Development of an aptasensor to target metallo-β-lactamase through Förster resonance energy transfer

The escalating issue of antibiotic resistance in bacteria necessitates innovative detection methods to identify resistance mechanisms promptly. In this study, we present a novel approach for detecting resistance in Pseudomonas aeruginosa, a bacterium known for its metallo-β-lactamase production during the development of antibiotic resistance. We have designed an aptasensor employing Förster resonance energy transfer utilising two distinct methodologies. Initially, indium phosphide quantum dots with a zinc sulphide shell, and gold nanoparticles were utilised as the Förster resonance energy transfer donor-acceptor pair. Although this system demonstrated a response, the efficiency was low. Subsequently, optimisation involved relocating the donor and acceptor in close proximity and incorporating two quantum dots with varying emission wavelengths as the acceptor and donor. This optimisation significantly enhanced the Förster resonance efficiency, resulting in a novel method for detecting metallo-β-lactamase. Förster resonance energy transfer efficiency was increased from 31% to 63% by optimising the distance and donor using a quantum dot-quantum dot pair. Our findings showcase a cheap, rapid and versatile aptasensor with potential applications beyond antibiotic resistance, highlighting its adaptability for diverse scenarios.


Introduction
The rapid progress of nanotechnology has extended the use of nanomaterials as promising tools for diagnostic biosensors, drug delivery, and biomedical imaging [1].In particular, nanoparticles employed in Förster resonance energy transfer (FRET) based technologies have played a significant role in biosensing, where FRET is used as a 'molecular ruler' to measure and image changes in biomolecular conformations, nucleic acid hybridisation, enzyme activity, and environmental parameters [2][3][4].Here, we apply this to develop a sensor that can be used to detect antimicrobial resistance.
Any fluorescent moiety, including organic small-molecule dyes, fluorescent proteins, lanthanide dyes, and fluorescent nanoparticles, can be used as a FRET donor [5].However, the intrinsic photophysical properties of fluorescent dyes and proteins, which generally have broad absorption/emission profiles and low photobleaching thresholds, have limited their effectiveness in long-term imaging FRET devices [6].The choice of acceptor molecules in FRET sensing devices is particularly important.Two types of molecules can be employed: quenchers (non-fluorescent energy acceptors) employed in turn-on sensors or fluorophores employed in ratiometric sensors [7,8].
Metal nanoparticles and organic molecules can play the role of a quencher, while fluorescent acceptors are typically organic dyes or fluorescent proteins [4].Quantum dots (QDs) offer an alternative fluorophore with high stability, decreased photobleaching and broad absorption profile [3,9].In addition to the advantageous photophysical characteristics of QDs, their nanoparticle structure also presents a large, biochemically accessible surface area making them perfect candidates as both donor and acceptor molecules [10].
For biosensing purposes, the donor and the acceptor can be linked via targeting moieties such as antibodies or aptamers [11,12].The presence of the target causes a conformational change in the targeting moieties, which leads to the separation of the donor and acceptor.Thus, a change in the donors emission is observed leading to the detection and sensing of molecules.Aptamers possess the ability to form a particular secondary structure producing loops and stems that can bind to specific target molecules, making them suitable for FRET-based biosensors.
Since their discovery in 1990, interest in aptamers has grown as alternatives to antibodies.These oligonucleotide molecules have been applied to many applications, from environmental sensors and clinical reagents to biophysical discoveries.Compared to antibodies, aptamers show higher affinity and exquisite specificity for any target, including ions and small molecules.They are more stable than antibodies, and their thermal denaturation is reversible, extending the range of assay conditions and making them perfect candidates for the fabrication of FRET biosensors.
Recent reports have highlighted the advantages of using QD-dye FRET for sensitive sensing platforms However, little work has been done using QD-QD as a FRET pair and there is no mention of QD-QD FRET aptasensors in the literature.Here we studied the fluorescence resonance energy transfer efficiency within these nanoassemblies.Indium phosphide QDs with a zinc sulphide shell (InP/ZnS QDs) were synthesised and conjugated to base specific DNA sequences.This led to the development of a QD FRET aptasensor that can be potentially used for the detection of metallo-β-lactamases. First, a FRET turn-on aptasensor is investigated by quenching the fluorescence of the QDs when attached to AuNPs via an aptamer (and its complementary strand).In the second part, the AuNPs are replaced by a QD; the fluorescence is therefore shifted from green to red due to FRET.
This study details the development of an aptasensor detector for metallo-β-lactamases (see figure 1).The introduction of the enzyme attracts the aptamer Lac30 resulting in the physical separation of the donor and the acceptor.Thus, a fluorescent signal is observed in the case of a turn-on sensor or a shift of the fluorescence back to green, so a change in the intensity ratio of QDs emission peaks can be detected in the case of a ratiometric sensor.These assays yield qualitative insights into potential biosensors developed through this methodology.It's noteworthy that this technology extends beyond sensing proteases, demonstrating its adaptability for monitoring various enzymatic modifications or other biomedical applications [13].

