Veterinary Immunology and Immunopathology, 28 ( 1991 ) 57-65
57
Elsevier Science Publishers B.V., Amsterdam
Isolation and purification of a diagnostic antigen for bovine cysticercosis by hydrophobic chromatography* Eugene G. Hayunga mand Mary P. Sumner 2 Serology Branch, Pathology and Serology Division, Science and Technology Program, Food Safety and lnspection Service, U.S. Departmem of Agricuiture, Beltsville, MD 20705, LTA (Accepted 26 April 1990)
ABSTRACT Hayunga, E.G. and Sumner, M.P., 1991. Isolation and purification of a diagnostic antigen for bovine cysticercosis by hydrophobic chromatography. Vet. Immunol. ImmunopathoL, 28: 57-65. An ammonium sulfate fraction of Taenia hydatigena cyst fluid (ThFAS) was further fractionated by hydrophobic interaction chromatography, using alkylagarose and co-amino alkylagarose columns, in an effort to.isolate and purify a specific diagnostic antigen in the ThFAS preparation. The < 12 kDa antigen was found to have an affinity for immobilized alkanes with chain length of six carbons or greater. The antigen was recovered in an ethylene glycol eluate from a hexylagarose column then analyzed by Western blot; it reacted with bovine and human cysticercosis infection sera and with specific monoclonal antibodies but not with control sera or fascioliasis infection scra. When the eluate was used as coating antigen in a plate ELISA assay no false positive reactions were seen in sera from cattle infected with Fasciola hepatica; false positive reactions were observed for the unfractionated ThFAS antigen preparation.
INTRODUCTION
Serological assays offer the possibility of rapid diagnosis of helminthic diseases using non-invasive means. Immunoassays can detect infection prior to *The opinions or assertions contained herein are the private ones o f the authors and are not to be construed as official or reflecting the views o f the U.S. Department o f Agriculture, the Food Safety and Inspection Service or the Agricultural Research Service. The mention of a particular trade name or manufacturer is made for descriptive purposes only and is not intended to imply endorsement of that product over similar products from other commercial sources. The experiments reported herein were conducted according to the principles set forth in the "'Guide for the Care and Use o f Laboratory Animals," Institute o f Animal Resources, National Research Council, D H E W Publication No. ( N I H ) 78-23. 1Present address: National Institutes of Health, Westwood Building, Bethesda, MD 20892, USA. 2present address: Kirkegaard and Perry Laboratories, 2 Cessna Court, Gaithersburg, MD 20879, USA.
0165-2427/91/$03.50
© 1991 m Elsevier Science Publishers B.V.
58
E.G. HAYUNGAAND M.P. SUMNER
the onset of other clinical signs (Hayunga et al., 1987a), and may prove more sensitive than time-consuming organoleptic methods of food inspection (Dewhirst et al., 1967). One highly promising reagent for the immunodiagnosis of bovine cysticercosis is a < 12 kDa antigen found in an ammonium sulfate soluble fraction of Taenia hydatigena cyst fluid, ThFAS (Rhoads et al., 1985; Kamanga-Sollo et al., 1987; Hayunga et al., 1991 ). However, this fraction also contains components that react with sera from uninfected cattle (Hayunga et al., 1991 ) and therefore may contribute to false positive reactions when ThFAS is employed in a conventional plate ELISA. In the present paper, we report a liquid chromatographic method for purifying the diagnostic antigen from ThFAS. MATERIALS AND METHODS
Antigen preparation Taenia hydatigena cysts were collected from the mesenteries of naturally infected sheep at the time of slaughter; fluid was removed by incision, then filtered through two Nitex brand nylon screens, 110/zm and 44/tm exclusion size, to remove gross particulate contamination. An ammonium sulfate soluble fraction of the cyst fluid, ThFAS, was prepared as described previously (Rhoads et al., 1985). ThFAS was dialyzed against 0.05 M Tris-HCL, pH 8.0, and adjusted to a protein concentration of I mg/ml by the addition of buffer. Alkylagarose and oJ-amino alkylagarose were obtained commercially from Miles Scientific (Naperville, Indiana) as kits of six pre-packed 1 ml analytical columns containing 0-, 2-, 4-, 6-, 8- and !0-carbon chain !igands cova!ent!y bound to agarose. After equilibration of the columns with Tris-HCl, aliquots of approximately 1 mg ThFAS were applied and allowed to enter the gel bed at room temperature at a flow rate of approximately 0.45 ml/min (by gravity), then washed with additional buffer. The Tris-HCl "fall through" fraction, 1 M NaCI and 50% ethylene glycol fractions were collected as described previously (Hayunga et al., 1987b). Column fractions were dialyzed against several changes of distilled water, lyophilized and resuspended in SDS sample buffer. SDS polyaerylamide gel electrophoresis was conducted using 5-15 % 1.5 mm gradient slab gels run under reducing conditions in an SDS, Tris-glycine continuous buffer system, as described previously (Sumner and Hayunga, 1986). Proteins were then transferred to nitrocellulose by electroblotting at 0°-I0°C, 50 V constant voltage for 1 h, then 25 V overnight, using a Tris-glycine-methanol transfer buffer, as described previously (Hayunga et al., 1991 ). Blots were blocked with gelatin, then reacted with infection serum, alkaline phosphatase conjugated anti-IgG, and BCIP/NBT substrate. The enzyme-linked immunosorbent assay (ELISA) was performed using
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
59
peroxidase conjugate and ABTS substrate, as described previously (Hayunga et al., 1987a). Serum samples diluted 1: 100 were run in triplicate with corresponding blanks; threshold absorbance value for a positive test was calculated as the mean of the uninfected control sera plus three standard deviations (X+ 3s.d.).
