I eterinat3' Parasitolog)', 41 ( 1992 ) 109-120 Elsevier Science Publishers B.V., A m s t e r d a m
109
Specificity of affinity-purified Trichinella spiralis antigens Shiu-Wan Chan and Ronald C. Ko ~ Department o fZooh~:t'. University of tlong Kong, tlong Kong (Accepted 18 July 1991 )
ABSTRACT Chan, S.-W. and Ko, R.C.. 1992. Specificity of affinity-purified Trichinella spiralis antigens. I'et. Parasitol., 41 : 109-120. An affinity-purified fraction (APF) was obtained by passing crude somatic antigens of lrtchinelhz ~piralis muscle larvae through an Aft-Gel 10 column coupled with anti-7)'ichuris suis IgG. The fraclion contained seven antigens with molecular weights ranging from 28 to 55 kDa. When tested with antiserum against other common nematodes of pigs from China, the APF was found to be markedly more specific than $3 antigens (prepared by a combination of cell fractionation and differential centrifugation according to Despommier and Lacetti, 1981 ) and fractions produced by Scphacryl S-200 gel filtration ( FI to F12 ). When the APF was used in an indirect IgG--enzyme-linked immunosorbenl assay ( IgG-ELISA ) to screen serum samples from 2000 pigs imported from China, a positive rate of 7.5% was obtained. Similar screenings using the crude somatic antigens F I and $3 gave a large number of cross-reactions and false positive reactions. Positive rates of 48%, 39% and 59.5% respectively were obtained lbr the three antigens.
INTRODUCTION
Although trichinellosis is a major parasitic problem of pigs in many regions of China, few attempts have been made to develop an effective immunoscreening system which could be used in the local abattoirs ( Ko, 1984; Chan, 1986 ). The development of such a system is greatly hampered by the fact that pigs imported from China are commonly infected with numerous species of nematodes which share common antigens with Trichinella spiralis. Hong Kong imports as many as one million live pigs annually for local consumption, about 90% of which come from China. Outbreaks of human trichinellosis were only first documented in Hong Kong in 1981 (Ko, 1984). Since 1984, a research programme has been initiated in our laboratory to develop a sensitive and specific immunodiagnostic method to screen for T. spiralis antibodies in pigs slaughtered at the local abattoirs. This paper compares the ~Author to whom correspondence should be addressed.
~3 1992 Elsevier Science Publishers B.V. All rights reserved 0 3 0 4 - 4 0 1 7 / 9 2 / $ 0 5 . 0 0
II()
~;,-v, ('II,XN \ N I ) R ( . K ~
specificity of an affinity-purified fraction (APF) with that of antigens prepared by Sephacryl S-200 gel filtration and by a combination of cell fractionation and differential centrifugation as described by Despommier and Lacetti (1981). M-\TERI..\LS A N I ) M E T H ( ) I ) S
Crude somatic antigens were prepared from the following common nematodes of pigs from China: 7~ spiralis (muscle larvae), Metastrongyhts apri, Trichuris suis, Gnathostoma hispidum and Stephanurus dentatus. The strain of T. spiralis used was originally isolated from a pig in Ontario, Canada, in 1967 and has since been maintained in the laboratory in ICR mice. Adult worms of the other species were collected from local abattoirs. Crude somatic antigens were prepared by sonication or homogenization by grinding, followed by extraction at 4~-C ['or two days and then centrifugation for 30 rain at 20 000 ×g. The supernatants were used as the sources of antigens. The total protein in the supernatants was determined according to Lowry et al. ( 1951 ). Positive control serum was obtained by infecting one one-month-old piglet weighing 20 kg with 10 000 muscle larvae recovered from mice. Non-infected pigs of similar age and body weight served as the source of the negative control serum. Hyperimmune serum to T. spiralis was produced by infecting nineweek-old albino New Zealand rabbits (Oo'ctolagus cuniculus). Each animal was given 5 000 larvae initially, and then a further 20 000 larvae at one, two and three months after the first infection. Antisera against the other species of nematodes listed above and normal pig, mouse and rat sera were produced in 16-week-old rabbits. This was done by an intramuscular injection of 0.5 ml ( 1 mg) total protein emulsified with an equal volume of Freund's complete adjuvant (Difco), followed by alternating subcutaneous and intramuscular injections of the same amount of protein emulsified in Freund's incomplete adjuvant on Days 7, 14 and 21. The animals were bled a week after the last injection. lmmunodiffusion (ID), immunoelectrophoresis (IEP), rocket-immunoelectrophoresis (RIEP), crossed-immunoelectrophoresis (CIEP), sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting were conducted as outlined by Chan and Ko ( 1990 ).
Indirect enz)'me-linked immunosorbent assay (EL1SA) This was performed according to the methods described by Voiler et al. (1976) and Chan and Ko ( 1988 ) but with minor modifications. Total protein (4/Lg, ml-~ ) was used as the antigen source. Rabbit antisera were used at dilutions of 1:3 200 whereas a pig antiserum dilution of l: 800 was used. The dilutions of horseradish peroxidase conjugate (Miles Lab) used were
SPECIFI('ITY OF IRI('IIINELL.-t SI'IR..ILIS ANTIGENS
[11
1 : 14000 for anti-rabbit IgG and 1 : 10000 for anti-porcine IgG. Linbro plates (Flow Lab) were used. The absorbance was read by a spectrophotometer (ELISA minireader, Dynatech) at 490 nm. The cut-offvalue was set at three times the mean optical density (O.D.) value of the negative control samples using crude somatic antigens of T. spiralis ( i.e.O.D. 0.105 for rabbit antisera and O.D. 0.462 for pig antisera).
AJfinity chromatography IgG fractions from T. suis antibodies were precipitated by an equal volume of saturated a m m o n i u m sulphate solution (Sigma), pH 7.2, held at 4°C overnight. The recovered IgG fractions (300 mg) were coupled to the Affi-Gel 10 (Bio-Rad) for 3 h at room temperature while being subjected to constant agitation. Active ester groups in the Affi-Gel were blocked by adding 25 ml of buffered I M ethanolamine (Sigma)-HCl, pH 8.0, and gently mixing at room temperature for 1 h. Total protein (20 rag) in 10 ml of phosphate buffer (0.05M, pH 7.0) was poured onto a 1 × 5 0 column (Bio-Rad Econo-Column Selection Pack-A) packed with T. suis IgG coupled Affi-Gel 10 (Bio-Rad) (Kum and Ko, 1986). Elution was carried out using PBS, pH 7.0, under a flow rate of 10 ml h r - t using a peristaltic p u m p (P1, Pharmacia).