Chemicals and materials
All chemicals and materials were purchased from Sigma Aldrich unless specified otherwise.

Oligonucleotide probes
All oligonucleotides (aptamers and complimentary strands) were ordered from Integrated DNA Technologies (IDT) (table 1).The Lac30 aptamer was ordered with a thiol primer at 5 attached through C6.The thiol primer on the 30-base oligonucleotide sequence allowed the conjugation of AuNPs.The Lac30 was ordered with an amino primer at 5 end through C6.The complementary oligonucleotide Comp2 was ordered with an amino primer at 5 end through C6.The complementary oligonucleotide, Comp1 was also ordered with an aminoterminal at 3 end through C6.

Synthesis of InP/ZnS quantum dots and oligonucleotide conjugation
InP/ZnS QDs were synthesised using a previously published procedure by Tessier et al, 0.45 mmol of indium (III)chloride (99.99%) and 2.2 mmol of zinc(II)chloride (98%) were first mixed in 5 ml of oleylamine (70%) and heated to 120 °C under a vacuum [14].After 60min the reaction was put under a nitrogen atmosphere and heated to 180 °C.Once the desired temperature was reached, 1.6 mmol of tris-(diethylamino) phosphine (97%) was injected quickly into the mixture to allow the InP core growth.This step is considered our zero time point (t = 0 min).After 20 min, 1 ml of 2.2 M sulphur (99.998%) in tri-n-octylphosphine (97%, TOP-S) was slowly injected over 10 min.Subsequently, at 60 min, the temperature was increased from 180 °C to 200 °C.At 120 min, 1 g of zinc stearate in 4 ml of 1-octadecene (ODE, 99%) was slowly injected dropwise for 10 min, after which the temperature was increased from 200 °C to 220 °C.At 150 min, 0.7 ml of TOP-S was injected slowly for 10 min, after which the temperature was increased to 240 °C.0.5 g of zinc stearate in 2 ml of ODE was slowly added at 180 min, after which the temperature was increased to 260 °C.The reaction was stopped after 210 min by cooling the temperature to 70 °C.The InP/ZnS QDs were diluted with toluene (99.8%), precipitated in ethanol, and resuspended in toluene.Further purification was done using a size exclusion column in toluene.This yielded red emitting (λ em = 650 nm) toluene soluble QDs.QDs emitting at 500 nm were synthesised using InI and ZnI precursors, and QDs emitting at 580 nm were synthesised using InBr and ZnBr precursors, as reported by Lauferskey et al [15].0.30 g of mercaptosuccinic acid (MSA, 97%) was mixed with 5 ml of toluene and stirred for 10-15 min to ensure the MSA was dissolved.Next, 1 ml of InP/ZnS QDs (20 mg/mL) was added to the mixture and stirred for 1 min.1 ml of ammonium hydroxide and 1 ml of Milli-Q water were added, and the reaction was left to stir for 5 h or until all the QDs migrated to the water phase.The coloured aqueous layer was collected and purified using ethanol precipitation and centrifugation.The supernatant was discarded, and the pellet was re-dispersed in 1 ml of Milli-Q water.

