English

• Cell secretions play a crucial role in cell metabolism, serving as vital products that contribute to the intricate regulatory mechanisms within cells and facilitate communication between cells, ranging from ions, micromolecules, nucleic acids, and proteins to even micro-scale extracellular vesicles [1,2]. Recently, the investigation of cell secretion and its function has aroused great interest of researchers, which has led to the development of various detection techniques for cell secretion biomarkers [3,4]. Among these techniques, nanochannel technology has found extensive applications in cell biology and life sciences due to its label-free properties, ease of sample processing, high spatial and temporal resolution, and high sensitivity [5–7]. By combining solid-state nanochannels with DNA technology through the precise structural design of nanochannels and the molecular design of DNA, many strategies have achieved accurate detection of proteins, peptides, DNA, RNA, etc. [8–10]. Jiang et al. developed a nanofluidic sensor to detect DNA and ATP (adenosine triphosphate) in intricate mixtures by incorporating DNA supersandwich structures into the cylindrical PET nanochannels [11]. Ensinger and colleagues constructed a biosensor based on conical PET nanochannels for the unmarked identification of phosphoprotein (PPn). They utilized a Zn²⁺ chelator fixed on the nanochannel's inner surface as a ligand to specifically bind with phosphoproteins [12]. Existing nanochannel sensors mainly focused on investigating the impact of the inner walls of nanochannel platforms and the alteration of probes situated on these inner walls. This strategy necessitates the biomarkers enter the nanochannels for subsequent detection, which not only imposes restrictions on the size of the targets but may also be subject to non-specific blockages by interfering molecules, leading to the generation of erroneous signals, especially in complex biological systems [13,14].

  In recent years, a series of solid-state nanochannels with functional outer surfaces have been designed to enhance the sensitivity, specificity, and interference resistance of sensors [15–17]. Through the manipulation of the functional components on the outer surface of solid-state nanochannels, precise regulations over the symmetry of the channel structure, charge distribution, and the hydrophilic and hydrophobic properties have been achieved [18–20]. This outer surface detection strategy, which is capable of detecting large analytes, utilizes a sensing interface with high surface accessibility, allowing for the effective capture of various analytes from the bulk solution. In comparison to probes on the inner wall, probes on the outer surface are easier to modify and characterize, making them particularly suitable for in-situ analysis of large items such as proteins, viruses, and cells. Consequently, a range of highly sensitive and specific sensing systems have been engineered for diverse biological markers [21,22]. Lou and colleagues developed a sensing platform capable of distinguishing folded and unfolded protein states based on an outer surface modified nanochannel array by maleimide tetrastylanyl (MI-TPE) probes [23]. Xia et al. proposed a strategy for analyzing the cancer target matrix metalloproteinase-2 (MMP-2) based on the variation of wettability of the outer surface of nanochannels, which was immobilized with an amphiphilic peptide probe (CRRRR) composed of hydrophilic units [24]. However, the function of these biosensors was single and only single target was analyzed, which was far enough for the precise diagnosis of disease.

  Recently, scientists have proposed hybrid membranes with nanochannel-ionchannel geometric asymmetry structure to obtain the unique characteristic of ionic current rectification (ICR) [25–27]. The asymmetrical chemical composition, channel structure, and surface charge polarization endow the hybrid membrane with excellent ion selectivity, which exhibits superior surface charge-controlled ion transport behavior [28–32]. Numerous studies have reported the use of hybrid membranes for osmotic energy harvesting and bioanalysis [33–40]. Aptamers can bind to a great variety of targets such as small molecules, proteins, cells, and even tissues with high affinity and selectivity, making it possible to detect multiple biomarkers of varying sizes.

