Eur. J. Epidemiol. 0392-2990
EUROPEAN
Vol. 8, No. 5
JOURNAL
September 1992, p. 677-682
Ov
EPIDEMIOLOGY
EVALUATION OF AN IMMUNOENZI £ TIC TEST (ELISA) FOR THE DIAGNOSIS OF LEPTOSPIROSIS IN ITALY M. CINCO 1, D. BALANZ1N and E. BANFI Istituto di Microbiologia - Universitgt degli Studi - Via Fleming, 22 - Trieste - Italy.
Key words: ELISA - Leptospirosis - A u s t r a l i s
-
Epidemiology
An enzyme-linked immunosorbent assay (ELISA) was assessed to detect in humans IgM and IgG against the main serogroups/serovars of Leptospira present in Italy. Sonicated antigens from strain Wijnberg, serogroups Icterohaemorrhagiae and Riccio 2, serogroup Australis, were used in ELISA and the results were compared to microagglutination test (MAT). IgM were confirmed to be the predominant class of antibodies; the total sensitivity obtained by Wijnberg/ELISA with the sera of patients infected by different serogroups was 81.5% in relation to MAT, better sensitivity was achieved by the use of Riccio 2/ELISA; no reaction occurred with the control sera. Cross-reactivity was noted with some sera of patients with Lyme disease (IgM) and with few sera of patients with autoimmune disease (IgM and IgG). To improve the sensitivity of the ELISA both antigens Wijnberg and Riccio 2 must be employed in the immunoenzimatic assay, since serogroup Australis must be considered as a new emerging serogroup causing human leptospirosis in our country.
INTRODUCTION
Leptospirosis is an infection which may manifest itself in humans with a multiform symptomatology which ranges from influenza-like pyrexia to severe clinical pictures of Weil disease, characterized by icterohaemorrhagia with nephritis (10). The diagnosis of this disease offers some difficulties for two reasons: the first is the aspecific clinical symptomatology, which may mislead clinicians in the diagnosis in the majority of cases (13, 19); the second is the variety of leptospiral serovars circulating in a determinate territory. Leptospiral infection in humans may be diagnosed in the laboratory either by isolating the microorganism and/or by demonstrating a rise in specific serum antibodies. Direct diagnosis, i.e., the isolation of the infecting strain and its identification, is seldom executed as a routine procedure in hospitals whereas serological tests are widely employed. The 1 Corresponding author.
most important serological procedure is the microscopic agglutination test (MAT) (3, 14), which is still the standard reference test for leptospiral diagnosis (27) and detects antibodies at serovar level. This test involves the use of a battery of leptospires belonging to the different serovars and, since MAT requires the maintenance of stock cultures, it remains a specialized test which is not generally performed in routine diagnostic laboratories. In addition to MAT other kinds of test are available for the diagnosis of human leptospirosis in routine diagnostic laboratories. Slide tests with antigen coated on a microcapsule carrier (4) are often available, they are forward, qualitative and quantitative, but peculiar to false positivity, thus reducing the specificity. Different enzyme-linked immunosorbent assays (ELISA) have been proposed by Reference Leptospira Laboratories in various countries because they are suitable for automation and standardization. Terpstra (23) introduced ELISA for the serodiagnosis of leptospirosis, employing a soluble antigen prepared from heated cultures of a
677
Cinco M. et al.
Eur. J. Epidemiol.
single leptospiral strain (strain Wijnberg, serovar copenhageni). The microscopic agglutination test (MAT) was employed as the reference test. ELISA and MAT results agreed in 95% of the sera. One false positive was found in the controls. Adler et al. (2) employed homologous whole cell sonicated strains for ELISA to detect leptospiral IgM and IgG in sera from patients suffering from leptospirosis, IgM remained the predominant class of antibodies in human leptospirosis (1, 16, 20, 25). These authors employed for antigen preparations one or two serovars dominant in a particular region and tested human sera which originated from patients infected by homologous serovar. To our knowledge this does not ensure the detection, by ELISA, of Leptospira antibodies against heterologous serovars, that is at the genus level. In Italy there is a complex and heterogeneous epidemiology: 8 main serovars and 6 minor serovars have been found to circulate (6, 9). The aim of this research was to evaluate if an immunoenzymatic assay could detect antibodies against the main serogroups - serovars present in Italy. To this purpose we evaluated the genus specificity of the antigens employed in our ELISA test for antigen preparation, strain Wijnberg, serovar copehnageni, and strain Riccio 2, serovar brat&lava. Immunodominant antigens present in cross-reacting autoimmune disease sera were also identified by Western blot and the correlations are discussed. MATERIALS AND METHODS
Straina and media - The following Leptospira strains, representing the serovars known to occur in Italy (in parentheses), were used: Bianchi (icterohaemorrhagiae), Wijnberg (copenhageni), Alarik (canicola), Mezzano (pomona), Pavia (bataviae), Zanoni (zanoni), Poi (po/), Topino 1 (seyroe), Riccio 2 (bratislava), Riccio 37 (lora), Castellon 3 (castellonis), Mitis Johnson (tarassovi), Hardjopraijtno (hardjo), Sari (mimi), Ballico (australis). The strains were cultivated for 6-7 days in PML5 leptospire medium Reheis (Armour) at 30°C without agitation. Sera - 103 positive human sera from different regions of Italy were employed. The sera were from patients in the acute and the convalescent state of disease. In addition, 32 sera collected among humans affected by an Austral& infection occured in an outbreak (7) were also tested. The control sera were from 92 healthy blood donors, 19 patients with autoimmune disease and 14 sera from patients with Lyme disease. Microscopic agglutination test (MAT) - Serial twofold dilutions of the sera were made starting from 1:100. MAT was applied by incubating the sera at 37°C for 90 min with suspensions of live leptospires (106/ml) of each strain. The titre was the highest dilution which
