Enzyme-Immobilized 3D-Printed Reactors for Online Monitoring of Rat

Article pubs.acs.org/ac

Enzyme-Immobilized 3D-Printed Reactors for Online Monitoring of Rat Brain Extracellular Glucose and Lactate Cheng-Kuan Su,*,† Shuo-Chih Yen,‡ Tzu-Wen Li,‡ and Yuh-Chang Sun*,‡ †

Department of Bioscience and Biotechnology, National Taiwan Ocean University, Keelung, 20224, Taiwan Department of Biomedical Engineering and Environmental Sciences, National Tsing-Hua University, Hsinchu, 30013, Taiwan

‡

S Supporting Information *

ABSTRACT: In this study we constructed a highly sensitive system for in vivo monitoring of the concentrations of rat brain extracellular glucose and lactate. This system involved microdialysis (MD) sampling and fluorescence determination in conjunction with a novel sample derivatization scheme in which glucose oxidase and lactate oxidase were immobilized in ABS flow bioreactors (manufactured through low-cost threedimensional printing (3DP)), via fused deposition modeling, for online oxidization of sampled glucose and lactate, respectively, in rat brain microdialysate. After optimizing the experimental conditions for MD sampling, the manufacture of the designed flow reactors, the enzyme immobilization procedure, and the online derivatization scheme, the available sampling frequency was 15 h−1 and the system’s detection limits reached as low as 0.060 mM for glucose and 0.059 mM for lactate, based on a 20-μL conditioned microdialysate; these characteristics were sufficient to reliably determine the concentrations of extracellular glucose and lactate in the brains of living rats. To demonstrate the system’s applicability, we performed (i) spike analyses of offline-collected rat brain microdialysate and (ii) in vivo dynamic monitoring of the extracellular glucose and lactate in living rat brains, in addition to triggering neuronal depolarization by perfusing a high-K+ medium from the implanted MD probe. Our analytical results and demonstrations confirm that postprinting functionalization of analytical devices manufactured using 3DP technology can be a powerful strategy for extending the diversity and adaptability of currently existing analytical configurations.

S

tered, making it difficult to ensure that the acquired signals are derived entirely from the desired analyte.8 The resulting analytical performance (e.g., accuracy, method’s stability) of these implanted biochemical sensors can be biased by the coexistence of redox-active species and biofouling of the sensing components.13−15 Continuous microdialysis (MD) sampling can overcome some of these problems by onlineexcluding almost all of large biomolecules during sampling, thereby allowing explicit identification, as well as reliable calibration, of the chemical substances sampled from the targeting regions when coupled to relevant analytical equipment and systems.8,16−18 Alternatively, enzyme-based potentiometric biosensors hyphenated with MD sampling devices can also greatly decrease interference from biomolecules or fouling of the reactor’s sensing components, thereby providing the ability to resolve the dynamic variations of chemical substances in living biological systems.19,20 Although enzyme-based flow-through biosensors are attractive for their high specificity, reusability, and convenient fitting

tudying energy metabolites, including glucose, lactate, and pyruvate, in brain extracellular fluid (ECF) can be a relatively straightforward approach toward understanding the chemical regulation process, how neurons communicate with each other, and how the brain regions orchestrate physiological functions in our bodies.1−5 Nevertheless, the question of how the brain extracellular homeostasis of these metabolites is balanced by the energy consumption from brain activities and supplemented from local cerebral blood flows has rarely been ascertained for various physiological and special pathological conditions.3,5−7 Consequently, the need remains to develop practical analytical strategies enabling real-time monitoring of the dynamic fluctuations of brain extracellular metabolites in living animals, especially in response to specific physiological events.8−12 Among conventional methods for invasively neurochemical monitoring, implanted electrochemical biosensors that function based on enzymatic biochemical sensing and recognition have been applied quite widely to resolve the real-time dynamics of brain extracellular glucose/lactate in anesthetized or freely moving animals, taking advantage of the biosensors’ small dimensions (causing minimal tissue damage) and high temporal responses.10−12 Practical difficulties are, however, still encoun© XXXX American Chemical Society

Received: January 20, 2016 Accepted: May 18, 2016

A

DOI: 10.1021/acs.analchem.6b00272 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. (A, B) CAD drawings of (A) the flow reactor and (B) a layer of ordered cuboids in the reaction chamber. (C) Photograph of the printed reactor; two flat-bottom female connectors with a piece of polytetrafluoroethylene (PTFE) tubing were fitted to allow connection to an flow injection analysis (FIA) interface.