  Inspired by the envisage, herein, we developed an ordered mesoporous carbon-silicon/anodic aluminum oxide hybrid membrane (MCS/AAO) with outer surface probes for in situ detecting living cells released secretions with a wide size range (from nano-scale to micron-scale). As shown in Scheme 1, MCS/AAO was fabricated with an interface super-assembly strategy, then aptamers were modified on the outer surface as capture probes (MCS/AAO-aptamer). Ascribing to the asymmetric nanochannel structure and charge distribution, MCS/AAO-aptamer exhibited cation selectivity and high ionic current rectification up to 29.21, which was approximately 4.5-fold higher than that reported in the literature [15]. ATP, VEGF (vascular endothelial growth factor), and HepG2-MVs (micro vesicles) were chosen as model secretions representing different sizes: ATP represents micromolecules, VEGF represents macromolecules, and HepG2-MVs represent micro vesicles. After the living cells' secretions were specifically captured by the outer surface probes, the ion transport behaviors changed. By real time monitoring the ion current through hybrid membrane using the homemade device, different cell secretions could be selectively and sensitively detected. The detection limits were 0.64 fmol/L for ATP, 3.31 fg/mL for VEGF, and 5.37 × 10⁴ particles/mL for HepG2-MVs. Furthermore, the C/Si two-component interface super-assembly of the MCS layer provided the MCS/AAO hybrid membrane with exceptional mechanical stability, the detection interface could be regenerated at least 5 times with unchanged ionic transport performance. The prepared hybrid membrane with outer surface probes is promising and has potential for future applications in clinical diagnosis and target screening.

  Scheme 1

  

  Scheme 1.  Schematic illustration of the preparation and modification of MCS/AAO hybrid membrane, which was used for in situ and noninvasive monitoring of cell secretions with different sizes (ATP, VEGF and MVs).

  DownLoad: Full-Size Img PowerPoint

  MCS/AAO hybrid membrane was established by a previously reported interface super-assembly strategy (Fig. S1 in Supporting information) [28]. AAO membrane with regular channels of 70 ± 20 nm in diameter was used as substrate (Fig. 1a). The AAO membrane was coated with the MCS precursor solution using spin coating. Following evaporation-induced self-assembly (EISA), the resol (carbon precursor) and tetraethyl orthosilicate (TEOS, silica precursor) underwent co-assembly with the F127 micelles, resulting in the formation of a MCS membrane containing fine pores (Fig. 1b). The Brunauer-Emmett-Teller (BET) was used to calculate the surface area and pore volume to be as high as 491 m²/g and 0.34 cm³/g, respectively, indicating the uniform mesopores of the MCS membrane (Fig. S2a in Supporting information). There was a narrow pore size distribution and the pore size calculated from the adsorption branch was ~3.77 nm (Fig. S2b in Supporting information). MCS membranes with different carbon-silica ratios were obtained by adjusting the ratio of Resol to TEOS in precursor solutions (Table S1 in Supporting information), which would affect the charge distribution in the membrane. The cross-sectional structure of final MCS/AAO membrane was displayed in Fig. 1c, Figs. S3 and S4 (Supporting information), a 3.78 ± 0.5 µm-thick MCS membrane was densely grown on the and 55 ± 5 µm-thick AAO membrane. In Fig. 1d, the scanning transmission electron microscope (STEM) and energy dispersive spectrometer (EDS) mapping spectroscopy analysis exhibited that the elements of carbon, oxygen, and silicon were evenly dispersed in the MCS. This observation suggested the robust property of the co-assembly process on the AAO substrate. Transmission electron microscope (TEM) image of MCS along the [110] zone axis (Fig. 1e) and the [001] zone axis (Fig. 1f), signifying its 2D hexagonal channel structure. The small angle X-ray scattering (SAXS) pattern apparently displayed three peaks, corresponding to the 10, 11 and 20 reflections of a 2D hexagonal crystal structure with the space group of P6 mm [28], further proving the presence of a well-ordered pore structure in the MCS, which could offer a network of nanochannels essential for efficient ion transport (Fig. 1g). The TEM image in Fig. 1h also revealed distinct regular mesoporous hexagonal structures and the pore size measured was ~3.7 nm, in accordance with the pore size value calculated from BET (Fig. S2). The Fourier transform infrared (FT-IR) pattern showed that there was a hydroxyl peak at ~3400 cm^-1, originating from the Si-OH and phenolic hydroxyl groups. The peak at ~1075 cm^-1 corresponded to the strong vibration of the Si-O-Si bond. Peaks at ~1607 cm^-1 and ~1475 cm^-1 were attributed to the carbonyl (-COOH) and skeletal vibration of benzene, respectively (Fig. 1i). We deduced that the carbonyl groups on the surface of MCS was attributed to the oxidative transformation of Resol during the process of pyrolysis. The existence of hydroxyl and phenolic hydroxyl groups improved the mechanical stability and water stability of the MCS/AAO hybrid membrane.

