Immunoblot evaluation of the 100 and 130 kDa antigens in camel hydatid cyst fluid for the serodiagnosis of human cystic echinococcosis in Libya

276 TRANSACTIONS

OF THE ROYAL

SOCIETY OF

TROPICAL MEDICINE

AND HYGIENE (1995) 89, 276279

lmmunoblot evaluation of the 100 and 130 kDa antigens in camel hydatid fluid for the serodiagnosis of human cystic echinococcosis in Libya

cyst

M.K. Shambesh1j2, P.S. Craig I*, A.M. Gusbi3, E.F. Echtuish4 and H. Wen 1S ‘Department of Biological Sciences, University of Sal ord, Sa&d, MS 4W, UK; 2Department of Community Medicine, Faculty of Medicine, Al Fateh Universi~, Tripoli, Libya; f Faculty o Veterinary Medicine, Al Fateh University, Tripoli, Libya; 4Department of Surgey, Tripoli Central Hospital, Tripoli, Libya; f Department of Surgery, Xinjiang Medical College, Urumqi, Xinjiang, People’s Republic of China

Abstract Two high molecular weight antigens with molecular massesapproximately 100 kDa and 130 kDa were identified bv immunoblottine camel hvdatid cvst fluid, with 94% sensitivitv in sera from sureicallv confirmed Libyan cystic echino&ccosis cases.40% of sera’from surgically confirmed alveolar echmocdccosis casescross-reactedwith the 100 and 130 kDa antigens, as did 5.3% of sera from human Taenia solium cysticercosis patients. No cross-reaction occurred with sera from human schistosomiasismansoni or onchocerciasis patients. In addition, all serafrom patients with non-hydatld space-occupying lesions (i.e. simple liver cysts, kidney cysts, lung tuberculosis, pulmonary carcinoma, pulmonary empyema, and lung abscess)were seronegative against the same antigens, as were control serum samples from healthy individuals. The 100 and 130 kDa antigens were strongly recognized by sera from cystic echinococcosis patients when camel or horse hydatid cyst fluid was used in immunoblotting but were only weakly recognized if sheepor human hydatid cyst fluid was used. Camel hydatid cyst fluid could be an important source of diagnostic antigens for human cystic ecbinococcosisin the Middle East endemic region. Keywords:

cystic echinococcosis,Echinococcusgranulosus, serodiagnosis,Libya

Introduction Human cystic echinococcosis (CE) is highly endemic in most of the Mediterranean basin including North Africa and the Middle East (MATOSSIAN et al., 1977). Echinococcusgranulosus occurs over much of northern Libya, with high prevalence rates in sheep (8.4%) and dogs (31.5%) (GUSBI et al., 1987a,1987b). The parasite also readily infects camels (Camelus dromedarius), with a prevalence of 31.9% recorded by GUSBI et al. (1990). Human CE is an important public health problem in Libya, which requires both surgical and chemotherapeutic intervention (EL-MUFTI et al., 1993). A retrospective hospital review from 1971 to 1979in the surgical departments of hospitals in the north of Libya showed that 291 casesof cystic hydatid diseasewere confirmed surgicallv among 43 979 admissions, an incidence of 0.7% (DAR & TA&JRI, 1978; ABOUDAYA, 1985). A community orevalence rate of 2% for human CE has recentlv been recorded in north Libya in an ultrasound scanning survey (SHAMBESH et al., 1992).

