Profiling Antibodies to Mycobacterium tuberculosis by Multiplex Microbead Suspension Arrays for Serodiagnosis of Tuberculosis

M. tuberculosis antigens.Early secretory antigenic target 6 (ESAT-6; Rv3875), culture filtrate protein (CFP-10; Rv3874), HspX (Rv2031C), MPT53 (Rv2878C), and MPT63 (Rv1926C) were expressed in Escherichia coli as recombinant histidine-tagged products and purified to near-homogeneity as previously described (5). These recombinant antigens were characterized for purity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (data not shown). Peptides representing Ag85B protein were obtained from Genemed Synthesis Inc. (San Antonio, TX). These Ag85B peptides have been previously used as antigens in measuring frequencies of anti-M. tuberculosis CD8⁺ T cells in TB patients (16).

Antigen coupling to microbeads.Microbead sets, individually identifiable by unique fluorescence (Luminex Corp., Austin, TX), were coupled through carbodiimide linkages with each of the five M. tuberculosis recombinant antigens and Ag85B synthetic peptides as previously described (12, 13).

Nonhuman primates and plasma samples.Four specific-pathogen-free (SPF) rhesus monkeys (numbers 21200, 21300, 21222, and 21229), housed at the Oregon National Primate Research Center (Beaverton, OR), were experimentally inoculated as previously described (17). Briefly, M. tuberculosis strain H37Rv was inoculated by instillation of 1,000 CFU into the right lower lobe of each subject. Disease progression was monitored, and animals were characterized for infection and disease by computer tomography (CT) scanning, chest X-rays, and M. tuberculosis culture in bronchoalveolar lavage fluid (17). In these animals, disease progression varied. CT scans revealed early signs of small focal lung lesions in three of four animals at week 4 (21229, 21222, and 21200); lesions became progressively more pronounced at 8 and 12 weeks postinoculation in all four animals. Results from chest X-rays and M. tuberculosis cultures from bronchoalveolar lavage fluid followed a similar pattern as for the CT scans with the exception that animal 21203 did not become culture positive during the 12 weeks of monitoring (17). Necropsies were done on these animals, and the specimens were subjected to histopathological examination (17). Plasma samples were obtained from these animals at preinoculation and postinoculation time points as follows: 4 weeks, 8 weeks, and 12 weeks.

For latent and active-chronic infection studies, cynomolgus macaques, housed at the Primate Facility at the University of Pittsburgh (Pittsburgh, PA), were inoculated with M. tuberculosis strain Erdman (4). Animals were inoculated with 25 CFU per animal in 2 ml of sterile saline via bronchoscopic instillation into the right caudal or middle lung lobe. Latent infection was characterized as the presence of granulomas and low bacterial loads (compared to active disease) in the lung at necropsy but no radiographic involvement for 4 weeks postinfection or clinical signs of TB for at least 6 months postinfection (4). Samples at several time points, spanning a period over 2 years postinoculation, from two animals representing latent disease were analyzed by MMIA (see Fig. 2B, below). Active-chronic disease was characterized as previously described (4) and was based on evidence of persistent disease, radiographic involvement, persistent culture positivity, or other clinical signs of active disease. Samples at several time points spanning about 8 months to 1 year postinoculation from four active-chronic disease animals were analyzed by MMIA (see Fig. 2B, below). In addition, samples drawn at 1 year postinoculation from five active-chronic infected animals were analyzed (see Fig. 2A, below). Preinoculation plasma samples from four of these animals were used as controls (see Fig. 2A, below). Animal protocols were approved by the Institutional Animal Care and Utilization committees at the University of California, Davis.