Synthesis of oligonucleotide conjugated gold nanoparticles
Gold nanoparticles (AuNPs) stabilised with citrate were synthesised following the method previously reported by Piella et al [16].All glassware was cleaned with aqua regia to remove any gold residue that would hinder nanoparticle formation.The glassware was thoroughly rinsed with milli-Q water and then dried in an oven.In a three-necked round-bottom flask, a mixture of 150 ml of sodium citrate (99%) solution (2.2 mM), 0.1 ml of tannic acid (2.5 mM), and 1 ml of potassium carbonate (150 mM) was heated to 70 °C.Once the temperature reached 70 °C, 1 ml of tetrachloroauric acid (25 mM, 99.97%) was injected, resulting in a rapid colour change from clear to red-orange within 2min.The reaction continued for an additional 5 min at 70 °C.Subsequently, the solution was cooled to room temperature, and the 5 nm AuNPs were separated by centrifugation at 25,000 g for 20 min.
Oligonucleotide conjugation to the AuNPs followed the approach outlined by Peng et al [17].Initially, the thiol-modified oligonucleotides (100 μM) underwent thermal treatment, being heated to 90 °C to unfold its structure and subsequently cooled to room temperature.To enhance the efficiency of thiol-aptamer immobilising onto the AuNPs, a solution containing 1 μl of Tris(2-carboxyethyl) phosphine (TCEP) (1 M) was combined with 99 μl of the heated oligonucleotide.This mixture was incubated for 30 min at room temperature to cleave disulphide linkages.Subsequently, 10 μl of the treated oligonucleotide solution was blended with 60 μl of pre-prepared AuNPs.The resulting amalgamation was incubated for 16 h at room temperature.

Nuclear magnetic resonance
The oligonucleotide sample was analysed by 1H nuclear magnetic resonance (NMR).Oligonucleotide samples were dissolved in 500 μl tris(hydroxymethyl)aminomethane hydrochloride buffer (2 mM Tris-HCl, 10 mM NaCl, 0.5 mM KCl, 0.2 mM MgCl 2 and 0.1 mM CaCl 2 ; pH 7.5) containing 20 % D 2 O.The final concentration of DNA was 2 mM.NMR experiments were performed on 600 MHz NMR spectrometer type JNM-ECZ600R with data recorded at 20 °C.1D proton NMR spectra were acquired with 3500 data points using 1280 scans due to low sample concentration.

Photoluminescent spectroscopy
FRET samples were prepared following the published study by Shi et al [18].For the QD-AuNP FRET pair: A fixed amount (0.02 mg/mL) of QDs-Comp1 was incubated for 2 h at room temperature with different concentrations of AuNPs (AuNPs-Lac30 for FRET and citrate capped AuNPs as control samples).For the QD-QD FRET pair: a fixed amount of QDs500 (0.08 mg/mL) was incubated with increasing concentrations of QDs650 from 0.2 mg/mL to 0.8 mg/mL.Samples were incubated for 2 h at room temperature to allow for the binding of the oligonucleotide.Photoluminescence data were obtained using a 1 cm quartz cuvette loaded in an Edinburgh Instruments FLS980 spectrophotometer equipped with a continuous wave Xenon lamp, single-grating excitation and emission monochromators, and a Hamamatsu R928P thermoelectrically-cooled photomultiplier tube detector.
Typical emission spectra resulted from 3 scans of 0.2 s dwell time each, utilising moderate excitation (Δλ = 1 − 4 nm) and fine emission (Δλ < 0.5 nm) bandwidths, optimising peak signal to ∼700,00 counts per second (cps) with sample optical density of less than 0.1 au at the excitation wavelength.Quantum yields for these QDs have been previously measured [14,19,20].

Lifetime measurements
The lifetime fluorescence of QDs was determined through time-correlated single photon counting for each sample using a 380 nm ps pulsed laser and an Edinburgh Instruments FLS980 spectrophotometer.Signals were collected from 1 or 2 μs pulse rates divided across 8192-time bins until one bin registered 1000 counts.Calculations of lifetime fluorescence were undertaken using the exponential decay model.All the decays show multiexponential decay kinetics and fit well with a biexponential decay function.The tail was fit to a biexponential decay function following the equation: Where τ is the lifetime, and α is the pre-exponential factor.Instrument response functions (IRF) were recorded for each sample using a scattering solution at the same pulse rates and ∼1,000 cps laser intensity.The IRF was subtracted from each sample, and the tail was fit to several decay functions.FRET efficiencies were calculated using: Where τ D is the lifetime fluorescence of the donor, τ DA is the lifetime of the donor in the presence of the acceptor.

UV-vis absorption spectroscopy
The absorbance was measured using a Varian Cary Bio-50 spectrometer equipped with a Xenon flash lamp.Samples were scanned using a 1 cm gap two-sided quartz cuvette from 300-700 nm with a 0.1 s dwell time and background-subtracted with blank depending on the sample.