Gel filtration Gel filtration was performed in a K26/70 glass column (Pharmacia) using Sephacryl S-200 superfine (Pharmacia). Twelve fractions (F1-F12 ) were obtained by eluting total protein ( 150 mg) with 0.1M phosphate buffer, pH 7.0, at a flow rate of 3.88 ml rain- 1 at room temperature using a peristaltic p u m p (P 1 Pharmacia).
Preparation of the S3 fraction This was performed according to the method described by Despommier and Lacetti ( 1981 ), using 2 ml of muscle larvae recovered from male ICR mice as the starting material for cell fractionation and differential centrifugation.
Screening of serum and muscle samples The specificity of the various antigens was compared by using indirect IgGELISA to screen serum samples collected from 2000 pigs. The samples werc collected in the Kennedy Town Abattoir in the following manner: blood samples from ten pigs from the same area of origin were pooled. A total of 200 lots, i.e. from 2000 pigs, were collected from pigs from different areas ofori-
[ 12
s.-w, ( H A N
A N I ) R,('. K( )
gin: northern China (n = 87 ), southern China (n = 60), coastal areas of China ( n = 12), south-east Asia (mainly Thailand, n = 3 2 ) and the local region (n=9). Muscle samples (n = 410) from pigs were also screened by the pooled digestion method. A total of 20 g of diaphragm muscles from ten pigs were pooled, minced and digested in 200 mi of 0.6% pepsin/1% HC! held at 37°C overnight in a shaking incubator (Gallenkamp). The digest was checked under a stereomicroscope for the presence of larvae. To trace the origin of the infected pigs, samples from the individual carcasses that comprised a positive pooled lot were compressed between two glass plates and examined for the presence of larvae under a stereomicroscope. RESULTS
ID, IEP, RIEP and CIEP Using ID, the APF was resolved into two precipitin arcs when anti-T, spiralis serum was employed as the developing serum (Fig. 1 ). No arc was observed when it was tested against anti-T, suis, M. apri, G. hispidum, S. dentatus or the normal pig and mouse sera. Using the APF in IEP, one migrating arc was noted when anti-T, spiralis
10
11
""
'
"
"~';.,;:',3~.
•
¼
Fig. 1. lmmunodiffusion study of reactions between the affinity purified fraction ( A P F ) ( 1 ) and rabbit antisera against 7~ spiralis (2), F. suis ( 3 ), .~I. apri (4), G. hispidum ( 5 ), S. dentatus (6), normal pig serum ( 7 ) and normal mouse serum ( 8 ). The anti-sera were in turn developed with homologous antigens: 7~ sptralis (9), 1~ suis (10), M. apri (11), G. hispidum (12). N. dentatus (13), normal pig serum (14) and normal mouse serum (15).
SPE('IFICITY
OF
I I3
I'RI('IIINEI.I.A SPIRALIS A N I I G E N S
was employed as the developing serum but no arc was noted when it was tested against anti-T, suis and S. dentatus sera. T h e results o f RIE P showed that the A P F reacted with anti-T, spiralis serum but not with anti-T, suis serum. One precipitin arc close to the origin anodally was detected only when anti- T. spiralis was used as the developing serum in RIEP. CIEP yielded two arcs when A P F and anti-T, spiralis serum were used but no arc was observed when anti7". suis was used as the developing serum.
ELISA Using the APF as the source of antigens, positive O.D. values of 0.65 and 0.63 were obtained when tested against rabbit anti-T, spiralis and T. suis sera respectively. The O.D. values were comparatively lower than the O.D. value of 1.58 obtained when the crude extract was used as the source o f antigens. No cross-reaction was detected when the APF was tested with antisera against M. apri, G. hispidum, S. dentatus and the normal pig (NPS), mouse and rabbit sera. When the 12 fractions obtained by gel filtration were used as antigens, fractions F1-F6 and F8-F12 cross-reacted with rabbit anti-T, suis serum. In addition, the fraction cross-reacted with anti-M, apri and S. dentatus sera whereas 16 -
[ ] .re p,g lsOIrOIiS serum !
1.2
_
[ ] N0rmnl pig serum
;
o
i
.,,i
-
i
i
°°
_
_
.
.
.
.
.
.
.
.
.
.
.
.
.
O/.
++ i.
Cruc~ FI
F2
F3
F4
FS
F6
F7
FB
Fg
FIO FII
FI2 $2
$3
APF
Antigens Fig. 2. Comparison of the O.D. ELISA values obtained when the antigenic fractions were tested against positive control and normal pig serum. The cut-off value was set al O.D. 0.462. The $2 fraction was the supernatant obtained after thc second centrifugation during the production of thc $3 fraction.
114
S_~x,. ('H-\N
a
"~NI) R (.
K()
b
.. i
1.1¸¸:.!?¸!1:51:.11 ,:7;i: ¸ :i
..
.i
,,
.i...: . •
: .
Fig. 3. SDS-PAGE profile of APF. Lanes: ( a ) crude antigens of 7] spiralis: ( b ) A I)F. M W markcrs (arrows from top to bottom: 66, 45, 24, 18, 14 k D a ) . A gradient gel of 7.5-20% and rabbit a n t i - l ] spiralis serum were used.
the F2 and F3 fractions cross-reacted with the former. The F7 fraction gave a low positive O.D. value (0.26) with rabbit anti-T, spiralis serum. The F I I fraction cross-reacted with all heterologous antibodies (Fig. 2 ). The $3 frac-
SI'FI( IFI('II'Y O I I R I ( I I I . V I : I I . - I .'~I'IR.II.I,S" ,'~NTI(It-NS
I 15
tion cross-reacted with anti-T, suis and M. apri sera, giving O.D. values of 1.595 and 0.469 respectively. When the various purified fractions were used as antigens to test against positive and negative control swine sera, the A P F gave the highest positive : negative ratio.
Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SD,%PAGE) and electro-immunotran,~fer blot (EITB) SDS-PAGE showed that the A P F contained six polypeptides ( MW 2 8 - 6 6 k D a ) (Fig. 3). EITB resolved the A P F into a m i n i m u m of seven antigens, with MW ranging from 18 to 55 kDa. One highly reactive band with a M W of 28 kDa also cross-reacted with anti-T, suis serum (Fig. 4). When the F I and F4 fractions separated from the Sephacryl-200 column a
b
D
Fig. 4. EITB showing cross-reactions between APF and heterologous antisera. Lanes: (a) rabbit anti-7] spiralisserum: (b) rabbit anti-71 suis serum. MW markers (arrows from top to bottom: 66,45,24. 18.14kDa).
116
S,.-W ('II-~,N -\ND R.(' k()
were analysed by EITB, the FI fraction cross-reacted with the rabbit anti-l. suis, anti-M, apri. anti-NPS, anti-G, hispidum, and anti-S, dentatus sera. The F4 fraction cross-reacted with the first three antisera as well as with normal rabbit serum. The $3 fraction was shown b v EITB to share twelve, six and three common antigens with 7: suis, M. apri and normal mouse serum respectively. SCREENING
DATA
The ELISA results showed that two of the purified antigens (APF and F4) yielded significantly lower O.D. values than those of the crude antigens (Table 1 ). In the latter case, 48% of the 200 pooled serum samples screened were found to be positive. Using the FI and $3 fractions, the positive rates were 39% and 59.5% respectively. The APF gave a substantially lower positive rate (7.5%). None of the samples screened by the F4 fraction yielded O.D. values greater than the cut-off value. The positive rates of the serum samples from five geographical regions are compared in Fig. 5. Using the APF as the source of antigens, serum samples from northern and southern China showed markedly higher positive rates than samples from the southeast Asian region; samples from Hong Kong and the coastal regions of China were negative. When screened by $3 antigens, the samples from all the different regions generally yielded high positive rates (42-65%); these rates were even higher than those obtained by the crude antigens of the F 1 fraction. TABLE I
('omparison of mean O.D. values and positive rates o f 200 p o o l e d s e t u r n samples using crude a n d purificd 71 sptralis antigens in indirect I g G - E L I S A Antigens ( 4 j,g m l (rude
I-I I-4 -M'F $3
~)
M e a n ().I). 0.541 - . 0 . 2 5 8 (0.131-1.365) 0 . 4 4 4 _- 0 . 2 4 6 (0.140-1.605) 0.103 , 0.059* ( I).014-0.285 ) 0 . 2 8 9 ~- 0 . 1 0 3 " * (0.075-0.584) 0.571 _" 11.236 (0.150-1.210)
No. + ~ e
% 4 ~c
96
48
78
39
I) 15
7.5
119
59.5
* P < 0.1 * * P < 0.5 ( Student's t-test ) when compared against the mean ().I). of crude antigens. Protein at a concentration of 4 j,g ml b was used as antigens in indirect IgG-EI.ISA to screen 200 pooled pig serum samples collected from the local abattoirs. Thc cut-offvalue was set at 0.462. T h e results are tabulated as mean O . D . + d e v i a t i o n (range) in the second column. The number and perc c n t a g e positive are indicated in the third and fourth columns.
SPECIFICITY
-
-J lad
60
40
Northern Chino (N=870) 1773Southern Chlna (N=600) 17;73Coast regions of China (N=120) S.C..~la (N=320) Hong Kong (N=90)
7 / / /
>
+
-o
l [7
O F 7t¢I('tlL.VELL.! SI'IR.II.15; . ' ~ N T I ( i E N S
\/
/N
\/
\ ,',17 ~ ,.'4/ w,xJ
0
~ VI
20
,
\
.~
N
k~,V
~k-I
k~V k'9' L~d~
0
Crude
F1
$5
APF
Antigens Fig. 5. Comparison of the percentage of positive samples obtained by using indirect IgG-ELISA, crude. FI, $3 and APF antigens to screen serum samples from pigs originating from different regions of China, SE .Asia and Hong Kong. \ = s a m p l e size of pigs screened.
Of the 410 pigs examined by the pooled digestion method, larvae were recovered from two animals (0.5%). The numbers of larvae per gram of muscle were 87 and 69 respectively. Both pigs were from Henan Province (northern China ). DISCUSSION
Numerous methods have been used to isolate specific antigens from extracts of 7". spiralis muscle larvae (see reviews by Ruitenberg et al.. 1983: Despommier, 1986; Ko, 1988, 1989). However, there is still no universal source of somatic antigens which is satisfactory for use in the immunodiagnosis of trichinellosis in swines harbouring various indigenous nematode fauna. The present study shows that antigens purified by affinity chromatography were markedly more specific. Most of the fractions produced by gel filtrations cross-reacted with rabbit anti-T, suis serum. The $3 fraction, which was shown by Despommier and Lacetti ( 1981 ) to contain all the antigens associated with an excretory/secretory antigen preparation, was also observed to cross-react with antisera against T. suis, M. apri and normal mouse serum. Moreover, the APF gave a high positive O.D. value and the highest positive: negative ratio with the positive control pig sera in ELISA. Therefore, the APF seems to have good potential for use as the basic source of antigens for screening trichinellosis. According to the SDS-PAGE profile, only six polypeptides were present in
1 Ig
~,.-~,~,. ( I { \ N ~ , N I ) R . ( K t )
APF. However, seven active antigens were demonstrated by EITB. Therefore, one of the antigens resolved by the latter method was probably non-protein in nature. Although the results of ID and IEP showed that common antigens of /i spiredi.s and G. hispidum, and of :11. apri and S. dc,~lalu.s could be removed by affinity chromatography, the results obtained from EITB revealed that one of the seven antigens present in the APF still cross-reacted with rabbit anti-7~ ,sttis serum. However. it seems that removal of the cross-reacting factors to anti-7 i ,sui.s serum may also remove most of the non-specific factors to rabbit anti-T, spiralis serum. Our previous work has shown that 41 of 47 antigens of the crude T..~pirali.s extract cross-reacted with rabbit anti-T..sttis serum (Chan and Ko, 1990). Since both T. spirali.~ and "!~ suis arc phylogenetically related trichuroids possessing a stichosome, they must share many common somatic and excretory/secretory antigens as suggested by Despommier and Muller (1970a, b, 1976). Therefore, an affinity column coupled with anti-7" .sttis antibodies may serve as a useful step in the production of specific antigens. Among the five sources of antigens used for screening pig serum samples, the APF seemed to elicit the least number of false positive reactions and has good potential for development as a source of screening antigens. Further studies will locus on studying the detection limit of the APF using pigs experimentally infected with different numbers of parasites. The practical use of the APF will be evaluated on a large number of known positive cases of trichinellosis. The APF can easily be produced and with further purification its specificity and sensitivity may become similar to those of the excretory/secretory antigens as reported by Gamble et al. ( 1983, 1988). The 7.5% positive rate produced by the APF is higher than the rate of 0.5% obtained by the pooled digestion method. However, given the relative insensitivity of the latter method and the inclusion of some false positives in the former method, the true prevalence of trichinellosis in pigs slaughtered in Hong Kong will probably fall within the above range. Since Hong Kong annually imports and slaughters one million pigs, even a 1% positive rate should be considered as very significant. All the samples screened by the F4 fraction were negative. This is probably a result of its low sensitivity to the swine antibodies. The $3 fraction appeared to produce the largest number of false positive reactions. Non-specific factors were probably enriched during the fractionation and centrifugation steps (Seawright et al., 1983 ). However, an affinity column purified antigen of the $3 fraction has been suggested as a potential source of antigens for use in epidemiological studies in slaughterhouse screenings (Seawright et al., 1983: Kolata, 1985 ). Since both pigs found to be positive using the pooled digestion method originated from the Henan Province, trichinellosis may well be common there.
