You are on page 1of 8

Serodiagnosis of Tuberculosis: Comparison of Immunoglobulin A (IgA) Response to Sulfolipid I with IgG and IgM Responses to 2,3-Diacyltrehalose, 2,3,6-Triacyltrehalose, and

Cord Factor Antigens
Esther Julián, Lurdes Matas, Andrés Pérez, José Alcaide, Marie-Antoinette Lanéelle and Marina Luquin J. Clin. Microbiol. 2002, 40(10):3782. DOI: 10.1128/JCM.40.10.3782-3788.2002. Downloaded from on October 9, 2013 by guest

Updated information and services can be found at: These include:

This article cites 29 articles, 6 of which can be accessed free at: Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more»


Information about commercial reprint orders: To subscribe to to another ASM Journal go to:

No. i. IgM. and trehalose6. A total of 228 serum samples were studied. in a wide population of patients affected by TB and non-TB pulmonary infections. tuberculosis: diacyltrehaloses. Badalona. The results indicated a significantly elevated IgG and IgA antibody response in tuberculosis patients. (i) TB patients. the acylated trehalose family has been the most frequently investigated group of glycolipids. 3782 the use of detergent in washing buffers) (15. To control tuberculosis (TB). MATERIALS AND METHODS Study subjects. to date no analysis has been undertaken to determine the presence of specific IgA against the glycolipids mentioned above.JOURNAL OF CLINICAL MICROBIOLOGY. However. proteins. American Society for Microbiology. healed. all from Spain.1 Lurdes Matas. as well as in healthy people. including purified protein derivative-negative. and sulfolipid I (SL-I). 24–26) was obtained.3. compared with controls. 11. test sensitivities were as different as 11 to 88% for the DAT antigen (11. 25) and 51 to 93% for the TAT antigen (11.3-diacyltrehalose (DAT). Hospital Universitari Germans Trias i Pujol. the study was carried out in parallel with four M. and vaccinated individuals.00ϩ0 DOI: 10. thus.5%. and glycolipidic molecules from the mycobacterial cell wall (31). in combination with other antigenic molecules. Vol. The demographic and clinical characteristics of the individuals involved are listed in Table 1. could be a useful antigen for tuberculosis serodiagnosis.2002 Copyright © 2002. and Cord Factor Antigens Esther Julia ´n. In addition to the lipoarabinomannan antigen.6Ј-dimycolate (cord factor [CF]). 30). Toulouse. 08193 Bellaterra (Barcelona). 18. it is widely accepted that it is not possible to achieve sensitivities above 80% in tuberculosis serodiagnosis when using one antigen alone. Sera were collected from adult patients who. it is still necessary to find a diagnostic method that is both more rapid to carry out and more sensitive than traditional methods (smear and culture) but which is simpler and less expensive than the new molecular diagnostic tests that are based on the amplification of nucleic acids. In spite of the presence of high IgA antibody levels in sera from TB patients (9) and although different commercialized assays exist that detect IgA antibodies against cellular extracts and purified proteins (14. Thus.. and of 66 and 87. the present study analyzed the IgA antibody response to mycobacterial glycolipids. or to the different sera groups analyzed (the number and type of patients and control subjects) (10).5% for IgA. These contradictory results are due to the ELISA protocol used (the different antigen concentrations or * Corresponding author. 3782–3788 0095-1137/02/$04. with all the antigens used. For the first time. Spain. high test specificity was achieved (93. Facultat de Cie `ncies. p. and IgA antibody responses to four trehalose-containing glycolipids purified from M. Facultat de Cie `ncies i Institut de Biotecnologia i Biomedicina. France4 Received 30 January 2002/Returned for modification 30 June 2002/Accepted 25 July 2002 Nonpeptidic antigens from the Mycobacterium tuberculosis cell wall are the focus of extensive studies to determine their potential role as protective antigens or serological markers of tuberculous disease. 