gave 50% agglutination. Titre 1:100 was considered positive. Enzyme-linked immunosorbent assay (ELISA) - As antigens cultures of strain Wijnberg (copenhageni) and Riccio 2 (bratislava) were employed; they were grown as described, centrifuged at 10000 g for 30 min and washed three times with phosphate buffered saline (PBS), pH 7.2. Finally the pellets were resuspended in carbonate buffer (0.1M, pH 9.6) to a density of 5 107 leptospires/ml and sonicated at 20 KHz for three times, 15 sec each. Tests performed with antigen prepared from strain Wijnberg were indicated as Wijnberg/ELISA; when antigen was prepared from Riccio 2, the test was indicated as Riccio 2/ELISA. Polystyrene microtiter plates (Labsystem) were coated overnight at 4°C with 100 lal of sonicated leptospiral suspension. After three whashing with PBS containing Tween 20 (Merk) 0.05% (v/v), each well was coated with 1% bovine serum albumin (BSA) in PBS. The wells were washed as described above and stored at 4°C in plastic bags. The antigen covered plates were stable for at least six months. The test was performed according to the following procedure: twofold dilutions of the sera were made in PBS/BSA, starting from 1:40; volumes of 100 lal were employed. After lh incubation the plates were washed 3 times with PBS/Tween and alkaline phosphates conjugated antihuman IgM (la chain) or IgG (1~ chain) (Sigma, USA), at a dilution of 1:800 in PBS/BSA, was added. After an additional 1 h incubation period, the plates were washed as described above and 100 pl ofsubstrate (p- nitrophenil phosphatase disodium [Sigma] 1 mg/ml in 0.1 M glycine buffer, pH 10.4) were added and incubated at 37°C for 30 rain. The enzyme-substrate reaction was stopped by adding 50 Ill of NaOH 1M to each well. The optical density (O.D.) was read at 405 nm with a Uniskan spectrophotometer (Labsystem). To determine the cutoff points, 92 sera were employed from blood donors, diluted at 1:40 and 1:80. The cutoff points for IgM and IgG were expressed by means of O.D. of all negative sera plus two standard deviations, 0.245 and 0.251, respectively. SDS-PAGE and Western blotting - Strains Riccio 2 and Wijnberg, washed and resuspended at 5.107 cells/ml, were sonicated before electrophoresis. The protein concentration was determined in the sonicated suspensions by the method of Lowry et al. (18). SDSPAGE was carried out as described by Laemmli (17) with a 12.5°/0 (w/v) acrylamide running gel and a 3.65% (w/v) stacking gel, loading 1.25-1.5 lag/ml protein for each sample. Standard proteins of known molecular masses wer employed (bovine serum albumin [67 KDa], ovalbumin [45 KDa], chymotrypsinogen [25 KDa], soybean trypsin inhibitor [21 KDa], cytochrome c [12.5 KDa] [Boeringer] and myoglobulin I [8 KDa], [BDH]). Gels were blotted on nitrocellulose
678
Vol. 8, 1992
ELISA diagnosis of leptospirosis in Italy
(Schleicher and Schuell) in a Semidry Transblott Cell (Biorad) 30 min at 1- 15 mV following the method described by Towbin et al. (26). After transfer the nitrocelluloses were blockked with 3% (w/v) gelatin in 20 mM Tris- buffered saline (TBS), pH 8.3, washed and incubated with different sera at dilutions of 1:50 in TBS containing 1% (w/v) gelatin for 90 rain on the rocking platform. After incubation and three washings the nitrocelluloses were incubated with 1:1000 horseradish peroxidase conjugated rabbit antihuman IgG (Biorad) in TBS gelatin. The washing steps were repeated and the reaction developed according to Towbin et al. (26).
in the order of 29% in relation to the 41.7% detected by MAT, while sera positive for Javanica and Serjoe were 3.8% versus 7.7% and 6.8% versus 8.7%, respectively. The total agreement between MAT and ELISA obtained with the sera of leptospirosis patients was 75.8% and 45.4% for IgM and IgG, respectively. It can be noted that the most represented class of antibodies within the leptospirosis patients, as identified by ELISA, was IgM. The total positivity obtained by ELISA with sera of patients infected by different serogroups was 81.5% in relation to MAT. Figures 1 and 2 show the values of O.D. obtained in ELISA for IgM and IgG, respectively. Column A reports the results obtained for the sera of leptospirosis patients. Column B reports the results of the control sera (healthy blood donors) and column C reports the results obtained for patients with autoimmune disease and for patients with Lyme disease (round and square symbols, respectively). It can be noted that none of the sera from healthy blood
RESULTS
Reactivity of the sera with MAT and Wijnberg/ELISA
Table 1 reports the results obtained with 103 sera from leptospirosis patients, examined by MAT and Wijnberg/ELISA and expressed as percent of positivity for serogroup.
I gM
Qr
.o~ o m
19
TABLE 1. - Percent reactivity* of human sera in MAT and Wijnberg/ELISA. Serogroup
MAT/serogroup
1.7 1,6
Wijnberg/ELISA Tot. pos.f IgM IgG
13 1.4
Icterohaem.
35.2
35.2
35.2
20.4
Australis
41.7
29
23.2
14.5
Javanica
7.7
3.8
3.8
1.9
Seyroe
8.7
6.8
6.8
3.8
/;I
Canicola
2.9
2.9
2.9
2.9
1.C
Pomona
2.9
2.9
2.9
1.9
0.~
Hardjo
0.9
0.9
0.9
0
O~
81.5
75.8
45.4
Total positivity
100
1.2
* values calculated on the total number of sera (103): positivities were expressed toward the serogroup/serova, reacting in MAT; t positive for almost one immunoglobulin is•type.
°..
..:..
o°
• .': .... 02 01 [~
By MAT assay the highest positivities of the sera were found for serogroups Australis (41.7%) and Icterohaemorrhagiae (35.20/0). The results indicated that the ELISA test was able to recognize all the sera reacting within the homologous serogroup Icterohaemorrhagiae. Sera positive for heterologous serogroups Canicola, Pomona, Hardjo were fully recognized; sera positive for heterologous serogroups Javanica, Seyroe and Australis were only partially recognized. In fact, the sera positive for Australis were recognized by ELISA
o[][] •°°c]
~!.i~::.'" :__~..~ .... °,-.
..2i'~..
•...::-..-..
~.go
[].-." []
°°°
A
B
C
Figure 1. - Optical density values (O.D., in ordinate) obtained employing Wijnberg/ELISA IgM conjugate. Column A: patients with leptospirosis; column B: blood donors; column C: Lyme disease patients ([:3) and autoimmune disease patients (0). Horizontal line indicates cut-off value: 0.245.