S P E ( I F I ( I I Y ~)F 7 R I ( I t I . V I : L L . !
.',TIR.II.IS a, N T I ( i E N S
1l9
Most of the pigs imported from China into Hong Kong come from the following provinces: Hubei, Hunan, Henan, Zhejiang, Jiangxi, Fujiang, Guangdong. Swine trichinellosis is known to be endemic in some parts of China, especially in the northeastern and southwestern regions (Xu, 1975; China Co-ordinating Group for Prevention and Treatment of Trichinosis, Harbin City, 1981 ; Wang, 1981; Niu et al., 1985; Tang and Tang, 1987). In Gansu province, positive rates of 13.5% and 68.24% were detected by both the compression-slide method and ELISA (Niu et al., 1985). The apparently common occurrence of this parasite in Chinese pigs may perhaps be explained by the persistence of the garbage-feeding practice and poor management methods. About 95% of pigs in China are reared as a sideline activity by people living in rural regions. ACKNOWLEI)GEMENTS
We are very grateful to the Croucher Foundation, Hong Kong, for financial support. REFERENCES Chan, S.-W.. 1986. Studies on the prevalence of 7Hchinella ,spira/is infection in pigs imported into l-tong Kong. M. Phil. Thesis, Univ. Hong Kong, 295 pp. Chan, S.-W. and Ko, R.C.. 1988. Comparison between standard ELISA and dot-ELISA for scrodiagnosis of human trichinosis. Trans. R. Soc. Trop. Mcd. Hyg., 82: 802-894. Chan, S.-W. and Ko, R.(?., 1990. Cross-reactions between crude antigens of 7)'ichinella .spiralis and other nematodes of Chinese pigs. Trop. Biomed.. 7:143-150. China Co-ordinating Group for Prevention and Treatment of Trichinosis, Harbin City, 1981. A survey of trichinelliasis acquired through eating scalded mutton. Chin. J, Epidcmiol.. 2: p. 103. Dcspommicr, I).D.. 1986. Trichincllosis. In: K.M. Walls (Editor), Immunodiagnosis of Parasitic Diseases. Academic Press. New York. pp. 163-81. I)cspommier. D.D. and Lacetti, A.. 1981. l)'ichine/la spiralis: proteins and antigens isolated from a large-particle fraction derived from the muscle larva. Exp. Parasitol., 51: 279-295. I.)cspommier. D.D. and Muller, M., 1970a. Functional antigens of 7)'ichine//a spirah,s. J. Parasitol.. 54: p. 76. Dcspommier, D.D. and Muller, M., 1970b. The stichosomc of "l)'ichinella ,v~lralis: its structure and function. J. Parasitol.. 54: 76-77. l)cspommicr, D.D. and Muller, M., 1976. The stichosomc and ils secretion granules in the mature muscle larva of Trichim'lla spiralt,s. J. Parasitol., 62: 775-785. Gamble, I I.R.. Anderson, W.R., Graham. C.E. and Murrell, K.D., 1983. Diagnosis of swine trichinosis by enzyme-linked immunosorbent assay (ELISA) using an excretory-secretory antigen. Vet. Parasitol., 13: 349-361. Gamble, i I.R.. Rapic. D., Marinculic, A. and Murrell, K.D., 1988. Influence of cultivation condilions on specificity of excretory-secretory antigens for the diagnosis of Iriehinellosis. Vet. Parasilol., 30: 131-137. Ko. R.('., 1984. Are therc parasitic disease problems in man in Hong Kong? In: R.C. Ko (Edilot), Current Perspeetivcs in Parasilic Diseases. Depl. Zool., Univ. Hong Kong, pp. 29-51.