Antisera Reference antisera were collected from cattle that had been experimentally infected with either Taenia saginata or Fasciola hepatica as part of other studies. Human infection sera were obtained from clinically confirmed cases of cysticercosis; control human sera were obtained from laboratory personnel with no history of infection. Monoclonal antibodies derived from mouse lymphocyte hybridomas were prepared against the ThFAS antigen fraction by commercial contract (Berkeley Antibody Company, Berkeley, California). Conjugated antibodies and corresponding substrates were obtained from Kirkegaard and Perry Laboratories (Gaithersburg, MD). RESULTS
Equal aliquots of all three fractions from each of the alkylagarose columns were analyzed by Western blot, using immune serum from a heavily infected calf (Fig. l ). Virtually all of the ThFAS applied to the agarose control column ( C = 0 ) was recovered in the non-adsorbed or "fall through" fraction. Increasing the length of the carbon chain (C) of the alkylagarose columns resulted in increased binding of ThFAS and, consequently, reduced quantities in the nnn-adgnrhad fractinn E x a m i n a t i n n n f e n l n m n eluate~ revealed that a small amount of the < 12 kDa antigen was recovered in the NaCI eluate from the hexylagarose (C = 6) columns; maximum recovery occurred in the ethylene glycol eluate from the hexylagarose (C = 6 ) column (arrow, Fig. 1). Aliquots of the ethylene glycol eluate from the hexylagarose column, designated C6Fx3, were further analyzed by Western blot using a battery of experimental antisera (Fig. 2). It was found that the < 12 kDa antigen (arrow) retained its reactivity with both bovine and human Taenia infection sera and with the monoclonal antibody prepared against the ThFAS fraction, but not with uninfected controls or with sera from cattle with Fasciola infections. Examination of the nitrocellulose strip that had been reacted with human Taenia solium infection serum revealed the presence of an additional reactive band with a relative mobility (Mr) of approximately 77 kDa; however, this band did not appear on any of the control strips. Next, the ethylene glycol eluate, C6Fx3, was compared with the ThFAS fraction for efficacy as coating antigen in a plate ELISA for the diagnosis of bovine cysticercosis. Both the ELISA using ThFAS and that using C6Fx3 could correctly distinguish between experimental T. saginata infections and the un-
Fig. 1. Western blot analysis of fractions obtained following the application of ThFAS to alkylagarose columns of varying carbon chain length. The nitrocellulose was reacted with serum from an experimentally infected calf with a heavy T. saginata parasite burden. Arrow indicates isolation of the < 12 kDa diagnostic antigen.
t'rl
Z
C ~r
~r
>
z
-r
m
O~
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
61
Fig. 2. Western blot analysis of the ethylene glycol eluate following application of ThFAS to the hexylagarose column. The diagnostic antigen (arrow) r~acted with bovine and human cysticercosis infection sera and with a specifc mo~oclonal antibody (6B7-G 12 ), but not with control sera or fascioliasis infection sera. B = blank,
infected controls, as depicted in Figs. 3 and 4, respectively. However, one of the F. hepatica infected cattle was found to be seropositive when ThFAS was used as coating antigen. (The threshold for a positive test was calculated as
E.G.HAYUNGAANDM.P.SUMNER
(3")
ThFAS Antigen E~
1.0OD . 0.6~ 0.4°-"
0
Con#als
T. sag/nala
+
F. hepa//ca +
Fig. 3. Plate ELISA absorbance values for cattle with experimental T. saginata and F. hepatica infections when the ThFAS preparation was used as coating antigen. Threshold value for a positive test (dotted line) was calculated at ,~+ 3s.d.