679
Cinco M. et al.
Eur. J. Epidemiol.
Ig
M
405nm
le G
405
Results of Riccio 2/ELISA on
nm
Australis
sera
Since the antigen prepared from strain Wijnberg was not very reactive with the sera of patients definitely infected by Leptospires belonging to serogroup A u s t r a l i s , which is prevalent in Italy, we tested an homologous antigen preparation 32 human sera which were collected during an A u s t r a l i s infection. Figure 3 shows the results obtained. The cut-off points were 198 for IgM and 169 for IgG. All sera gave a positive reaction in the Riccio 2/ELISA, 84% of the sera were positive for IgM and 74% for IgG. Consequently, if this antigen preparation is used in addition to Wijnberg/ELISA, the sensitivity of the test should reach 93.3%.
a°i
Western blot (WB)
Using WB technique we tested 3 sera from patients affected by autoimmune disease which reacted positively with IgG in the Elisa test (Fig. 3
I gG 4 0 ~ O.
nnl
i--
0.~-i
"•
i
0.4!-
I 0:~_ I
0:~_ t" m 0.1_ !
1.
•
eole
,m" •, ,
m
•o 1A • O • 1.-
,.•
•
it
O
O
•
O O O
A
0.!--
B
05 0." 03
..2-..~
02
Figure 2. - Optical density values (O.D., in ordinate) obtained employing Riccio 2/ELISA for IgM, column A and for IgG, column B. Horizontal line indicates cut-offvalue: 0.198 for IgM and 0.169 for IgG.
• ,•°
0.' i 02
0.',
o.!
-....
•
..":..
• "-..(;: ...:-:..... •
..~.:7":..':'.i'. .. :.~........'~:.:..
%1]
•O°o•
:'-".... •
donors was positive in ELISA (column B in Figure 1 and 2) and that the sensitivity in ELISA was greater for IgM than for IgG (column A in Figure 1 and 2, respectively). In addition, 6 of the 14 sera belonging to patients with Lyme disease were positive for IgM (column C in Figures 1, square symbols) but all sera were negative for IgG (column C in Figure 2, square symbols). It can be noted that 2 of the 19 sera of patients with autoimmune disease were positive for IgM and 5 for IgG (column C in Figure 1 and 2, round symbols). 680
A
13
e
c
Figure 3. - Optical density values (O.D., in ordinate) obtained employing Wijnberg/ELISA IgG conjugate. Column A: patients with leptospirosis; column B: blood donors; column C: Lyme disease patients (El) and autoimmune disease patients (O). Horizontal line indicates cut-off value: 0.251.
Vol. 8, 1992
ELISA diagnosis of leptospirosis in Italy
column C, round symbols). The sera recognized different antigens in two strains examined (L. interrogans strains Wijnberg and strain Riccio 2) corresponding to bands at 92, 60, 59, 51, 32, 30 and 22 Kd (data not shown). The reactivities at 92 and 60 kd were observed in two of the three sera examined. DISCUSSION Immunoenzymatic assays have been employed by different authors in the past (2, 23, 24) to detect specific immunoglobulins of leptospiral infections, using different antigenic preparations such as sonicated antigen or soluble antigen prepared from a heated culture of leptospires. Besides the classic ELISA technique, Dot-ELISA (28) and Dig-ELISA (12) were also experimented with. These techniques were effective, but did not available simultaneous detection of several antigens. A new, simple immunoenzimatic test, the Line Blot, was also set up (22) but it did not seem to be quantitative nor very specific, giving a 13% false positive rate. In our investigation we used a sonicated antigen prepared from L. interrogans strain Wijnberg (serogroup Icterohaemorrhagiae) and have employed it to detect specific antibodies in the sera of leptospirosis patients in Italy infected by various serovars/ serogroups. It was noted that the sensitivity of an immunoenzimatic assay was greater for sera reacting with the homologous serogroup Icterohaemorrhagiae than with heterologous serogroups. In particular, sensitivity for serogroups Australis, the most represented serogroup in this survey, was low. In fact, only 29% of the total sera reacted in the enzyme assay in relation to 41.7% of those sera which were recognized by MAT. However, it is clear from our research that sensitivity can be improved for IgM for IgG in the sera of patients infected by serogroups Australis if we add an antigen prepared with strain Riccio 2 (serogroup Australis, serovar bratislava). A combination of both antigens Riccio 2 and Wijnberg would increase the sensitivity of the ELISA to 93.3%, marking the test genus specific. The detection of antibodies by ELISA confirmed that IgM remained the predominant class of antibodies. Therefore, the demostration of specific IgM is a valid diagnostic tool in the early stage of the disease. No false positive reaction was found in the control group of blood donors. We have found cross-reactivity with sera of patients infected by other spirochetes such as Borrelia burgdooeeri (IgM class) and with sera of patients with autoimmune disease (IgM and IgG). The crossreactivity between Leptospires and Borrelia burgdotferi, already reported by Raoult (21), occurs in sera from leptospirosis patients. Cross-reactivity with sera from patients with autoimmune disease is demonstrated here for the first time. Some of these samples analyzed by western blotting reacted with bands 92, 60, 59, 51, 32, 30 ad 22 kd. It is known that
the 60 Kd protein ofBorrelia burgdorferi belongs to the family of heat shock (Hsp) proteins (15) which are widespread among bacteria and eucariotic ceils; its presence in leptospires was recently demonstrated (5) and, consequently, this can be one element of crossreactivity. Recent epidemiological trends of human leptospirosis in Italy (6, 9), have shown that the frequency of the infecting serovars has significantly changed when compared to the years when rice field leptospirosis was predominant. In fact, besides the persistence of infections caused by serovars belongint to the Icterohaemorrhagiae serogroups, a strong reduction in infections due to the Bataviae group, and mostly the emergence of serovars of the Australis serogroup, have been noted. Since Australis must be considered a new emerging serogroups causative of human leptospirosis in our country (6, 7, 9, 11), to improve the sensitivity of the ELISA we have come to the conclusion that both antigens from strains Wijnberg and Riccio 2 must be imployed in the immunoenzimatic assay. This would ensure the detection of, with more effectiveness, antibodies antileptospira in sera from leptospirosis in patients in Italy.
