Actively Controllable Solid Phase Microextraction in a Hierarchically Organized Block Copolymer-Nanopore Electrode Array Sensor for Charge-Selective Detection of Bacterial Metabolites

Abstract

Pseudomonas aeruginosa produces a number of phenazine metabolites, including pyocyanin (PYO), phenazine-1-carboxamide (PCN), and phenazine-1-carboxylic acid (PCA). Among these, PYO has been most widely studied as a biomarker of P. aeruginosa infection. However, despite its broad-spectrum antibiotic properties and its role as a precursor in the biosynthetic route leading to other secondary phenazines, PCA has attracted less attention, partially due to its relatively low concentration and interference from other highly abundant phenazines. This challenge is addressed here by constructing a hierarchically organized nanostructure consisting of a pH-responsive block copolymer (BCP) membrane with nanopore electrode arrays (NEAs) filled with gold nanoparticles (AuNPs) to separate and detect PCA in bacterial environments. The BCP@NEA strategy is designed such that adjusting the pH of the bacterial medium to 4.5, which is above the pKa of PCA but below the pKa of PYO and PCN, ensures that PCA is negatively charged and can be selectively transported across the BCP membrane. At pH 4.5, only PCA is transported into the AuNPs-filled NEAs, while PYO and PCN are blocked. Structural characterization illustrates the rigorous spatial segregation of the AuNPs in the NEA nanopore volume, allowing PCA secreted from P. aeruginosa to be quantitatively determined as a function of incubation time using square-wave voltammetry and surface-enhanced Raman spectroscopy. The strategy proposed in this study can be extended by changing the nature of the hydrophilic block and subsequently applied to detect other redox-active metabolites at low concentration in complex biological samples and, thus, help understand metabolism in microbial communities.

Graphical Abstract

Cross-sectional SEM image and schematic illustration of the electrochemical SERS sensor with BCP gate closed at pH 7.0 (middle) and BCP gate open at pH 4.5 (right) for the permselective transport and detection of negatively charged molecules by square wave voltammetry and SERS.

The bacterium Pseudomonas aeruginosa is an opportunistic pathogen that accounted for approximately 32,600 healthcare-associated infections and contributed to 2,700 estimated deaths in the United States in 2017.¹ P. aeruginosa is a metabolically diverse microbe that grows readily in many natural and host-associated environments. Phenazines are one variety of the many types of secondary metabolites produced by P. aeruginosa under different growth conditions. These phenazines are redox-active pigments that are toxic to many other microbes and enable P. aeruginosa to survive in otherwise hostile environments due to the reactive oxygen species (ROS) generation.² Furthermore, it has been proposed that phenazines play a role in altering antibiotic susceptibility.³ Lastly, because of their association with virulence, the phenazines are attractive as biomarkers that might be suitable to identify P. aeruginosa infection.⁴

Several P. aeruginosa phenazines have been identified, including phenazine-1-carboxylic acid (PCA) which can be derivatized to phenazine-1-carboxamide (PCN), 1-hydroxyphenazine (1-OHP), 5-methylphenazine-1-carboxylic acid (5-MCA), and pyocyanin (PYO).^3–5 PYO is the best-known and most widely studied blue redox-active secondary metabolite within the phenazine family secreted by P. aeruginosa.^{6, 7} Studies have shown that PYO interferes with cellular respiration, gene expression, electron transport, and energy metabolism in host cells.⁸ In addition, the PYO precursor PCA has shown antifungal and antibacterial activities, which are thought to arise from its oxidative activity and inactivation of important proteins in the oxidative stress response.⁹ The antagonistic effects and redox transformations of phenazines greatly affect other microbes, and a more complete knowledge about PCA is necessary to understand the behavior of multi-organism assemblies involving P. aeruginosa.