Diagnosis of human CE commonly involves serological confirmation based on specific antibody detection by indirect haemagglutination tests or enzyme-linked immunosorbent assays(ELISA). Both tests have high sensitivity but variable specificity with crude hydatid cyst fluid antigens, which are usually derived from sheep or cattle cysts (RICKARD & LIGHTOWLERS,1986; CRAIG, 1993). In order to improve specificity for immunodiagnosis of human CE, hydatid cyst fluid antigens have been evaluated by immunoblotting, predominantly with respect to the low molecular weight subunits (i.e 8-24 kDa) of Echinococcus antigen B (MADDISON et al., 1989; LEGGATTet al., 1992). The sensitivity of immunoblotting for hydatid cyst fluid antigen B recognition is, however, subject to variation with reported values of SO-90%(CRAIG, 1993). At present neither ELISA nor immunoblot confirmatory testing are routinely used in Libya for immunodiagnosis of human CE. The current study was undertaken to assessthe notential use of camel hvdatid cvst fluid, which is readily available in Libya, as-antigen; in cornparison with sheep, human and horse hydatid cyst fluids for immunoblot confirmation of human CE. In doing this, 2 new candidate diagnostic antigens with molecular massesof 100 and 130 kDa were identified in camel hydatid cyst fluid.

*Author for correspondence.

Materials and Methods Antigens

Camel and sheep hydatid cyst fluid extracts were collected from infected animals at slaughter in Libyan abattoirs. Human hydatid fluid was collected from a Libyan patient with hepatic CE after surgery. Horse hydatid cyst fluid was collected from an abattoir in the UK. All cyst fluids were clarified by centrifugation at 2000g for 10 min, separated into aliquots and stored at -20°C until used. Serum samples

Blood samples were collected from 99 surgically proven CE patients (84 from Libyan patients and 15 Chinese patients) and from 15 Chinese alveolar echinococcosis(AE) patients. Heterologous non-Echinococcus human serum controls were derived from 183 persons, comprising 38 Taenia sohum cysticercosis patients (from Guatemala), 28 onchocerciasis patients (from Sierra Leone), 17 patients infected with Schistosoma mansoni (from Ken a), 71 Libyan patients with non-hydatid simple cysts (4Y patients with simple kidney cysts and 26 with simple liver cysts), 19 Libyan patients with lung abscess,empyema, pleural effusion or pneumonia, 5 Libyan patients with lung cancer, and 5 patients with pulmonary tuberculosis. Twenty serum samples for use as negative controls were collected from healthy Libyan volunteers. Electrophoresis and immunoblotting

Hydatid cyst fluids (camel, sheep, horse or human) were separated under reducing conditions on sodium dodecyl- sulphate-polyacrylamide gel electrophoresis (SDS-PAGE)using 8% or 12.5% mini-gels as described &eviouslv ~SHEP~~ERD& MCMANUS, 1987; LIGHTOWiERS et ai., i986,1989).

Total oroteins were stained with Coomassie brilliant blue. Aniigenic comnonents in camel, sheep and human hydatid cyst fluids resolved by SDS-PAGEwere electroohoreticallv transfered at 200mA for 2.5 h to nitrocellulose paper-(NCP) using a mini-blot apparatus (Hoefer, USA). Non-binding sites on the NCP were blocked using 0.15M phosphate buffered saline (PBS) containing 0.3% Tween 20@in 5% milk (pH 7.2) for one hour. Following blocking, the NCP was probed with sera from cystic echinococcosis cases,sera from other infections, or normal seradiluted 1:200 in PBS+ 5% milk (pH 7.2) and incubated for one hour using a mini-trough system (Biometra, Germany). After incubation, the NCP was washed 3 times with

277

PBS containing 0.3% Tween 20@(pH 7.2) and incubated for 1.5 h with alkaline phosphatase-conjugated rabbit antihuman immunoglobulin G (IgG) (Fe specific, Sigma) diluted 1:lO 000 in PBS +5% milk (pH 7.2). The reaction was visualized using 5-bromo-4-chloro-3-indolyl phosphate /nitro blue tetrazolium as substrate.

Fig. 1. Proteins of Echinococc~gmnulosu~cyst fluids separatedby sodium dodecyl sulphate-polyacrylamide gel electrophoresis and stained with Coomassie brilliant blue. Hydatid cyst fluid samples from camel (l), horse (2), human (3) and sheep(4). (K and k indicate molecular massesin kDa.)