to most analytical interfaces,9,21 their applicability, when coupled with MD sampling devices, remains restricted by the limited designs/configurations of flow bioreactors, the need for sophisticated enzyme immobilization procedures, and low conversion efficiencies, all of which increase the method’s complexity when analyzing brain extracellular substances. Therefore, a challenge remains to develop more-efficient interfaces for enzyme-incorporated flow bioreactors for use with MD sampling to investigate biochemical processes occurring in living animals. Recently, emerging additive manufacturing (three-dimensional printing; 3DP) technologies have accelerated the rapid prototyping of research components,22−24 the construction of complicated devices in multilayer working domains,25,26 and the customization of flowinjection-analysis (FIA) devices in general laboratories.27,28 Fused deposition modeling (FDM) prototyping, which relies upon extruding thermoplastics (e.g., acrylonitrile butadiene styrene (ABS), polylactic acid (PLA)), is the most accessible 3DP technology because of its uncomplicated instrument setup, simplicity, and low cost. In addition, the ABS and PLA filaments are inexpensive and easily recycled, making them useful for researchers wishing to lower the costs of prototyping experimental devices. The prior incorporation of desirable compounds into the raw printing materials or the postprinting functionalization of these chemically inert thermoplastic polymers can impart printed devices with additional physical and chemical functionality.29−34 With the aim to develop more-efficient analytical methods for probing the dynamic variations of brain extracellular substances in or during neurochemical reactions, in this study we used a FDM-type 3D printer and ABS thermoplastics to manufacture a flow reactor designed for a system operating in an online FIA configuration. Postprinting immobilization of glucose oxidase (GOx) and lactate oxidase (LOx) individually onto the surfaces of the reactors’ inner structures formed enzyme-immobilized printed devices that functioned as key components interfaced between an MD sampling probe and an online fluorescence spectrophotometer. When rat brain microdialysate passed through either the GOx- or LOx-immobilized flow bioreactors, biochemical oxidation of the sampled glucose/ lactate led to the continuous online generation of hydrogen peroxide (H2O2), which transformed a H2O2-sensing molecular probe (Amplex UltraRed (AUR), a fluorogenic substrate for horseradish peroxidase (HRP)) into a brightly fluorescent emitting product. After we had optimized the parameters of the MD sampling, the enzyme immobilization procedure, the

online automatic enzymatic derivatization scheme, and the online fluorescence determination system, we evaluated the analytical performance of the resulting glucose/lactate monitoring system. Subsequently, we tested the system’s applicability through spike analyses of off-line collected rat brain microdialysate samples as well as in vivo dynamic monitoring of rat brain extracellular glucose/lactate following a depolarization simulation in which we perfused a high-K+ medium through the implanted MD probe.

■

EXPERIMENTAL SECTION Chemicals. Glutaraldehyde (GA) solution (G5882, grade I, 25% in H2O), D-(+)-glucose (G8270), GOx (G7141, from Aspergillus niger), L-(+)-lactic acid (L1750), LOx (L0638, from Pediococcus sp.), HRP (P8250), sodium chloride (NaCl, S5886), potassium chloride (KCl, 05257), trisodium citrate dihydrate (C8532), sodium bicarbonate (NaHCO3, S5761), sodium carbonate (Na2CO3, S7795), sodium dihydrogen phosphate (NaH2PO4, S8282), disodium hydrogen phosphate (Na2HPO4, S7907), and urethane (U2500) were all purchased from Sigma-Aldrich. AUR Reagent (A36006) was purchased from Thermo Fisher Scientific. All chemical solutions were prepared with water purified through a Milli-Q Integral water purification system (Merck Millipore). The AUR working solution was keep in the dark as much as possible. To avoid bubble formation during fluorescence analysis, all of the system’s carrier solutions were purged in advance with highpurity N2 gas. Design and Fabrication of the Flow Reactor. SolidWorks 2013 (Dassault Systèmes) computer-aided design software was used to construct the whole 3D object. The designed all-in-one flow reactor included two parallel reaction chambers and two fittings for standard 10-32 flat-bottom male connectors (see Figure 1A and Figure S1).28 To shorten the fabrication time and prevent blockage of the printed device, these two reaction chambers were aligned in series and connected by a hollow channel. The reaction chambers were filled with ordered cuboids (1 × 1 × 0.8 mm3, L × W × H) that overlapped by a square of 0.1 × 0.1 mm2 (L × W) at each edge (Figure 1B). The two ends of the chamber, designed in the form of square pyramids, were arranged with the same ordered cuboids to yield a robust monolithic structure and offered the additional functions of a flow distributor and collector.35,36 The designed 3D object was exported to an STL file format, digitally sliced into multiple 2D-layer images using “UP!” software (v. 2.0), and then fabricated using a portable 3D printer (UP Plus B


Article


Figure 2. Schematic representation of the proposed online glucose/lactate monitoring system. (A) Determination of lactate: The microdialysate was mixed online with two streams of AUR and HRP solutions and then loaded into sample loop A, while the conditioned microdialysate in sample loop B was transferred via the LOx-immobilized bioreactor to the fluorescence spectrometer. (B) Determination of glucose: The microdialysate was conditioned online and then loaded into sample loop B, while the conditioned microdialysate in sample loop A was transferred via the GOximmobilized bioreactor to the fluorescence spectrometer.

with 10 cm-long polytetrafluoroethylene (PTFE) tubing (0.02 in. i.d.) for connection to conventional FIA interfaces (Figure 1C). Enzyme Immobilization Procedure. For immobilization of GOx, the reactor was (i) filled with a GA solution to allow polymerization on the reactor’s surface, (ii) evacuated and cleaned with a coating buffer solution (0.1 M phosphate− citrate buffer solution, pH 6.5) to remove the residual GA solution, (iii) refilled with a GOx-containing coating buffer solution to allow coupling of the enzyme to the GA-activated surfaces, and (iv) evacuated and cleaned using a coating buffer solution to remove the uncoated enzymes.37 For immobilizing

