Development of a sensitive competitive enzyme-linked immunosorbent assay for serodiagnosis of Burkholderia mallei, a Tier 1 select agent

Abstract

Glanders is a highly contagious and potentially serious disease caused by Burkholderia mallei, a Tier 1 select agent. In this study, we raised a monoclonal antibody (mAb) against the lipopolysaccharide (LPS) of B. mallei and developed a competitive enzyme-linked immunosorbent assay (cELISA) for B. mallei infection. Using the titrated optimal conditions of B. mallei-LPS (2 ng) for microtiter plate coating, sample serum dilution at 1:20 and 3.5 ng/μL anti-LPS mAb B5, the cutoff value of the cELISA was determined using serum samples from 136 glanders-free seronegative horses in Hong Kong. All calculated percentage inhibition (PI) values from these seronegative samples were below 39.6% inhibition (1.5 standard deviations above mean PI) and was used as the cutoff value. The diagnostic sensitivity of the developed LPS-based cELISA was first evaluated using sera from donkeys and mice inoculated with B. mallei. An increasing trend of PI values above the defined cELISA cutoff observed in the donkey and mouse sera suggested positive detection of anti-LPS antibodies. The sensitivity and specificity of the LPS-based cELISA was further evaluated using 31 serologically positive horse sera from glanders outbreaks in Bahrain and Kuwait, of which 30 were tested positive by the cELISA; and 21 seronegative horse sera and 20 seronegative donkey sera from Dubai, of which all were tested negative by the cELISA. A cELISA with high sensitivity (97.2%) and specificity (100%) for the detection of B. mallei antibodies in different animals was developed.

Fig 1. Reactivity of anti-LPS B5 to a variety of bacterial strains.

Western blot (top panel) showing reactivity of the mouse monoclonal anti-LPS antibody B5 produced in this study against the bacterial lysate of seven Burkholderia species and eight common Gram-negative pathogens. The bottom panel shows the protein present in each lane as stained by Coomassie blue.

Fig 2. Specificity and optimization of anti-LPS B5.

The specificity and optimal concentration of the mouse monoclonal anti-LPS antibody (mAb) B5 produced in this study was determined using indirect ELISA with varying amounts of B. mallei lipopolysaccharide (LPS; black lines) or bovine serum albumin (BSA; grey lines), and detecting antigen with varying concentrations (1.75, 3.5, and 7.0 ng/μL) of mAb B5.

Fig 3. Lipopolysaccharide (LPS)-based competitive ELISA (cELISA) in the presence of 2 ng of B. mallei LPS and 3.5 ng/μL mAb B5 with varying dilutions of B. mallei-positive and negative sera. cELISA values calculated as percentage inhibition (PI) of mAb B5 binding.

Serum from a glanders-seropositive horse (SE1719.4/19) from a glanders outbreak in Kuwait was used as the B. mallei-positive sample. Serum from a glanders-seronegative horse (ED18-0081-0001) from AFCD in Hong Kong was used as the B. mallei-negative sample. Dashed line indicates 2 standard deviations above the mean PI of the B. mallei-negative serum at each serum dilution.

Fig 4. Evaluation of lipopolysaccharide-based competitive ELISA (cELISA) using horse and donkey sera obtained from Hong Kong and the Middle East (Dubai, Bahrain and Kuwait).

Data indicate the distribution of cELISA values calculated as percentage inhibition (PI) of mAb B5 binding. The defined cELISA cutoff value (PI of 39.6%) is indicated by the dashed line. Solid lines represent the median and interquartile range. Circle, seronegative sera from glanders-free horses and donkeys; triangle, horse seropositive sera from a glanders outbreak in Bahrain; square, horse seropositive sera from a glanders outbreak in Kuwait; empty triangle, horse seropositive serum from a glanders outbreak in Bahrain which was tested as negative by the developed LPS-based cELISA.

Fig 5. Using the lipopolysaccharide-based competitive ELISA (cELISA) to detect anti-LPS antibodies in seven donkeys and two mice inoculated with B. mallei.

The donkeys were challenged via the following routes: oral injection at inoculum of 1.0 × 10⁹ CFU/mL (donkeys 1 and 2), oral injection at inoculum of 1.0 × 10⁸ CFU/mL (donkey 3), direct feeding through water (donkey 4), direct feeding through feed (donkey 5), intranasal spray (donkey 6), and subcutaneous injection (donkey 7). The two mice (mouse 1 and 2) were inoculated twice with non-viable B. mallei via intraperitoneal injection. Numbers on top of bars indicate the day post infection (donkey) or the day after the first inoculation (mouse) on which sera was collected. Data indicate cELISA values calculated as the percentage inhibition (PI) of mAb B5 binding. The defined cELISA cutoff value (PI of 39.6%) is indicated by the dashed line.