Immunoassay with Novel Paired Antibodies for Detection of Lipoarabinomannan in the Pleural Fluid and Plasma of Patients with Tuberculous Pleurisy

Abstract

Tuberculous pleurisy (TP) is one of the most common forms of extrapulmonary tuberculosis, but its diagnosis is challenging. Lipoarabinomannan (LAM) antigen is a biomarker for Mycobacterium tuberculosis (Mtb) infection. LAM detection has potential as an auxiliary diagnostic method for TP. We have successfully generated five rabbit anti-LAM monoclonal antibodies (BJRbL01, BJRbL03, BJRbL20, BJRbL52, and BJRbL76). Here, anti-LAM antibodies were tested to detect LAM in the pleural fluid and plasma of patients with TP by sandwich enzyme-linked immunosorbent assays (ELISAs). The results revealed that all of the anti-LAM antibodies were successfully used as capture and detection antibodies in sandwich ELISAs. The BJRbL01/BJRbL01-Bio pair showed better performance than the other antibody pairs for detecting mycobacterial clinical isolates and had a limit of detection of 62.5 pg/mL for purified LAM. LAM levels were significantly higher in the pleural fluid and plasma of patients with TP than in those of patients with malignant pleural effusion or the plasma of non-TB, and LAM levels in the pleural fluid and plasma were positively correlated. Moreover, LAM levels in the pleural fluid sample were significantly higher in confirmed TP patients than in clinically diagnosed TP patients. Our studies provide novel LAM detection choices in the pleural fluid and plasma of TP patients and indicate that LAM detection assay has an auxiliary diagnostic value for TP, which may help to improve the diagnosis of TP.

Keywords: lipoarabinomannan; plasma; pleural fluid; sandwich ELISA; tuberculous pleurisy.

Figure 1

Screening paired anti-LAM antibodies in sandwich ELISAs. Unlabeled and biotin-labeled anti-LAM antibodies were used as capture and detection antibodies to detect purified LAM from M.tb H37Rv.

Figure 2

Sensitivities of paired anti-LAM antibodies for purified LAM by sandwich ELISA. Each anti-LAM antibody was used as a capture antibody and a detection antibody in the sandwich assay. Purified LAM from M.tb H37Rv was used as the detection antigen.

Figure 3

Reactivity of the paired anti-LAM mAbs to multiple mycobacterial species in sandwich ELISAs. Eleven slow-growing mycobacterial species (left) and twelve rapid-growing NTM strains (right) were tested for each antibody combination. Slowly growing mycobacterium species included M.tb H37Rv, M. bovis, M. kansasii, M. marinum, M. scrofulaceum, M. gordonae, M. xenopi, M. avium, M. intracellulare, M. gastri, and M. malmoense; rapidly growing NTM strains included M. smegmatis, M. fortuitum, M. aurum, M. neoaurum, M. abscessus, M. parafortuitum, M. salmoniphilum, M. nonchromogenicum, M. vaccae, M. phlei, M. confluentis, and M. gilvum.

Figure 4

Reactivity of the paired anti-LAM mAbs to multiple mycobacterial clinical isolates in sandwich ELISAs. Thirty M. tuberculosis isolates (left) and ten M. abscessus isolates (right) were used to evaluate the performance of each anti-LAM antibody pair in sandwich ELISAs.

Figure 5

Reactivity of the anti-LAM mAbs to common pneumonia-causing pathogenic bacteria. Reactivity of the anti-LAM mAbs to the inactivated supernatants of pathogenic bacteria that commonly cause pneumonia as measured by indirect ELISA. SP: Streptococcus pneumoniae; SA: Staphylococcus aureus; PA: Pseudomonas aeruginosa; HI: Haemophilus influenzae.

Figure 6

LAM detection in the pleural fluid of patients with tuberculous pleurisy (TP). (A) LAM detection in the pleural fluid of TP patients (n = 160) and malignant pleural effusion (MPE) samples (n = 50) by sandwich ELISA. Horizontal lines indicate the cut-off based on the mean value from the MPE group plus two standard deviations. (B) The receiver operating characteristic (ROC) curve analysis for the evaluation of LAM detection capacity in TP patients. (C) The purified LAM standard curve as detected with the BJRbL01 and BJRbL01-Bio antibodies by sandwich ELISA. (D) The concentration of LAM in the pleural fluid. Differences were assessed by a Mann–Whitney test.

Figure 7

LAM detection in the plasma of patients with tuberculous pleurisy (TP). (A) LAM detection in the plasma of patients with TP (n = 45) and non-TB (n = 30) by sandwich ELISA. Horizontal lines indicate the cut-off based on the mean value from non-TB plus two standard deviations. (B) The receiver operating characteristic (ROC) curve analysis for the evaluation of LAM detection capacity in TP patients. (C) The concentration of LAM in the plasma. (D) The correlation analysis of LAM detection in the pleural fluid and plasma of TP patients. Differences were assessed by a Mann–Whitney test.

Figure 8

Analysis of LAM level in pleural fluid and plasma in patients with confirmed and clinically diagnosed tuberculous pleurisy (TP). (A) LAM level detection in the pleural fluid of patients with confirmed TP (n = 88), clinically diagnosed TP (n = 72), and malignant pleural effusion (MPE, n = 50) by sandwich ELISA. (B) The receiver operating characteristic (ROC) curve analysis for the evaluation of pleural fluid LAM detection capacity in the two TP subgroups. (C) LAM level detection in the plasma of subjects with confirmed TP (n = 19), clinically diagnosed TP (n = 26), and non-tuberculosis (non-TB, n = 30) by sandwich ELISA. (D) The ROC curve analysis for the evaluation of plasma LAM detection capacity in the two TP subgroups. Horizontal lines indicate the cut-off based on the mean value from MPE or non-TB plus two standard deviations, respectively. Differences were assessed by a Mann–Whitney test.