The redox-active and Raman-active properties of PCA make electrochemical and Raman characterization ideal for its study.^{3, 5, 10, 11} However, directly detecting PCA in bacterial cultures is challenging, since there are other abundant metabolites, such as PCN and PYO, which interfere with detection of the less-abundant PCA. One straightforward approach would involve isolation of PCA directly from bacterial cultures. However, conventional separation methods to isolate and purify PCA require time-consuming sample pretreatment. High-performance liquid chromatography, free flow electrophoresis, mass spectrometry, and nuclear magnetic resonance spectroscopy have all been employed to detect phenazines due to their high selectivity, but they either require large sample volumes and sample pretreatment, or the instrumentation is expensive.^12–15 Therefore, it would be very useful to develop a low-cost, low-volume, and efficient isolation method to separate PCA from bacterial cultures containing P. aeruginosa.

Recently, heterogeneous block copolymer (BCP) membranes have been applied for the separation of species ranging from small ions to proteins, owing to their tunable nanoscale pore size and charge and the ability to produce highly ordered and diverse nanoarchitectures.^16–18 In particular, poly(styrene)-b-poly(4-vinyl pyridine), i.e., PS-b-P4VP, has attracted considerable attention, because it exhibits pH-responsive and charge-selective dual gating characteristics, and it can be prepared in a variety of structural motifs.¹⁹ For example, Peinemman and co-workers demonstrated size-selective separation of proteins as well as selective separation of similarly sized proteins by exploiting different isoelectric point values of the proteins.¹⁶ More recently, our laboratory demonstrated that PS-b-P4VP BCP membranes can be combined with nanopore electrode arrays (NEAs) to achieve pH-gated, charge-selective electrochemical signal amplification of ions transported into NEAs through the hierarchically organized BCP@NEA membrane.²⁰

Here, we extend this strategy to selectively collect low concentration, negatively charged PCA (pKa = 4.2) from positively charged PYO (pKa = 4.9) and neutral PCN by adjusting the pH of the bacterial culture samples to 4.5.^{21, 22} In order to enable both electrochemical and SERS dual mode detection, the NEAs are filled with AuNPs to generate a SERS-active volume within the solid phase microextraction architecture. The introduction of a BCP membrane introduces pH-responsive and charge-selective properties capable of effecting the efficient separation of PCA from other phenazine metabolites produced by P. aeruginosa, as shown in Figure 1. The negatively charged PCA is transported across the anion-permselective PS-b-P4VP into the NEA nanopores where it can be detected using square-wave voltammetry (SWV) and surface-enhanced Raman scattering (SERS). This actively controllable, hierarchically-organized solid phase microextraction architecture was then used to assay PCA produced by P. aeruginosa grown in planktonic culture. Interestingly, only a negligible amount of PCA was detected at 8 h incubation, but rapid production was observed from 16 h to 24 h. The integration of pH-gated, charge-selective solid-phase microextraction to isolate PCA with electrochemical and SERS detection in a single monolithic nanostructure produces a biosensor which can detect low-abundance metabolites, and the design flexibility makes it possible to extend the range of application by changing the nature of the hydrophilic block, rendering it applicable to the detection of other, e.g., cationic, redox-active metabolites in complex biological samples. The nanoscale integrated, actively controllable structures developed here have the potential to discover new metabolites and pathways involving low-abundance secreted species.

Figure 1.

Architecture and scheme of the hierarchically-organized BCP@NEA electrochemical SERS sensor. (A) Phenazine biosynthetic pathway in P. aeruginosa. PCA, phenazine-1-carboxylic acid; 1-OHP, 1-hydroxyphenazine; 5-MCA, 5-methylphenazine-1-carboxylic acid; PCN, phenazine-1-carboxamide; PYO, pyocyanin. Phenazine-modifying genes for the conversion reactions are labelled along with arrows.^{5, 29–31} (B) Schematic illustration of the electrochemical SERS sensor with BCP gate closed at pH 7.0. (C) Schematic illustration of the electrochemical SERS sensor with BCP gate open at pH 4.5 for the permselective transport and detection of negatively charged molecules by square wave voltammetry and SERS.

EXPERIMENTAL SECTION

Materials.