  Figure 1

  

  Figure 1.  Characterization of the MCS/AAO hybrid membrane. SEM image of the (a) bare AAO surface. SEM images of (b) the MCS membrane and (c) cross-section of MCS/AAO hybrid membrane. (d) STEM image and EDS mapping of the element distribution on the surface of the MCS membrane. TEM image of MCS membrane film along (e) the [110] zone axis and (f) the [001] zone axis. (g) SAXS pattern of the MCS membrane. (h) TEM image of the MCS membrane of MCS/AAO hybrid membrane with two-dimensional (2D) hexagonal symmetry structure. Inset: the magnified image of the MCS membrane. (i) The FT-IR spectra of MCS membrane.

  DownLoad: Full-Size Img PowerPoint

  Above results showed that the MCS/AAO hybrid membrane was successfully prepared, which had asymmetric structure, including asymmetric channel size and membrane thickness.

  In order to specifically capture the targets, the outer surface of MCS/AAO was modified with the corresponding aptamer (MCS/AAO-aptamer). All DNA aptamer sequences were displayed in Table S2 (Supporting information), and the modification process was shown in Fig. S5 (Supporting information). CLSM images and EDS spectra demonstrated that the aptamer was successfully modified on the MCS/AAO surface (Figs. S6 and S7 in Supporting information). As we known, the ion transport capability of nanochannel was affected by shape, charge and wettability. The asymmetric structure of the MCS/AAO hybrid membrane was shown in Fig. 1. So before studying ion transport behavior, the charge distribution and wettability of the hybrid membrane were firstly investigated. The solid zeta potential measurements were conducted (Fig. 2a). It was observed that the existence of amino groups on the surface of anodic aluminum oxide imparted a positive charge, whereas the carboxyl groups on the surface of MCS and modified aptamer contributed to a substantial negative charge. This finding showed that the hybrid membrane had asymmetric charge polarity distribution (the MCS layer was negatively charged, while the AAO layer was positively charged). The contact angles of AAO, MCS/AAO and MCS/AAO-aptamer were presented in Fig. 2b, compared with AAO and MCS/AAO, MCS/AAO-aptamer exhibited increased hydrophilicity due to the introduction of amino and carboxyl groups, which can promote the ion permeation capability.

  Figure 2

  

  Figure 2.  Electrochemical detection and wettability characterization of the MCS/AAO hybrid membrane. (a) The zeta potential of the bare AAO, AAO—NH₂, MCS/AAO, and MCS/AAO-aptamer. (b) The contact angle of the bare AAO, MCS/AAO and MCS/AAO-aptamer. (c) The ionic current curves of the bare AAO, MCS/AAO and MCS/AAO-aptamer, n = 5. (d) Schematic illustration of charge distribution of the MCS/AAO hybrid membrane. CV measurements of the bare AAO (black curve) and MCS/AAO hybrid membrane (red curve) electrode in 10 mmol/L KCl containing 0.5 mmol/L (e) Ru(NH₃)₆³⁺ and (f) Fe(CN)₆³⁻. The scan rate was 50 mV/s.

  DownLoad: Full-Size Img PowerPoint

  Then the ion transport behavior of the modified hybrid membrane was investigated with a homemade electrochemical cell (Fig. S8 in Supporting information), and the ionic current was recorded by Ag/AgCl electrodes. An ohmic linear I-V curve was observed for the bare AAO. While for MCS/AAO and MCS/AAO-aptamer, the I-V displayed a diode-like curve attributed to asymmetric channel structure and surface charge distribution (Fig. 2c). The ICR mechanism of the hybrid membrane was illustrated schematically in Fig. 2d [18,28,41]. Under pH 7.2 environment, the amine-functionalized AAO was expected to carry a positive charge due to the pK_a of the primary amine group being approximately 10.0. When the direction of electric field was from top (MCS side) to bottom (AAO side), cations (K⁺) were migrated from top to bottom, and anions (Cl⁻) were from bottom to top. K⁺ can freely traverse the cation-selective MCS layer due to electrostatic attraction, Cl⁻ would be hindered by MCS and store within the AAO nanochannels. To keep electrically neutral, some K⁺ ions remained to balance the presence of Cl⁻ ions, leading to an increase of ionic strength. This elevated ion concentration inside the AAO nanochannels resulted in a decrease in the thickness of the electrical double layer (EDL), enabling more K⁺ ions can pass through the MCS/AAO-aptamer membrane without hindrance. Consequently, the current of K⁺ (positive voltage region) was higher than Cl⁻ (negative voltage region). A triangle wave voltage was also applied to the hybrid membrane. The nonzero cross-point in Fig. S9 (Supporting information) indicated that the I-V curve under this periodic voltage resulted from the effect of surface charge in an asymmetric channel, according to Chua's theory.