681
REFERENCES
1. Adler B. and Faine S. (1978): The antibodies involved in the human immune response to leptospiral infection - J. Med. Microbiol. 11: 387-400. 2. Adler B., Murphy A.M., Locarnini S.A. and Faine S. (1980): Detection of specific anti-leptospiral immunoglobulins M and G in human serum by solid-phase enzyme-linked immunosorbent assay - J. Clin. Microbiol. 11: 452-457. 3. Alexander A.D. (1974): Leptospira. In: Manual of Clinical Microbiology, II ed., Lennette E.H., Spaulding E.H. and Traunt G.B. - American Society for Microbiology, Washington D.C. 347-354. 4. Arimitsu Y., Kobayashi S., Akama K. and Matuhasi T. (1982): Development of a simple serological method for diagnosis of leptospirosis: a microcapsule agglutination test - J. Clin. Microbiol. 15: 835-841. 5. Ballard S., Faine S. and Adler B. (1990): Purification and characterization of a protein antigen from L. interrogans serovar hardjo common to a wide range of bacteria - J. Gen. Microbiol. 136: 1849-1857. 6. Banff E., Castagnari L. and Cinco M. (1989): Serosurvey of Leptospira agglutinins among railway employeed in the National Railway Compartments of Bologna and Venice - Eur. J. Epidemiol. 5: 468473. 7. Cacciapuoti B., Ciceroni L., Maffei C., Stanislao F., Strusi P., Calegari L., Lupidi R., Scalise G., Cagnoni G. and Renga G. (1987): A waterborne outbreak of leptospirosis - Amer. J. Epidemiol. 128: 535-545.
Cinco M. et al.
Eur. J. Epidemiol.
19. Nityananda K. and Harvey T. (1971): Leptospirosis in Ceylon. Epidemiological and laboratory studies Ceylon J. Med. Sci. 20: 5-14.
8. Chernukha YU. G., ShisMna Z.S., Baryshev P.M. and Kokovin I.L. (1976): The dynamics of IgM and IgGantibodies in leptospiral infection in man - Zbl. Bakt. Hyg. I Abt. Orig. A 236: 336-343.
20. Pike R.M., McBrayer H.L., Schulze M.L. and Chandler C.H. (1965): Chromatographic analysis and sulfhydryl sensitivity of antileptospira agglutinins in rabbit and human sera - Proc. Soc. Exp. Biol. Med. 120: 786-789.
9. Ciceroni L., Pinto A. and Caccapuoti B. (1988): Recent trends in human leptospirosis in Italy - Eur. J. Epidemiol. 4: 49-54. 10. Cinco M. (1989): La leptospirosi: Generalith e diagnostica - Clin. Microbiology Comer 5: 7-11.
21. Raoult D., Hechemy K.E. and Baranton G. (1989): Cross-reaction with Borrelia burgdorferi antigen of sera from patients with human immunodeficiency virus infection, syphilis and leptospirosis - J. Clin. Microbiol. 27: 2152-2155.
11. Cinco M., Banff E., Stornello C., Campo F and Korver H. (1989): First human isolate of Leptospira interrogans as serovar bratislava in Italy - FEMS Microbiol. Immunol. 47: 499-504.
22. Raoult D., Bres P. and Baranton G. (1989): Serological diagnosis of leptospirosis: comparison of line blot and immunofluorescence techniques with the Genus specific microscopic agglutination test - J. Infect. Dis. 160 (4): 734-735.
12. Cursons R.T.M. and Pyke P.A. (1981): Diffusion in gelenzyme-linked immunosorbent assay a new serological test for leptospirosis - J. Clin. Pathol. 34: 840-842.
23. Terpstra W.J., Ligthart G.S. and Schoone G.J. (1980): Serodiagnosis of human leptospirosis by enzymelinked immunosorbent assay (ELISA) - Zbl. Bakt. Hyg. Abt. Orig. A 247: 400-405.
13. Feigin R.D. and Anderson D.C. (1975): Human leptospirosis - CRC Crit. Rev. Clin. Lab. Sci. 5: 413467. 14. Galton M.M., Menger R. IV., Shotts E.B., Nahmias A.J. and Heath C.W. (1962): Leptospirosis: Epidemiology clinical manifestations in man and animals and methods in laboratory diagnosis - U.S. Public Health Serv. Publ. 51: 1-70.
24. Terpstra W.J., Ligthart G.S. and Schoone G.J. (1985): Elisa for the detection of specific IgM and IgG in human leptospirosis - J. Gen. Microbiol. 131: 377385. 25. To~ngM.J., Rosenberg E.B., Votteri B.A. and Che C.T. (1971): Immunological response in leptospirosis. Report of three cases - Am. J. Trop. Med. Hyg. 20: 625-630.
15. Hansen K., Bangborg J.M., Fjordvang H., Strandberg (1988): Pedersen N. and Hindersson P. Immunochemical characterization and isolation of the gene for Borrelia burgdorferi immunodominant 60 Kd Antigen common to a wide range of bacteria Infect. Immun. 56: 2047-2053. 16. Hartmann L., Filitti-Wurmser S., Jacquot-Armand Y., Mailloux M., Hurez D. and Fauvert R. (1964): Macromolecular nature of an antibody of leptospirosis australis - Biochim. Biophys. Acta 82: 249-259. 17. Laemmli V.K. (1970): Cleavage of structural proteins during the assembly of the head of bacteriophage T4 - Nature 227: 680-685. 18. Lowry O.H., Rosenbrough N.J., Favor A.L. and Randall R.J. (1951): Protein measureament with the folin phenol reagent - Biol. Chem. 193: 265-275.
26. Towbin H., Staehafin T. and Gordon J. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications - Proc. Natl. Acad. Sci. U.S.A. 76: 4350-4354. 27. Turner L.H. (1968): Leptospirosis II - Trans. Royal Soc. Trop. Med, Hyg. 62: 880-899. 28. Watt G., Alquiza L.M., Padre L.P., Tuazon M.L. and Laughlin L.W. (1988): The rapid diagnosis of leptospirosis: a prospective comparison of the dot enzyme-linked immunosorbent assay and the genus specific microscopic agglutination test at different stages of illness - J. Infect. Dis. 57: 840-842.