Poly(styrene)-b-poly(4-vinyl pyridine) (P18248-S4VP, 48.4-b-21.3, M_w/M_n = 1.09) was purchased from Polymer Source. Phenazine-1-carboxylic acid was obtained from SynQuest Laboratories. Phenazine-1-carboxamide, pyocyanin, Rhodamine 6G, potassium ferricyanide (III), hexaammineruthenium (III) chloride, potassium chloride, ethyl alcohol (200 proof), gold nanoparticles (AuNPs) (150 nm), chromium etchant and latex polystyrene beads (1.1 μm) were purchased from Sigma Aldrich. Sylgard 184 silicone Elastomer Kit was obtained from Dow Corning Corporation. Thermal release tape (REVALPHA) was bought from Nitto Denko Corporation. Dulbecco’s Phosphate Buffered Saline (DPBS, 1x, without calcium and magnesium) was purchased from Lonza, and silicon wafers were purchased from University Wafer. Deionized (ρ ~ 18.2 MΩ cm, DI) water was prepared using a Milli-Q Gradient water purification system (Millipore). All reagents were used as received without further purification.

Fabrication of the Sensor.

A schematic diagram of the fabrication process is given in Figure S1. A gold (Au) layer (400 nm) was deposited in an electron-beam vacuum evaporator (Oerlikon 450B evaporator) using 10 nm titanium (Ti) as an adhesion layer. An additional glass (SiO₂) layer (1 μm) was then deposited by the plasma-enhanced chemical vapor deposition system (Unaxis 790 series) using Ti (10 nm) as an adhesion layer. A close-packed monolayer of latex polystyrene (PS) beads (1.1 μm) was transferred onto the SiO₂ surface at an air-water interface by emersing the edge of the wafer through the PS bead layer. Reactive ion etching (RIE) (3 min) in O₂ gas was processed to reduce the bead size to around 950 nm by using an inductive coupled plasma (ICP) reactor (Oerlikon ICP-RIE system). A chromium (Cr) layer (350 nm) was coated as a protective layer on and between the beads. Then, the beads were physically removed in acetone solution by using laundered polyurethane foam swab lab-tips (Berkshire Corp). Finally, RIE (26 min) in O₂ and CF₄ gas was used to etch the SiO₂ between the protective Cr layers to fabricate SiO₂ nanopores. The device was then immersed into Cr etchant (120 s) to remove the remaining Cr layer.

AuNP-containing solution was centrifuged at 10000 rpm for 10 min (RS-200 microcentrifuge, REVSCI). The solvent was decanted and the AuNPs were dispersed in DI water and sonicated for 5 min. This procedure was repeated twice, and the AuNPs were finally resuspended in DI water to a concentration of 1.44 × 10¹⁰ particles mL⁻¹. The AuNPs solution was then drop cast onto the SiO₂ nanopore wafer, air dried in a fume hood, and stored for future use.

A 2% (w:v) poly(styrene)-b-poly(4-vinyl pyridine) solution (in dioxane) was drop cast on a pre-cleaned silicon wafer and spin coated at 500 rpm for 1 min. Then, the membrane was immersed in ethanol and incubated for 30 min. After drying with N₂ gas, the membrane was transferred from silicon wafer onto the AuNP-containing SiO₂ nanopore wafer using thermal release tape. The tape was separated and easily peeled off from the wafer after the device was heated on a hot plate (120 °C) for 10 s leaving the membrane attached to the AuNPs-contained SiO₂ nanopore wafer. SEM images were acquired to confirm the success of each step.

Characterization.

Scanning electron microscopy (SEM) analysis was performed using a Helios G4 Ux dual beam SEM/FIB workstation operating at an acceleration voltage of 5 kV for the electron beam and 25 kV for the ion beam. Raman spectroscopy experiments were conducted by using an Alpha 300R confocal Raman microscope (WITec Instruments Corporation). Excitation radiation at 785 nm from a solid-state diode pumped laser was focused and collected through a 40× (NA = 0.6) objective. WITec project 2.1 and MATLAB (Mathworks Inc.) were used to perform initial data analysis. All Raman experiments were repeated on three replicates per sample condition.