Circular dichroism spectroscopy
Aptamer samples with and without enzymes and complementary strands were analysed with a circular dichroism (CD) instrument.1.5 μM solutions of each oligonucleotide were prepared in water and in folding buffer (1 mM MgCl 2 in PBS (137 mM NaCl, 2.7 mM KCl, 8 mM H 3 Na 2 O 5 P, 2 mM KH 2 PO 4 )).Oligonucleotides were heated to 90 °C for 5 min and cooled to room temperature before analyses.The enzyme concentration was 40 mM, and CD studies were performed on a Chirascan Spectrometer in a 1 mm path-length quartz cuvette.The spectra were recorded from 380 to 180 nm at 20 °C at a rate of 200 nm/min and corrected by subtraction of the background scan of the buffer.Samples containing the enzyme were baselined to a buffer containing the buffer and the same enzyme concentration as the analysed sample.

Secondary structure and hybridisation mechanism
The aptamer used in this study was originally identified by Kim et al [21] through the SELEX method to target metallo-β-lactamases isolated from Bacillus cereus.Initially, a 30-residue aptamer (Lac30) was isolated.Subsequent analysis revealed a 10-residue sequence forming a crucial loop within Lac30.Interestingly, the inhibition results with the 10-residue aptamer (Lac10) mirrored those of the longer sequence.
Lac10 was not chosen for this work due to its sequence encompassing the enzyme-interacting region, potentially interfering with nanoparticle conjugation and specificity.Thus, starting with the extended version allows for adequate space for conjugations.Subsequently, shorter sequences can be explored, offering advantages such as enhanced FRET efficiency and biosensor sensitivity through reduced inter-nanoparticle distances.To validate previous findings, we utilised UNAfold to predict secondary structures of selected aptamers (see figure S1).Lac30, under conditions of 10 mM NaCl and 0.2 mM MgCl 2 , exhibited two low-energy structures mimicking hairpin loops.This common loop, isolated as Lac10, is presumed to be the interactive region responsible for metallo-β-lactamase inhibition.
Attention must be paid when selecting the complementary strand.The binding at the 3' end could leave the proposed interactive site accessible, necessitating complementary strand displacement for enzyme interaction.Thus, it's crucial that the complementary strand binds to loop-forming bases, ensuring aptamer detachment before enzyme targeting.This process is vital for separating the donor and acceptor, where the aptamer with higher enzyme affinity unbinds from its complementary strand, increasing the QD-acceptor nanoparticle distance, thereby activating fluorescence and indicating enzyme presence.A previous study showed that the aptamer tends to hybridize more strongly with complementary DNA than a target if the length of complementary DNA is the same as that of an aptamer [22].Thus, the preferred complementary strand should contain less than 30 bases and link to the main aptamer through 6 bases (less than the number of bases binding the loop).
CD spectroscopy was done to investigate the secondary structure of the selected aptamer or complimentary strands (figure 2(A)).Distinctive peaks associated with DNA polymers were observed at (1) 220 nm, indicating H-bonding; (2) 246 nm, corresponding to handedness or helicity; and (3) 275 nm, signifying base-pair stacking [23,24].The CD spectra revealed a peak at 220 nm, indicative of H-bonding, suggesting the presence of a hairpin loopa structure formed when bases on the same strand are linked through hydrogen bonds.At 246 nm, a negative peak in the CD spectra suggested the formation of a helix structure, providing further evidence of the hairpin loop structure.The 275 nm peak was indicative of base pair stacking, and considering Lac30ʼs inclusion of G bases known for their stacking properties, this peak confirmed the presence of stacking [25].Additionally, the CD spectra of Lac30 in the presence of the metallo-β-lactamases exhibited an increased magnitude at 220 nm, affirming the interactive nature of the hairpin loop in the aptamer (figure 2).A CD spectra of Lac30 and the complementary strand can be shown in figure S2.
Oligonucleotides form complexes with their targets and can also hybridise with complementary strands [26,27].The interaction between an aptamer and its target is significantly influenced by the sequence of the complementary strand [28][29][30].Ideally, the complementary strand should partially bind to the aptamer in the absence of a target, without competing with the target for binding.This requires the aptamer's interactive site to be inaccessible, promoting the displacement of the complementary strand.The number of bases linking the aptamer and its complement is crucial, with the aptamer having a higher affinity for the enzyme than for its complement.
When selecting the complementary oligonucleotide, the length of the complementary oligonucleotide plays a crucial role [22].For instance, a previous study showed stronger hybridisation with complementary oligonucleotides when the lengths are equal.Therefore, the preferred complementary strand should have fewer than 30 bases and link to the main aptamer through 6 bases.For Lac30, a complementary oligonucleotide sequence, named Comp1, was selected based on high helicity and minimal influence on the aptamer's conformational changes [31].CD analysis of Lac30 and Comp1 confirmed their interaction and hybridisation, with a decrease in helicity upon hybridisation, indicating interference with the aptamer's hairpin loop (figure 2(A) and 3(A)).
To confirm the hybridisation mechanism between the aptamer and its complementary strand, 1H NMR studies were conducted at high aptamer concentrations (2 mM) in a folding buffer (figure 3(B)).Lac30 and Comp1 were individually analysed, and their spectra were compared to those of the mixture.The 9-15 ppm peaks in the spectra correspond to imino protons in thymine and guanine bases [32].Watson-Crick imino protons appear in the 1215ppm region, while non-Watson-Crick imino protons appear in the 912 ppm region due to proton exchange with solvent [33].Lac30, with three Watson-Crick base pairs, exhibited peaks at 13.4 ppm, consistent with guanine imino protons involved in Watson-Crick hydrogen bonding [34,35].The observed increase in imino protons in the oligonucleotide mixture compared to individual spectra suggests hydrogen bonding through base pairing, confirming hybridisation.These findings align with CD spectra results, indicating that Lac30 is indeed hybridising with Comp1.