Results The major hydatid fluid proteins revealed were primarily antigen B subunits (8,16 and 24 kDa) and antigen 5 subunit (38 kDa) in all 4 hydatid cyst fluid samples. In addition, 2 higher molecular massproteins (100 kDa and 130 kDa) were detected, especially in camel and horse hydatid fluids (Fig. 1). Immunoblotting the hydatid cyst fluid antigens separated by SDS-PAGEon 12.5% els and screening with sera from human CE patients resu! ted in strong recognition of the antigen B subunits (8,16 and 24 kDa) and the major antigen 5 subunit (38 kDa) in the 3 hydatid cyst fluid samplesused (camel, sheep, human) (Fig.2). In addition, CE patients sera also strongly recognized the 2 higher molecular mass proteins (approximately 100 and 130 kDa) in camel and horse hydatid cyst fluid. These 2 high molecular mass antigens were recognized on immunoblots transferred from either 12.5% or 8% gels, but the recognition was more marked with 8% gels (Fig.3).

Fig. 2. Immunoblots on 12.5% gels of hydatid cyst fluids probed with sera from patients infected with Echinococcusgranulosus. Hydatid cyst fluid antigen from camel (panel 1,6 sera), human (panel 2, 5 sera) and sheep(panel 3,s sera); control (negative) sera with camel cyst fluid antigen are shown in panel 4. (K and k indicate molecular massesin kDa.)

Table. Immunoblot reactivity of the high (loo/130 kDa) and low molecular weight bands (8,16,24 and 38 kDa) with human hydatid sera and other human sera using camel hydatid cyst fluid antigen Infection Echinococcus granulosus E. multilocularis Taenia soliuma Schistosoma mansoni Onchocerca volvulus

Tested ?T 38 is7

Simple liver cysts Simple kidney cysts 2 Tuberculosis (lung) 5 Carcinoma (lun$ 5 Chest infection Controls (negative) :; aCysticercosis. bLung abscess,empyema, pleural effusion, pneumonia.

Numbers of cases With 100/130 With antigen kDa antigen 5 (38 kDa)

With anti en B (8, 16, 24gkDa)

93 94% 6 r 40% j

93 94% 6 g40% j

5 Y4%) 0

2 T3%)

i i z

El 8 8 0

278

Fig. 3. Immunoblotson 8%gelsof camelhydatidcystfluid probedwith serafrom patientsinfectedwith Echinococcusgranulosus (panel1,5 sera),E. multi1onJaris(panel3,8 sera),Taenias&m (cysticercosis; panel4,6 sera)andSchistoratammsoni (panel 5, 10sera).Sevencontrol(negative)seraareshownits panel2. (K andk indicatemolecularmassesin kDa.)

Human CE sera only very weakly recognized the 100 and 130 kDa antigens in sheep or human hydatid cyst fluids, even though antibodies bound well to both antigen B and the antigen 5 subunits in these fluids (Fia.2). The 2 urotein bands of 100 and 130 kDa therefore gave posit&e binding reactivity with surgically confirmed human cystic hydatid sera using camel hydatid cyst fluid, but not when using human or sheep hydatid cyst fluids. The sensitivity of immunoblotting for recognition of these high molecular mass bands (100 and 130 kDa) in camel hydatid cyst fluid was 94% with sera from surgically confiied hepatic CE cases.Binding to the 100and 130 kDa bands did not occur with control negative sera (Table). Cross-reactivity of the 100 and 130 kDa antigens in camel hydatid cyst fluid with human sera from other taeniid larval cestodeinfections occurred with 40% of the alveolar echinococcosis sera and 5.3% of the cysticercosis sera. Sera from donors infected with S. mansoni or onchocerciasis patients did not cross-react with the 100 or 130 kDa antigens, and no cross-reactivity was observed with sera from patients with simple liver or kidney cysts, lung tuberculosis, lung carcinoma, lung abscess,or lung empyema. The overall specificity of the 100and 130 kDa antigens in camel cyst fluid was 96% (190/198). The sensitivitv of soecific immtmoblot bindine to the antigen B subunits in’camel hydatid cyst fluid was 94%, as observed for the 100 and 130 kDa antigens. Only 5.3% of cysticercosis sera, and none of the sera from patients with schistosomiasis, onchocerciasis and nonhydatid cystic lesions, cross-reacted with the major antigen B subunits of 8,16 or 24 kDa. Discussion Immunodiagnosis of human CE is usually basedon detection of specific serum antibodies against E. granulosus hydatid cyst fluid antigens, primarily of sheephost origin (RICKARD & LIGHTOWLERS, 1986; CRAIG, 1993). The 2 major antigens in hydatid fluid, the lipoprotein antigen 5 and antigen B, have been characterized by SDS-PAGE and Western blotting (LIGHTOWLERS et al., 1989). Antigen 5 consists of 2 major subunits of molecular mass 38 and 20 kDa (under reducing conditions) and, although highly sensitive in immunoblotting for human