2. according to clinical parameters. were suspected of having TB and who had been admitted to the “Germans Trias i Pujol” University Hospital (HUGTiP) in Badalona.1 Servei de Microbiologia.40. and Institut de Pharmacologie et de Biologie Structurale du CNRS. respectively. Phone: Departament de Salut i Seguretat Social.7%) with a sensitivity of 67. Mailing address: Departament de Gene `tica i de Microbiologia. Currently. 16. Serological methods seem to be the ideal choice and. E-mail: marina. triacyltrehaloses.6Ј-tetraacyltrehalose 2Ј-sulfate (sulfolipid I [SL-I]). They are 2.2 Jose ´ Alcaide.2 Andre ´s Pe ´rez. Spain. Universitat Auto `noma de Barcelona.3782–3788. and 52 sera from nontuberculous pneumonia patients). 26).4 1 and Marina Luquin * Departament de Gene `tica i de Microbiologia.luquin@uab. tuberculosis glycolipids. 10 Serodiagnosis of Tuberculosis: Comparison of Immunoglobulin A (IgA) Response to Sulfolipid I with IgG and IgM Responses to 2. 2. Using the enzyme-linked immunosorbent assay (ELISA) technique. using an ELISA method specifically optimized for them (15). and an extensive variability in immunoglobulin G (IgG) or IgM titers (6. compared with the IgG and IgM response. many mycobacterial antigens have been evaluated: cellular extracts. Universite ´ Paul Sabatier. several studies were performed with these antigens.6.6%.2 and Programa de Control i Prevencio ´ de la Tuberculosi.3. were studied. 20. Fax: 34-93-5812387. we conclude that SL-I. Universitat Auto `noma de Barcelona.6-triacyltrehalose (TAT). Using this antigen and combining IgA and IgG antibody detection. 7.3 Marie-Antoinette Lane ´elle. showing test sensitivities and specificities for IgG of 81 and 77. no definite conclusions as to their utility as serodiagnostic antigens have been reached. we have made a comparative study of the immunoglobulin G (IgG).10. 23). Sera from 92 tuberculosis patients (taken before starting antituberculosis treatment) and a wide group of control individuals (84 sera from healthy donors. Barcelona. Thus. Regarding this latter role and using an enzyme-linked immunosorbent assay. Oct. Moreover.3. -positive. cord factor.1128/JCM. 2. the only serum samples included in this group were from patients in whom the disease had been confirmed by isolation of the tuberculous bacillus in cultures (Lo ¨wenstein-Jensen and the nonradiometric MB/Bact system [Organon . SL-I was the best antigen studied. Bellaterra.3 Spain. 2002. 40.3-Diacyltrehalose.e. All Rights Reserved.6-Triacyltrehalose.

9 had been BCG vaccinated in the past. the reactivity of pulmonary TB.4 52 5. they had received the standard treatment for TB and had completed it correctly. and cutaneous (one) cases. without any antigen) were included for each serum tested and were used as a second negative control. These diseases were originated by Chlamydia spp. RESULTS The demographic and clinical data for the TB patients and controls studied are shown in Table 1. All statistical tests were conducted using SPSS for Windows statistical software (version 9. no significant differences were observed between the mean OD values of the TB patients and those of the overall control groups for any antigen or antibody assayed (P Ͼ 0. This table shows that higher test sensitivities were obtained in smear-positive TB patients than in smear-negative patients. Fifty of these were from non-TB pneumonia patients. and Streptococcus pneumoniae (8 samples). Bio-Tech Instruments. and high levels (standards) and a blank (blocking buffer) were included in each plate.5 20.D.). Although only four sera from children with non-TB pneumonia were studied.05). Final calculations were performed as previously described (6) with slight modifications.3 23 4. six sera were taken from children upon termination of standard treatment. In the child group. They were all purified protein derivative (PPD) negative. pleural TB (two).5 42. Mich.1 31. There was no statistically significant difference in the mean age Ϯ SD of TB patients and control subjects (35. two serum samples included in this group were from patients with other pulmonary mycobacterial diseases: one produced by Mycobacterium kansasii and the other by Mycobacterium xenopi.0. SPSS. plates (Immulon I. tuberculosis clinically isolated strain (22) was grown for 6 weeks at 37°C on Middlebrook 7H10 supplemented with oleic acid-albumin-dextrose-catalase enrichment (Difco Laboratories. Dynatech Laboratories) were coated with purified DAT. plus 3 SD for each antigen-Ig combination. extrapulmonary TB. Mean and individual antibody levels. wells treated with solvent alone (i. Three serum samples were included from adults who had suffered from TB more than 5 years prior to sampling. the highest test sensitivity was Teknika. but not when using the SL-I and CF antigens.2 23 11. Briefly. 