120
S.-W, ('HAN ,\NI) RC. K~)
Ko, R.C., 1988. A review of the seroepidcmiological techniques in Trichincllosis. Proc. Working Group on Seroepidemiological Tools for Programmcs in Tropical Diseases, UNPI)/World Bank/WHO Special TDR Programme, Kuala kumpur. Malaysia. 1987. Trop. Biomed., 5: 61-73. Ko, R.C., 1989. Application of serological techniques for the diagnosis oflrichinellosis. In: R.C. Ko (Editor), ('urrcnt Concepts in l'arasitology. Hong Kong University Press, Hong Kong, pp. 81 - 10(I. Kolata, G., 1985. Testing |or trichinosis. Science, 227:621-624. Kum, W.S. and Ko, R.C., 1986. Isolation of specific antigens from .,lngiostr
109
Specificity of affinity-purified Trichinella spiralis antigens Shiu-Wan Chan and Ronald C. Ko ~ Department o fZooh~:t'. University of tlong Kong, tlong Kong (Accepted 18 July 1991 )
ABSTRACT Chan, S.-W. and Ko, R.C.. 1992. Specificity of affinity-purified Trichinella spiralis antigens. I'et. Parasitol., 41 : 109-120. An affinity-purified fraction (APF) was obtained by passing crude somatic antigens of lrtchinelhz ~piralis muscle larvae through an Aft-Gel 10 column coupled with anti-7)'ichuris suis IgG. The fraclion contained seven antigens with molecular weights ranging from 28 to 55 kDa. When tested with antiserum against other common nematodes of pigs from China, the APF was found to be markedly more specific than $3 antigens (prepared by a combination of cell fractionation and differential centrifugation according to Despommier and Lacetti, 1981 ) and fractions produced by Scphacryl S-200 gel filtration ( FI to F12 ). When the APF was used in an indirect IgG--enzyme-linked immunosorbenl assay ( IgG-ELISA ) to screen serum samples from 2000 pigs imported from China, a positive rate of 7.5% was obtained. Similar screenings using the crude somatic antigens F I and $3 gave a large number of cross-reactions and false positive reactions. Positive rates of 48%, 39% and 59.5% respectively were obtained lbr the three antigens.
INTRODUCTION
Although trichinellosis is a major parasitic problem of pigs in many regions of China, few attempts have been made to develop an effective immunoscreening system which could be used in the local abattoirs ( Ko, 1984; Chan, 1986 ). The development of such a system is greatly hampered by the fact that pigs imported from China are commonly infected with numerous species of nematodes which share common antigens with Trichinella spiralis. Hong Kong imports as many as one million live pigs annually for local consumption, about 90% of which come from China. Outbreaks of human trichinellosis were only first documented in Hong Kong in 1981 (Ko, 1984). Since 1984, a research programme has been initiated in our laboratory to develop a sensitive and specific immunodiagnostic method to screen for T. spiralis antibodies in pigs slaughtered at the local abattoirs. This paper compares the ~Author to whom correspondence should be addressed.
~3 1992 Elsevier Science Publishers B.V. All rights reserved 0 3 0 4 - 4 0 1 7 / 9 2 / $ 0 5 . 0 0
II()
~;,-v, ('II,XN \ N I ) R ( . K ~
specificity of an affinity-purified fraction (APF) with that of antigens prepared by Sephacryl S-200 gel filtration and by a combination of cell fractionation and differential centrifugation as described by Despommier and Lacetti (1981). M-\TERI..\LS A N I ) M E T H ( ) I ) S
Crude somatic antigens were prepared from the following common nematodes of pigs from China: 7~ spiralis (muscle larvae), Metastrongyhts apri, Trichuris suis, Gnathostoma hispidum and Stephanurus dentatus. The strain of T. spiralis used was originally isolated from a pig in Ontario, Canada, in 1967 and has since been maintained in the laboratory in ICR mice. Adult worms of the other species were collected from local abattoirs. Crude somatic antigens were prepared by sonication or homogenization by grinding, followed by extraction at 4~-C ['or two days and then centrifugation for 30 rain at 20 000 ×g. The supernatants were used as the sources of antigens. The total protein in the supernatants was determined according to Lowry et al. ( 1951 ). Positive control serum was obtained by infecting one one-month-old piglet weighing 20 kg with 10 000 muscle larvae recovered from mice. Non-infected pigs of similar age and body weight served as the source of the negative control serum. Hyperimmune serum to T. spiralis was produced by infecting nineweek-old albino New Zealand rabbits (Oo'ctolagus cuniculus). Each animal was given 5 000 larvae initially, and then a further 20 000 larvae at one, two and three months after the first infection. Antisera against the other species of nematodes listed above and normal pig, mouse and rat sera were produced in 16-week-old rabbits. This was done by an intramuscular injection of 0.5 ml ( 1 mg) total protein emulsified with an equal volume of Freund's complete adjuvant (Difco), followed by alternating subcutaneous and intramuscular injections of the same amount of protein emulsified in Freund's incomplete adjuvant on Days 7, 14 and 21. The animals were bled a week after the last injection. lmmunodiffusion (ID), immunoelectrophoresis (IEP), rocket-immunoelectrophoresis (RIEP), crossed-immunoelectrophoresis (CIEP), sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting were conducted as outlined by Chan and Ko ( 1990 ).
Indirect enz)'me-linked immunosorbent assay (EL1SA) This was performed according to the methods described by Voiler et al. (1976) and Chan and Ko ( 1988 ) but with minor modifications. Total protein (4/Lg, ml-~ ) was used as the antigen source. Rabbit antisera were used at dilutions of 1:3 200 whereas a pig antiserum dilution of l: 800 was used. The dilutions of horseradish peroxidase conjugate (Miles Lab) used were
SPECIFI('ITY OF IRI('IIINELL.-t SI'IR..ILIS ANTIGENS
[11
1 : 14000 for anti-rabbit IgG and 1 : 10000 for anti-porcine IgG. Linbro plates (Flow Lab) were used. The absorbance was read by a spectrophotometer (ELISA minireader, Dynatech) at 490 nm. The cut-offvalue was set at three times the mean optical density (O.D.) value of the negative control samples using crude somatic antigens of T. spiralis ( i.e.O.D. 0.105 for rabbit antisera and O.D. 0.462 for pig antisera).
AJfinity chromatography IgG fractions from T. suis antibodies were precipitated by an equal volume of saturated a m m o n i u m sulphate solution (Sigma), pH 7.2, held at 4°C overnight. The recovered IgG fractions (300 mg) were coupled to the Affi-Gel 10 (Bio-Rad) for 3 h at room temperature while being subjected to constant agitation. Active ester groups in the Affi-Gel were blocked by adding 25 ml of buffered I M ethanolamine (Sigma)-HCl, pH 8.0, and gently mixing at room temperature for 1 h. Total protein (20 rag) in 10 ml of phosphate buffer (0.05M, pH 7.0) was poured onto a 1 × 5 0 column (Bio-Rad Econo-Column Selection Pack-A) packed with T. suis IgG coupled Affi-Gel 10 (Bio-Rad) (Kum and Ko, 1986). Elution was carried out using PBS, pH 7.0, under a flow rate of 10 ml h r - t using a peristaltic p u m p (P1, Pharmacia).