TST and Primagam.TST and Primagam assays were performed as previously described (23). For TST, animals were skin tested using mammalian old tuberculin as follows: (i) SPF animals (n = 20), Mycobacterium bovis bacillus Calmette-Guérin (BCG) inoculated (n = 9), and Mycobacterium avium-positive monkeys (n = 10) were tested once a year; (ii) rhesus macaques (see Fig. 1, below) were tested prior to and 1 month after inoculation with H37Rv; (iii) cynomolgus macaques (see Fig. 2, below), infected with the Erdman strain, were tested by skin test and Primagam assay prior to infection and at various time points during the infection.

• Open in new tab
• Download powerpoint

FIG. 1.

Analysis of M. tuberculosis antibodies in early stages after experimental inoculation of four rhesus macaques. (A) Detection of antibodies to M. tuberculosis antigens by MMIA. All bead sets with positive results are shown. Results are shown as the MFI for each microbead set coated with an antigen. The assay background for each microbead set is shown in red (average MFI ± standard deviation), representing the reactivity of each to plasma samples from uninfected rhesus macaques (n = 20). (B) Detection of antibodies by Western blotting to ESAT-6, CFP-10, and HspX. Animal numbers are indicated, and samples A, B, C, and D represent 0, 4, 8, and 12 weeks postinoculation. In addition, two negative and two positive control plasma samples (from rhesus macaques naturally infected in a TB outbreak at Tulane National Primate Research Center, LA) were included in the analysis (-Ve Cntrl and +Ve Cntrl, respectively). Positive control animals used for ESAT-6 and CFP-10 are numbers BC64 and EV64 (lanes 1 and 2, respectively), and those for HspX are BH33 and EV64 (lanes 1 and 2, respectively). Antibodies are indicated by arrows. HspX antibodies were only detected in one of two positive control samples and in none of the experimentally infected animals.

• Open in new tab
• Download powerpoint

FIG. 2.

Detection of antibodies to M. tuberculosis antigens by MMIA from latent and active-chronic infection samples (cynomolgus macaques). (A) All bead sets with positive results are shown. Plasma samples from five animals experimentally infected with M. tuberculosis for 1 year and samples from four of the same animals before inoculation (control group I) are shown as red closed circles. Samples from uninfected animals collected at six time points over more than 2 years are shown as green closed circles (n = 10 animals; control group II). Animals infected with M. avium (n = 10 animals; control group III) and those inoculated with BCG vaccine (n = 9 animals; control group IV) are shown in blue and purple closed circles, respectively. (B) Time course analysis of M. tuberculosis antibody profiles by multiplex microbead immunoassay. Results for all six bead sets are shown; two samples (110 and 111) were an exception, where the Ag85 bead set was not included in the experiments. Animals 110 and 111 were latently infected for about 2 years. Animals 15704 and 15204 and animals 10105 and 10605 were infected for about 1 year and about one-half year, respectively. Immune responses measured by Primagam assay are shown as the optical density at 450 nm for comparison. The Primagam assay cutoff was an optical density of 0.025.

Multiplex detection of anti-M. tuberculosis antibodies.The MMIA, containing a mixture of six microbead sets, one for each of six coated antigens described above, was used in the analysis of plasma samples, which were diluted 1:200 in Prionex (bio-WORLD, Dublin, OH), as previously described (12). Plasma samples from 20 SPF rhesus macaques provided the baseline (background) median fluorescence intensity (MFI) for the MMIA (see Fig. 1A, below). Positive control plasma samples for MMIA and Western blot analysis (see Fig. 1B, below) were obtained from three naturally infected rhesus macaques (from a TB outbreak at the Tulane National Primate Research Center, Corington, LA; P. J. Didier, personal communication). All three of these animals (EV64, BH33, and BC64) were skin test positive and Primagam positive. In addition, these animals were found positive for TB pneumonia (EV64) and TB lymphadenitis (BH33 and BC64) at necropsy. Additional negative control plasma samples, obtained at the California National Primate Research Center (Davis, CA), were as follows: six serial samples taken over 2 years from 10 SPF animals (control group II) (see Fig. 2A, below), 10 animals infected with a nontuberculous mycobacterium (M. avium) (control group III), and 9 animals vaccinated with BCG (control group IV) (see Fig. 2A, below).