2, equipped a 0.4 mm copper nozzle) with ABS plastic filaments (C-01-01, 1.75 mm i.d., 0.7 kg/white) from Delta Micro Factory (PP3DP) and a 260 °C nozzle temperature. Before printing an exported device, the platform was preheated to 80 °C. To avoid fluid permeation and leakage, printing reactors was conducted under a z-axis (layer) thickness of 0.15 mm, and the most densely filling mode with six compact surface layers for outer boundary structures. The printing speed varied from structure to structure and was in the range from 10 to 100 cm3 h−1. As the reactors were fabricated, their inlets and outlets were fitted with two conventional flat-bottom nuts with ferrules (P-840 and P-844, IDEX Health & Science) equipped C


Article


Figure 3. Effects of (A) GA activation concentration, (B) GOx concentration, (C) LOx concentration, and (E) coating pH for GOx/LOx on the resulting fluorescence intensities when using the proposed enzyme-immobilized derivatization method. Concentrations of glucose and lactate: 1 mM. Each plotted signal intensity was the difference between the signals measured from the sample containing analyte and its respective blank sample; all data have been normalized to their respective maxima for each parameter. Error bars represent standard deviations (n = 4 for each parameter).

fold online with the two streams of AUR and HRP through a cross (ZX1MFPK, Valco) and loaded into the sample loops. The conditioned samples were then transferred into either the GOx- or LOx-immobilized reactors for online flow-through derivatization of glucose or lactate. The proposed online automatic derivatization and fluorescence determination system (Figure 2) was configured in a parallel dual-channel sample loading and detection mode, managed by two 20-μL sample collection loops, two 10-port/two-position valves (C22Z3180D, Valco), and a peristaltic pump. As illustrated in Figure 2A,B, when one microdialysate was being diluted and loaded online into a sample loop for lactate determination, the other conditioned microdialysate was transported through the GOximmobilized bioreactor and introduced directly into a liquidchromatography flow cell accessory (L2250138, PerkinElmer; inner volume, 22.5 μL) equipped in a fluorescence spectrometer (LS-55, PerkinElmer, IL) for time-resolved scanning (emission wavelength, 581 nm; excitation wavelength, 560 nm). All of these steps were manipulated in the closed FIA system to avoid reactive oxygen species contributed from the used labware and surrounding environment. The two valves were synchronized and programmed using a single laptop, connected through a serial valve interface (SIV-110, Valco, Lucerne, Switzerland), to eliminate possible errors from manual

LOx, the reactor was (i) cleaned with a coating buffer solution (0.1 M citrate buffer, pH 6), (ii) filled with an LOx-containing coating buffer solution to allow the attachment of enzymes to the reactor’s surfaces, and (iii) evacuated and cleaned using a coating buffer solution to remove the uncoated enzymes. A peristaltic pump (Miniplus 3, Gilson) equipped with a black/ black PVC peristaltic tube (0.76 mm i.d.) was used to deliver all of the solutions for the immobilization steps (flow rate, 100 μL min−1). The enzyme-immobilized bioreactors were preserved with the coating buffer solutions until the experiments were performed. Online Fluorescence Monitoring System. The continuous sampling apparatus contained an MD probe with a 4 mmlong, 500-μm-diameter polyarylethersulfone (PAES) membrane having a molecular weight cutoff (MWCO) of 20 kDa (8010435; CMA 20, CMA Microdialysis), a dual-channel syringe pump (KDS260, KD Scientific), and two plastic syringes (4606051 V, B. Braun, Melsungen, Germany). A sixport, two-position valve (C22Z-3186, Valco) arranged between the two outlets of the syringes and the inlet of the MD probe was used to manually switch the dialysis perfusion solutions (either 0.9% NaCl or 1.15% KCl) for local triggering of acute neuronal depolarization. For online derivatization of glucose/lactate sampled from rat brain ECF, the microdialysates were immediately diluted 30D


Article


Figure 4. Effects of (A) derivatization pH, (B) number of layers of cuboids in the reaction chamber, (C) the system’s carrier flow rate, and (D) activities of immobilized enzymes on the resulting fluorescence intensities when using the proposed enzyme-immobilized derivatization method. Concentrations of glucose and lactate: 1 mM. Each plotted signal intensity was the difference between the signals measured from the sample containing analyte and its respective blank sample; all data have been normalized to their respective maxima for each parameter. Error bars represent standard deviations (n = 4 for each parameter).

■

operation. Quantification was performed through integration of the peak areas of the time-derived fluorescence intensities. In Vivo Experiments. Male Sprague−Dawley rats (250 ± 20 g; n = 4, specifically pathogen-free) from BioLASCO (Taiwan) were acclimatized to their environmentally controlled quarters (25 °C; 12-h light/12-h dark cycle). Water and food were available ad libitum. All animal treatments and experimental protocols were conducted in conformity with the guidelines and approval of the Institutional Animal Care and Use Committee at National Tsing-Hua University (approval number 10212). The rats were first anesthetized with urethane (1.6 ± 0.1 g kg−1 body weight via I.P. administration), and then each rat’s head was mounted on a stereotaxic apparatus (Davis Kopf Instruments, Tujunga). After an MD probe had been precisely implanted into the brain (2.0 mm anterior-posterior and 2.0 mm laterally from the bregma to target the hippocampus38), continuous measurements (15 cycles, 8 min per cycle, 120 min total) were conducted to reach physiological hemostasis of the brain tissues around the insertion region and acquire the basal brain extracellular glucose/lactate concentrations. The perfusion solution was then switched from 0.9% NaCl to 1.15% KCl for 32 min (four measurement cycles) to locally induce neuronal depolarization. Each rat was euthanatized when the experiment was complete.