682
EUROPEAN
Vol. 8, No. 5
JOURNAL
September 1992, p. 677-682
Ov
EPIDEMIOLOGY
EVALUATION OF AN IMMUNOENZI £ TIC TEST (ELISA) FOR THE DIAGNOSIS OF LEPTOSPIROSIS IN ITALY M. CINCO 1, D. BALANZ1N and E. BANFI Istituto di Microbiologia - Universitgt degli Studi - Via Fleming, 22 - Trieste - Italy.
Key words: ELISA - Leptospirosis - A u s t r a l i s
-
Epidemiology
An enzyme-linked immunosorbent assay (ELISA) was assessed to detect in humans IgM and IgG against the main serogroups/serovars of Leptospira present in Italy. Sonicated antigens from strain Wijnberg, serogroups Icterohaemorrhagiae and Riccio 2, serogroup Australis, were used in ELISA and the results were compared to microagglutination test (MAT). IgM were confirmed to be the predominant class of antibodies; the total sensitivity obtained by Wijnberg/ELISA with the sera of patients infected by different serogroups was 81.5% in relation to MAT, better sensitivity was achieved by the use of Riccio 2/ELISA; no reaction occurred with the control sera. Cross-reactivity was noted with some sera of patients with Lyme disease (IgM) and with few sera of patients with autoimmune disease (IgM and IgG). To improve the sensitivity of the ELISA both antigens Wijnberg and Riccio 2 must be employed in the immunoenzimatic assay, since serogroup Australis must be considered as a new emerging serogroup causing human leptospirosis in our country.
INTRODUCTION
Leptospirosis is an infection which may manifest itself in humans with a multiform symptomatology which ranges from influenza-like pyrexia to severe clinical pictures of Weil disease, characterized by icterohaemorrhagia with nephritis (10). The diagnosis of this disease offers some difficulties for two reasons: the first is the aspecific clinical symptomatology, which may mislead clinicians in the diagnosis in the majority of cases (13, 19); the second is the variety of leptospiral serovars circulating in a determinate territory. Leptospiral infection in humans may be diagnosed in the laboratory either by isolating the microorganism and/or by demonstrating a rise in specific serum antibodies. Direct diagnosis, i.e., the isolation of the infecting strain and its identification, is seldom executed as a routine procedure in hospitals whereas serological tests are widely employed. The 1 Corresponding author.
most important serological procedure is the microscopic agglutination test (MAT) (3, 14), which is still the standard reference test for leptospiral diagnosis (27) and detects antibodies at serovar level. This test involves the use of a battery of leptospires belonging to the different serovars and, since MAT requires the maintenance of stock cultures, it remains a specialized test which is not generally performed in routine diagnostic laboratories. In addition to MAT other kinds of test are available for the diagnosis of human leptospirosis in routine diagnostic laboratories. Slide tests with antigen coated on a microcapsule carrier (4) are often available, they are forward, qualitative and quantitative, but peculiar to false positivity, thus reducing the specificity. Different enzyme-linked immunosorbent assays (ELISA) have been proposed by Reference Leptospira Laboratories in various countries because they are suitable for automation and standardization. Terpstra (23) introduced ELISA for the serodiagnosis of leptospirosis, employing a soluble antigen prepared from heated cultures of a
677
Cinco M. et al.
Eur. J. Epidemiol.
single leptospiral strain (strain Wijnberg, serovar copenhageni). The microscopic agglutination test (MAT) was employed as the reference test. ELISA and MAT results agreed in 95% of the sera. One false positive was found in the controls. Adler et al. (2) employed homologous whole cell sonicated strains for ELISA to detect leptospiral IgM and IgG in sera from patients suffering from leptospirosis, IgM remained the predominant class of antibodies in human leptospirosis (1, 16, 20, 25). These authors employed for antigen preparations one or two serovars dominant in a particular region and tested human sera which originated from patients infected by homologous serovar. To our knowledge this does not ensure the detection, by ELISA, of Leptospira antibodies against heterologous serovars, that is at the genus level. In Italy there is a complex and heterogeneous epidemiology: 8 main serovars and 6 minor serovars have been found to circulate (6, 9). The aim of this research was to evaluate if an immunoenzymatic assay could detect antibodies against the main serogroups - serovars present in Italy. To this purpose we evaluated the genus specificity of the antigens employed in our ELISA test for antigen preparation, strain Wijnberg, serovar copehnageni, and strain Riccio 2, serovar brat&lava. Immunodominant antigens present in cross-reacting autoimmune disease sera were also identified by Western blot and the correlations are discussed. MATERIALS AND METHODS
Straina and media - The following Leptospira strains, representing the serovars known to occur in Italy (in parentheses), were used: Bianchi (icterohaemorrhagiae), Wijnberg (copenhageni), Alarik (canicola), Mezzano (pomona), Pavia (bataviae), Zanoni (zanoni), Poi (po/), Topino 1 (seyroe), Riccio 2 (bratislava), Riccio 37 (lora), Castellon 3 (castellonis), Mitis Johnson (tarassovi), Hardjopraijtno (hardjo), Sari (mimi), Ballico (australis). The strains were cultivated for 6-7 days in PML5 leptospire medium Reheis (Armour) at 30°C without agitation. Sera - 103 positive human sera from different regions of Italy were employed. The sera were from patients in the acute and the convalescent state of disease. In addition, 32 sera collected among humans affected by an Austral& infection occured in an outbreak (7) were also tested. The control sera were from 92 healthy blood donors, 19 patients with autoimmune disease and 14 sera from patients with Lyme disease. Microscopic agglutination test (MAT) - Serial twofold dilutions of the sera were made starting from 1:100. MAT was applied by incubating the sera at 37°C for 90 min with suspensions of live leptospires (106/ml) of each strain. The titre was the highest dilution which
