHAtest.
RESULTS
Antigen titration.
Initial experiments
con-firmed
that a heat extractantigen of
P.multo-cida
P-1256 or P-1235 wasadsorbed
ontofresh
SRBC and that the sensitized SRBC
werespecif-ically agglutinated
byrabbit
antiserum. TheGA-SRBC sensitized with
heatextractantigen of
P.multocida
P-1256 or P-1235 showed the sameresults
in box titration
as were shown byfresh
SRBC. Typical results of a box titration of heat
extract
antigen-sensitized GA-SRBC
are shownin Table 1. An excess
of antigen (dilutions of
1:16 and
below)
slightly inhibited
IHA (P.mul-tocida
P-1235and
P.haemolytica
L-101[Table
1]). Almost identical titers
wereobtained with
the same
antisera when several different lots of
sensitized GA-SRBC
wereprepared and tested
during the
courseof
the experiment. TheGA-SRBC
werestable
for
at least 7 months andadsorbed the
heat extract antigens ofPasteur-ella
strains. Controls of unsensitized GA-SRBCin each
dilution
of the antiserum and controls ofGA-SRBC sensitized with antigen
indiluent
showed definite negative patterns.
Typical
results of a box titration of KSCNextract
antigen-sensitized
T-GA-SRBC areshown in Table 2. The antibody titer of rabbit anti-P. multocida group A serum was low
against
the KSCN extract antigen of P.multo-cidaP-3827, and this result differed from that for
the
antigen of
other Pasteurella strains. Identi-calantibody
titers were obtained with the sameserawhen several different lots of sensitized
T-GA-SRBC
were prepared and tested. For at 5,on February 7, 2020 by guest
http://jcm.asm.org/
TABLE 1. Typical results of box titration of
GA-SRBC sensitized withheatextract antigens and
homologous rabbitantisera
IHAtiter inantiserum againsttheantigen ofa
Antigen P.multocida P.
dilution haemolytica
P-3827(A)b P-1256 (B) P-1235 (E) L-101 (1)
1:1 1,024 NDc ND 512
1:2 1,024 ND ND 512
1:4 1,024 ND ND 1,024
1:8 512 1,024 2,048 1,024
1:16 256 1,024 2,048 2,048
1:32 64 1,024 4,096 4,096
1:64 <4 1,024 4,096 4,096
1:128 <4 512 4,096 4,096
1:256 <4 <4 4,096 4,096
1:512 <4 <4 <4 1,024
1:1,024 <4 <4 <4 64
aTitersareexpressedasthereciprocals oftheinitial
dilutions ofserum.
bLetter inparentheses indicates the capsulargroup;
number indicates theserotype.
c ND, Notdone.
TABLE 2. Typicalresults ofbox titration of
T-GA-SRBC sensitized with KSCNextractantigenand
homologous rabbit antisera
IHAtiter in antiserumagainst theantigenofa:
Antigen P.multocida P.
dilution haemolytica
P-3827(A)b P-1256(B) P-1235 (E) L-101(1)
1:1 32 1,024 4,096 2,048
1:2 32 1,024 4,096 2,048
1:4 32 1,024 4,096 2,048
1:8 32 1,024 4,096 2,048
1:16 16 1,024 4,096 1,024
1:32 8 512 4,096 64
1:64 <4 256 <4 64
1:128 <4 <4 <4 32
1:256 <4 <4 <4 <4
aTitersareexpressedasthereciprocalsoftheinitial
dilutions ofserum.
bLetter inparenthesesindicatesthecapsulargroup;
number indicates the serotype.
least 6 months duringthe courseof the
experi-ment,GA-SRBCwereabletoadsorb the KSCN
extractantigen of Pasteurella strains satisfacto-rily after the treatment with tannic acid. Con-trols in the IHA test always showed definite
negativepatterns.Antigen-sensitized GA-SRBC
orT-GA-SRBC could be used foratleast 1week inthe IHAtestwithoutanylossof
agglutinabili-tyby homologous rabbitantiserum.
Specificityof the IHAreaction. IHAcrosstests
were performed to determine the specificity of
thereaction in the IHAtestwithheatorKSCN
extract antigen. Capsular group-specific
reac-TABLE 3. IHA cross testof heatextract
antigen-sensitizedGA-SRBC with rabbit antiserum
IHA titer inantiserumofb:
Antigena P.multocida P.