RESULTS AND DISCUSSION

Fabrication of ABS Flow Reactor. Our flow reactors designed for sample derivatization and system’s operation in an online FIA configuration featuring two parallel reaction chambers filled with ordered cuboids and two female fittings. It has been confirmed that reaction chambers filled with ordered cuboids as small as possible can enhance reaction efficiencies, due to the effects of improved mixing and larger surface areas.28,35,36 As illustrated in Figure S2, the FDM-type 3D printer manufactured the cuboids with dimensions of approximately 0.8 × 0.8 × 0.8 mm3 (L × W × H); more defects and blocked interstitial volumes appeared in smaller printed cuboids.39,40 Subsequently, 118 ordered cuboids (7 + 5 layer) having dimensions of 1 × 1 × 0.8 mm3 and overlapping by a square of 0.1 × 0.1 mm2 at each edge were arranged inside the two reaction chambers. The total fabrication time for each reactor was approximately 38 min (100 μm per layer). The printed reactor had a weight of approximately 3.6 g and an inner volume of 360 ± 14 μL (n = 4). Figure S1 provides the detailed dimensions of the designed reactor. GA Activation of 3D-Printed ABS Reactor. The fluorescence determination strategy, following an online enzymatic biochemical derivatization scheme, has received much attention because of its specificity, adaptability, diversity, E


Article

Analytical Chemistry and cost-effectiveness.41,42 For online in vivo monitoring of rat brain extracellular glucose and lactate, we have, for the first time, immobilized GOx and LOx onto the inner surfaces of 3Dprinted ABS flow reactors to convert glucose and lactate, respectively, into H2O2 and related derivatives during the passage of the microdialysate. We applied GA, a cross-linker commonly used to facilitate fixing of biomolecules to electrode surfaces or sensing components,43,44 to link the enzymes to the ABS surface of the 3D-printed object. The immobilization procedure involved prior activation of the bare ABS surface through GA polymerization (only for GOx), followed by coupling of the oxidases. The first step of GA activation, which remarkably affected both the subsequent activity and stability of the immobilized enzymes, we found (Figure 3A) that the concentration of GA strongly influenced the cross-linking between the polymerized GA and GOx, with 10% GA required to immobilize GOx. In contrast, LOx attached to the bare ABS surface without the need for preactivation with GA. Indeed, the polymerized GA hindered the attachment of LOx to the ABS surface, decreasing the derivatization efficiency when the GA molecules were not clearly removed (Figure S3A). We suspect that the coupling of GOx to the GA-preactivated support occurred through interactions of the amino groups of GOx with the polymerized GA’s residual aldehyde groups.37,45,46 We also suppose that the immobilization of LOx was assisted through its hydrophobic interactions with the styrene units of the ABS structure.37,47,48 Thus, we optimized the immobilization of GOx and LOx using 10% GA for the former but no prior GA activation for the latter. For the GA-preactivation of the bare ABS surface for GOx coupling, the optimal activation pH and time were 9.5 and 24 h, respectively (Figure S3B,C). Immobilization of GOx and LOx onto ABS Reactor. Prior to performing GOx immobilization, the flow reactors were cleaned with the coating buffer (0.1 M phosphate−citrate buffer) for 10 min to remove any residual GA; this step was unnecessary for LOx. The coating solution containing either GOx or LOx was then infused using a peristaltic pump; we optimized the following immobilization conditions: enzyme concentration, coating pH, and coating time. Figure 3B,C indicates that the derivatization efficiencies increased as the coating concentrations of GOx and LOx increased. Considering the cost of the enzymes and the possibility of their inducing background contributions, we selected immobilization concentrations for GOx and LOx of 500 and 8 U mL−1, respectively. Figure S3D and Figure 3D reveal that the optimal coating time was 24 h and that the optimal values of pH when immobilizing GOx and LOx were 6.5 and 6, respectively. The resulting GOxand LOx-immobilized reactors were preserved with the coating buffer solutions until required for experiments. Having optimized the procedures for immobilizing the oxidases in the 3D-printed ABS reactors, the SEM images in Figure S4 demonstrated that GA, GOx, and LOx were possibly attached to their respective ABS surfaces, setting the stage for online derivatization of the sampled glucose and lactate during passage of the microdialysate. Online Derivatization of Glucose and Lactate. During delivery of the glucose- and lactate-containing microdialysates through our proposed enzyme-immobilized 3D-printed ABS flow-through bioreactors, the H2O2 molecules generated through online oxidation of glucose and lactate converted AUR stoichiometrically into corresponding fluorescent species. To facilitate online fluorescence determination of the derived