22). In addition. In TB adult patients. Sera were diluted in blocking buffer at predetermined optimal dilutions (1/400 for measuring IgG antibodies. ELISA was performed as .7 Ϯ 22. (ii) Control subjects. Coxiella burnetii (6 samples). Arithmetic means and standard deviations (SD) from the mean were calculated for corrected optical density (OD) values. TAT. 40. and 22 were PPD positive (8 adults and 14 children). smear-positive. Durham. Absorbance was determined at 405 nm with a microtiter reader (ELx 800 automated microplate reader. medium. pneumoniae Healthy subjects PPD-negative adults PPD-positive Adults Children Vaccinated Healed Adults Children Total 92 48 10 13 3 18 52 2 12 6 11 9 4 8 84 44 8 14 9 3 6 228 62:30 37:11 5:5 7:6 2:1 11:7 34:18 0:2 7:5 5:1 7:4 4:5 3:1 8:0 52:32 26:18 5:3 12:2 3:6 1:2 5:1 148:80 35. Fifty-eight of these adult patients were suffering from pulmonary TB (10 of them were coinfected with human immunodeficiency virus [HIV]).3 63. Considering all the adult populations in this study. If these data were not satisfactory (slope below 98%). 2002 TABLE 1. However. Moreover. The patients had not yet started the antituberculosis treatment when the serum samples were taken. and 1/100 for IgA). IgM. 1). To adjust the cutoff value. The extrapulmonary localizations comprised disseminated TB (seven). we selected the mean plus 2 SD measured in the overall control population. N. Inc. lymphatic (two).5 Ϯ 23. xenopi and M. CF.e. the mean values of IgG and IgA antibodies to each of the four glycolipid antigens tested were significantly elevated (P Ͻ 0. and 16 had extrapulmonary TB (three of them were HIV positive).C. no significant differences were observed between the mean values of the different groups of healthy controls within each test. to detect nonspecific absorption. or SL-I (1.1 26. All points were duplicated. (12 serum samples). These healthy control sera were obtained from employees of the HUGTiP or Ph. Comparison between several groups was made using the Kruskal-Wallis analysis (12).5 20 83 40 54.8–17 1–86 28–78 62–86 33–45 26–74 20–28 1–7 33–75 1–80 21–50 21–39 3–16 23–51 40–80 1–12 1–87 described elsewhere (15). and CF were purified using column chromatography as previously described (21. A curve was drawn for each plate. All sera were aliquoted and stored at Ϫ40°C until use. Statistical treatment of results. and in pulmonary TB with respect to extrapulmonary TB (Table 2). ELISA glycolipids. and IgA alkaline phosphatase conjugates (Southern Biotechnology Associates.) at a 1/3. The significance of the difference between means was calculated using the Mann-Whitney U test. Ala. Demographic and clinical data for TB patients and controls Source of serum samples Patients (n) Sex (men:women) HUMORAL RESPONSE TO MYCOBACTERIAL GLYCOLIPIDS 3783 Age Mean Range TB patients Pulmonary adult TB HIV Ϫ HIV ϩ Extrapulmonary adult TB HIV Ϫ HIV ϩ Pulmonary child TB Non-TB pneumonia patients M. The sera from these non-TB pneumonia patients were obtained from the Microbiology Service Seroteca at the HUGTiP. pus abscess (one). and the comparison of their slopes was carried out. One hundred thirty-six HIV-seronegative serum samples were also included as negative controls (Table 1). Eighteen serum samples from children with a clinical diagnosis of TB were also included in the study.001) than the rest of the controls in each test (Fig. respectively). The normalized data were then calculated to establish the corrected ⌬405 values by using the curve of standards. 