QD-AuNP as a FRET Pair
InP/ZnS QDs were synthesised following established protocols with oleylamine ligands, rendering them soluble in organic solvents [14,15].This method enables the synthesis of QDs with varying sizes by manipulating the precursors, thereby allowing for tunable photoluminescence.Specifically, InI and ZnI precursors were utilised to produce QDs emitting at 500nm (QDs500), InBr and ZnBr for QDs emitting at 580 nm (QDs580), and InCl and ZnCl for QDs emitting at 650 nm (QDs650).The normalised emission spectra of the synthesised QDs are presented in figure 3(A).
For this study, a ligand exchange was performed to impart water solubility to the QDs, facilitating their conjugation to amino-primer aptamers.Detailed procedures for this ligand exchange were previously reported by Ayed et al [36].Notably, ligand exchange processes can potentially etch the QD surface and, in some cases, reduce the quantum yield.To ensure the continued functionality of the QDs as fluorescent labels, emission spectra were recorded after each step of QD functionalisation.The emission spectra of green-emitting QDs before and after ligand exchange are displayed in figure 3(A).The emission of the as-prepared QDs in toluene (QDs-OA) is λ 530 nm, which shifts to higher energy after the ligand exchange with mercaptosuccinic acid (QDs-MSA) to λ 500 nm.A slight emission shift after the aptamers are conjugated to the surface occurs under 500 nm.
AuNPs with a diameter of approximately 8 nm were synthesised and subsequently conjugated to Comp1, as illustrated in the TEM image and histogram provided in figure S3.The absorbance profile of the AuNPs after conjugation with Comp1 is depicted in figure 3(B).This also illustrates the overlap of the absorbance spectra of AuNPs and the QDs.An overlap of the emission spectrum of the donor and the absorption spectrum of the acceptor ensures that the energy released from the donors transition from excited to ground state could excite the acceptor group.Thus, the more overlap of spectra, the better a donor can transfer energy to the acceptor.The overlap integral J(λ) was calculated using equation (3) and values presented in table 2.
Where, F D (λ) is the normalised emission spectrum of the donor, ò A standards for the molar absorption coefficient of the acceptor, and λ is the wavelength.The Förster radius (R 0 ), representing the distance at which FRET is 50 % efficient, was calculated for the FRET pair QDs580-AuNP (see equation ( 4)) [37].Long Förster Radius R 0 : can cause high FRET efficiency.Based on Försters analysis, R 0 is a function of quantum yield of the donor chromophore θ D , the spectral overlap of donor-acceptor J(λ) the directional relationship of transition dipoles κ 2 , and the refractive index of the medium n.The calculations were modelled in water (refractive index η = 1.333), and the orientation factor k 2 was assumed as 2 3 , commonly used in similar studies [38].The TEM image and size histogram of AuNPs are provided in figure S3.The R 0 value for this FRET pair was determined as 11nm.Aptamers are typically 3-7 nm long, and the addition of an amino primer attached via a 6-carbon chain is estimated to be less than 2 nm [39].Thus, the total separation distance between donor-acceptor is estimated to A fixed amount (0.02 mg/mL) of Comp1-conjugated QDs was incubated with various concentrations of Lac30-functionalised AuNPs (QDs = 0 M, QDs-AuNPs.4= 4.17 × 10 −10 M, QDs-AuNPs.8= 8.34 × 10 −10 M, QDs-AuNPs.12= 12.51 × 10 −10 M).Following incubation, fluorescence intensities were recorded under the same conditions.In figure 2(C), higher AuNP concentrations reduce the QD donor emission peak, confirming decreased QD fluorescence.Lifetime fluorescence measurements were conducted to distinguish between radiative and non-radiative processes, such as FRET, which results in faster decay.Comparing donor lifetime fluorescence in the absence and presence of the acceptor helps differentiate these mechanisms (figure 3(D).
To address potential photon re-absorption, a control experiment was conducted using citrate-capped AuNPs (without aptamers) incubated with the same QDs580 concentration (0.02 mg/mL).All the controls are shown in figure S4, such as citrate-capped AuNPs with no aptamer conjugation.Figure 3(C) demonstrates a significant decrease in fluorescence intensity between control and FRET samples, confirming QDs580 fluorescence quenching due to energy transfer.However, definitive confirmation of the observed mechanism requires lifetime measurements.