H2O2, as well as eliminate biological matrix effects, the microdialysates were immediately diluted online (in total 30fold) using the two streams of AUR and HRP solutions, because the AUR molecular probe was extremely sensitive to the presence of H2O2 (allowing determination at concentrations down to the subnanomolar level).49,50 Using our optimized immobilization scheme and the goal of approaching maximal sensitivity, we sought to optimize several of the online system’s parameters, namely, the derivatization pH, the number of layers of ordered cuboids (volume of reaction chamber), and the carrier flow rate (derivatization time). Figure 4A reveals that the fluorescence intensities were maximized when the derivatization pH was near 7.5 for both GOx and LOx. With the derivatization pH set at 7.5, Figure 4B,C reveals that the system’s carrier flow rate, rather than the areas of the enzymeimmobilized surfaces, critically influenced the derivatization efficiencies and resulting fluorescence intensities. With consideration of the system’s sensitivity and temporal resolution, we selected a carrier flow rate of 75 μL min−1 to allow delivery of the conditioned microdialysate through a 12layer flow bioreactor within 4 min. Employing the optimized immobilization procedures and system parameters, we tested the interdevice repeatability and interday variation (device stability) of our enzyme-immobilized bioreactors. The signals obtained revealed that the variations from the four reactors were all less than 15%, meaning that the proposed oxidaseimmobilized bioreactors, prepared using our designed configuration, manufacturing process, and immobilization procedure, would ensure the derivatization of glucose and lactate with constant quality and repeatability. Figure 4D also reveals that, when preserved at a temperature of 4 °C, the immobilized GOx in the ABS bioreactors maintained its activity for up to 42 days; in contrast, the immobilized LOx lost 60% of its activity of within 1 week postimmobilization. Thus, although each enzyme’s stability was dependent on its storage conditions,51,52 the LOx-immobilized bioreactors should be used as soon as possible. Analytical Performance. The flow rate of the MD sampling was set at 1 μL min−1 (recoveries of glucose and lactate, 29.7 ± 0.4 and 43.5 ± 4.6% (mean ± standard deviation, n = 4), respectively) for in vivo sampling of rat brain extracellular glucose and lactate. The time interval for microdialysate loading was set at 4 min, corresponding to a temporal resolution of 15 h−1 (7.5 h−1 for glucose or lactate alone). Table 1 summarizes the optimized operating parameters; the signal profiles for calibrations of glucose and lactate under the system’s optimization were provided in Figure S5. The detection limits (DLs, defined as 3 times the standard deviation of the baseline noise (n = 7)) for glucose and lactate, when operating the system under the optimized conditions, were 0.060 and 0.059 mM, respectively (Table 2); these values are quite sufficient for simultaneous analysis of these two analytes in the ECF of living rat brains.42,53−56 In addition, the relative standard deviations of the obtained fluorescence signals were acceptable: less than 8.3% (n = 42) after online monitoring of a glucose/lactate solution (1 mM) for 5.6 h. Thus, we expected our developed monitoring system to be capable of measuring the concentrations of rat brain extracellular glucose and lactate with sufficient sensitivity and negligible signal fluctuations. To confirm the analytical reliability of our proposed method, we performed spike analyses of offline-collected rat brain microdialysates. The measured recoveries for glucose and F


Article

Analytical Chemistry Table 1. Optimized Conditions for the Established Online Glucose/Lactate Monitoring System MD Sampling perfusate 0.9% NaCl sampling flow rate 1 μL min−1 sampling frequency 15 h−1 GA Activation and Oxidase Immobilization Procedure GA concentration (for GOx) 10% (0.2 M carbonate buffer, pH 9.5) GA activation time (for 24 h (room temperature) GOx) GOx concentration 500 U mL−1 (0.1 M phosphate−citrate buffer, pH 6.5) LOx concentration 8 U mL−1 (0.1 M citrate buffer, pH 6) immobilization time of 24 h (room temperature for GOx; 4 °C for GOx/LOx LOx) Online Derivatization and Fluorescence Determination System

Figure 5. Time course of rat brain extracellular glucose and lactate concentrations following implantation of the MD probe and treatment through perfusion with a high-K+ medium. Error bars represent standard deviations (n = 4). Black bar: Time period of perfusing highK+ medium.

AUR concentration 25 μM HRP concentration 0.2 U mL−1 mixing flow rate of AUR/HRP 14.5/14.5 μL min−1 dilution factor 30 system’s carrier solution 0.01 M phosphate−citrate buffer, pH 7.5 system’s carrier flow rate 75 μL min−1 Fluorescence Spectrometer excitation/emission wavelength excitation/emission slit data interval