1). Eighty-four sera were taken from healthy controls: 44 of these were PPD negative. DAT.5 years. In Table 2. students. three titrated sera having low.).. Differences in the IgM antibody levels were observed between the TB patients and the overall control groups for the DAT and TAT antigens. For all the antigens and Igs tested. the illness was subsequently confirmed by isolating Mycobacterium tuberculosis from gastric lavage or from respiratory specimens.8 51. Chicago. Sensitivity and specificity.1 43. C. and smearnegative samples are represented. the results were higher than those of TB children in all the tests. puncture (two). TAT. burnetii L. Detroit.. an M. or they were collected from the Barcelona Tuberculosis Prevention and Control Programme. pneumophila M. Mycoplasma pneumoniae (9 samples from adults and 4 from children). SL-I. the plate was rejected.3 9.000 ng each in 50 ␮l of n-hexane/well). The values of tested sera were corrected as follows: the difference between absorbance of serum and nonspecific absorption (zero) was taken and the mean value was calculated. Ill.001) above those of the overall control sera (healthy and non-TB pneumonia) (Fig. The cutoff points chosen were equal to the means of the OD corrected readings obtained with sera from all of the healthy individuals.000 dilution in blocking buffer were added. Sensitivity and specificity were calculated by standard methods. Goat anti-human IgG.3 33. Data analysis.).1 years versus 31. pneumoniae Adults Children S. Moreover. kansasii Chlamydia spp. the non-TB pneumonia patients did show higher levels (P Ͻ 0.VOL.3 33. Legionella pneumophila (11 samples). To correlate the data for day-to-day variations. for six children. For glycolipid isolation.5 1–87 26–77 26–48 19–87 27–35 0.7 44. Birmingham.]). 1/200 for IgM. bone (one). Inc.


Extra TB. healthy PPD negative. Results of the ELISA for glycolipids in sera from adult subjects. Vaccin. healthy vaccinated. pulmonary TB. healthy PPD positive. 40. PPDϩ. 2002 HUMORAL RESPONSE TO MYCOBACTERIAL GLYCOLIPIDS 3785 FIG. extrapulmonary TB. The dotted line indicates cutoff values above which a test is positive (mean ϩ 3 SD of the healthy population values). 1. . Mycob. and Other resp dis. other respiratory diseases. Individual absorbance values for each test are shown. other mycobacterial diseases. PPDϪ. Each dot indicates an individual serum sample.VOL. Healed. healthy healed. Pulm TB.

5% for IgA. respectively. so test specificities between 96.5) 1 (6.14) 7 (50) 6 (42.75) 3 (18. SL-I.33) 2 (4. 20) and for purified proteinaceous antigens (29).28) 1 (7. very low sensitivity values were obtained in all tests (between 0 and 22. The results obtained by Rojas-Espinosa et al. the low quantity of antigen used by them (only 100 ng per well) while we found 1. The results obtained throughout in the child group and when detecting IgM antibodies in adult sera are particularly bleak. these discouraging results (Table 2) have already been described for three of the four glycolipids (DAT. whereas IgA was found to be more specific.14) 20 (41. and 27 and present results).2%). and TAT IgA tests (Table 2). no antigen tested would seem to be useful as a diagnostic tool in child TB patients. 1 and Table 2). as we did.16) 2 (4.25) 8 (50) 8 (50) 7 (43. MICROBIOL.75) 3 (18. Test used (antigen/Ig) Pulmonary TB patients Total (n ϭ 58) Smear positive (n ϭ 42) Smear negative (n ϭ 16) Total (n ϭ 16) Extrapulmonary TB patients Smear positive (n ϭ 2) Smear negative (n ϭ 14) Non-TB pneumonia patients (n ϭ 48) Healthy subjects (n ϭ 64) DAT/IgG TAT/IgG SL-I/IgG CF/IgG DAT/IgA TAT/IgA SL-I/IgA CF/IgA DAT/IgM TAT/IgM SL-I/IgM CF/IgM 33 (56.34) 41 (70.57) 6 (42.6%). On the other hand.7%). the best results were obtained when detecting IgG and IgA in adult TB patients. tuberculosis strains (8).42) 4 (28. SL-I showed the best relation between test sensitivity and specificity: 81 and 77. In the present study. Thus.58) 40 (68.56) 2 (3. Thus. TAT IgG.14) 2 (14. and SL-I IgA (66. that adult TB patients have a specific IgA response against these glycolipids. taking all the adult populations of this study into account. never- .25) 1 (6.3%). we have accurately analyzed the complete antibody response to each one of these antigenic molecules: DAT.75) 7 (12.12) 2 (3. (7) found. In the case of IgM detection in adults. respectively.25) 