  The cation selectivity of MCS/AAO-aptamer was further confirmed by cyclic voltammetry (CV) with three-electrode system (Fig. S10 in Supporting information), Ru(NH₃)₆³⁺ and Fe(CN)₆^3- as the cationic and anionic electrochemical probes, respectively. The redox peaks of Ru(NH₃)₆³⁺ and Fe(CN)₆^3- could be observed in the AAO membrane (Figs. 2e and f), showcasing that both anions and cations could successfully pass through the channels of bare AAO. In comparison to the bare AAO, the CVs of the MCS/AAO hybrid membrane exhibited an observable redox current for the Ru(NH₃)₆³⁺ cation, while no redox current was observed for the Fe(CN)₆^3- anion. This demonstrated that this hybrid membrane selectively permitted the passage of the cation Ru(NH₃)₆³⁺ through its channels, indicating the cation permselectivity of MCS, which endowed the MCS/AAO-aptamer hybrid membrane high ion rectification ratio (ICR).

  To obtain the highest ICR, a series of optimization experiments were conducted. The ionic current curves of MCS/AAO membranes at different carbon-silica ratios and the diameter of AAO channel were shown in Figs. S11 and S12 (Supporting information), and Table S1. The MCS/AAO-4 under a TEOS/Resol ratio of 1:5 was selected as the upper layer material, and 70 ± 20 nm AAO was selected as the lower layer material. As shown in Fig. S13 (Supporting information), the presence a barrier layer (PMMA) can improve the rectification effect. ICR ratios of the MCS/AAO hybrid membrane at different concentrations of PBS (phosphate buffer solution) were measured (Fig. S14 in Supporting information). The results indicated that MCS/AAO membrane could still maintain a high ICR under the conditions of cell incubation (10 mmol/L PBS). During the process of tumor cell growth, a large amount of lactic acid was produced, leading to an acidic tumor microenvironment. This acidic characteristic was considered to be one of the important factors contributing to tumor growth and metastasis. This study further investigated how pH fluctuations in the extracellular matrix influence the ion rectification performance of MCS/AAO, ICR ratios of MCS/AAO membrane at different pH was observed (Fig. S15 in Supporting information). The rectification ratio was not affected when the pH value was in the range of 6.0–7.6, and it was affected when the pH value fluctuated greatly (pH > 11 or pH < 3). Notably, the ICR of MCS/AAO-aptamer was as high as 29.21, far exceeding the level of non-hybrid membranes reported in previous literature, with an astonishing increase by approximately 4.5-fold (Table S3 in Supporting information). Thus, the MCS/AAO-aptamer hybrid membrane has made a huge breakthrough in the ionic current response range, opening up unlimited possibilities for quantitative detection and cell analysis.