A CHI842C or CHI760A electrochemical workstation (CH Instruments, USA) was used for all electrochemical measurements. Typically, the Au layer was used as working electrode in 3-electrode configuration with Pt counter and Ag/AgCl reference electrode. Prior to electrochemical measurements, a poly(dimethylsiloxane) (PDMS) well was mounted onto the nanopore array to contain 100 μL of an electrolyte solution or bacteria growth medium. SWV experiments were conducted using a potential amplitude of 25 mV and a potential increment of 4 mV at a frequency of 15 Hz and cyclic voltammograms were obtained at a scan rate of 0.1 V s⁻¹. For the Raman and electrochemical measurements on planktonic culture, the bacteria-free supernatant solution was adjusted to pH 4.5 using hydrochloric acid (1%). The supernatant (pH 4.5) was then transferred onto the device and incubated for 30 min to allow membrane equilibration with the pH 4.5 solution before testing.

Analysis of the SERS Performance of AuNP-filled SiO₂ Nanopores.

Rhodamine 6G (10 mM) water solution was drop cast onto a gold-coated silicon wafer and also onto a AuNP-filled BCP@NEA wafer. SERS spectra were recorded under excitation with 785 nm laser as shown in Figure S2. Raman enhancement factor (EF) of the AuNPs-containing BCP@NEA structure was calculated from the following equation:

$$\text{EF}=\left({\text{I}}_{\text{SERS}}{\text{/N}}_{\text{SERS}}\right)/\left({\text{I}}_{\text{Raman}}{\text{/N}}_{\text{Bulk}}\right)$$

where N_SERS and N_Bulk represent the number of molecules on the substrate surface and in the bulk solution in the probe volume, respectively, and I_SERS and I_Raman represent the intensities in SERS and Raman, respectively. N_SERS was estimated from the molecular density of a compact monolayer of Rhodamine 6G on half of a monolayer of AuNPs, to account for the fact that AuNPs are thick enough to block incident laser radiation from the backward facing hemisphere, meaning that only half of the AuNP surface was effectively irradiated. I_SERS and I_Raman were determined from the peak intensities of the band of Rhodamine 6G at ~1595 cm⁻¹ and yielded an EF of 1.3 × 10⁵.

Supernatant of Planktonic Culture.

P. aeruginosa strain PA14 and its isogenic Δphz (phenazinenull) strain were used in this study.²³ Bacteria were streaked on Luria-Bertani (LB) agar plates and incubated at 37 °C for 16 h. A single colony was inoculated in 10 mL LB medium grown at 37 °C with shaking at 240 rpm. The optical density at 600 nm, OD₆₀₀, of the P. aeruginosa culture was measured to monitor the bacterial growth as shown in Figure S3. The bacterial culture was collected and centrifuged at 15000 rpm for 5 min. The supernatant was filtered (0.2 μm filter pore size) and this bacterial-free supernatant was collected for future use.

Detection of PCA in planktonic culture.

The bacteria-free supernatant (0.5 mL) was mixed with chloroform (1 mL), vortexed for 30 s and incubated for another 20 min. After allowing chloroform to settle to the bottom, this phase became green-blue. The aqueous phase was removed. The phenazines were recovered by air-drying the solvent in a fume hood, after which the analytes were redispersed in DI water (pH 4.5). The solution (pH 4.5) was drop cast on the device and incubated for 30 min before testing. Because it was difficult to completely remove all PCA from the device after a measurement cycle, all devices were used for single measurements only.

RESULTS AND DISCUSSION

Functional overview of electrochemical SERS sensor design.

The fabrication of the electrochemical SERS sensor is described in the Experimental Section and is shown in Figure S1. The resulting AuNP-filled BCP@NEA structures were characterized with SEM imaging. As shown in Figure 2, 150 nm AuNPs occupied the majority of the nanopore volume after filling, and the physical integrity of the BCP membrane is clearly evident in Figures 2A and 2B (top- and tilted-view, respectively). The AuNP filling of a typical nanopore is shown in a cross-sectional image in Figure 2C. At pH 7.0, above the pKa of the P4VP block (pKa = 4.8), the BCP membrane is hydrophobic and dewetted, and thus no mass transport, including water, occurs, as depicted schematically in Figure 1B. However, at pH 4.5, below the P4VP pKa, the P4VP domains of the BCP membrane become protonated, rendering them positively charged, hydrophilic, and anion permselective, permitting negatively charged molecules to pass, as shown in Figure 1C.²⁰ These properties of the BCP portion of the BCP@NEA make it possible to separate molecules based on charge and thus to be used for actively-controllable solid phase micro-extraction. This is particularly useful in the isolation and detection of PCA, for example. The phenazine biosynthetic pathway in Figure 1A shows PCA and four other important metabolites secreted by P. aeruginosa. Among them, PYO is the dominant secreted factor in planktonic culture, which makes the other phenazines difficult to detect. However, because the phenazines exhibit pH-dependent charge states, PCA can be isolated by careful manipulation of the charge state of the P4VP domain of the BCP membrane.