0 (0) 7 (43.25) 2 (100) 2 (100) 2 (100) 1 (50) 2 (100) 2 (100) 2 (100) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 5 (35.89) 41 (70. Considering all the adult populations. We think this is due to differences in the ELISA procedure.57) 16 (38.09) 33 (78. CLIN. Even if using a different cutoff from that selected.72) 25 (59. Around 50% of non-TB pneumonia patients reacted to the SL-I IgG. we achieved a test sensitivity of 81%. (24) are discrepant in the sense that they found a high reactivity for the IgM detection. The SL-I IgG test was the most efficient for detecting smearnegative TB cases. for IgG. we have achieved the best relation between sensitivity (67.06) 1 (1.14) 1 (7.25) 1 (6. with a sensitivity of 66. TAT. including the non-TB pneumonia patients.68) 45 (77. in contrast to only 33% reported by those authors. IgG was comparatively more reactive.52) 11 (26. the SL-I IgA test was the most specific (87. for the first time. elevated test specificity values were obtained in the healthy control group (between 96 and 100%).75) 13 (81. TAT IgG (68. with a specificity of 77.56) 0 (0) 1 (1.75) 35 (60.000 ng per well for SL-I to be an optimum coating (15).5%. by combining the detection of IgA and IgG antibodies against SL-I and adjusting the cutoff value.25) 8 (50) 10 (62.14) 10 (71. these percentages dropped in non-TB pneumonia patients. IgA is less sensitive. a test sensitivity of 58. and CF) (7. that the reaction of sera from TB patients was significantly higher than that from healthy controls when detecting IgG antibodies to SL-I. However.25) 6 (37.52) 30 (71.09) 4 (9.8 and 100% were obtained.9%).85) 8 (57. In our case.71) 8 (57. SL-I. Despite it being an exclusive antigen to the virulent M. Seropositive adult subjects with each of the 12 serological tests No. Test sensitivities lower than 25% were obtained in all the tests when detecting IgM antibodies in adult patients.12) 1 (1.06) 16 (27. Taking this control population (non-TB pneumonia patients) into account.75) 10 (62. 9.5) 12 (75) 5 (31.42) 35 (83.2%) (data not shown). Among healthy adult subjects. specifically.56) 1 (1. 24).25) 9 (56.25) 2 (12.5) 1 (6. only 29% of these patients reacted to SL-I IgA.33) 24 (50) 12 (25) 24 (50) 25 (52. Among the antigens.12) 0 (0) 2 (3. the observed humoral response is lower than that with adult patients (refer- ences 3.2 and 87.16) 0 (0) 4 (8. and that IgM reactivity was negligible. but its attractiveness lies in that it is more specific.66) 28 (58. IgA was not tested in either of these previous serological studies using SL-I or in the other studies using the other glycolipids.08) 14 (29.19) 6 (14. In general. We have shown. and 66.90) 33 (78.96) 19 (32.38) 8 (50) 11 (68. One way of improving these results is to adjust the cutoff. only 1 or 2 serum samples out of 64 were positive for some of the tests.1% was obtained for detecting both IgG and IgA antibodies against SL-I.85) 1 (7. including the non-TB pneumonia patients.56 and 93. were achieved by adjusting for the SL-I IgA and IgG tests and combining them.57) 37 (88.68) 48 (82.6%. and CF. DISCUSSION To determine the best test for potential use in TB serodiagnosis. With both Ig’s. test sensitivity and specificity values of 67.16) 1 (2.3786 ´ N ET AL.56) obtained when detecting IgG antibodies against the SL-I antigen (81%.5%) and specificity (93.58) 7 (12. In comparison with IgG. Cruaud et al. and no differences between infected and sick children have been reported (references 27 and 28 and the present work).2%). there are only two previous serological evaluations of IgG and IgM antibodies against SL-I (7. Sensitivity values above 65% were obtained for the following test combinations: TAT IgA (74.33) 26 (61.75%.75) 5 (31. However. using a wide population.56) 1 (1. JULIA TABLE 2. (%) seropositive (n ϭ 186 serum samples) J.1% with a specificity of 90.28) 1 (2. In child populations few studies have been carried out throughout the history of TB serodiagnosis.5) 9 (56.08) 0 (0) 0 (0) 1 (1. giving the highest absorbance values and highest sensitivities (Fig. it is not possible to discriminate between true and false positives in any test. With respect to child TB.