Gel filtration Gel filtration was performed in a K26/70 glass column (Pharmacia) using Sephacryl S-200 superfine (Pharmacia). Twelve fractions (F1-F12 ) were obtained by eluting total protein ( 150 mg) with 0.1M phosphate buffer, pH 7.0, at a flow rate of 3.88 ml rain- 1 at room temperature using a peristaltic p u m p (P 1 Pharmacia).
Preparation of the S3 fraction This was performed according to the method described by Despommier and Lacetti ( 1981 ), using 2 ml of muscle larvae recovered from male ICR mice as the starting material for cell fractionation and differential centrifugation.
Screening of serum and muscle samples The specificity of the various antigens was compared by using indirect IgGELISA to screen serum samples collected from 2000 pigs. The samples werc collected in the Kennedy Town Abattoir in the following manner: blood samples from ten pigs from the same area of origin were pooled. A total of 200 lots, i.e. from 2000 pigs, were collected from pigs from different areas ofori-
[ 12
s.-w, ( H A N
A N I ) R,('. K( )
gin: northern China (n = 87 ), southern China (n = 60), coastal areas of China ( n = 12), south-east Asia (mainly Thailand, n = 3 2 ) and the local region (n=9). Muscle samples (n = 410) from pigs were also screened by the pooled digestion method. A total of 20 g of diaphragm muscles from ten pigs were pooled, minced and digested in 200 mi of 0.6% pepsin/1% HC! held at 37°C overnight in a shaking incubator (Gallenkamp). The digest was checked under a stereomicroscope for the presence of larvae. To trace the origin of the infected pigs, samples from the individual carcasses that comprised a positive pooled lot were compressed between two glass plates and examined for the presence of larvae under a stereomicroscope. RESULTS
ID, IEP, RIEP and CIEP Using ID, the APF was resolved into two precipitin arcs when anti-T, spiralis serum was employed as the developing serum (Fig. 1 ). No arc was observed when it was tested against anti-T, suis, M. apri, G. hispidum, S. dentatus or the normal pig and mouse sera. Using the APF in IEP, one migrating arc was noted when anti-T, spiralis
10
11
""
'
"
"~';.,;:',3~.
•
¼
Fig. 1. lmmunodiffusion study of reactions between the affinity purified fraction ( A P F ) ( 1 ) and rabbit antisera against 7~ spiralis (2), F. suis ( 3 ), .~I. apri (4), G. hispidum ( 5 ), S. dentatus (6), normal pig serum ( 7 ) and normal mouse serum ( 8 ). The anti-sera were in turn developed with homologous antigens: 7~ sptralis (9), 1~ suis (10), M. apri (11), G. hispidum (12). N. dentatus (13), normal pig serum (14) and normal mouse serum (15).
SPE('IFICITY
OF
I I3
I'RI('IIINEI.I.A SPIRALIS A N I I G E N S
was employed as the developing serum but no arc was noted when it was tested against anti-T, suis and S. dentatus sera. T h e results o f RIE P showed that the A P F reacted with anti-T, spiralis serum but not with anti-T, suis serum. One precipitin arc close to the origin anodally was detected only when anti- T. spiralis was used as the developing serum in RIEP. CIEP yielded two arcs when A P F and anti-T, spiralis serum were used but no arc was observed when anti7". suis was used as the developing serum.
ELISA Using the APF as the source of antigens, positive O.D. values of 0.65 and 0.63 were obtained when tested against rabbit anti-T, spiralis and T. suis sera respectively. The O.D. values were comparatively lower than the O.D. value of 1.58 obtained when the crude extract was used as the source o f antigens. No cross-reaction was detected when the APF was tested with antisera against M. apri, G. hispidum, S. dentatus and the normal pig (NPS), mouse and rabbit sera. When the 12 fractions obtained by gel filtration were used as antigens, fractions F1-F6 and F8-F12 cross-reacted with rabbit anti-T, suis serum. In addition, the fraction cross-reacted with anti-M, apri and S. dentatus sera whereas 16 -
[ ] .re p,g lsOIrOIiS serum !
1.2
_
[ ] N0rmnl pig serum
;
o
i
.,,i
-
i
i
°°
_
_
.
.
.
.
.
.
.
.
.
.
.
.
.
O/.
++ i.
Cruc~ FI
F2
F3
F4
FS
F6
F7
FB
Fg
FIO FII
FI2 $2
$3
APF
Antigens Fig. 2. Comparison of the O.D. ELISA values obtained when the antigenic fractions were tested against positive control and normal pig serum. The cut-off value was set al O.D. 0.462. The $2 fraction was the supernatant obtained after thc second centrifugation during the production of thc $3 fraction.
114
S_~x,. ('H-\N
a
"~NI) R (.
K()
b
.. i
1.1¸¸:.!?¸!1:51:.11 ,:7;i: ¸ :i
..
.i
,,
.i...: . •
: .
Fig. 3. SDS-PAGE profile of APF. Lanes: ( a ) crude antigens of 7] spiralis: ( b ) A I)F. M W markcrs (arrows from top to bottom: 66, 45, 24, 18, 14 k D a ) . A gradient gel of 7.5-20% and rabbit a n t i - l ] spiralis serum were used.
the F2 and F3 fractions cross-reacted with the former. The F7 fraction gave a low positive O.D. value (0.26) with rabbit anti-T, spiralis serum. The F I I fraction cross-reacted with all heterologous antibodies (Fig. 2 ). The $3 frac-
SI'FI( IFI('II'Y O I I R I ( I I I . V I : I I . - I .'~I'IR.II.I,S" ,'~NTI(It-NS
I 15
tion cross-reacted with anti-T, suis and M. apri sera, giving O.D. values of 1.595 and 0.469 respectively. When the various purified fractions were used as antigens to test against positive and negative control swine sera, the A P F gave the highest positive : negative ratio.
Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SD,%PAGE) and electro-immunotran,~fer blot (EITB) SDS-PAGE showed that the A P F contained six polypeptides ( MW 2 8 - 6 6 k D a ) (Fig. 3). EITB resolved the A P F into a m i n i m u m of seven antigens, with MW ranging from 18 to 55 kDa. One highly reactive band with a M W of 28 kDa also cross-reacted with anti-T, suis serum (Fig. 4). When the F I and F4 fractions separated from the Sephacryl-200 column a
b
D
Fig. 4. EITB showing cross-reactions between APF and heterologous antisera. Lanes: (a) rabbit anti-7] spiralisserum: (b) rabbit anti-71 suis serum. MW markers (arrows from top to bottom: 66,45,24. 18.14kDa).
116
S,.-W ('II-~,N -\ND R.(' k()
were analysed by EITB, the FI fraction cross-reacted with the rabbit anti-l. suis, anti-M, apri. anti-NPS, anti-G, hispidum, and anti-S, dentatus sera. The F4 fraction cross-reacted with the first three antisera as well as with normal rabbit serum. The $3 fraction was shown b v EITB to share twelve, six and three common antigens with 7: suis, M. apri and normal mouse serum respectively. SCREENING
DATA
The ELISA results showed that two of the purified antigens (APF and F4) yielded significantly lower O.D. values than those of the crude antigens (Table 1 ). In the latter case, 48% of the 200 pooled serum samples screened were found to be positive. Using the FI and $3 fractions, the positive rates were 39% and 59.5% respectively. The APF gave a substantially lower positive rate (7.5%). None of the samples screened by the F4 fraction yielded O.D. values greater than the cut-off value. The positive rates of the serum samples from five geographical regions are compared in Fig. 5. Using the APF as the source of antigens, serum samples from northern and southern China showed markedly higher positive rates than samples from the southeast Asian region; samples from Hong Kong and the coastal regions of China were negative. When screened by $3 antigens, the samples from all the different regions generally yielded high positive rates (42-65%); these rates were even higher than those obtained by the crude antigens of the F 1 fraction. TABLE I
('omparison of mean O.D. values and positive rates o f 200 p o o l e d s e t u r n samples using crude a n d purificd 71 sptralis antigens in indirect I g G - E L I S A Antigens ( 4 j,g m l (rude
I-I I-4 -M'F $3
~)
M e a n ().I). 0.541 - . 0 . 2 5 8 (0.131-1.365) 0 . 4 4 4 _- 0 . 2 4 6 (0.140-1.605) 0.103 , 0.059* ( I).014-0.285 ) 0 . 2 8 9 ~- 0 . 1 0 3 " * (0.075-0.584) 0.571 _" 11.236 (0.150-1.210)
No. + ~ e
% 4 ~c
96
48
78
39
I) 15
7.5
119
59.5
* P < 0.1 * * P < 0.5 ( Student's t-test ) when compared against the mean ().I). of crude antigens. Protein at a concentration of 4 j,g ml b was used as antigens in indirect IgG-EI.ISA to screen 200 pooled pig serum samples collected from the local abattoirs. Thc cut-offvalue was set at 0.462. T h e results are tabulated as mean O . D . + d e v i a t i o n (range) in the second column. The number and perc c n t a g e positive are indicated in the third and fourth columns.
SPECIFICITY
-
-J lad
60
40
Northern Chino (N=870) 1773Southern Chlna (N=600) 17;73Coast regions of China (N=120) S.C..~la (N=320) Hong Kong (N=90)
7 / / /
>
+
-o
l [7
O F 7t¢I('tlL.VELL.! SI'IR.II.15; . ' ~ N T I ( i E N S
\/
/N
\/
\ ,',17 ~ ,.'4/ w,xJ
0
~ VI
20
,
\
.~
N
k~,V
~k-I
k~V k'9' L~d~
0
Crude
F1
$5
APF
Antigens Fig. 5. Comparison of the percentage of positive samples obtained by using indirect IgG-ELISA, crude. FI, $3 and APF antigens to screen serum samples from pigs originating from different regions of China, SE .Asia and Hong Kong. \ = s a m p l e size of pigs screened.
Of the 410 pigs examined by the pooled digestion method, larvae were recovered from two animals (0.5%). The numbers of larvae per gram of muscle were 87 and 69 respectively. Both pigs were from Henan Province (northern China ). DISCUSSION
Numerous methods have been used to isolate specific antigens from extracts of 7". spiralis muscle larvae (see reviews by Ruitenberg et al.. 1983: Despommier, 1986; Ko, 1988, 1989). However, there is still no universal source of somatic antigens which is satisfactory for use in the immunodiagnosis of trichinellosis in swines harbouring various indigenous nematode fauna. The present study shows that antigens purified by affinity chromatography were markedly more specific. Most of the fractions produced by gel filtrations cross-reacted with rabbit anti-T, suis serum. The $3 fraction, which was shown by Despommier and Lacetti ( 1981 ) to contain all the antigens associated with an excretory/secretory antigen preparation, was also observed to cross-react with antisera against T. suis, M. apri and normal mouse serum. Moreover, the APF gave a high positive O.D. value and the highest positive: negative ratio with the positive control pig sera in ELISA. Therefore, the APF seems to have good potential for use as the basic source of antigens for screening trichinellosis. According to the SDS-PAGE profile, only six polypeptides were present in
1 Ig
~,.-~,~,. ( I { \ N ~ , N I ) R . ( K t )
APF. However, seven active antigens were demonstrated by EITB. Therefore, one of the antigens resolved by the latter method was probably non-protein in nature. Although the results of ID and IEP showed that common antigens of /i spiredi.s and G. hispidum, and of :11. apri and S. dc,~lalu.s could be removed by affinity chromatography, the results obtained from EITB revealed that one of the seven antigens present in the APF still cross-reacted with rabbit anti-7~ ,sttis serum. However. it seems that removal of the cross-reacting factors to anti-7 i ,sui.s serum may also remove most of the non-specific factors to rabbit anti-T, spiralis serum. Our previous work has shown that 41 of 47 antigens of the crude T..~pirali.s extract cross-reacted with rabbit anti-T..sttis serum (Chan and Ko, 1990). Since both T. spirali.~ and "!~ suis arc phylogenetically related trichuroids possessing a stichosome, they must share many common somatic and excretory/secretory antigens as suggested by Despommier and Muller (1970a, b, 1976). Therefore, an affinity column coupled with anti-7" .sttis antibodies may serve as a useful step in the production of specific antigens. Among the five sources of antigens used for screening pig serum samples, the APF seemed to elicit the least number of false positive reactions and has good potential for development as a source of screening antigens. Further studies will locus on studying the detection limit of the APF using pigs experimentally infected with different numbers of parasites. The practical use of the APF will be evaluated on a large number of known positive cases of trichinellosis. The APF can easily be produced and with further purification its specificity and sensitivity may become similar to those of the excretory/secretory antigens as reported by Gamble et al. ( 1983, 1988). The 7.5% positive rate produced by the APF is higher than the rate of 0.5% obtained by the pooled digestion method. However, given the relative insensitivity of the latter method and the inclusion of some false positives in the former method, the true prevalence of trichinellosis in pigs slaughtered in Hong Kong will probably fall within the above range. Since Hong Kong annually imports and slaughters one million pigs, even a 1% positive rate should be considered as very significant. All the samples screened by the F4 fraction were negative. This is probably a result of its low sensitivity to the swine antibodies. The $3 fraction appeared to produce the largest number of false positive reactions. Non-specific factors were probably enriched during the fractionation and centrifugation steps (Seawright et al., 1983 ). However, an affinity column purified antigen of the $3 fraction has been suggested as a potential source of antigens for use in epidemiological studies in slaughterhouse screenings (Seawright et al., 1983: Kolata, 1985 ). Since both pigs found to be positive using the pooled digestion method originated from the Henan Province, trichinellosis may well be common there.
