Detection of buried explosives with immobilized bacterial bioreporters

Abstract

The unchecked dispersal of antipersonnel landmines since the late 19th century has resulted in large areas contaminated with these explosive devices, creating a substantial worldwide humanitarian safety risk. The main obstacle to safe and effective landmine removal is the identification of their exact location, an activity that currently requires entry of personnel into the minefields; to date, there is no commercialized technology for an efficient stand-off detection of buried landmines. In this article, we describe the optimization of a microbial sensor strain, genetically engineered for the remote detection of 2,4,6-trinitrotoloune (TNT)-based mines. This bioreporter, designed to bioluminescence in response to minute concentrations of either TNT or 2,4-dinitotoluene (DNT), was immobilized in hydrogel beads and optimized for dispersion over the minefield. Following modifications of the hydrogel matrix in which the sensor bacteria are encapsulated, as well as their genetic reporting elements, these sensor bacteria sensitively detected buried 2,4-dinitrotoluene in laboratory experiments. Encapsulated in 1.5 mm 2% alginate beads containing 1% polyacrylic acid, they also detected the location of a real metallic antipersonnel landmine under field conditions. To the best of our knowledge, this is the first report demonstrating the detection of a buried landmine with a luminescent microbial bioreporter.

Fig. 1

Luminescent response of alginate–PAA beads (2% w/v alginate, 1% w/v PAA) when placed on LB–agar containing 5 mg l⁻¹ DNT. The beads were prepared with various concentrations of the BS02 bioreporter strain. The response is presented as the difference in luminescent signal intensity between the sample and the control, in the plate reader’s arbitrary relative light units (RLU).

Fig. 2

(A) Effect of PAA content on the rate of water loss from alginate–PAA beads. Beads were allowed to desiccate on an open Petri dish at room temperature. Error bars represent the standard deviation of three samples, each containing 10 beads of ca. 3 mm diameter. (B) Effect of PAA on the response to DNT. A monolayer of alginate beads (including 0%–1.5% PAA), encapsulating reporter strain BS02 (1.5 × 10⁵ cells per bead, 1.55 mm diameter), was placed on LB–agar with DNT (5 mg l⁻¹) in a 24‐well plate. The response is presented as the maximal difference (in the plate reader’s arbitrary relative light units) between the sample and the control over a period of 10 h. Error bars represent the standard deviation over three different samples.

Fig. 3

Bioluminescent response of bead‐encapsulated sensor strains BS01, BS02 and BS03 to DNT buried in sand (15 μg per well, 5 mg kg_sand ⁻¹). All three strains carry a ΔpykF deletion and the pACYC_yhaJ_G2 plasmid, along with one of three lux variants; C55_Pl (BS01), C55_Af (BS02), and C55_Pleio (BS03). A single layer of 1.55 mm beads (1.51 × 10⁵ cells per bead, 2% w/v alginate, 1% PAA) was placed on each well of an opaque white 24‐well microtitre plate. Error bars represent the standard deviation over at least 3 measurements.

Fig. 4

Bioluminescent response to DNT (2 mg l⁻¹ in LB–agar) of immobilized bioreporters (strain BS03), subjected to lyophilization and stored for three months at −80°C and −20°C. The signal intensity (A) and signal to background ratio (B) were monitored every 2–3 weeks. Error bars represent the standard deviation of 3 independent measurements.

Fig. 5

Response of the immobilized BS01 reporter strain to buried explosives. DNT (50 g) was buried under 8 cm of sand in a 15 cm × 25 cm container, and incubated for 1 year at room temperature. Luminescent response of the biosensors was monitored in an emulator chamber (70% humidity, 25°C) by a custom‐made detection system, consisting of a photomultiplier device that traversed the target area at constant time intervals.

Fig. 6

Luminescent response of BS03 reporter strain, immobilized in 1.55 mm alginate–PAA beads (1.51 × 10⁵ cells/bead, 2% w/v alginate, 1% PAA), 5.5 h following exposure to an empty sand target (left) and to an aged target containing a PRB‐M35 antipersonnel landmine buried in sand (right).