and H. 1989. Wiley & Sons. Papa. S. de Kantor. T. 21.. M. Luh. J. 1995. 2000. Enferm... and M. 271:481–485. F. Nakagawa. Yamamura. 156:906–911. Clinical evaluation of rapid serodiagnosis of pulmonary tuberculosis by ELISA with cord factor (trehalose-6. M. M. Res. tuberculosis.. Bai. Microb. Recognition of phenolic glycolipid-I (Mycobacterium leprae) and sulfolipid-I (M. Simonney. 1993. Sempere. Clin. Yang. 2000. Lane ´elle. Hiraga. Glender. D. M. 148:997–1001. Diagnostic utility of the estimation of mycobacterial antigen A60 specific immunoglobulins IgM. Lung Dis. L. Colangeli. 26. and I. Festestein. Occurrence of an antigenic triacyl trehalose in clinical isolates and reference strains of Mycobacterium tuberculosis.. J. Tuber. S. and M. Very few studies (17) have analyzed the efficacy of absorbents in diminishing the cross-reactions. M. Kumari. F. the specificity drops dramatically when sera from patients with non-TB pneumonia are considered.. M. L. J. Evaluation of synthetic pseudo cord-factor-like glycolipids for the serodiagnosis of tuberculosis. Berlie. Rudd. Tuberc. C. Chest 111:601–605. 10. R. O. A. the specificity went down markedly (5. J. G. 3:1106–1112. and with few (if any) patients infected by other agents. Lung Dis. J. E. Faraone. A. Phadtare. C. Serodiagnosis of tuberculosis. N. 18. and I. S. K. P. L. consequently. When the same test was carried out with a larger number of non-TB pneumonia patients. S. García-Barcelo ´. Kashima. L. Laszlo. A. Dis. 4. 7:240–246. 143:217–223. Nevertheless.. M. J. Evaluation of a new serodiagnostic tuberculosis test based on the immunoglobulin A detection against Kp-90 antigen.VOL. Analysis of circu- . Respir. 8. and L. E. An ELISA for five glycolipids from the cell wall of Mycobacterium tuberculosis: Tween 20 interference in the assay. Human IgG antibodies immunoreacting with specific sulfolipids from Mycobacterium tuberculosis. Alcaide. Ustac Arikan. P. Lett. 1997. J. Daffe ´. Methods 251:21–30. Adv. 141:679–694. V. Leclerc. Eur. Lin. 1990. Chiang. Respir. H. and P. and M. Cruaud. N. M. Bakteriol. 14. Luquin. Am. Infecc. Lyashchenko. Julia ´n. Lab. and H.. the inclusion of this population is crucial for clinical evaluation of new tests. Bothamley. M. Heterogeneous antibody responses in tuberculosis. Rudd. Rodriguez-Gonza ´lez. and S. Respir. Infect. Samanich. could be a useful antigen for inclusion in a TB serodiagnostic test. N. Microbiol. Immunol. C. REFERENCES 1. Lung Dis. Reports showing high test specificities (90 and 95%) studied few samples belonging to non-TB pneumonia patients (1. Sa ´nchez-Monto ´n. A. Udani. 5. J. Bothamley. 7. Mun ˜ oz. He. Us. Zentbl. Clin. J. tuberculosis) by serum from Mexican patients with leprosy or tuberculosis. Respir. tuberculosis. 38:2227–2231. S. By including such a high percentage of non-TB pneumonia patients in the control population. Yano. Y. Daniel. 135:1137–1151. S. M. 158:678–680. M. J. M.Y. a positive result using the SL-I IgA test would not indicate starting an antituberculosis treatment. In this way SL-I. Human humoral responses to antigens of Mycobacterium tuberculosis: immunodominance of high-molecular-mass antigens. R. in a clinical setting. and R. Yamashita. Biostatistics: a methodology for health sciences.. Crit. and P. B. Luna-Herrera. and third. P.3-diacyl-trehalose-2Ј-sulphate (SL-IV) antigen for case finding and diagnosis of leprosy and tuberculosis.. Ueda. David. S. L.. 76:418–424. Baer. M. 23. Infect. Sofia. Tuberc. Gennaro. U ¸elebi. New York. Res. and A. 13:292–296. Ariza. Cruaud. Clinical evaluation of a serological assay using a monoclonal antibody (TB72) to the 38 kDa antigen of Mycobacterium tuberculosis. D. P. E. Bourrillon. 