3 Anligen E,m 1.2 1.0
.
O.B 0.6
"i I).2
o
i--
Conl~ls
i
lnil.. r. mNjmta +
F . ~ +
Fig. 4. Plate ELISA absorbance values for cattle with experimental T. saginata and F. hepatica infections when the ethylene glycol eluate from the hexylagarose column (C6Fxa) was used as coating antigen. Threshold value for a positive test (dotted line) was calculated as .~+ 3s.d.
.(+ 3s.d., or three standard deviations greater than the mean of the uninfected controls.) In contrast, both F. hepatica infections were found to lie below the threshold value when C6Fx~ was used as coating antigen. DISCUSSION
A major obstacle to the development of immunoassays for helminthic diseases has been the inherent cross-reactivity between antigens of even distantly
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
63
related species (Hillyer and Serrano, 1983; Aronstein et al., 1986). Indeed, it was precisely by exploiting antigenic cross-reactivity that c.arlier investigators were able to diagnose bovine cysticercosis using antigenic material from Taenia hydatigena, a relatively abundant parasite of sheep, or Taenia crassiceps, which can be readily maintained in the laboratory, instead of Taenia saginata, which is more difficult to collect in quantity (Geerts et al., 1981; Wong et al., 1983). The challenge, then, in serodiagnostic test development is to select and isolate useful antigens from those with undesired cross-reactivity. Antigen isolation is critical, for even when sophisticated gene cloning techniques are used to produce large quantities of synthetic antigen, the recombinant peptides of interest must still be isolated from other bacterial products. Failure to remove unwanted cross-reacting components would be expected to result in reduced specificity of the assay. Of particular concern in the serodiagnosis of bovine cysticercosis are false positive reactions due to infection with Fasciola hepatica, a common trematode parasite of cattle (Craig and Rickard, 1980). Although reactivity with F. hepatica infection sera was greatly reduced in the ammonium sulfate soluble fraction of Taenia hydatigena cyst fluid, ThFAS (Rhoads et al., 1985), even residual cross-reactivity is likely to complicate test interpretation. Analysis of ThFAS by Western blot revealed the presence of both a prominent diagnostic antigen band that reacted very specifically with cysticercosis infection sera, and other bands that reacted with both cysticercosis and fascioliasis infection sera (Hayunga et al., 1991 ). Thus, when the ThFAS fraction was used directly as coating antigen in a plate ELISA, one of the two F. hepatica infected cattle examined presented an absorbance value that was significantly higher than the uninfected controls (Fig. 3 ). While this cross-reactivity may be of no consequence in identifying cattle heavily infected with fully developed cysts, which present greatly elevated absorbance values, it would probably be difficult to differentiate between the fascioliasis cases and light or early T. saginata infections. In contrast, there was no significant difference between F. hepatica infected cattle and uninfected controls when the column eluate was used as coating antigen (Fig. 4 ). Clearly, more work will be needed to determine the actual sensitivity and gpecificity of an assay using the purified antigen, particularly under field conditions where low grade infections are typical. Nevertheless, reduction of the unwanted reaction with F. hepatica infection sera to background levels is extremely promising. A number of methods had been employed in earlier attempts to isolate the diagnostic antigen, both in our laboratory and elsewhere. A major difficulty frequently encountered was the co-elution of sheep serum albumin, which is a major component of crude cyst fluid and which is not removed by the ammonium sulfate precipitation used to prepare ThFAS. Albumin is known to bind with many organic compounds, particularly fatty acids, and its physiological role as a scavenger molecule may help explain its apparent affinity for
64
E.G HAYUNGAAI';DM.P~SUMNER
the < 12 kDa parasite antigen (see review by Peters, 1985 ). We tried hydrophobic chromatography as an isolation technique because a preliminary analysis suggested that the < 12 kDa antigen was a lipoprotein (Rhoads et al., 1985) and because we had found that the antigen appeared to have affinity for the hydrophobic transfer membrane, Immobilon-P (Hayunga et al., 1991 ). We observed that the molecule has a reversible affinity for hexylagarose and used this property for its isolation; with alkylagarose columns of longer carbon chain length the affinity appears to be irreversible. Previously, Hayunga et al. ( 1991 ) isolated the diagnostic antigen by Western blot and found that it was sufficiently separated from other components to allow for unambiguous results in a dipstick type assay. The chief disadvantage of that method, however, was the time and expense involved in the Western blot separation. In the present paper, we have described a column chromatographic method to recover the diagnostic antigen in sufficient purity to allow for direct application in a conventional plate ELISA assay. The plate ELISA offers the advantage of compatibility with other immunoassays developed internally or available commercially, the capability of greater throughput, and the increased facility for system automation in the future. ACKNOWLEDGEMENTS
Taenia hydatigena cysts were obtained through the courtesy of Dr. E. Fajardo, Inspector-in-Charge, Monfort Lamb Company, Dixon, CA. Sera from experimentally infected cattle were provided by Dr. K. MurreU and Ms. M. Rhoads, Agxicultural Research Service, Beltsville, MD, and by Dr. G. Zimmerman, Oregon State University, Corvallis, OR. Human sera were collected by Dr. J. Cross, Uniformed Services University, Bethesda, MD, Dr. L. Rickman, Naval Hospital, Bethesda, MD, and Dr. J. Keiser, George Washington University Medical Center, Washington, DC. Monoclonal antibodies were pre:_~a~edby Dr. K. Balakrish~an, Berkeley Antibody Company, Berkeley, CA. We are also grateful to Mr. C. Watson, USDA, FSIS, Information Office, Washington, DC, for assisting in preparing illustrations, and to Ms. M. DiBari, Serology Branch, for editorial assistance.