Figure 2.

Scanning electron microscopy (SEM) images of the hierarchically-organized BCP@NEA electrochemical SERS sensor. Top-view (A) and tilted-view (B) SEM images of the electrochemical SERS sensor showing membrane-covered (left side of both A and B) and uncovered (right side of both A and B) NEA with AuNPs. (C) Cross-sectional SEM image of a single Au nanoparticle-filled nanopore from the sensor.

Electrochemical characterization of AuNP-filled NEAs.

Before attaching the BCP membrane onto the AuNP-filled NEAs, the electrochemical function of the NEAs was characterized by comparing the cyclic voltammetry (CV) of the reversible redox couple, Fe(CN)₆^3/4− at 1 mM on AuNP-filled NEAs and a planar gold electrode, as shown in Figure S4. The AuNP-filled NEAs produce slightly greater cathodic and anodic peak currents, an observation likely due to the larger active electrode area of AuNP-filled NEAs. Importantly, filling AuNPs into NEA nanopores does not alter the qualitative characteristics of the electrochemical response of Fe(CN)₆³⁻, but the resulting physical architecture has the potential to significantly enhance Raman scattering signals by generating SERS hot spots between AuNPs, thereby enabling simultaneous electrochemical/SERS dual detection of phenazine metabolites.

Selective electrochemical detection of PCA on BCP-coated planar gold electrode.

Initially, the pH-responsive permselective behavior of the PS-b-P4VP BCP was characterized on a planar gold electrode by electrochemical detection of phenazine metabolites produced by P. aeruginosa. As shown in Figure S5A, single phenazine species, i.e., PCA, PCN, and PYO, at pH 4.5 in LB medium all generate a distinct peak current in oxidative square wave voltammetry (SWV). However, the three peaks partially overlap in the potential range ca. −0.36 V ⩽ E_appl ⩽ −0.1 V. Different peak currents for the three phenazines tested suggest that each species exhibits different electrochemical kinetics. In addition, a mixture composed of PCA, PCN, and PYO (each 33.3μM) was also measured by SWV, and consistent with the significant current overlap observed in the individual responses, a single broad peak was obtained, as shown in Figure S5B.

In order to characterize anion permselective transport across the PS-b-P4VP BCP membrane, a BCP membrane was spin-coated onto a planar gold electrode. The pH of all phenazine-containing LB medium solutions was adjusted to pH 4.5, which is lower than the pKa of PYO (4.9) and PCN, but higher than the pKa of PCA (4.2).^{21, 22} When the BCP membrane-coated gold electrode was exposed to PCN and PYO, oxidative SWV showed no current response beyond a broad background, as shown in Figure S6A, consistent with the anion permselective BCP membrane preventing transport of neutral/positive PCN and PYO to the electrode. In contrast, anionic PCA (at pH 4.5) produced a peak response resulting from the anion-selective transport properties of the membrane at pH 4.5. In contrast, no current response was obtained from PCA at pH 7.0, consistent with the P4VP domains being in a collapsed, hydrophobic, dewetted state. The membrane permselectivity exhibited in these planar gold experiments mimics the designed pH-dependent selectivity for PCA on AuNP-filled BCP@NEA structures.

Clearly there is a range of pH over which the desired permselectivity can be obtained. For instance, if pH 4.7 is chosen, it is advantageous in terms of the ionization percentage of PCA, but pH 4.7 being close to the P4VP pKa would diminish the anion permselectivity of the BCP membrane. On the other hand, at pH 4.3, the ionized fraction of PCA is lower, leading to less efficient PCA collection. Thus, pH 4.5 was chosen as a compromise.