gave 50% agglutination. Titre 1:100 was considered positive. Enzyme-linked immunosorbent assay (ELISA) - As antigens cultures of strain Wijnberg (copenhageni) and Riccio 2 (bratislava) were employed; they were grown as described, centrifuged at 10000 g for 30 min and washed three times with phosphate buffered saline (PBS), pH 7.2. Finally the pellets were resuspended in carbonate buffer (0.1M, pH 9.6) to a density of 5 107 leptospires/ml and sonicated at 20 KHz for three times, 15 sec each. Tests performed with antigen prepared from strain Wijnberg were indicated as Wijnberg/ELISA; when antigen was prepared from Riccio 2, the test was indicated as Riccio 2/ELISA. Polystyrene microtiter plates (Labsystem) were coated overnight at 4°C with 100 lal of sonicated leptospiral suspension. After three whashing with PBS containing Tween 20 (Merk) 0.05% (v/v), each well was coated with 1% bovine serum albumin (BSA) in PBS. The wells were washed as described above and stored at 4°C in plastic bags. The antigen covered plates were stable for at least six months. The test was performed according to the following procedure: twofold dilutions of the sera were made in PBS/BSA, starting from 1:40; volumes of 100 lal were employed. After lh incubation the plates were washed 3 times with PBS/Tween and alkaline phosphates conjugated antihuman IgM (la chain) or IgG (1~ chain) (Sigma, USA), at a dilution of 1:800 in PBS/BSA, was added. After an additional 1 h incubation period, the plates were washed as described above and 100 pl ofsubstrate (p- nitrophenil phosphatase disodium [Sigma] 1 mg/ml in 0.1 M glycine buffer, pH 10.4) were added and incubated at 37°C for 30 rain. The enzyme-substrate reaction was stopped by adding 50 Ill of NaOH 1M to each well. The optical density (O.D.) was read at 405 nm with a Uniskan spectrophotometer (Labsystem). To determine the cutoff points, 92 sera were employed from blood donors, diluted at 1:40 and 1:80. The cutoff points for IgM and IgG were expressed by means of O.D. of all negative sera plus two standard deviations, 0.245 and 0.251, respectively. SDS-PAGE and Western blotting - Strains Riccio 2 and Wijnberg, washed and resuspended at 5.107 cells/ml, were sonicated before electrophoresis. The protein concentration was determined in the sonicated suspensions by the method of Lowry et al. (18). SDSPAGE was carried out as described by Laemmli (17) with a 12.5°/0 (w/v) acrylamide running gel and a 3.65% (w/v) stacking gel, loading 1.25-1.5 lag/ml protein for each sample. Standard proteins of known molecular masses wer employed (bovine serum albumin [67 KDa], ovalbumin [45 KDa], chymotrypsinogen [25 KDa], soybean trypsin inhibitor [21 KDa], cytochrome c [12.5 KDa] [Boeringer] and myoglobulin I [8 KDa], [BDH]). Gels were blotted on nitrocellulose
678
Vol. 8, 1992
ELISA diagnosis of leptospirosis in Italy
(Schleicher and Schuell) in a Semidry Transblott Cell (Biorad) 30 min at 1- 15 mV following the method described by Towbin et al. (26). After transfer the nitrocelluloses were blockked with 3% (w/v) gelatin in 20 mM Tris- buffered saline (TBS), pH 8.3, washed and incubated with different sera at dilutions of 1:50 in TBS containing 1% (w/v) gelatin for 90 rain on the rocking platform. After incubation and three washings the nitrocelluloses were incubated with 1:1000 horseradish peroxidase conjugated rabbit antihuman IgG (Biorad) in TBS gelatin. The washing steps were repeated and the reaction developed according to Towbin et al. (26).
in the order of 29% in relation to the 41.7% detected by MAT, while sera positive for Javanica and Serjoe were 3.8% versus 7.7% and 6.8% versus 8.7%, respectively. The total agreement between MAT and ELISA obtained with the sera of leptospirosis patients was 75.8% and 45.4% for IgM and IgG, respectively. It can be noted that the most represented class of antibodies within the leptospirosis patients, as identified by ELISA, was IgM. The total positivity obtained by ELISA with sera of patients infected by different serogroups was 81.5% in relation to MAT. Figures 1 and 2 show the values of O.D. obtained in ELISA for IgM and IgG, respectively. Column A reports the results obtained for the sera of leptospirosis patients. Column B reports the results of the control sera (healthy blood donors) and column C reports the results obtained for patients with autoimmune disease and for patients with Lyme disease (round and square symbols, respectively). It can be noted that none of the sera from healthy blood
RESULTS
Reactivity of the sera with MAT and Wijnberg/ELISA
Table 1 reports the results obtained with 103 sera from leptospirosis patients, examined by MAT and Wijnberg/ELISA and expressed as percent of positivity for serogroup.
I gM
Qr
.o~ o m
19
TABLE 1. - Percent reactivity* of human sera in MAT and Wijnberg/ELISA. Serogroup
MAT/serogroup
1.7 1,6
Wijnberg/ELISA Tot. pos.f IgM IgG
13 1.4
Icterohaem.
35.2
35.2
35.2
20.4
Australis
41.7
29
23.2
14.5
Javanica
7.7
3.8
3.8
1.9
Seyroe
8.7
6.8
6.8
3.8
/;I
Canicola
2.9
2.9
2.9
2.9
1.C
Pomona
2.9
2.9
2.9
1.9
0.~
Hardjo
0.9
0.9
0.9
0
O~
81.5
75.8
45.4
Total positivity
100
1.2
* values calculated on the total number of sera (103): positivities were expressed toward the serogroup/serova, reacting in MAT; t positive for almost one immunoglobulin is•type.
°..
..:..
o°
• .': .... 02 01 [~
By MAT assay the highest positivities of the sera were found for serogroups Australis (41.7%) and Icterohaemorrhagiae (35.20/0). The results indicated that the ELISA test was able to recognize all the sera reacting within the homologous serogroup Icterohaemorrhagiae. Sera positive for heterologous serogroups Canicola, Pomona, Hardjo were fully recognized; sera positive for heterologous serogroups Javanica, Seyroe and Australis were only partially recognized. In fact, the sera positive for Australis were recognized by ELISA
o[][] •°°c]
~!.i~::.'" :__~..~ .... °,-.
..2i'~..
•...::-..-..
~.go
[].-." []
°°°
A
B
C
Figure 1. - Optical density values (O.D., in ordinate) obtained employing Wijnberg/ELISA IgM conjugate. Column A: patients with leptospirosis; column B: blood donors; column C: Lyme disease patients ([:3) and autoimmune disease patients (0). Horizontal line indicates cut-off value: 0.245.