1998. J. Oka. 40... and it would be very interesting to investigate this by testing a wide range of them. FEMS Microbiol. Martínez. Al Jahdali. 39:131–203. Int. F. K. M. 4:1082– 1085. 1995. Ito. M. Evaluation of a sulpholipid antigen. these disappointing results have been borne out by other authors using other antigens. Dis. Laal. Dis. since it would be necessary to have other clinical or microbiological suggestions of TB. Pottumarthy. 2002 HUMORAL RESPONSE TO MYCOBACTERIAL GLYCOLIPIDS 3787 theless. since they are susceptible to being confused with TB patients. C.. and M. L. The unspecific hyperglobulinemia associated with some non-TB respiratory diseases (4) and the fact that the majority of adults have been in contact with environmental mycobacteria all their lives could be reasons for the false positive results in some non-TB pneumonia patients. Evaluation of a novel 2. M. avoiding possible antigen contamination during the purification process. Wells. Papa. L. in the daily routine of a microbiology laboratory a high percentage of these patients will be correctly diagnosed using specific microbiological tests for each of the remaining infectious agents. Mun ˜ oz. K. 2000. Care Med. We think that three strategies could be used to improve the specificity of TB serodiagnosis: first. 7:639–645. Rev. Care Med. Luquin. Ausina. Julia ´n. Humoral response to defined epitopes of tubercle bacilli in adult pulmonary and child tuberculosis. M. Evaluacio ´n de aciltrehalosas de Mycobacterium fortuitum en el diagno ´stico serolo ´gico de tuberculosis pulmonar. Anti-Kp 90 antibodies in the diagnosis of active tuberculosis. Res. Vincent. Clin. 25. 157: 251–259. Cama. Bakteriol. 1997. Gupta.. However. Julia ´n. 148:405– 412. Debanne. 24. Banwalikar. T. 1997.. We have observed that this is a constant in TB serological studies. M. Antibody and antigen detection for the immunodiagnosis of tuberculosis: why not? What more is needed? Where do we stand today? J. K. Microbiol. M. Microbiol. 13. Tuncer. 6. Martín-Luengo. H. H. and S. A. J. or with molecules that share epitopes with the specific antigens. van Belle. Alifano. Yu. J. De Pascalis. Gennaro.. J. As mentioned. H. R. The serodiagnosis of tuberculosis and other mycobacterial diseases by enzyme-linked immunosorbent assay. and P. L. Am.. 1997. Thus. M. and H. and S. H. 4:49–56. Suo. Microbiol. J. D. Physiol. Lagrange.. V. 278:49–57. J. with other bacteria. V. 1998. L. C. Dis. 16). Daffe ´. 1999. Infect. is one of the common defects in TB serology studies (31). Menzies. Covelli. Gupta. Serodiagnosis of tuberculosis. Diagn. 66:3936–3940. Valero-Guille ´n. S. 1993. and P. ACKNOWLEDGMENT This work was supported by grants from the Fondo de Investigaciones Sanitarias de la Seguridad Social (96/1422). 12. and M. Y. A comparison of seven tests for serological diagnosis of tuberculosis. T. David. 22. Morris. 11. Evaluation of IgA-mediated humoral immune response against 2. Rojas-Espinosa. Draper. Respir. Eur. W. 3. Microbiol. It has been clearly accepted that TB patients produce antibodies against more than one antigen (19) and. M. S. 17. 