FRET efficiency depends on factors such as the distance between the FRET pair, spectral overlap of the donor's emission and acceptor's absorption profiles (J(λ)), and the luminescence lifetime of the donor (τ D ).Despite a calculated R 0 of 11 nm for the QDs-AuNP pair, the spectral overlap (J = 2.0647 × 10 17 M −1 cm −1 nm 4 ) can be enhanced.QDs, with tunable emissions based on particle size, were synthesised for this purpose.QDs500 (λ em = 500 nm) were chosen for improved spectral overlap with the AuNPs' SPR band (figure 3(B)), yielding J = 2.425 362 × 10 18 M −1 cm −1 nm 4 , surpassing the previously used QDs580.Calculations for the R 0 in the new FRET pair QDs500-AuNPs resulted in 16.27 nm.
Pulsed laser lifetime measurements were performed on QDs500 with and without AuNPs to validate the quenching mechanism, confirming FRET in this pair (figure 3(D)).A two-exponential decay tail fitting exhibited superior fit quality over a single exponential fit, assessed through chi-squared (χ2) and residual plots.Results indicate a FRET efficiency (EFRET) of 0.31, lower than the efficiency at R 0 (EFRET = 0.5), suggesting a greater FRET pair distance than initially estimated.
Several factors may contribute to this discrepancy: (1) Initial distance estimation considered only the oligonucleotides and carbon chains, excluding the QD shell and ligands, which should be included in FRET distance calculations.(2) Assumption of a dipole angular orientation (κ 2 ) equal to 2/3, based on prior studies, may not accurately apply to QDs tethered through an oligonucleotide.However, determining the precise κ 2 value is challenging without detailed knowledge of the oligonucleotides structure and nanoparticle positions.Previous CD studies indicated a B-form structure for the aptamer, where FRET efficiency correlates with the number of DNA bases (n) between donor and acceptor.For this work, the number of bases equals the length of the complementary strand (18 bases), resulting in the observed low FRET efficiency for this pair.
Despite this, the FRET pair distance can be reduced by altering the orientation of the oligonucleotide strands.Initial FRET results utilised a complementary strand with an amino primer at the 5 end (QDsAApt in table 1).To minimise the distance between nanoparticles, an amino primer was added to the aptamers 3 end (QDsAptA in table 1).This adjustment increased FRET efficiency from 31% to 63% for the same AuNP concentration (C = 13.8 × 10 −10 ), as shown in figure 4(B) and summarised in table S1.
In this study, we aimed to quench QD emission with AuNPs, creating a visible 'turn-on sensor' response.Figure 3(C) illustrates reduced QD fluorescence intensity, but finite FRET efficiency permits observable QD donor emission even at high acceptor concentrations (25.3 × 10 −10 M).Lifetime measurements remain the sole confirmation of the FRET mechanism.However, inconsistent emission intensities and persistent unquenched QD signals make this FRET pair unsuitable for precise biosensing, impacting sensor sensitivity for low enzyme concentrations.
The calculated R 0 for QDs500-AuNP appears unusually high compared to common FRET pairs (typically 1-10 nm), ranging from 11 nm to 20 nm depending on AuNP size (table 2).This discrepancy suggests a lessexplored mechanism, nanoparticle surface energy transfer (NSET) [40].NSET, a nonradiative dipoledipole energy transfer involving metallic nanoparticles (e.g., AuNPs), is less studied than FRET.Similar to FRET, R 0 in NSET is tunable based on nanoparticle size, shape, and spectral overlap with the QD.Notably, NSET efficiency inversely depends on the fourth power of the distance between the donor and acceptor surface, enabling longerdistance energy transfer.
Studies indicate that fluorophores attached to plasmonic nanoparticles exhibit partial fluorescence retention in NSET [40,41].If the observed mechanism is NSET rather than FRET, total quenching of donor fluorescence is not achievable, precluding the development of an aptasensor.Exploring alternative FRET pairs, such as QD-QD, may lead to changes in emission peak ratios, enabling the design of a ratiometric sensor.Ratiometric sensing offers self-calibration and increased resilience to fluctuations in the sensor environment.The non-single exponential behaviour observed in the photoluminescence decay for QD500-Aptamer without the presence of AuNPs suggests the involvement of multiple recombination pathways.This phenomenon can be attributed to various factors such as surface ligands, which in this case are long DNA strands that are bulky.We hypothesis that the aptamer attachment to the QDs will have some varying distribution in size and the number of ligands attached [42].The presence of surface ligands or capping agents on the QDs can influence their electronic structure and charge carrier dynamics, contributing to non-single exponential decay behaviour.