679
Cinco M. et al.
Eur. J. Epidemiol.
Ig
M
405nm
le G
405
Results of Riccio 2/ELISA on
nm
Australis
sera
Since the antigen prepared from strain Wijnberg was not very reactive with the sera of patients definitely infected by Leptospires belonging to serogroup A u s t r a l i s , which is prevalent in Italy, we tested an homologous antigen preparation 32 human sera which were collected during an A u s t r a l i s infection. Figure 3 shows the results obtained. The cut-off points were 198 for IgM and 169 for IgG. All sera gave a positive reaction in the Riccio 2/ELISA, 84% of the sera were positive for IgM and 74% for IgG. Consequently, if this antigen preparation is used in addition to Wijnberg/ELISA, the sensitivity of the test should reach 93.3%.
a°i
Western blot (WB)
Using WB technique we tested 3 sera from patients affected by autoimmune disease which reacted positively with IgG in the Elisa test (Fig. 3
I gG 4 0 ~ O.
nnl
i--
0.~-i
"•
i
0.4!-
I 0:~_ I
0:~_ t" m 0.1_ !
1.
•
eole
,m" •, ,
m
•o 1A • O • 1.-
,.•
•
it
O
O
•
O O O
A
0.!--
B
05 0." 03
..2-..~
02
Figure 2. - Optical density values (O.D., in ordinate) obtained employing Riccio 2/ELISA for IgM, column A and for IgG, column B. Horizontal line indicates cut-offvalue: 0.198 for IgM and 0.169 for IgG.
• ,•°
0.' i 02
0.',
o.!
-....
•
..":..
• "-..(;: ...:-:..... •
..~.:7":..':'.i'. .. :.~........'~:.:..
%1]
•O°o•
:'-".... •
donors was positive in ELISA (column B in Figure 1 and 2) and that the sensitivity in ELISA was greater for IgM than for IgG (column A in Figure 1 and 2, respectively). In addition, 6 of the 14 sera belonging to patients with Lyme disease were positive for IgM (column C in Figures 1, square symbols) but all sera were negative for IgG (column C in Figure 2, square symbols). It can be noted that 2 of the 19 sera of patients with autoimmune disease were positive for IgM and 5 for IgG (column C in Figure 1 and 2, round symbols). 680
A
13
e
c
Figure 3. - Optical density values (O.D., in ordinate) obtained employing Wijnberg/ELISA IgG conjugate. Column A: patients with leptospirosis; column B: blood donors; column C: Lyme disease patients (El) and autoimmune disease patients (O). Horizontal line indicates cut-off value: 0.251.
Vol. 8, 1992
ELISA diagnosis of leptospirosis in Italy
column C, round symbols). The sera recognized different antigens in two strains examined (L. interrogans strains Wijnberg and strain Riccio 2) corresponding to bands at 92, 60, 59, 51, 32, 30 and 22 Kd (data not shown). The reactivities at 92 and 60 kd were observed in two of the three sera examined. DISCUSSION Immunoenzymatic assays have been employed by different authors in the past (2, 23, 24) to detect specific immunoglobulins of leptospiral infections, using different antigenic preparations such as sonicated antigen or soluble antigen prepared from a heated culture of leptospires. Besides the classic ELISA technique, Dot-ELISA (28) and Dig-ELISA (12) were also experimented with. These techniques were effective, but did not available simultaneous detection of several antigens. A new, simple immunoenzimatic test, the Line Blot, was also set up (22) but it did not seem to be quantitative nor very specific, giving a 13% false positive rate. In our investigation we used a sonicated antigen prepared from L. interrogans strain Wijnberg (serogroup Icterohaemorrhagiae) and have employed it to detect specific antibodies in the sera of leptospirosis patients in Italy infected by various serovars/ serogroups. It was noted that the sensitivity of an immunoenzimatic assay was greater for sera reacting with the homologous serogroup Icterohaemorrhagiae than with heterologous serogroups. In particular, sensitivity for serogroups Australis, the most represented serogroup in this survey, was low. In fact, only 29% of the total sera reacted in the enzyme assay in relation to 41.7% of those sera which were recognized by MAT. However, it is clear from our research that sensitivity can be improved for IgM for IgG in the sera of patients infected by serogroups Australis if we add an antigen prepared with strain Riccio 2 (serogroup Australis, serovar bratislava). A combination of both antigens Riccio 2 and Wijnberg would increase the sensitivity of the ELISA to 93.3%, marking the test genus specific. The detection of antibodies by ELISA confirmed that IgM remained the predominant class of antibodies. Therefore, the demostration of specific IgM is a valid diagnostic tool in the early stage of the disease. No false positive reaction was found in the control group of blood donors. We have found cross-reactivity with sera of patients infected by other spirochetes such as Borrelia burgdooeeri (IgM class) and with sera of patients with autoimmune disease (IgM and IgG). The crossreactivity between Leptospires and Borrelia burgdotferi, already reported by Raoult (21), occurs in sera from leptospirosis patients. Cross-reactivity with sera from patients with autoimmune disease is demonstrated here for the first time. Some of these samples analyzed by western blotting reacted with bands 92, 60, 59, 51, 32, 30 ad 22 kd. It is known that
the 60 Kd protein ofBorrelia burgdorferi belongs to the family of heat shock (Hsp) proteins (15) which are widespread among bacteria and eucariotic ceils; its presence in leptospires was recently demonstrated (5) and, consequently, this can be one element of crossreactivity. Recent epidemiological trends of human leptospirosis in Italy (6, 9), have shown that the frequency of the infecting serovars has significantly changed when compared to the years when rice field leptospirosis was predominant. In fact, besides the persistence of infections caused by serovars belongint to the Icterohaemorrhagiae serogroups, a strong reduction in infections due to the Bataviae group, and mostly the emergence of serovars of the Australis serogroup, have been noted. Since Australis must be considered a new emerging serogroups causative of human leptospirosis in our country (6, 7, 9, 11), to improve the sensitivity of the ELISA we have come to the conclusion that both antigens from strains Wijnberg and Riccio 2 must be imployed in the immunoenzimatic assay. This would ensure the detection of, with more effectiveness, antibodies antileptospira in sera from leptospirosis in patients in Italy.