REFERENCES Aronstein, W.S., Lewis, S.A., Norden, A.P., Dalton, J.P. and Strand, M., 1986. Molecular identity of a major antigen of Schistosoma mansoni which cross-reacts with Trichinella spiralis and Fasciola hepatica. Parasitology, 92:133-151. Craig, P.S. and Rickard, M.D., 1980. Evalution of"crude" antigen prepared from Taenia saginata for the serological diagnosis of 7". sagianta cysticercosis in cattle using the enzyme-linked immunosorbent assay ( ELISA ).Z. Parasitenk., 61: 287-297.
A Y-,~AGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
65
Dewhirst, L.W., Cramer, J.D. and Sheldon, J.J., 1967. An analysis of current inspection procedures for detecting bovine cysticercosis. J. Am. Vet. Med. Assoc., 150:412-417. Geerts, S., Kumar, V., Ceulemans, F. and Mortelmans, J., 1981. Serodiagnosis of Taenia saginata cysticercosis in experimentally and naturally infected cattle by enzyme linked immunosorbent assay. Res. Vet. Sci., 30: 288-293. Hayunga, E.G., M6llegird, I., Duncan, Jr., J.F., Sumner, M.P., Stek, Jr., M. and Hunter, Jr., K.W., 1987a. Early diagnosis of Schistosoma mansoni in mice using assays directed against cercarial antigens isolated by hydrophobic chromatography. J. Parasitol., 73:351-362. Hayunga, E.G., M~lleg~ird, I. and Sumner, M.P., 1987b. Analysis ofSchistosoma mansoni tegumental proteins by hydrophobic chromatography. Proc. Helminthol. Soc. Wash., 54:162165. Hayunga, E.G., Sumner, M.P., Rhoads, M.L., Murrell, K.D. and Isenstein, R.S., 1991. Development of a sergological assay for cysticercosis using an antigen isolated from Taenia cyst fluid. Am. J. Vet. Res. (in press) Hillyer, G.V. and Serrano, A.E., 1983. The antigens ofParagonimus westermani, Schistosoma mansoni and Fasciola hepatica adult worms. Am. J. Trop. Med. Hyg., 32: 350-358. Kamanga-Sollo, E.I.P., Rhoads, M.L. and Murrell, K.D., 1987. Evaluation of an antigenic fraction of Taenia hydatigena metacestode cyst fluid for immunodiagnosis of bovine cysticercosis. Am. J. Vet. Res., 48: 1206-1210. Peters, T., Jr., 1985. Serum albumin. Adv. Protein Chem., 37: 161-245. Rhoads, M.L., Murrell, K.D., Dilling, G.W., Wong, M.M. and Baker, N.F., 1985. A potential diagnostic reagent for bovine cysticercosis. J. Parasitol., 7 h 779-787. Sumner, M.P. and Hayunga, E.G., 1986. Characterization ofschistosome antigens by SDS-polyacrylamide gel electrophoresis. In: E.G. Hayunga and M. Stek, Jr. (Ed~itors), Nuclear and Related Techniques in Parasitology. International Atomic Energy Agency, Vienna, Austria, IAEA-TECDOC-371, pp. 133-144. Wong, M.M., Baker, N.F., Dilling, G.W., Rhoads, M.L. and Murrell, K.D., 1983. Comparison of various cestode antigens in serodiagnosis of bovine cysticercosis by ELISA. Fourth Annual Western Conference for Food Animal Veterinary Medicine, 28-30 March 1983.