CE, cross-reactivity with sera from other hehninthic diseasesis a roblem (LIGHTOWLERS et al., 1989; MADDISON et a P.) 1989). Antigen B subunits of E. granulosus hydatid cyst fluid comprise 8 kDa (or 12 kDa), 16 kDa and 24 kDa antigens under reducing (and non-reducing) conditions and recognition of these low molecular mass subunits appears to be highly genus-specific for human echinococcosis (including both E. granulosus and E. muZrilocularis), with a low level of cross-reactivity (about 5%) with T. solium cysticercosis (MADDISON et al., 1989; LEGGATTet al., 1992). However, reports of the sensitivity of E. granulosus B antigen subunit recognition by human CE sera in immunoblottina have indicated some variability (SIRACLJSANOet al., 199‘i; CRAIG, 1993). Although several higher molecular mass E. granulosus cyst fluid antigens have been observed after SDS-PAGE separation and immunoblotting (CRAIG, 1986; GOTTSTEIN et al., 1987), the recognition of high molecular mass proteins (>65 kDa) in E. granulosus cyst fluid by sera from CE patients has not generally been considered of diagnostic importance and has been ignored. On the contrary, the present stud showed that 2 proteins with high molecular masses oY approximately 100 and 130 kDa in camel and horse E. granulosus hydatid cyst fluid are recognized by human CE sera, with the samedegree of sensitivity (94%) and specificity as obtained with the low molecular mass subunits of antigen B. The sensitivity of total IgG antibody recognition of the 100 and 130 kDa antigens in camel hydatid cyst fluid by sera from the 99 CE patients was 94%, which was slightly higher than that of the 8 kDa subunit of antigen B in sheep hydatid cyst fluid (91%) reported by MADDISON et al. (1989). It was also higher than the sensitivity of recognition of the 12 kDa subunit of sheep B antigen (90.9%) reported by LEGGA?Tet al. (1992). However, our study also showed a higher level of sensitivity (94%) for the antigen B subunits. The cross-reactivity rate between sera from T. solium cysticercosis casesand the 12 kDa band in sheep hvdatid cyst fluid was re orted to be 5% (LEGGATTet al.,-1992; VERASTEGUI et aP. , 1992). which is similar to the level of cross-reactivity found in- the current study for the 100 and 130 kDa antigens in camel hydatid cyst fluid. The cross-reactivity rate between serafrom patients with alveolar echinococcosis and the 100 and 130 kDa antigens,