681
REFERENCES
1. Adler B. and Faine S. (1978): The antibodies involved in the human immune response to leptospiral infection - J. Med. Microbiol. 11: 387-400. 2. Adler B., Murphy A.M., Locarnini S.A. and Faine S. (1980): Detection of specific anti-leptospiral immunoglobulins M and G in human serum by solid-phase enzyme-linked immunosorbent assay - J. Clin. Microbiol. 11: 452-457. 3. Alexander A.D. (1974): Leptospira. In: Manual of Clinical Microbiology, II ed., Lennette E.H., Spaulding E.H. and Traunt G.B. - American Society for Microbiology, Washington D.C. 347-354. 4. Arimitsu Y., Kobayashi S., Akama K. and Matuhasi T. (1982): Development of a simple serological method for diagnosis of leptospirosis: a microcapsule agglutination test - J. Clin. Microbiol. 15: 835-841. 5. Ballard S., Faine S. and Adler B. (1990): Purification and characterization of a protein antigen from L. interrogans serovar hardjo common to a wide range of bacteria - J. Gen. Microbiol. 136: 1849-1857. 6. Banff E., Castagnari L. and Cinco M. (1989): Serosurvey of Leptospira agglutinins among railway employeed in the National Railway Compartments of Bologna and Venice - Eur. J. Epidemiol. 5: 468473. 7. Cacciapuoti B., Ciceroni L., Maffei C., Stanislao F., Strusi P., Calegari L., Lupidi R., Scalise G., Cagnoni G. and Renga G. (1987): A waterborne outbreak of leptospirosis - Amer. J. Epidemiol. 128: 535-545.
Cinco M. et al.
Eur. J. Epidemiol.
19. Nityananda K. and Harvey T. (1971): Leptospirosis in Ceylon. Epidemiological and laboratory studies Ceylon J. Med. Sci. 20: 5-14.
8. Chernukha YU. G., ShisMna Z.S., Baryshev P.M. and Kokovin I.L. (1976): The dynamics of IgM and IgGantibodies in leptospiral infection in man - Zbl. Bakt. Hyg. I Abt. Orig. A 236: 336-343.
20. Pike R.M., McBrayer H.L., Schulze M.L. and Chandler C.H. (1965): Chromatographic analysis and sulfhydryl sensitivity of antileptospira agglutinins in rabbit and human sera - Proc. Soc. Exp. Biol. Med. 120: 786-789.
9. Ciceroni L., Pinto A. and Caccapuoti B. (1988): Recent trends in human leptospirosis in Italy - Eur. J. Epidemiol. 4: 49-54. 10. Cinco M. (1989): La leptospirosi: Generalith e diagnostica - Clin. Microbiology Comer 5: 7-11.
21. Raoult D., Hechemy K.E. and Baranton G. (1989): Cross-reaction with Borrelia burgdorferi antigen of sera from patients with human immunodeficiency virus infection, syphilis and leptospirosis - J. Clin. Microbiol. 27: 2152-2155.
11. Cinco M., Banff E., Stornello C., Campo F and Korver H. (1989): First human isolate of Leptospira interrogans as serovar bratislava in Italy - FEMS Microbiol. Immunol. 47: 499-504.
22. Raoult D., Bres P. and Baranton G. (1989): Serological diagnosis of leptospirosis: comparison of line blot and immunofluorescence techniques with the Genus specific microscopic agglutination test - J. Infect. Dis. 160 (4): 734-735.
12. Cursons R.T.M. and Pyke P.A. (1981): Diffusion in gelenzyme-linked immunosorbent assay a new serological test for leptospirosis - J. Clin. Pathol. 34: 840-842.
23. Terpstra W.J., Ligthart G.S. and Schoone G.J. (1980): Serodiagnosis of human leptospirosis by enzymelinked immunosorbent assay (ELISA) - Zbl. Bakt. Hyg. Abt. Orig. A 247: 400-405.
13. Feigin R.D. and Anderson D.C. (1975): Human leptospirosis - CRC Crit. Rev. Clin. Lab. Sci. 5: 413467. 14. Galton M.M., Menger R. IV., Shotts E.B., Nahmias A.J. and Heath C.W. (1962): Leptospirosis: Epidemiology clinical manifestations in man and animals and methods in laboratory diagnosis - U.S. Public Health Serv. Publ. 51: 1-70.
24. Terpstra W.J., Ligthart G.S. and Schoone G.J. (1985): Elisa for the detection of specific IgM and IgG in human leptospirosis - J. Gen. Microbiol. 131: 377385. 25. To~ngM.J., Rosenberg E.B., Votteri B.A. and Che C.T. (1971): Immunological response in leptospirosis. Report of three cases - Am. J. Trop. Med. Hyg. 20: 625-630.
15. Hansen K., Bangborg J.M., Fjordvang H., Strandberg (1988): Pedersen N. and Hindersson P. Immunochemical characterization and isolation of the gene for Borrelia burgdorferi immunodominant 60 Kd Antigen common to a wide range of bacteria Infect. Immun. 56: 2047-2053. 16. Hartmann L., Filitti-Wurmser S., Jacquot-Armand Y., Mailloux M., Hurez D. and Fauvert R. (1964): Macromolecular nature of an antibody of leptospirosis australis - Biochim. Biophys. Acta 82: 249-259. 17. Laemmli V.K. (1970): Cleavage of structural proteins during the assembly of the head of bacteriophage T4 - Nature 227: 680-685. 18. Lowry O.H., Rosenbrough N.J., Favor A.L. and Randall R.J. (1951): Protein measureament with the folin phenol reagent - Biol. Chem. 193: 265-275.
26. Towbin H., Staehafin T. and Gordon J. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications - Proc. Natl. Acad. Sci. U.S.A. 76: 4350-4354. 27. Turner L.H. (1968): Leptospirosis II - Trans. Royal Soc. Trop. Med, Hyg. 62: 880-899. 28. Watt G., Alquiza L.M., Padre L.P., Tuazon M.L. and Laughlin L.W. (1988): The rapid diagnosis of leptospirosis: a prospective comparison of the dot enzyme-linked immunosorbent assay and the genus specific microscopic agglutination test at different stages of illness - J. Infect. Dis. 57: 840-842.
682