Electrochemical quantitation of P. aeruginosa-derived PCA.

Next, the BCP membrane was combined with AuNP-filled NEAs for quantitation of PCA secreted by P. aeruginosa. Figure 3A shows the oxidative SWVs obtained from −0.8 V to +0.2 V vs. Ag/AgCl from PCA standards at varying concentration in LB medium at pH 4.5. To generate a calibration curve, peak current was integrated using the peak finding function on the CHI potentiostat software, which showed a linear relationship with PCA concentration in the range 1 μM ⩽ [PCA] ⩽ 100 μM. The 3.3σ/S limit of detection (LOD) was estimated to be 3.3 μM, which is comparable with other reported LOD values of phenazine metabolites obtained by electrochemical techniques including those obtained by SWV, CV, and differential pulse voltammetry.^{6, 24–27}

Figure 3.

Electrochemical detection of PCA. (A) Representative oxidative SWVs as a function of PCA concentration ranging from 1 μM to 100 μM in LB medium at pH 4.5. (B) Oxidative SWVs obtained at different incubation times from P. aeruginosa cultured in LB medium. Prior to electrochemical detection, the collected sample was adjusted to pH 4.5 for the selective separation and transport of PCA into the interior of the BCP@NEAs. (C) The calibration plot for peak SWV current vs. PCA concentration. The mean values and standard deviations were calculated from three independent measurements. (D) Histograms of the measured concentration of PCA secreted from P. aeruginosa based on the PCA calibration plots. The error bars indicate standard deviation obtained from three independent measurements.

Next, PCA produced by P. aeruginosa in planktonic culture was directly monitored as a function of time from 0 to 24 h at 8 h increments. Before transferring bacteria-free supernatant onto the BCP-coated AuNP-filled NEAs, the growth medium was adjusted to pH 4.5 to ensure that the PCA was net negatively charged. As shown in Figure 3B, no PCA was detected in the 0–8 h incubation window, but a distinct peak was observed at 16 h that further increased in intensity at 24 h. PCA concentration was determined to be 36.8 μM and 56.8 μM at 16 and 24 h incubation time, respectively, based on the calibration curve in Figure 3C. Samples obtained from an equivalently cultured P. aeruginosa Δphz null mutant strain (that cannot produce phenazines) yielded no detectable phenazine signal by SWV on AuNP-filled BCP@NEA structures, thus serving as a biological negative control (Figure S7).

SERS detection of PCA.

In order to explore the possibility that the BCP@NEA structures could support simultaneous electrochemical and SERS dual detection, SERS spectra were collected at varying concentrations of PCA ranging from 100 nM up to 100 μM at pH 4.5 with 785 nm laser excitation for quantitation, as shown in Figure 4A. Figure 4C shows that within the PCA concentration ranging from 1 μM to 100 μM, PCA concentration can be expressed quantitatively by I = 331 ln C - 104 (R² = 0.95), where I is the SERS intensity obtained at 1396 cm⁻¹, and C is the PCA concentration (in μM). The BCP@NEA structure was immersed in the PCA solution for ca. 30 min to ensure that the P4VP domains in the BCP membrane were hydrophilic and positively charged before SERS analysis. Since the pH-responsive and charge-selective BCP membrane can separate PCA from other more abundant phenazines by selective transport into the BCP@NEA nanopores, the PCA produced by P. aeruginosa bacteria can be isolated from the background interferents and selectively detected. To avoid the strong interference from LB medium, the organic metabolites produced in P. aeruginosa culture were extracted and redispersed in DI water (at pH 4.5), as shown in Figure S8 (see Experimental Section for detailed procedure). Figure 4B shows the SERS spectra of the samples extracted from PA14 culture at 0, 8, 16, and 24 h incubation times. Importantly, the SERS spectra obtained at 16 h and 24 h are identical to those obtained with commercial PCA but do not match any of the other phenazines (Figure S9 and Table S1), confirming the presence of PCA after 16 h incubation with an OD₆₀₀ of 2.0 as isolated by permselective transport into the BCP@NEA structure. Based on the calibration curve, the PCA concentration in P. aeruginosa culture was determined to be 3.6 μM and 15.4 μM at 16 and 24 h, respectively, as shown in Figure 4D. No PCA was detected in the extracted samples at 0 or 8 h incubation, consistent with the electrochemical results. It is notable that the PCA concentrations measured from the SERS spectra are smaller than those measured with oxidative SWV, a result which can be attributed to mass loss during the extraction process. SERS measurements were also performed for the Δphz mutant and no phenazine signal was detected, as expected, consistent with the SWV results for the Δphz mutant.