279

and the antigen B subunits in camel and horse hydatid cvst fluid was the same (40%) as that reuorted for the antigen B subunits in shiep hydatid fluid (LEGGATT et al.. 1992). However. E. multilocularis is not endemic

in

L&y;. do cross-reaction occurred between the 100 and 130 kDa antigens and any of the sera from schistosomiasis, onchocerciasis, tuberculosis or cancer patients. The current study also confirmed previous immunoblot studies which found high sensitivity but low specificity of recognition of the 39-40 kDa subunits of antigen 5 in E. granulosus hydatid cyst fluid (SIRACUSANO et al., 1991; SHAPIRO et al., 1992). The 100 and 130 kDa antigens reported here were resent in camel, horse, sheep and human hydatid cyst 8.mds, but were strongly antigenic only in the camel and horse fluids. These antigens are possibly the sameas the 110-120 kDa and the 150 kDa antigens described in sheep hydatid cyst fluid by SHAPIRO et al. (1992) and the 116 kDa antigen in sheep, goat, pig and human cyst fluids (KANWAR et al., 1992). The present results with camel hydatid cyst fluid are consistent with the reported high itiunoblot sensitivity for recognition by serafrom human CE natients of the 110-120 kDa antigens (89%). the 150 kca (81%) antigen, and the 116 kba hidati; cyst fluid antigen (100%) (KANWAR et al., 1992; SHAPIRO et al., 1992). In conclusion, the current study indicated that 2 high molecular mass Echinococcus antigens, of approximately 100 kDa and 130 kDa present in camel hydatid cyst fluid, are strongly recognized by human CE sera, with sensitivity of 94% and overall specificity of 96%. These values are at least as good as the sensitivity and specificity described here or in other studies for the 8 and 12 kDa subunits of E. granulosus cyst fluid antigen B. It is suggested that in Libya, and possibly in other Middle East countries with endemic hydatid disease, human seropositivity for CE could be confirmed using camel hydatid cyst fluid in immunoblotting for recognition of the 100 and 130 kDa antigens in addition to the 8-12 kDa subunits of antigen B. Acknowledgements We thank Dr M.T. Rogan for useful advice and comments, Dr David Taylor (Department of Pathology, University of Cambridge, UK) for the gift of schistosomiasisand onchocerciasis sera, and Dr Jose Garcia (University of San Carlos, Guatemala) for cysticercosis sera. This work was supported by the government of Libya and by the Wellcome Trust. References Aboupay?, M,. A. (1985). Prevalence of human hydatidosis in T;;poh region of Libya. Garyoums MedxalJoumal, 9, 307Craig, I’. S. (1986). Detection of specific circulating antigen, immune complexes, and antibodies in human hydatidosis from Turkana (Kenya) and Great Britain by enzyme-immunoassay.ParasiteImmunology, 8,171-181. Craig, I’. S. (1993). Immunodiagnosis of Echinococcusgranulosus. In: Compendium on Cystic Echinococcosis (with special referenceto the Xinjiang Uygur autonomous region, the People’s Republic ofchina), Andersen, F. L., Chai, J. & Liu, D. (editors). Provo, Utah: Brigham Young University, pp. 85-118. Dar, F. K. & Taguri, S. (1978). Human hydatid diseasein eastern Libya. Transactions of the Royal Society of Tropical Medicine and Hygiene, 72,313-314.

El-Mufti, M., Kamag, A., Ibrahim, H., Taktuk, S., Swaisi, I., Zaidan, A., Sameen, A., Shimbish, F., Bouzghaiba, W. & Hassi, S. (1993). Albendazole therapy of hydatid disease: 2-

year follow-up of 40 cases. Annals of Tropical Medicine and Parasitology, 87,241-246. Gottstein, B., Eckert, J., Michael, S. A. & Thompson, R. C. A. (1987). Echinococcus granulosus antigens: immunoelectro-

phoretic and Westernblot analysisof hydatid cyst fluids.

ParasitologyResearch, 73,186-189. Gusbi, A. M., Awan, M. A. Q. & Beesley, W. N. (1987a). Echinococcus granulosus in dogs with particular referenceto the role of the dog in Libyan society. Annals of Tropical Medicine and Parasitology, 81,29-34.