values of 1.74 ± 0.05 and 0.79 ± 0.03 mM (n = 4), respectively. These values agree well with those previously reported.42,53−56 To further demonstrate the system’s capability for real-time monitoring of the dynamics of extracellular glucose and lactate in living rat brains, we replaced the original perfusion solution (0.9% NaCl) with 1.15% KCl (both solutions have equal osmotic pressure) for 32 min to locally induce acute neuronal depolarization. The resulting profiles (Figure 5) indicate that the concentration of extracellular glucose decreased to 0.83 ± 0.07 mM (48% of the basal value) at 16 min post-treatment, but the concentration of extracellular lactate increased to a higher plateau of 1.18 ± 0.08 mM (149% of the basal value) at approximately 32 min post-treatment, similar to the profiles acquired from the same animal model.58,59 After switching the perfusion solution back to 0.9% NaCl, their concentrations returned to the original basal values at time points of 24 min post-treatment for glucose and 40 min post-treatment for lactate. The abnormally high content of K+ ions that diffused into the brain ECF for depolarizing of surrounding neurons had to be eliminated immediately and transported into the intracellular fluids to extracellularly and intracellularly restore Na+/K+ homeostasis. Therefore, the glycolysis process, which dominates the energy support for removal of extracellular K+ ions, would intensify, the main reason for our observation of rapid depletion of extracellular glucose and delayed increase of extracellular lactate.60−62 This effect implies that the response to a massive release of intracellular K+ ions into the brain ECF would profoundly influence the local synaptic signaling, the dynamic actions of neurotransmitters, and the regulation of neural energy metabolism,55,63,64 especially in terms of acute traumatic brain injury.65,66 Therefore, the ability to describe the dynamic variations in the concentration of glucose and lactate would be a significant boost to studies of their physiological functions in living rat brains, because the anaerobic glycolysis pathway tends to dominate the energy consumption process that maintains Na+/K+ homeostasis. According to its analytical performance and demonstrated utility in animal studies, we suspect that our proposed system will have high applicability for online monitoring of extracellular glucose and lactate in living rat brains and that it provides sufficient sensitivity and temporal

560/581 nm 2.5 nm 3s

Table 2. Analytical Performance of the Proposed Online Glucose/Lactate Monitoring System

glucose lactate a

working range, mM

R

MDL, mM

stability (5.6 h), %

spike recoverya, %

0.1−5 0.1−5

0.9927 0.9993

0.060 0.059

8.3 7.7

105.2 ± 2.5 94.0 ± 0.7

Spike concentration: 1 mM.

lactate were 105.2 ± 2.5 and 94.0 ± 0.7%, respectively (Table 2), suggesting that this method was tolerant of any interference caused by the extracellular biological matrix when coupling the MD sampling technique to a fluorescence determination scheme. Accordingly, we suspected that our developed monitoring system would accurately determine the concentrations of rat brain extracellular glucose and lactate and that our proposed strategy of online enzymatic derivatization, using enzyme-immobilized 3D-printed ABS flow bioreactors, and subsequent fluorescence determination would be highly reliable, sensitive, and applicable to other chemical substances after immobilizing their respective specific oxidases. In addition, we have demonstrated that functionalizing inexpensive and readily available ABS thermoplastics can extend the applicability of FDM-type 3DP technology and renew the traditional designs of flow bioreactors for general analytical and bioanalytical applications. In Vivo Monitoring of Rat Brain Extracellular Glucose and Lactate. After evaluating its performance, we applied our online monitoring system to profile the dynamic changes in the concentrations of extracellular glucose and lactate in living rat brains (Figure 5). During the initial postimplantation period, we observed a slightly higher concentration of brain extracellular lactate, presumably the result of local disturbance of the normal physiological status after intrusion of the MD probe.57 At around 80 min postimplantation, the concentrations of brain extracellular glucose and lactate reached basal G



■

resolving power to investigate the rapid dynamic variations of the brain extracellular metabolites in response to physiological stimuli. Most importantly, a strategy of postprinting functionalization of 3D-printed analytical devices can extend their applicability and also diversify conventional analytical configurations.

CONCLUSIONS We have developed a novel sample derivatization scheme, involving GOx and LOx immobilized in 3D-printed ABS flow bioreactors for online oxidization of sampled glucose and lactate in microdialysates, to establish an enzyme-based biochemical assay strategy, coupled with fluorescence determination, for simultaneous online monitoring of rat brain extracellular glucose and lactate. For this proposed method, we manufactured the designed flow bioreactors using a commercial FDM-type 3D printer and developed a simple procedure to functionalize the printed ABS reactors to facilitate determination of glucose and lactate in biological samples. On the basis of the acquired results from animal experiments, we confirmed that variations in brain extracellular glucose and lactate occur postprobe implantation, measured their basal levels, and detected significant changes following perfusion of a high-K+ medium through the MD probe, all of which could be profiled dynamically. This highly sensitive analytical system for monitoring specific brain biomolecules, by adapting 3D printing technology to manufacture analytical devices in conjugation with postprinting functionalization, appears to be a valuable tool for facilitating investigations of how brain extracellular glucose and lactate are intricately manipulated in living animals. ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b00272. Detailed dimensions of the designed flow-through bioreactor; photograph of the printed ordered cuboids; effects of washing time for removing the residual GA, GA activation pH, GA activation time, and coating time for GOx/LOx on the resulting fluorescence intensities; SEM images of bare and enzyme-immobilized ABS surfaces (PDF)