S P E ( I F I ( I I Y ~)F 7 R I ( I t I . V I : L L . !
.',TIR.II.IS a, N T I ( i E N S
1l9
Most of the pigs imported from China into Hong Kong come from the following provinces: Hubei, Hunan, Henan, Zhejiang, Jiangxi, Fujiang, Guangdong. Swine trichinellosis is known to be endemic in some parts of China, especially in the northeastern and southwestern regions (Xu, 1975; China Co-ordinating Group for Prevention and Treatment of Trichinosis, Harbin City, 1981 ; Wang, 1981; Niu et al., 1985; Tang and Tang, 1987). In Gansu province, positive rates of 13.5% and 68.24% were detected by both the compression-slide method and ELISA (Niu et al., 1985). The apparently common occurrence of this parasite in Chinese pigs may perhaps be explained by the persistence of the garbage-feeding practice and poor management methods. About 95% of pigs in China are reared as a sideline activity by people living in rural regions. ACKNOWLEI)GEMENTS
We are very grateful to the Croucher Foundation, Hong Kong, for financial support. REFERENCES Chan, S.-W.. 1986. Studies on the prevalence of 7Hchinella ,spira/is infection in pigs imported into l-tong Kong. M. Phil. Thesis, Univ. Hong Kong, 295 pp. Chan, S.-W. and Ko, R.C.. 1988. Comparison between standard ELISA and dot-ELISA for scrodiagnosis of human trichinosis. Trans. R. Soc. Trop. Mcd. Hyg., 82: 802-894. Chan, S.-W. and Ko, R.(?., 1990. Cross-reactions between crude antigens of 7)'ichinella .spiralis and other nematodes of Chinese pigs. Trop. Biomed.. 7:143-150. China Co-ordinating Group for Prevention and Treatment of Trichinosis, Harbin City, 1981. A survey of trichinelliasis acquired through eating scalded mutton. Chin. J, Epidcmiol.. 2: p. 103. Dcspommicr, I).D.. 1986. Trichincllosis. In: K.M. Walls (Editor), Immunodiagnosis of Parasitic Diseases. Academic Press. New York. pp. 163-81. I)cspommier. D.D. and Lacetti, A.. 1981. l)'ichine/la spiralis: proteins and antigens isolated from a large-particle fraction derived from the muscle larva. Exp. Parasitol., 51: 279-295. I.)cspommier. D.D. and Muller, M., 1970a. Functional antigens of 7)'ichine//a spirah,s. J. Parasitol.. 54: p. 76. Dcspommier, D.D. and Muller, M., 1970b. The stichosomc of "l)'ichinella ,v~lralis: its structure and function. J. Parasitol.. 54: 76-77. l)cspommicr, D.D. and Muller, M., 1976. The stichosomc and ils secretion granules in the mature muscle larva of Trichim'lla spiralt,s. J. Parasitol., 62: 775-785. Gamble, I I.R.. Anderson, W.R., Graham. C.E. and Murrell, K.D., 1983. Diagnosis of swine trichinosis by enzyme-linked immunosorbent assay (ELISA) using an excretory-secretory antigen. Vet. Parasitol., 13: 349-361. Gamble, i I.R.. Rapic. D., Marinculic, A. and Murrell, K.D., 1988. Influence of cultivation condilions on specificity of excretory-secretory antigens for the diagnosis of Iriehinellosis. Vet. Parasilol., 30: 131-137. Ko. R.('., 1984. Are therc parasitic disease problems in man in Hong Kong? In: R.C. Ko (Edilot), Current Perspeetivcs in Parasilic Diseases. Depl. Zool., Univ. Hong Kong, pp. 29-51.
120
S.-W, ('HAN ,\NI) RC. K~)
Ko, R.C., 1988. A review of the seroepidcmiological techniques in Trichincllosis. Proc. Working Group on Seroepidemiological Tools for Programmcs in Tropical Diseases, UNPI)/World Bank/WHO Special TDR Programme, Kuala kumpur. Malaysia. 1987. Trop. Biomed., 5: 61-73. Ko, R.C., 1989. Application of serological techniques for the diagnosis oflrichinellosis. In: R.C. Ko (Editor), ('urrcnt Concepts in l'arasitology. Hong Kong University Press, Hong Kong, pp. 81 - 10(I. Kolata, G., 1985. Testing |or trichinosis. Science, 227:621-624. Kum, W.S. and Ko, R.C., 1986. Isolation of specific antigens from .,lngiostr