Gusbi. A. M.. Awan. M. A. 0. & Beeslev. W. N. (1987b). Ecl&ococcdsis in Libya, II. Prevalence oihydatido& (E&inococcusgranulosus) in sheep. Annals of Tropical Medicine and Parasitology, 81,35-41. Gusbi, A. M., Awan, M. A. 0. & Beeslev. W. N. (1990). Echinocbccosisin Libia. IV. P;evalence ii hydatidisis @chinococcus granulosus) in goats, cattle and camels. Annals of Tropical Medicine and Parasrtology 84,477-482.

Kanwar, J. R., Kaushik, S. P., Sawhney, I. M. S., Kamboj, M. S., Mehta, S. K. & Vinayak, V. K. (1992). Specific antibodies in serum of patients with hydatidosis recognised by immunoblotting.Joumal ofMedical Microbiology, 36,46-S 1. Leggatt, G. R., Yang, W. & McManus, D. P. (1992). Serological evaluation of the 12 kDa subunit of antigen B in Echim coccus granulosus cyst fluid bv immunoblot analvsis. Transacsons of the R&al Society of-Tropical Medicine and.Hy-

p&86.189-192. LGhtowlers, M. W., Rickard, M. D. & Honey, R. D. (1986). Serum antibody res onse following parenteral immunization with hydatid cyst fvuld in sheep infected with Echinacoccus granulosus. AmericanJournal

of Tropical Medicine and Hygiene,

35,818-823. Lightowlers, M. W., Liu, D., Haralambous, A. & Ricard, M. D. (1989). Subunit composition and specdicity of the major cyst fluid antigens of Echinococcus granulosus. Molecular and BiochemicalParasitology, 37, 171-182. Maddison, S. E., Slemenda, S. B., Schantz, I’. M., Fried, J. A., Wilson, M. & Tsang, V.C. (1989). A specific diagnostic antigen of Echinococcusgranulosus with an apparent molecular weight of 8 kDa. American Journal of Tropical Medicine and Hygiene, 40,377-383.

Matossian, R. M., Rickard, M. D. 81Smyth, J. D. (1977). Hydatidosis a global problem of increasing importance. Bulletin of the World Health Organization, 55,49%507.

Rickard, M. D. & Lightowlers, M. W. (1986). Immunodiagnosis of hydatid disease. In: Biology of Echinococcus and Hydatid Disease, RCA. Thomson (editor). London: George Allen & Unwin, pp. 217-249. Shambesh, M. K., Macpherson, C. N. L., Beesley, W. N. & Gusbi, A. (1992). Prevalence of human hydatid disease in north-western Libya: a cross-sectionalultrasound study. Annals of Tropical Medicine and Parasitology, 86,381-386. Shapiro, S. Z., Bahr, G. M. & Hira, P. R. (1992). Analysis of host components in hydatid cyst fluid and immunoblot diagnosis of human Echinococcus granulosus infection. Annals of Tropical Medicine and Parasitology, 86,503-509.

Shepherd, J. C. & McManus, D. I’. (1987). Specific and cross reactive antigens of Echinococcusgranulosus hydatid cyst fluid. Molecular and Biochemical Parasitology, 25,143-154.

Siracusano,A., Ioppolo, S., Notargiacomo, S., Ortona, E., Rigano, R., Teggi, A., De Rosa, F. & Vicari, G. (1991). Detection of antibodies against Echinococcus granulosus major antigens and their subunits by immunoblotting. Transactions of the Royal Society of Tropical Medicine and Hygiene, 85,239-

243. Verastegui, M., Moro, P., Guevara, A., Rodriguez, T., Miranda, E. & Gilman, R. H. (1992). Enzyme-linked immunoelectrotransfer blot test for diagnosis of human hydatid disease.3oumal ofMedicalMicrobiologv, 67,129-143. Received 14 September 1994; accepted for publication 1.5 November 1995