57
Elsevier Science Publishers B.V., Amsterdam
Isolation and purification of a diagnostic antigen for bovine cysticercosis by hydrophobic chromatography* Eugene G. Hayunga mand Mary P. Sumner 2 Serology Branch, Pathology and Serology Division, Science and Technology Program, Food Safety and lnspection Service, U.S. Departmem of Agricuiture, Beltsville, MD 20705, LTA (Accepted 26 April 1990)
ABSTRACT Hayunga, E.G. and Sumner, M.P., 1991. Isolation and purification of a diagnostic antigen for bovine cysticercosis by hydrophobic chromatography. Vet. Immunol. ImmunopathoL, 28: 57-65. An ammonium sulfate fraction of Taenia hydatigena cyst fluid (ThFAS) was further fractionated by hydrophobic interaction chromatography, using alkylagarose and co-amino alkylagarose columns, in an effort to.isolate and purify a specific diagnostic antigen in the ThFAS preparation. The < 12 kDa antigen was found to have an affinity for immobilized alkanes with chain length of six carbons or greater. The antigen was recovered in an ethylene glycol eluate from a hexylagarose column then analyzed by Western blot; it reacted with bovine and human cysticercosis infection sera and with specific monoclonal antibodies but not with control sera or fascioliasis infection scra. When the eluate was used as coating antigen in a plate ELISA assay no false positive reactions were seen in sera from cattle infected with Fasciola hepatica; false positive reactions were observed for the unfractionated ThFAS antigen preparation.
INTRODUCTION
Serological assays offer the possibility of rapid diagnosis of helminthic diseases using non-invasive means. Immunoassays can detect infection prior to *The opinions or assertions contained herein are the private ones o f the authors and are not to be construed as official or reflecting the views o f the U.S. Department o f Agriculture, the Food Safety and Inspection Service or the Agricultural Research Service. The mention of a particular trade name or manufacturer is made for descriptive purposes only and is not intended to imply endorsement of that product over similar products from other commercial sources. The experiments reported herein were conducted according to the principles set forth in the "'Guide for the Care and Use o f Laboratory Animals," Institute o f Animal Resources, National Research Council, D H E W Publication No. ( N I H ) 78-23. 1Present address: National Institutes of Health, Westwood Building, Bethesda, MD 20892, USA. 2present address: Kirkegaard and Perry Laboratories, 2 Cessna Court, Gaithersburg, MD 20879, USA.
0165-2427/91/$03.50
© 1991 m Elsevier Science Publishers B.V.
58
E.G. HAYUNGAAND M.P. SUMNER
the onset of other clinical signs (Hayunga et al., 1987a), and may prove more sensitive than time-consuming organoleptic methods of food inspection (Dewhirst et al., 1967). One highly promising reagent for the immunodiagnosis of bovine cysticercosis is a < 12 kDa antigen found in an ammonium sulfate soluble fraction of Taenia hydatigena cyst fluid, ThFAS (Rhoads et al., 1985; Kamanga-Sollo et al., 1987; Hayunga et al., 1991 ). However, this fraction also contains components that react with sera from uninfected cattle (Hayunga et al., 1991 ) and therefore may contribute to false positive reactions when ThFAS is employed in a conventional plate ELISA. In the present paper, we report a liquid chromatographic method for purifying the diagnostic antigen from ThFAS. MATERIALS AND METHODS
Antigen preparation Taenia hydatigena cysts were collected from the mesenteries of naturally infected sheep at the time of slaughter; fluid was removed by incision, then filtered through two Nitex brand nylon screens, 110/zm and 44/tm exclusion size, to remove gross particulate contamination. An ammonium sulfate soluble fraction of the cyst fluid, ThFAS, was prepared as described previously (Rhoads et al., 1985). ThFAS was dialyzed against 0.05 M Tris-HCL, pH 8.0, and adjusted to a protein concentration of I mg/ml by the addition of buffer. Alkylagarose and oJ-amino alkylagarose were obtained commercially from Miles Scientific (Naperville, Indiana) as kits of six pre-packed 1 ml analytical columns containing 0-, 2-, 4-, 6-, 8- and !0-carbon chain !igands cova!ent!y bound to agarose. After equilibration of the columns with Tris-HCl, aliquots of approximately 1 mg ThFAS were applied and allowed to enter the gel bed at room temperature at a flow rate of approximately 0.45 ml/min (by gravity), then washed with additional buffer. The Tris-HCl "fall through" fraction, 1 M NaCI and 50% ethylene glycol fractions were collected as described previously (Hayunga et al., 1987b). Column fractions were dialyzed against several changes of distilled water, lyophilized and resuspended in SDS sample buffer. SDS polyaerylamide gel electrophoresis was conducted using 5-15 % 1.5 mm gradient slab gels run under reducing conditions in an SDS, Tris-glycine continuous buffer system, as described previously (Sumner and Hayunga, 1986). Proteins were then transferred to nitrocellulose by electroblotting at 0°-I0°C, 50 V constant voltage for 1 h, then 25 V overnight, using a Tris-glycine-methanol transfer buffer, as described previously (Hayunga et al., 1991 ). Blots were blocked with gelatin, then reacted with infection serum, alkaline phosphatase conjugated anti-IgG, and BCIP/NBT substrate. The enzyme-linked immunosorbent assay (ELISA) was performed using
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
59
peroxidase conjugate and ABTS substrate, as described previously (Hayunga et al., 1987a). Serum samples diluted 1: 100 were run in triplicate with corresponding blanks; threshold absorbance value for a positive test was calculated as the mean of the uninfected control sera plus three standard deviations (X+ 3s.d.).