G haemolytica
GroupA GroupB Group E serotype1
P-3827(A)c 2,048 <4 <4 <4 P-1256(B) <4 2,048 <4 <4 P-1235(E) <4 <4 8,192 <4
L-101(1) <4 <4 <4 8,192
aTheoptimal dilution of each antigenwasused for
the sensitization ofGA-SRBC.
bTitersareexpressed as the reciprocals ofthe final dilutions ofserum.
cLetter inparentheses indicates the capsular group;
number indicates the serotype.
TABLE 4. IHA crosstest of KSCNextract antigen-sensitizedT-GA-SRBC with rabbitantiserum
IHA titer in antiserumofb:
Antigena P.multocida P.
haemolytica
GroupA GroupB Group E serotype1
P-3827(A)c 64 <4 <4 <4 P-1256(B) 256 2,048 64 8,192 P-1235(E) 32 64 8,192 1,024 L-101 (1) 512 256 512 4,096
aTheoptimal dilution of each antigen was used for thesensitization ofT-GA-SRBC.
bTitersareexpressed asthe reciprocalsofthe final dilutions ofserum.
cLetter in parentheses indicates the capsular group; numberindicates the serotype.
tions were demonstrated for the Pasteurella
strains(Table 3).GA-SRBC sensitized withheat extract antigen ofP. haemolytica L-101
(sero-type 1) did not react with rabbit anti-type 2 serum(datanotshown).
The results of IHA cross tests of KSCN extract antigen-sensitized T-GA-SRBC are shown in Table 4. Capsulargroup-specific
reac-tions in IHA tests were not seen, with the
exception ofP. multocida P-3827. The KSCN
extract antigen ofP. multocida P-3827 showed
negative
reactionsagainst
heterologous
antise-rumandalowpositivereactionagainst
homolo-gous antiserum. The IHA titers for all strains
weregenerally
higher
inhomologous
reactionsthanin
heterologous
reactions.When GA-SRBC that had not been treated
with tannic acid were sensitized with KSCN
extractantigenof either P. multocida P-1256 or
P-1235, an IHA titer of1,024 or 2,048,
respec-tively,
wasobtainedagainst
homologous
antise-rum. No reactions occurred with heterologous
antisera.
on February 7, 2020 by guest
http://jcm.asm.org/
IHA TEST WITH GA-SRBC 755
DISCUSSION
The
GA-SRBC
can be used for at least 7months when stored at
4°C.
This saves time thatwouldbe spent in preparing fresh SRBC
suspen-sions
and increases
thereproducibility of
thetest because the same RBC lot
is
used.There-fore,
the useofGA-SRBC
in the IHAtestshouldbeuseful for the
serological
study ofpasteurello-sis.
Carter and Rappay(5) have described a
modi-fied IHA test that uses formalinized group 0
human RBC.
Although
SRBC were notcom-pared
with
human RBC in ourexperiment,
SRBC
always showed either distinct
hemagglu-tinating
ornon-hemagglutinating
patterns when-everfixed
orfresh
cells were used. We believe thatSRBC
can besubstituted
completely
forgroup 0
human
RBC.Lysis
of bovine
RBC at low serum dilutionshas been observed in the IHA test
for
theserotyping of
P.haemolytica strains (1).
Recent-ly, Frank and Wessman (7) reported that bovineRBC sensitized with the P.
haemolytica
type 12isolate
arehemolyzed in
theIHAreaction,
mak-ing
interpretation difficult. Frank
and Wessmannoted
thathemolysis
occurs even when theantiserum has
been
heatinactivated
andad-sorbed with bovine RBC.
Iffixed SRBC
areemployed in
the IHA testfor the
serotyping
of P.haemolytica
strains, hemolysis
canpossibly
be
eliminated.
Studies have shown that
formalinized
cellstend to
clump after
storagefor
6 months to 1year at
-20°C,
thereby becoming
unsuitablefor
antibody
assays. Inaddition,
uniformsuspen-sions of formalinized
cells aredifficult
to pro-duce(3, 6).
GA-SRBCprepared
and stored at4°C
in the presentexperiment
consistently
andeasily produced
auniform
cellsuspension
whenstirred,
andunsensitized GA-SRBC
orT-GA-SRBC
always produced definite negative
pat-terns
(smooth dot
in the centerof
thewell)
inantiserum
ordiluent.