■

REFERENCES

(1) Dienel, G. A.; Hertz, L. J. Neurosci. Res. 2001, 66, 824−838. (2) Chih, C. P.; Roberts, E. L., Jr. J. Cereb. Blood Flow Metab. 2003, 23, 1263−1281. (3) Bélanger, M.; Allaman, I.; Magistretti, P. J. Cell Metab. 2011, 14, 724−738. (4) Dienel, G. A. J. Cereb. Blood Flow Metab. 2012, 32, 1107−1138. (5) Magistretti, P. J.; Allaman, I. Neuron 2015, 86, 883−901. (6) Shah, K.; DeSilva, S.; Abbruscato, T. Int. J. Mol. Sci. 2012, 13, 12629−12655. (7) De la Monte, S. M.; Tong, M. Biochem. Pharmacol. 2014, 88, 548−559. (8) Parkin, M. C.; Hopwood, S. E.; Strong, A. J.; Boutelle, M. G. TrAC, Trends Anal. Chem. 2003, 22, 487−497. (9) Zhang, M.; Mao, L. Front. Biosci., Landmark Ed. 2005, 10, 345− 352. (10) Schultz, K. N.; Kennedy, R. T. Annu. Rev. Anal. Chem. 2008, 1, 627−661. (11) Robinson, D. L.; Hermans, A.; Seipel, A. T.; Wightman, R. M. Chem. Rev. 2008, 108, 2554−2584. (12) Zhang, M.; Yu, P.; Mao, L. Acc. Chem. Res. 2012, 45, 533−543. (13) Lowry, J. P.; Miele, M.; O’Neill, R. D.; Boutelle, M. G.; Fillenz, M. J. Neurosci. Methods 1998, 79, 65−74. (14) Vasylieva, N.; Maucler, C.; Meiller, A.; Viscogliosi, H.; Lieutaud, T.; Barbier, D.; Marinesco, S. Anal. Chem. 2013, 85, 2507−2515. (15) Stoytcheva, M.; Zlatev, R.; Velkova, Z.; Valdez, B.; Ovalle, M. Portugaliae Electrochimica Acta 2009, 27, 353−362. (16) De Lange, E. C. M.; de Boer, A. G.; Breimer, D. D. Adv. Drug Delivery Rev. 2000, 45, 125−148. (17) Matzneller, P.; Brunner, M. TrAC, Trends Anal. Chem. 2011, 30, 1497−1504. (18) Watson, C. J.; Venton, B. J.; Kennedy, R. T. Anal. Chem. 2006, 78, 1391−1399. (19) Boutelle, M. G.; Fellows, L. K.; Cook, C. Anal. Chem. 1992, 64, 1790−1794. (20) Yao, T.; Yano, T.; Nishino, H. Anal. Chim. Acta 2004, 510, 53− 59. (21) Mross, S.; Pierrat, S.; Zimmermann, T.; Kraft, M. Biosens. Bioelectron. 2015, 70, 376−391. (22) Pearce, J. M. Science 2012, 337, 1303−1304. (23) Gross, B. C.; Erkal, J. L.; Lockwood, S. Y.; Chen, C.; Spence, D. M. Anal. Chem. 2014, 86, 3240−3253. (24) Ho, C. M. B.; Ng, S. H.; Li, K. H. H.; Yoon, Y. J. Lab Chip 2015, 15, 3627−3637. (25) Bhargava, K. C.; Thompson, B.; Malmstadt, N. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 15013−15018. (26) Shallan, A. I.; Smejkal, P.; Corban, M.; Guijt, R. M.; Breadmore, M. C. Anal. Chem. 2014, 86, 3124−3130. (27) Su, C. K.; Hsia, S. C.; Sun, Y. C. Anal. Chim. Acta 2014, 838, 58−63. (28) Su, C. K.; Peng, P. J.; Sun, Y. C. Anal. Chem. 2015, 87, 6945− 6950. (29) Sandler, N.; Salmela, I.; Fallarero, A.; Rosling, A.; Khajeheian, M.; Kolakovic, R.; Genina, N.; Nyman, J.; Vuorela, P. Int. J. Pharm. 2014, 459, 62−64. (30) Wang, X.; Cai, X.; Guo, Q.; Zhang, T.; Kobe, B.; Yang, J. Chem. Commun. 2013, 49, 10064−10066. (31) Goyanes, A.; Buanz, A. B. M.; Basit, A. W.; Gaisford, S. Int. J. Pharm. 2014, 476, 88−92. (32) Goyanes, A.; Wang, J.; Buanz, A.; Martínez-Pacheco, R.; Telford, R.; Gaisford, S.; Basit, A. W. Mol. Pharmaceutics 2015, 12, 4077−4084. (33) Peterson, G. I.; Larsen, M. B.; Ganter, M. A.; Storti, D. W.; Boydston, A. J. ACS Appl. Mater. Interfaces 2015, 7, 577−583. (34) Jakus, A. E.; Secor, E. B.; Rutz, A. L.; Jordan, S. W.; Hersam, M. C.; Shah, R. N. ACS Nano 2015, 9, 4636−4648. (35) Fee, C.; Nawada, S.; Dimartino, S. J. Chromatogr. A 2014, 1333, 18−24.