  Under the optimal conditions, we investigated the detection ability of cell secretions. In our system, detection targets were captured on the outer surface of hybrid membrane, which had unique advantages: (1) The size of the target was not limited by the nanochannel size; (2) The outer surface modification was straightforward and the characterization was precise; (3) Heightened sensitivity, greater specificity, and improved resistance to interference. To verify the feasibility of the constructed hybrid membrane for detection cell secretions of different sizes, ATP, VEGF, and MVs (micromolecule, macromolecule, and micro vesicle) were used as model secretions. The customized electrochemical device filled with 10 mmol/L PBS solution (pH 7.2) and measured the ionic current by scanning voltage from –2 V to +2 V. For ATP detection, aptamer corresponding to ATP was modified on the outer surface of hybrid membrane [42,43]. With increasing amounts of ATP, the trapped ATP molecules covered the channel entrance of MCS efficiently, which will dramatically block the ionic flow through channels, resulting in a decreased ionic current at +2 V (Fig. 3a). The ICR (|I_{+2 V}/I_{−2 V}|) versus ATP concentration (1 fmol/L to 1 nmol/L) was shown in Fig. 3b, the linear curve was established by plotting the ICR ratio versus –lgC (C was the concentration of ATP). The linear regression equation was ICR = –2.78 lgC – 12.89 with a good correlation coefficient of 0.99, and the detection limit was 0.64 fmol/L (3σ/S). To evaluate the specificity of the MCS/AAO-aptamer nanodevice for cell secretions detection, several other secretions were employed to perform the same measurements including bovine serum albumin (BSA), VEGF, mucin 1 (MUC1) and glutathione (GSH). There was no significant change in the ICR, validating that only ATP could induce the subsequent reactions (Fig. 3c). Besides, a linear correlation (R² = 0.99) between the results from our platform and commercial luminescent ATP detection assay kit was obtained, validating the precision of the platform for ATP quantitative detection (Fig. S16 in Supporting information). Moreover, the sensitivity of our platform was over 10-fold higher than that of the commercial ATP assay kit. The hybrid membranes had excellent stability, which was assessed by intra-assays (Fig. S17 in Supporting information). The relative standard deviation (RSD) value for detecting ATP five times using the same MCS/AAO-aptamer was 2.6%, revealing acceptable stability. The detection platform could also realize the specific analysis of VEGF just through replacing the VEGF aptamer [44,45]. The calibration equation was ICR =1.59 lgC – 3.85 (C was the VEGF concentration in solution) with R² = 0.97, the detection limit was 3.31 fg/mL (Figs. S18 and S19 in Supporting information). Moreover, we established a strong linear correlation (R² = 0.99) between the outcomes obtained from our device and those from a commercial ELISA VEGF measurement kit, thus confirming the precision of our platform in quantitatively detecting VEGF (Fig. S20 in Supporting information). Compared with other VEGF detection methods (Table S4 in the Supporting information), the proposed platform exhibited a lower detection limit without the need for signal labeling and signal amplification strategy.

  Figure 3

  

  Figure 3.  The quantitative detection of cell secretion ATP. (a) I-V properties of the MCS/AAO-aptamer treated with different concentrations of ATP solution, n = 5. (b) Recetification ratio of the MCS/AAO-aptamer toward increasing ATP concentrations. Inset: The linear relationship between the ion rectification ratio and ATP concentration, n = 5. (c) Selectivity assessment of this MCS/AAO-aptamer toward ATP, BSA, MUC1, GSH, and VEGF. (d) The ATP release rate ((R₀–R)/R₀) of three cell lines (HepG2 cells, LX2 cells, and A549 cells) within 72 h. R₀ was the ICR of MCS/AAO-aptamer at 0 h, R was the ICR of MCS/AAO-aptamer treated with different cells at different times. (e) The ionic current curves of the MCS/AAO-aptamer after the HepG2 cells treated with PBS, 2, 4-dinitrophenol (DNP, ATP inhibitor), and glucose (ATP inducer), n = 5. (f) The concentrations of ATP secreted from HepG2 cells induced by PBS, DNP, and glucose was calculated by calibration equation, n = 5. ***P < 0.001, **P < 0.01, P < 0.05.

