Portable FRET-Based Biosensor Device for On-Site Lead Detection

Abstract

Most methods for measuring environmental lead (Pb) content are time consuming, expensive, hazardous, and restricted to specific analytical systems. To provide a facile, safe tool to detect Pb, we created pMet-lead, a portable fluorescence resonance energy transfer (FRET)-based Pb-biosensor. The pMet-lead device comprises a 3D-printed frame housing a 405-nm laser diode-an excitation source for fluorescence emission images (YFP and CFP)-accompanied by optical filters, a customized sample holder with a Met-lead 1.44 M1 (the most recent version)-embedded biochip, and an optical lens aligned for smartphone compatibility. Measuring the emission ratios (Y/C) of the FRET components enabled Pb detection with a dynamic range of nearly 2 (1.96), a pMet-lead/Pb dissociation constant (K_d) 45.62 nM, and a limit of detection 24 nM (0.474 μg/dL, 4.74 ppb). To mitigate earlier problems with a lack of selectivity for Pb vs. zinc, we preincubated samples with tricine, a low-affinity zinc chelator. We validated the pMet-lead measurements of the characterized laboratory samples and unknown samples from six regions in Taiwan by inductively coupled plasma mass spectrometry (ICP-MS). Notably, two unknown samples had Y/C ratios significantly higher than that of the control (3.48 ± 0.08 and 3.74 ± 0.12 vs. 2.79 ± 0.02), along with Pb concentrations (10.6 ppb and 15.24 ppb) above the WHO-permitted level of 10 ppb in tap water, while the remaining four unknowns showed no detectable Pb upon ICP-MS. These results demonstrate that pMet-lead provides a rapid, sensitive means for on-site Pb detection in water from the environment and in living/drinking supply systems to prevent potential Pb poisoning.

Keywords: FRET; Met-lead; groundwater lead; lead biosensors; portable Pb sensor; smartphone-based device; tap water lead.

Figure 1

Design and layout of the portable FRET-based Pb-sensing device (pMet-lead), and examples of images generated with pMet-lead compared to conventional FRET ratio imaging. (A) Side view (upper), exploded view (middle), and entity (bottom) of the optical design of the pMet-lead device to demonstrate the general setup of pMet-lead: (1) the top plate holds the smartphone; (2) the middle part of the apparatus contains the plano-convex lens and is linked to the filter turret (part 4); (3) the bottom part contains the illumination sources (LED for BF, LASER Diode for FL and light switch to turn off (O) or on to BF (I) or FL (II)), and is linked to the sample stage (part 5); (4) the filter turret contains two emission fluorescence filters that receive YFP and CFP emission signals (changeable according to various FRET pairs); (5) the sample stage receives the biochip containing the cells expressing the Met-lead biosensor. (B) Molecular structure design (left, originally from [8], and permission has been obtained to use the same here) and emission spectrum (right) of FRET-based Met-lead. (C) The representative bright-field (BF, left) and fluorescence (YFP, middle; CFP, right) images of the Pb-biosensor-expressing HEK293 cells obtained with pMet-lead (above) or under a FRET microscope (with 10 × objective; below). Scale bar, 30 μm.

Figure 2

Processing of fluorescence images in pMet-lead. (A) Flowchart of the computational window graphical user interface (GUI) for ratio color imaging. After image input (loading) of YFP and CFP, select ‘ratio’ to make and save the ratio images to temporary memory sites. The calculated ratios will be displayed in a rainbow color format (display update; color map; maximum mean). The ranges of selected regions of interest (ROIs) can be manually adjusted to show the best present status (specify ROI; contrast; threshold slider). (B) Screen shot of the GUI for the FRET-based device interface during Pb sensing. The ratio color bar represents YFP/CFP ratios from 2 (black) to 7 (white). (C) The whole procedure for on-site Pb sensing with pMet-lead.

Figure 3

Pre-processing of pMet-lead fluorescence images taken with a smartphone camera. (A) Met-lead-expressing cells taken from pMet-lead were first separated into YFP and CFP channels. The Y/C ratio image was obtained through the division of fluorescence from the YFP and CFP channels. Selected regions of YFP, CFP, and Y/C ratio from an image size of 400 × 400 pixels (left) were further magnified to 100 × 100 pixels (middle) and 10 × 10 pixels (right). Black spots can be observed from both the YFP and CFP channels. (B,C) A conceptual illustration of the function of the image processing step described in Section 2.5 to fill black spots with signals, from adjacent regions of fluorescent YFP or CFP signals (B), or Y/C ratios (C). (D) Comparison of images without (control) or with Pb (100 nM) taken from pMet-lead without (original) and with (pre-processed) corrections as described in Section 2.5. Scale bars, 30 μm. The ratio color bars are from 2 to 5 for the original images (left), and from 2 to 6 for the pre-processed images (right).

Figure 4

Sensing ability of pMet-lead. (A) The representative pMet-lead FRET ratio images of HEK293 cells expressing Met-lead 1.44 M1 taken at various concentrations of Pb from nM (1, 10, 20, 30, 40, 50, and 100 nM) to μM ranges (1 and 10 μM). The fluorescence images of Met-lead 1.44 M1 in the YFP and CFP channels, and the ratio color images (through the preprocessing procedure described in Figure 3) under CFP-specific illumination (405 nm). (B) Bar graphs of FRET ratios at various concentrations of Pb using pMet-lead analysis. The mean differences are described as significant at the 0.0005 (***) level. (C) Validation of pMet-lead (Y/C emission ratio) using ICP-MS (Pb measured in ppb) at various concentrations of Pb (upper: 10, 20, 30, 40, 50, and 100 nM; lower: 100 nM, 1 μM, and 10 μM). Scale bar, 30 μm. The ratio color bar is from 2 to 6.

Figure 5

Removal of zinc with tricine enables practical Pb sensing by pMet-lead. (A) Time-lapse of the emission ratio (YFP/CFP) recorded by the Met-lead biosensing system. Double-distilled water (distillation-distillation H₂O, ddH₂O) from the laboratory was pre-mixed with 11 μM zinc alone (Zn; solid line), 11 μM zinc and 10 mM tricine (Zn + tricine; dashed line), or 11 μM zinc, 10 mM tricine, and 10 μM Pb (Zn + tricine + Pb; dotted line), and introduced at the time point indicated by the arrow. (B) Bar graphs of the averaged Y/C ratio values from Met-lead under the conditions in (A). (C) Samples of tap water from a laboratory faucet, without (sample; solid line) or with 10 mM of tricine added (sample + tricine; dashed line), and 10 μM Pb (sample + tricine + Pb; dotted line) were placed in the Met-lead biosensing system at the time point indicated by the arrow. All experimental sample tests were with ionomycin (5 μM). (D) Bar graphs of the averaged Y/C ratio values from Met-lead under the conditions in (C).

Figure 6

Measurement of Pb levels in selected environmental areas using the portable FRET-based Pb-sensing device. (A) Geographical regions where water samples were randomly collected in selected cities of Taiwan: I. Taipei city; II. New Taipei; III. Taoyuan; IV. Miaoli; V. Nantou; VI. Yunlin. (B) Representative ratio color images of samples I–VI analyzed using pMet-lead. (C) Average ratio for each geographical region compared to control water samples. The mean differences are described as significant at the 0.0005 (***) level. (D) pMet-lead FRET ratio values of Pb measurements from various geographic sources of water were validated by the general standard method (ICP-MS). Scale bar, 30 μm. The ratio color bar is from 2 to 4.5.