Western blotting.Western blot analysis to confirm the detection of antibodies by MMIA against ESAT-6, CFP-10, and HspX was performed as previously detailed (14).

Detection of M. tuberculosis antibodies in early stages of infection.Profiles of antibodies against M. tuberculosis antigens in four rhesus macaques, at multiple time points postinfection, are shown in Fig. 1A. The MMIA included a mixture of six microbead sets, each coated with one of six antigens: ESAT-6, CFP-10, HspX, MPT53, MPT63, and Ag85B. In each sample, the MMIA was able to simultaneously detect antibodies to all but MPT63. Anti-M. tuberculosis antibody profiles produced by the MMIA allowed serodetection of infection in two animals as early as 4 weeks (21222 and 21229), in three animals by 8 weeks (21203, 21222, and 21229), and in all four by 12 weeks (Fig. 1A). Each of the four monkeys contained antibodies against at least two antigens during the 12-week period. One monkey (21203) contained antibodies to four antigens (ESAT-6, CFP-10, MPT53, and Ag85B), while another (21229) contained antibodies to three antigens (ESAT-6, CFP-10, and HspX) over the experimental period. Western blot analysis was performed to confirm detection of ESAT-6, CFP10, and HspX antibodies (Fig. 1B). Results obtained by the MMIA correlated with those obtained by Western blotting. However, the multiplex assay detected HspX antibodies that were undetectable by Western blotting (Fig. 1A and B). Antibodies against MPT63 were not detected in experimentally infected animals; the MPT63-coated microbead set detected antibodies in naturally infected rhesus macaques (data not shown).

Antibody profiles to M. tuberculosis in latent and active-chronic infection.To test the utility of antibody profiling in latent and active-chronic M. tuberculosis infection, cynomolgus macaques infected with M. tuberculosis for various lengths of time (4) were studied. In a group of five macaques at 1 year postinoculation, antibodies against M. tuberculosis antigens that were clearly detected were as follows: five of five animals were positive for ESAT-6, three of five animals were positive for HspX, and one of five animals was positive for CFP-10 antibodies (Fig. 2A). In preinoculation plasma samples (control group I), anti-M. tuberculosis antibodies were not detected (Fig. 2A). In addition, three other control groups (control groups II, III, and IV) were included in the analysis, all of which were negative for anti-M. tuberculosis antibodies (Fig. 2A), demonstrating specificity of the MMIA for serodetection of M. tuberculosis infection.

To investigate the kinetics and dynamics of anti-M. tuberculosis antibody profiles in latent and active-chronic infection, serial plasma samples from six cynomolgus macaques, experimentally infected with M. tuberculosis, were analyzed. Plasma samples from two monkeys were collected from each of three groups at several time points (Fig. 2B): (i) those latently infected for 2 years (numbers 110 and 111), (ii) those chronically infected for 1 year (15204 and 15704), and (iii) those infected for 0.5 year (10105 and 10605). In comparison to controls (control groups I, II, III, and IV) (Fig. 2A), all infected animals contained antibodies to multiple M. tuberculosis antigens (except animal 110, in which only ESAT-6 antibody was detected) (Fig. 2B). The Primagam test for cell-mediated immune responses, in contrast, displayed negative or barely detectable results at a number of time points (Fig. 2B). TST results were similarly inconsistent (data not shown). The following observations were made: (i) antibodies against ESAT-6 appeared early (6 to 7 weeks) and in most of the infected animals, (ii) antibodies against CFP-10 (animals 111, 15704, and 10105) and HspX (15204, 10105, and 10605) appeared in fewer animals, and (iii) antibodies against Ag85 and MPT53 were detected in only one animal (15204, at 41 weeks). Antibodies against MPT63 were not detected in any of the above samples.