Antisera Reference antisera were collected from cattle that had been experimentally infected with either Taenia saginata or Fasciola hepatica as part of other studies. Human infection sera were obtained from clinically confirmed cases of cysticercosis; control human sera were obtained from laboratory personnel with no history of infection. Monoclonal antibodies derived from mouse lymphocyte hybridomas were prepared against the ThFAS antigen fraction by commercial contract (Berkeley Antibody Company, Berkeley, California). Conjugated antibodies and corresponding substrates were obtained from Kirkegaard and Perry Laboratories (Gaithersburg, MD). RESULTS
Equal aliquots of all three fractions from each of the alkylagarose columns were analyzed by Western blot, using immune serum from a heavily infected calf (Fig. l ). Virtually all of the ThFAS applied to the agarose control column ( C = 0 ) was recovered in the non-adsorbed or "fall through" fraction. Increasing the length of the carbon chain (C) of the alkylagarose columns resulted in increased binding of ThFAS and, consequently, reduced quantities in the nnn-adgnrhad fractinn E x a m i n a t i n n n f e n l n m n eluate~ revealed that a small amount of the < 12 kDa antigen was recovered in the NaCI eluate from the hexylagarose (C = 6) columns; maximum recovery occurred in the ethylene glycol eluate from the hexylagarose (C = 6 ) column (arrow, Fig. 1). Aliquots of the ethylene glycol eluate from the hexylagarose column, designated C6Fx3, were further analyzed by Western blot using a battery of experimental antisera (Fig. 2). It was found that the < 12 kDa antigen (arrow) retained its reactivity with both bovine and human Taenia infection sera and with the monoclonal antibody prepared against the ThFAS fraction, but not with uninfected controls or with sera from cattle with Fasciola infections. Examination of the nitrocellulose strip that had been reacted with human Taenia solium infection serum revealed the presence of an additional reactive band with a relative mobility (Mr) of approximately 77 kDa; however, this band did not appear on any of the control strips. Next, the ethylene glycol eluate, C6Fx3, was compared with the ThFAS fraction for efficacy as coating antigen in a plate ELISA for the diagnosis of bovine cysticercosis. Both the ELISA using ThFAS and that using C6Fx3 could correctly distinguish between experimental T. saginata infections and the un-
Fig. 1. Western blot analysis of fractions obtained following the application of ThFAS to alkylagarose columns of varying carbon chain length. The nitrocellulose was reacted with serum from an experimentally infected calf with a heavy T. saginata parasite burden. Arrow indicates isolation of the < 12 kDa diagnostic antigen.
t'rl
Z
C ~r
~r
>
z
-r
m
O~
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
61
Fig. 2. Western blot analysis of the ethylene glycol eluate following application of ThFAS to the hexylagarose column. The diagnostic antigen (arrow) r~acted with bovine and human cysticercosis infection sera and with a specifc mo~oclonal antibody (6B7-G 12 ), but not with control sera or fascioliasis infection sera. B = blank,
infected controls, as depicted in Figs. 3 and 4, respectively. However, one of the F. hepatica infected cattle was found to be seropositive when ThFAS was used as coating antigen. (The threshold for a positive test was calculated as
E.G.HAYUNGAANDM.P.SUMNER
(3")
ThFAS Antigen E~
1.0OD . 0.6~ 0.4°-"
0
Con#als
T. sag/nala
+
F. hepa//ca +
Fig. 3. Plate ELISA absorbance values for cattle with experimental T. saginata and F. hepatica infections when the ThFAS preparation was used as coating antigen. Threshold value for a positive test (dotted line) was calculated at ,~+ 3s.d.