Figure 4.

SERS detection of PCA. (A) SERS spectra of PCA in DI water at pH 4.5 at different concentrations (from bottom to top: 0, 100 nM, 1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 75 μM, 100 μM). (B) SERS spectra of PCA produced by P. aeruginosa grown in planktonic cultures as a function of incubation time. (C) SERS intensity (at 1396 cm⁻¹) as a function of PCA concentration (1 μM, 5 μM, 10 μM, 25 μM, 50 μM, 75 μM, 100 μM). Dashed line is a logarithmic fit in the quantitative detection regions. The mean values and standard deviations were calculated from three independent measurements. (D) Histograms of measured PCA concentrations obtained from bacterial-free supernatant of P. aeruginosa cultures determined using the PCA calibration plot in panel (C) as a function of incubation time. Error bars indicate standard deviation.

CONCLUSIONS

Actively controllable hierarchically organized block copolymer nanopore electrode arrays have been developed to couple solid-phase microextraction with electrochemical and SERS dual-mode sensing. As a model system to demonstrate the capabilities of this approach, the low-abundance metabolite PCA, produced by P. aeruginosa, was measured in the presence of more abundant phenazine metabolites such as PYO and PCN. To fabricate the structure, single gold layer-embedded nanopores are filled with 150 nm AuNPs, to support simultaneous electrochemical and SERS detection. Then, a pH-responsive, charge-selective PS-b-P4VP BCP membrane is introduced as a covering layer on the AuNP-filled NEA. The anion permselective PS-b-P4VP BCP serves to isolate the less-abundant PCA from other more abundant phenazine metabolites. Thus, under the proper pH conditions, PCA can be selectively detected in the NEA without signal interference from the other more abundant phenazine metabolites. Adjusting solution pH to 4.5 was found to create conditions where negatively charged PCA is selectively transported thorough the BCP membrane to the NEA nanopores while neutral/positive PYO and PCN are excluded from the interior of the NEA nanopores. SWV produced a linear working curve from 1 μM to 100 μM PCA with a LOD of 3.3 μM and SERS was able to detect PCA down to 100 nM.

This approach is immediately applicable to the selective detection of specific virulence factors, such as the phenazines, produced by pathogenic bacteria, like P. aeruginosa. The actively controllable solid phase microextraction architecture with electrochemical and SERS sensing capabilities was demonstrated by detecting the minor metabolite PCA in the presence of abundant competing phenazine metabolites. Using this monolithic nanoscale isolation/sensing architecture, the production of PCA by PA14 P. aeruginosa was followed for 24 h using both SVW and SERS. Both methods returned comparable results, with no PCA detected from 0–8 h, followed by an onset and subsequent rapid production of PCA, producing increasing concentrations at 16–24 h, which is consistent with previous work on phenazine production.²⁸

The actively controlled hierarchically-organized solid phase microextraction/sensor can certainly be extended. One simple extension would substitute another polymer for the PS-b-P4VP BCP. Using poly(styrene-b-acrylic acid), for example, could produce a cation permselective layer, and manipulation of the polymer side chains would provide an entrée to tuning the pH-sensitivity. These structures could then be used to detect other metabolites produced by microbial communities on-demand. In addition, since SERS and SWV/CV provide complementary molecular information, e.g., quantitative assays of electroactive species from electrochemical measurements and molecular fingerprint information from SERS measurements, the combined isolation/detection capabilities could provide new ways to examine the secreted factors characteristic of metabolic pathways and those regulating the interaction with neighboring organisms in microbial assemblies.