1987. 1997. Ausina. J. always using the specific antigens present only in M.. and G. N.. Maekura. Zolla-Pazner. Julia ´n. V. Herna ´ndez. Lo ´pez-Marín. Am. L. Crit. we have included a large number of non-TB pneumonia patients (38. T. M. S. Immunologic diagnosis of tuberculosis. Arce-Paredes. T. Immunol. and J. 19. 20. 30(Suppl. Handzel. some authors (13) and ourselves propound the combination of specific purified antigens to increase the sensitivity of serodiagnosis. G. Clin. J. Fisher. 1996. 1998. we were able to choose the most specific antigen-antibody combination. 2000. 1998. and P. Int. 1992. preabsorbing the sera either with mycobacteria other than M.. T. 1994. T. Belisle. Savage.. H. Unquestionably. M. and J. Houde. Luquin. F. Matas. S. E.. 2). Dis. Chest 114:1253–1257. Zentbl. a wide spectrum of humoral responses exists in these patients. Rev. Ivanyi. K. Microbiol. M. R. second. Escamilla. Distribution of surface-exposed antigenic glycolipids in recent clinical isolates of Mycobacterium tuberculosis. 154:1864–1867. N. the mycobacterial antigen P-90 in diagnosis of pulmonary tuberculosis. P. Immun. 15. 1993. Thus. Am. E. Barrera. IgA and IgG in the sera of cases of adult human tuberculosis.. The envelope layers of mycobacteria with reference to their pathogenicity. Del Pozzo. and I. P. A study comparing three specific mycobacterial antigens. ¨ nal. combined with other specific antigens. Sonnenberg. I. T. Mycobacterium fortuitum glycolipids for the serodiagnosis of pulmonary tuberculosis. 27. The evaluation of new tests using mainly healthy subjects.6Ј-dimycolate) as antigen purified from Mycobacterium tuberculosis. 3):S243–S246. R. the usefulness of the SL-I IgA test will be greater than is shown by the overall results obtained in this study. 9. Infect. 16. Goren. Yano. M. G. Clin. Martín-Casabona. Daniel. 2001. Lane ´elle.2% with respect to overall controls) in the present study. Luquin. 1988. J. Mancilla.-M. For this reason. Torgal-Garcia. H.

In P. and J. Chapman & Hall. Antibody detection in tuberculosis. London. and M. Duvivier. E. Methods 177: 69–77. Humoral immune response in human tuberculosis: immuno- . Tuberc. 1998.. Clinical tuberculosis. J. van Nerom. Handzel. M. 4:152–160. O. G. Toppet. JULIA J. Davies (ed. CLIN. M. J. Care Med. 29.-P.-C. de Bruyn. Determination of humoral immunoglobulins M and G directed against mycobacterial antigen 60 failed to diagnose primary tuberculosis and mycobacterial adenitis in children. P. United Kingdom. and A. 81–96.3788 ´ N ET AL. 1994. p. L. Nyabenda. globulins G. Respir. 150:1508–1512. E. Laszlo. E. MICROBIOL. 31. Lung Dis. Development of an enzyme-linked immunosorbent assay (ELISA) combined with a streptavidinbiotin and enzyme amplification method to detect anti-2. 1988. Microbiol. Crit. Clin. Dierckx. van Vooren. Yernault. J. J. 26:1714–1719.3-di-o-acyltrehalose (DAT) antibodies in patients with tuberculosis. Turneer. Turneer. 1994. 28. J. Vera-Cabrera. Waelbroek. J. A. D. L. V.). A. lating immune complexes (CICs) in childhood tuberculosis: levels of specific antibodies to glycolipid antigens and relationship with serum antibodies.. Int. and M directed against the purified P32 protein antigen of Mycobacterium bovis bacillus Calmette-Gue ´rin.. Am. J. Immunol. Serruys. 30. Wilkins. A.