QD-QD as a FRET pair
A ratiometric sensor, utilising a colour change rather than emission intensity modulation, is exemplified by Medintz et al [43].Through FRET monitoring of proteolytic enzyme activity with QD-peptide conjugates.In their approach, a QD is functionalised with a peptide linker terminating in an energy-accepting fluorescent dye.Cleavage of the linker by the analyte alters the chromophores' intensity ratio, causing the dye to diffuse from the QD, resulting in enhanced QD emission.In our adaptation, a higher-wavelength-emitting QD is employed instead of a dye.
Green (λem = 500 nm) and red (λem = 650 nm) emitting donor and acceptor QDs, with minimal emission peak overlap, were synthesised for visual differentiation.The spectral overlap of the QDs500 emission spectra and QDs650 absorbance spectra, shown in 5(A), aligns with the exciton peak absorption, supporting the suitability of this QD pair for ratiometric sensing.
In QDs composed of the same material, the absorption cross-section of the larger, lower energy-emitting acceptor QD (QDs650) surpasses that of the smaller, higher energy-emitting donor QD (QDs500).Consequently, direct excitation of the acceptor can dominate the system, complicating the interpretation of photoluminescence (PL) intensity measurements.However, under conditions of moderate overall assay concentration and well-dispersed particles, the intensity of the donor peak should exhibit relatively minor changes upon addition of the acceptor.Thus, significant alterations in the donor peak intensity can be attributed to Förster Resonance Energy Transfer (FRET).
As illustrated in figure 5(B), escalating concentrations of acceptor QDs650 lead to a decline in the intensity of donor peak emission, accompanied by an increase in acceptor emission intensity, providing clear evidence of FRET between the two species.The concentration-dependent quenching of the donor by the acceptor further corroborates the occurrence of FRET.This is demonstrated in figure 5(C), where the shorter lifetime fluorescence of the donor in the presence of acceptor QDs650 underscores the FRET phenomenon.
Evaluation of the lifetime fluorescence of the donor in both the presence and absence of the acceptor reaffirms the FRET outcomes.This is depicted in figure 5(D), where the accelerated decay of QDs500 in the presence of acceptor QDs650 provides additional confirmation of the FRET process.This is similar to what was shown by Zhang et al. who showed a decrease in average lifetimes when the QD was linked [7].The non-single exponential behaviour observed in the photoluminescence decay for QD500-Aptamer without the presence of AuNPs suggests the involvement of multiple recombination pathways.This phenomenon can be attributed to various factors such as surface ligands, which in this case are long DNA strands that are bulky.We hypothesise that the aptamer attachment to the QDs will have some varying distribution in size and the number of ligands attached.The presence of surface ligands or capping agents on the QDs can influence their electronic structure and charge carrier dynamics, contributing to non-single exponential decay behaviour [42].
As shown in table 3, the highest FRET efficiency of this pair is 0.65 for the amino at the 5 end of the aptamer.This FRET efficiency is higher than that of the QD-AuNP pair of the same experiment (0.31 from table 3).Therefore, the QD-QD pair is proving to be a better pair than the QD-AuNP.Preliminary FRET efficiency results show a higher efficiency of the QD-QD FRET pair than the QD-AuNP, as expected (figure 4).If completed, optimising the FRET conditions of the QD-QD pair could lead to a visual sensor for antibiotic resistance.
The presented study serves as a proof of concept for the development of an aptasensor targeting metallolactamase through FRET.While our results demonstrate promising sensitivity and specificity in controlled laboratory conditions, it is important to acknowledge that real-world applications may involve complex biological samples containing various interferents that could potentially affect sensor performance.Therefore, further comprehensive studies are warranted to evaluate the aptasensor's robustness and reliability in practical settings.Future investigations will focus on systematically assessing the influence of interferents commonly encountered in clinical and environmental samples, including biological fluids, media components, and other molecules present in complex matrices.By addressing these critical aspects, we aim to enhance the sensor's applicability and utility for accurate and reliable detection of metallo-lactamase in diverse real-world scenarios [44].