■

■

Article

AUTHOR INFORMATION

Corresponding Authors

*Fax: +886-3-5723883. Phone: +886-3-5727309. E-mail: [email protected]. *Fax: +886-2-24622320. Phone: +886-2-24622192 ext. 5571. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Professor Mo-Hsiung Yang for providing helpful advice and the Ministry of Science and Technology of the Republic of China (Taiwan) for financial support (Grants 1022627-M-007-005-MY3 and MOST 104-2113-M-019-003MY2). H


Article

Analytical Chemistry (36) Vonk, R. J.; Vaast, A.; Eeltink, S.; Schoenmakers, P. J. J. Chromatogr. A 2014, 1359, 162−169. (37) Ukeda, H.; Fujita, Y.; Ohira, M.; Sawamura, M. J. Agric. Food Chem. 1996, 44, 3858−3863. (38) Paxinos, G.; Watson, C. The Rat Brain in Stereotaxic Coordinates, 6th ed.; Elsevier: Amsterdam, The Netherlands, 2007. (39) Polzin, C.; Spath, S.; Seitz, H. Rapid Prototyping J. 2013, 19, 37− 43. (40) Farzadi, A.; Solati-Hashjin, M.; Asadi-Eydivand, M.; Osman, N. A. A. PLoS One 2014, 9, e108252. (41) Su, C. K.; Chen, C. Y.; Tseng, P. J.; Sun, Y. C. Biosens. Bioelectron. 2015, 64, 535−541. (42) Wang, M.; Roman, G. T.; Schultz, K.; Jennings, C.; Kennedy, R. T. Anal. Chem. 2008, 80, 5607−5615. (43) Walt, D. R.; Agayn, V. I. TrAC, Trends Anal. Chem. 1994, 13, 425−430. (44) Asanomi, Y.; Yamaguchi, H.; Miyazaki, M.; Maeda, H. Molecules 2011, 16, 6041−6059. (45) Migneault, I.; Dartiguenave, C.; Bertrand, M. J.; Waldron, K. C. BioTechniques 2004, 37, 790−802. (46) Bhavani, P.; Sangeetha, D. Int. J. Plast. Technol. 2015, 19, 137− 152. (47) Qian, W.; Yao, D.; Yu, F.; Xu, B.; Zhou, R.; Bao, X.; Lu, Z. Clin. Chem. 2000, 46, 1456−1463. (48) Bayramoglu, G.; Karagoz, B.; Yilmaz, M.; Bicak, N.; Arica, Y. Bioresour. Technol. 2011, 102, 3653−3661. (49) Zhu, A.; Romero, R.; Petty, H. R. Anal. Biochem. 2010, 403, 123−125. (50) Cohn, C. A.; Simon, S. R.; Schoonen, M. A. A. Part. Fibre Toxicol. 2008, 5, 2. (51) Minagawa, H.; Nakayama, N. Biosens. Bioelectron. 1998, 13, 313−318. (52) Patel, N. G.; Erlenkötter, A.; Cammann, K.; Chemnitius, G. C. Sens. Actuators, B 2000, 67, 134−141. (53) Lin, B.; Busto, R.; Globus, M. Y. T.; Martinez, E.; Ginsberg, M. D. Stroke 1995, 26, 1634−1638. (54) McNay, E. C.; Gold, P. E. J. Neurochem. 1999, 72, 785−790. (55) Yao, T.; Yano, T.; Nishino, H. Anal. Chim. Acta 2004, 510, 53− 59. (56) Rocchitta, G.; Secchi, O.; Alvau, M. D.; Farina, D.; Bazzu, G.; Calia, G.; Migheli, R.; Desole, M. S.; O’Neill, R. D.; Serra, P. A. Anal. Chem. 2013, 85, 10282−10288. (57) Benveniste, H.; Huttemeier, P. C. Prog. Neurobiol. 1990, 35, 195−215. (58) Rogers, M. L.; Feuerstein, D.; Leong, C. L.; Takagaki, M.; Niu, X.; Graf, R.; Boutelle, M. G. ACS Chem. Neurosci. 2013, 4, 799−807. (59) Hopwood, S. E.; Parkin, M. C.; Bezzina, E. L.; Boutelle, M. G.; Strong, A. J. J. J. Cereb. Blood Flow Metab. 2005, 25, 391−401. (60) Peng, L.; Zhang, X.; Leif, H. Brain Res. 1994, 663, 168−172. (61) Hashimoto, M.; Takeda, Y.; Sato, T.; Kawahara, H.; Nagano, O.; Hirakawa, M. Brain Res. 2000, 872, 294−300. (62) Darbin, O.; Carre, E.; Naritoku, D.; Risso, J. J.; Lonjon, M.; Patrylo, P. R. Brain Res. 2006, 1116, 127−131. (63) Zheng, L.; Godfrey, D. A.; Waller, H. J.; Godfrey, T. G.; Chen, K.; Sun, Y. Neurochem. Res. 2000, 25, 823−835. (64) Wang, M.; Slaney, T.; Mabrouk, O.; Kennedy, R. T. J. Neurosci. Methods 2010, 190, 39−48. (65) Marion, D. W. Traumatic Brain Injury; Thieme: New York, NY, 1998. (66) Kimelberg, H. K.; Bowman, C.; Biddlecome, S.; Bourke, R. S. Brain Res. 1979, 177, 533−550.

I


Enzyme-Immobilized 3D-Printed Reactors for Online Monitoring of Rat

Recommend Documents