3 Anligen E,m 1.2 1.0
.
O.B 0.6
"i I).2
o
i--
Conl~ls
i
lnil.. r. mNjmta +
F . ~ +
Fig. 4. Plate ELISA absorbance values for cattle with experimental T. saginata and F. hepatica infections when the ethylene glycol eluate from the hexylagarose column (C6Fxa) was used as coating antigen. Threshold value for a positive test (dotted line) was calculated as .~+ 3s.d.
.(+ 3s.d., or three standard deviations greater than the mean of the uninfected controls.) In contrast, both F. hepatica infections were found to lie below the threshold value when C6Fx~ was used as coating antigen. DISCUSSION
A major obstacle to the development of immunoassays for helminthic diseases has been the inherent cross-reactivity between antigens of even distantly
A DIAGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
63
related species (Hillyer and Serrano, 1983; Aronstein et al., 1986). Indeed, it was precisely by exploiting antigenic cross-reactivity that c.arlier investigators were able to diagnose bovine cysticercosis using antigenic material from Taenia hydatigena, a relatively abundant parasite of sheep, or Taenia crassiceps, which can be readily maintained in the laboratory, instead of Taenia saginata, which is more difficult to collect in quantity (Geerts et al., 1981; Wong et al., 1983). The challenge, then, in serodiagnostic test development is to select and isolate useful antigens from those with undesired cross-reactivity. Antigen isolation is critical, for even when sophisticated gene cloning techniques are used to produce large quantities of synthetic antigen, the recombinant peptides of interest must still be isolated from other bacterial products. Failure to remove unwanted cross-reacting components would be expected to result in reduced specificity of the assay. Of particular concern in the serodiagnosis of bovine cysticercosis are false positive reactions due to infection with Fasciola hepatica, a common trematode parasite of cattle (Craig and Rickard, 1980). Although reactivity with F. hepatica infection sera was greatly reduced in the ammonium sulfate soluble fraction of Taenia hydatigena cyst fluid, ThFAS (Rhoads et al., 1985), even residual cross-reactivity is likely to complicate test interpretation. Analysis of ThFAS by Western blot revealed the presence of both a prominent diagnostic antigen band that reacted very specifically with cysticercosis infection sera, and other bands that reacted with both cysticercosis and fascioliasis infection sera (Hayunga et al., 1991 ). Thus, when the ThFAS fraction was used directly as coating antigen in a plate ELISA, one of the two F. hepatica infected cattle examined presented an absorbance value that was significantly higher than the uninfected controls (Fig. 3 ). While this cross-reactivity may be of no consequence in identifying cattle heavily infected with fully developed cysts, which present greatly elevated absorbance values, it would probably be difficult to differentiate between the fascioliasis cases and light or early T. saginata infections. In contrast, there was no significant difference between F. hepatica infected cattle and uninfected controls when the column eluate was used as coating antigen (Fig. 4 ). Clearly, more work will be needed to determine the actual sensitivity and gpecificity of an assay using the purified antigen, particularly under field conditions where low grade infections are typical. Nevertheless, reduction of the unwanted reaction with F. hepatica infection sera to background levels is extremely promising. A number of methods had been employed in earlier attempts to isolate the diagnostic antigen, both in our laboratory and elsewhere. A major difficulty frequently encountered was the co-elution of sheep serum albumin, which is a major component of crude cyst fluid and which is not removed by the ammonium sulfate precipitation used to prepare ThFAS. Albumin is known to bind with many organic compounds, particularly fatty acids, and its physiological role as a scavenger molecule may help explain its apparent affinity for
64
E.G HAYUNGAAI';DM.P~SUMNER
the < 12 kDa parasite antigen (see review by Peters, 1985 ). We tried hydrophobic chromatography as an isolation technique because a preliminary analysis suggested that the < 12 kDa antigen was a lipoprotein (Rhoads et al., 1985) and because we had found that the antigen appeared to have affinity for the hydrophobic transfer membrane, Immobilon-P (Hayunga et al., 1991 ). We observed that the molecule has a reversible affinity for hexylagarose and used this property for its isolation; with alkylagarose columns of longer carbon chain length the affinity appears to be irreversible. Previously, Hayunga et al. ( 1991 ) isolated the diagnostic antigen by Western blot and found that it was sufficiently separated from other components to allow for unambiguous results in a dipstick type assay. The chief disadvantage of that method, however, was the time and expense involved in the Western blot separation. In the present paper, we have described a column chromatographic method to recover the diagnostic antigen in sufficient purity to allow for direct application in a conventional plate ELISA assay. The plate ELISA offers the advantage of compatibility with other immunoassays developed internally or available commercially, the capability of greater throughput, and the increased facility for system automation in the future. ACKNOWLEDGEMENTS
Taenia hydatigena cysts were obtained through the courtesy of Dr. E. Fajardo, Inspector-in-Charge, Monfort Lamb Company, Dixon, CA. Sera from experimentally infected cattle were provided by Dr. K. MurreU and Ms. M. Rhoads, Agxicultural Research Service, Beltsville, MD, and by Dr. G. Zimmerman, Oregon State University, Corvallis, OR. Human sera were collected by Dr. J. Cross, Uniformed Services University, Bethesda, MD, Dr. L. Rickman, Naval Hospital, Bethesda, MD, and Dr. J. Keiser, George Washington University Medical Center, Washington, DC. Monoclonal antibodies were pre:_~a~edby Dr. K. Balakrish~an, Berkeley Antibody Company, Berkeley, CA. We are also grateful to Mr. C. Watson, USDA, FSIS, Information Office, Washington, DC, for assisting in preparing illustrations, and to Ms. M. DiBari, Serology Branch, for editorial assistance.