Conclusion
The structural analysis of the Lac30 aptamer facilitated the selection of an appropriate complementary strand, paving the way for the design of a FRET aptasensor with QDs500-AuNPs.However, comprehensive studies  revealed the inadequacy of such a system for detection-based sensors due to the persistent small signal from the unquenched QDs.Furthermore, Förster radius calculations indicated a potential shift towards a NSET mechanism rather than FRET, leading to the unavoidable partial preservation of fluorescence.
Addressing the challenges posed by the QD-AuNP FRET pair, alternative configurations were investigated, culminating in the selection of QD-QD FRET as a more efficient substitute for a ratiometric sensor.Initial investigations revealed significantly enhanced performance compared to the QD-AuNP pair under identical conditions, showcasing a notably higher FRET efficiency (0.6).These promising results lay the foundation for the development of machine-free, colour-changing sensors, poised to enable rapid and accurate detection of enzymes and, consequently, bacterial resistance to antibiotics.
The correlation between changes in emission ratios and alterations in solution colouration opens avenues for the creation of visual, colour-change aptasensors.Additionally, the exploration of shorter aptamer sequences presents supplementary benefits for the biosensor.This includes a diminished distance between the donor and acceptor nanoparticles, thereby augmenting FRET efficiency and bolstering biosensor sensitivity.

Figure 1 .
Figure 1.Schematic representation of the FRET process between a donor and acceptor where the switch is an emission shift from red to green in the presence of metallo-β-lactamases (target).

Figure 4 .
Figure 4. Schematic representation of donor and acceptor positions using amino 5 end (top) and amino 3 end (bottom).

Figure 5 .
Figure 5. (A) Spectral overlap of the absorbance of QDs500 (blue) and emission of QDs650 (black).(B) PL spectra of FRET samples.QD500 (0.02 mg/mL) donor emission decreases with increasing QD650 acceptor concentrations and the absorbance of the AuNPs.(C) Normalised exponential decay of donor QD500 in the absence and presence of QD650 acceptor.(D) Normalised exponential decay of the QD-QD pair.

Table 1 .
Aptamer and Complementary Strands Names and Sequences.

Table 2 .
FRET calculations of energy transfer from QDs500 to different size AuNP.Where R 0 is the FRET distance, and J(λ) is the spectral overlap of donor and acceptor.Calculations were done in MATLAB.

Table 3 .
Lifetime fluorescence and FRET efficiency of different concentration acceptor QDs650