Heat extract
antigen-sensitized GA-SRBC
re-acted
only
withhomologous
antiserum,
confirm-ing previous
resultsobtained
when sensitizedgroup 0 human RBCwere used (12). Therefore,
webelieve that an IHA test that uses GA-SRBC
would be useful for the detection of capsular
group-specific
P. multocida antibody. The typespecificity
of the IHA reaction with the heatextract
antigen
of P. haemolytica L-101 wasdetermined only against
anti-serotype
1 andse-rotype2sera. Ourfindings suggest that this test
would be useful for the detection of the
type-specific antibody against
P.haemolytica
sero-type 1 strains in cattle: P. haemolytica serotype
1 strainspredominate in cattle (1, 2, 7).
The KSCN extract
antigen-sensitized
T-GA-SRBC
reactedwith both homologous andheter-ologous group antisera. The negative reaction of
P. multocida P-3827
antigen
withheterologous
antiserum was considered tobe due tothe low
antigenicity of the strain, as determined by the
low IHA titer of the homologous antiserum.
The IHA test was capsular group-specific
whenantigens in the KSCN extract of P.
multo-cida P-1256 or P-1235 were adsorbed onto
GA-SRBC but not T-GA-SRBC. This suggests that
protein antigens, common to P. multocida
P-1256 and P-1235, in the KSCN extracts were
adsorbed only onto T-GA-SRBC.
In this study, the results of IHA tests that
usedT-GA-SRBC sensitized with KSCN extract
antigens indicated that cross-reactions between P.
multocida and
P.haemolytica took
place.Similar cross-reactions
havebeen
reported by
Mukkur (8), who used tanned SRBC sensitized
with P. haemolytica KSCN extract
antigens
todemonstrate antibody in calves inoculated with
P. multocida KSCN extract. These
cross-reac-tions suggest that there areantigens common to
these pasteurellae that are extractable with
KSCN.
Antigens
common to thesepasteurellae
have beendemonstrated
previously by
immuno-diffusion methods
(11). Commonantigens
ex-tracted
withKSCN
and used tosensitize
T-GA-SRBC canperhaps
be used in asingle
IHA test todemonstrate antibody against either
P.multo-cida
orP.haemolytica.
ACKNOWLEDGMENT
We thank William E. Brown and Lance L. Ferguson for technicalassistance.
LITERATURECITED
1. Biberstein, E. L., M. Gills, and H. Knight. 1960. Serologi-cal typesof Pasteurellahemolytica.Cornell Vet. 50:283-300.
2. Biberstein, E.L.,and D. A.Thompson. 1966. Epidemio-logical studieson Pasteurella haemolytica in sheep. J. Comp.Pathol. 76:83-94.
3. Bing,D.H.,J. G.M.Weyand,and A.B.Stavitsky.1967. Hemagglutination with aldehyde-fixed erythrocytes for assayofantigens and antibodies. Proc. Soc. Exp. Biol. Med.124:1166-1170.
4. Carter,G. R. 1955.StudiesonPasteurella multocida.1. A hemagglutination testfor theidentification of serological types. Am. J. Vet. Res. 16:481-484.
5. Carter, G.R., and D. E. Rappay. 1962. Formalinized erythrocytes in the haemagglutination test for typing Pasteurella multocida. Brit.Vet. J. 118:289-292. 6.Daniel, T. M.1965. Observationon the antibodyresponse
of rabbits tomycobacterialantigens.J. Immunol. 95:100-108.
7. Frank, G. H., and G. E. Wessman. 1978. Rapid plate agglutination procedure for serotypingPasteurella hae-molytica.J.Clin. Microbiol. 7:142-145.
8. Mukkur, T. K. S. 1978. Immunologic and physiologic responseofcalvesinoculatedwithpotassiumthiocyanate extractof Pasteurella multocidatype A. Am. J. Vet. Res. 39:1269-1273.
9. Mukkur, T. K. S., and P. R.Nilakantan. 1969.The rela-tionshipofhemagglutinating antibody with protectionin cattleimmunized againsthemorrhagic septicemia.Cornell Vet.59:643-648.
10. Namioka, S., and M. Murata. 1964.Serologicalstudies on
VOL. 15, 1982
on February 7, 2020 by guest
http://jcm.asm.org/
Pasteurellamultocida. V.Some epizootiological findings resulting from0antigenic analysis. CornellVet. 54:520-534.
11. Prince,G.H.,and J. E. Smith.1969.Antigenicstudieson Pasteurellamultocidausingimmunodiffusion techniques.
II. Relationships with other Gram-negative species. J. Comp. Pathol. 76:315-320.
12. Rimler,R. B.1978. Coagglutination test foridentification of Pasteurella multocida associated with hemorrhagic septicemia.J. Clin. Microbiol. 7:142-145.