REFERENCES Aronstein, W.S., Lewis, S.A., Norden, A.P., Dalton, J.P. and Strand, M., 1986. Molecular identity of a major antigen of Schistosoma mansoni which cross-reacts with Trichinella spiralis and Fasciola hepatica. Parasitology, 92:133-151. Craig, P.S. and Rickard, M.D., 1980. Evalution of"crude" antigen prepared from Taenia saginata for the serological diagnosis of 7". sagianta cysticercosis in cattle using the enzyme-linked immunosorbent assay ( ELISA ).Z. Parasitenk., 61: 287-297.
A Y-,~AGNOSTIC ANTIGEN FOR BOVINE CYSTICERCOSIS
65
Dewhirst, L.W., Cramer, J.D. and Sheldon, J.J., 1967. An analysis of current inspection procedures for detecting bovine cysticercosis. J. Am. Vet. Med. Assoc., 150:412-417. Geerts, S., Kumar, V., Ceulemans, F. and Mortelmans, J., 1981. Serodiagnosis of Taenia saginata cysticercosis in experimentally and naturally infected cattle by enzyme linked immunosorbent assay. Res. Vet. Sci., 30: 288-293. Hayunga, E.G., M6llegird, I., Duncan, Jr., J.F., Sumner, M.P., Stek, Jr., M. and Hunter, Jr., K.W., 1987a. Early diagnosis of Schistosoma mansoni in mice using assays directed against cercarial antigens isolated by hydrophobic chromatography. J. Parasitol., 73:351-362. Hayunga, E.G., M~lleg~ird, I. and Sumner, M.P., 1987b. Analysis ofSchistosoma mansoni tegumental proteins by hydrophobic chromatography. Proc. Helminthol. Soc. Wash., 54:162165. Hayunga, E.G., Sumner, M.P., Rhoads, M.L., Murrell, K.D. and Isenstein, R.S., 1991. Development of a sergological assay for cysticercosis using an antigen isolated from Taenia cyst fluid. Am. J. Vet. Res. (in press) Hillyer, G.V. and Serrano, A.E., 1983. The antigens ofParagonimus westermani, Schistosoma mansoni and Fasciola hepatica adult worms. Am. J. Trop. Med. Hyg., 32: 350-358. Kamanga-Sollo, E.I.P., Rhoads, M.L. and Murrell, K.D., 1987. Evaluation of an antigenic fraction of Taenia hydatigena metacestode cyst fluid for immunodiagnosis of bovine cysticercosis. Am. J. Vet. Res., 48: 1206-1210. Peters, T., Jr., 1985. Serum albumin. Adv. Protein Chem., 37: 161-245. Rhoads, M.L., Murrell, K.D., Dilling, G.W., Wong, M.M. and Baker, N.F., 1985. A potential diagnostic reagent for bovine cysticercosis. J. Parasitol., 7 h 779-787. Sumner, M.P. and Hayunga, E.G., 1986. Characterization ofschistosome antigens by SDS-polyacrylamide gel electrophoresis. In: E.G. Hayunga and M. Stek, Jr. (Ed~itors), Nuclear and Related Techniques in Parasitology. International Atomic Energy Agency, Vienna, Austria, IAEA-TECDOC-371, pp. 133-144. Wong, M.M., Baker, N.F., Dilling, G.W., Rhoads, M.L. and Murrell, K.D., 1983. Comparison of various cestode antigens in serodiagnosis of bovine cysticercosis by ELISA. Fourth Annual Western Conference for Food Animal Veterinary Medicine, 28-30 March 1983.