  DownLoad: Full-Size Img PowerPoint

  The viability of cells in this device was assessed using fluorescence microscopy. Intense green fluorescence and almost invisible red fluorescence were observed by FDA/PI double staining, indicating that the cells maintained high viability after insertion and removal from the chamber (Fig. S21 in Supporting information). For in situ monitoring of ATP released from living cells, three different types of cells (HepG2 cells (human hepatocellular carcinomas cells), A549 cells (non-small cell lung cancer cells) and LX2 cells (hepatic stellate cells)) were used. Cells were incubated on the surface of the PET membrane for 24 h, then the PET membranes were rolled up and inserted into the chamber of a self-made device (Fig. S22 in Supporting information). When the target was released from the living cell, it specifically combined with aptamer on the surface of the MCS/AAO, leading to the change of transmembrane ionic current signals. In the following, a kinetic study of the MCS/AAO-aptamer nanochannels in living cells was carried out. The I-V curves recorded the ATP concentration secreted from HepG2, A549, and LX2 cells within 72 h (Fig. S23 in Supporting information). The change in the ICR ratio (R₀-R/R₀) of the MCS/AAO-aptamer membrane reflected the rate of ATP secretion from cells in the left chamber. The ATP release rate curves presented differences and reached a plateau after 48 h, during the monitoring of HepG2 cells. Similarly, the ATP secretion of A549 cells and LX2 cells showed a trend of first fast and then slow, but it was lower than that of HepG2 cells at all stages, especially in normal cells where LX2 increased the slowest (Fig. 3d). Glucose and 2, 4-dinitrophenol (DNP) as ATP inducer and inhibitor can promote and inhibit ATP secretion, respectively. When the culture medium (control), Glucose and DNP were added in the left chamber for 2 h, the current change of secreted VEGF was monitored with nanochannel electrochemistry. Compared with the control group, the glucose treatment group showed a significant decrease in ion current rectification with an increase in ATP release. On the contrary, the ICR of the DNP treatment group increased, indicating that DNP has an inhibitory effect on ATP secretion (Figs. 3e and f). Rates of VEGF secretion and drug effects were also monitored in the three cell types (Figs. S24–26 in Supporting information). The results demonstrated that the MCS/AAO-aptamer hybrid membrane could accurately in situ analyze the secretion of various cells, indicating its potential for bioanalysis and clinical detection.

  Except for micromolecule (e.g., ATP) and macromolecule (e.g., VEGF), we evaluated whether the cancerous micro vesicles (MVs, 100–1000 nm) can be quantitatively detected with the MCS/AAO hybrid membrane. Cancerous MVs are lipid bimolecular membrane structures released from tumor cells into the extracellular substance and carry a large number of bioinformation such as proteins, nucleic acids, and lipid molecules. Cancerous MVs have been widely considered as potential biomarkers for various types of cancer. The separation and purification of HepG2-MVs from cell culture supernatant were carried by using a multi-step ultracentrifugation method. As shown in the TEM image (Fig. 4a), MVs displayed irregular circular shapes with an intact membrane structure, resembling the morphology and particle size reported in previous studies. Nanoparticle tracking analysis (NTA) characterization revealed that the concentration of MVs was 1.2 × 10¹¹ particles/mL and the diameter distribution was 40–750 nm with an average of 137.4 nm (Fig. 4b). For the efficient capture of HepG2-MVs, we utilized a dual aptamer modification strategy, aptamer-TLS11A and aptamer-EpCAM were co-modified on the outer surface of MCS/AAO [46–48]. HepG2-MVs in a wide range of concentrations from 1 × 10⁵ particles/mL to 5 × 10⁷ particles/mL could be effectively detected employing the dual-aptamer MCS/AAO membrane (Fig. 4c). There was a good linear correlation between the rectification ratio and lgC (C was the concentration of the HepG2-MVs) with a linear equation ICR = –2.82 lgC + 28.82 and R² = 0.97 (Fig. 4d), the limit of detection was 5.37 × 10⁴ particles/mL. The relative standard deviation of five assays of HepG2 MVs sample by the MCS/AAO-aptamer was 3.5%, indicating acceptable stability and reproducibility of the device (Fig. S27 in Supporting information). In order to in situ monitor the secretion of MVs from HepG2 cells, HepG2 cells were seeded in the left chamber and then incubated with the dual-aptamer MCS/AAO membrane for 72 h. The cumulative concentration of MVs secreted from HepG2 cells was about 5.063 × 10⁵ within 72 h, and 90% of them were completed within 24 h (Figs. 4e and f). SEM images recorded the surface morphology of the MCS/AAO membrane before and after capturing the HepG2-MVs. It could be clearly seen that the uniform flat membrane became bumpy after trapping the HepG2-MVs (Fig. S28 in Supporting information). To observe the captured MVs clearly, the HepG2-MVs were stained by DiO dye (green fluorescence) and imaged by confocal laser scanning microscopy (CLSM). As shown in Fig. S29 (Supporting information), the HepG2-MVs were successfully trapped on the outer surface of the dual-aptamer MCS/AAO membrane.

  Figure 4

  

  Figure 4.  Detection of MVs secreted by HepG2 cells. (a) The TEM images of MVs secreted by HepG2 cells. (b) The MVs concentration and size distribution determined by NTA, mean concentration was 1.2 × 10¹¹ particles/mL and the size distribution was between 40 nm and 750 nm with an average of 137.4 nm. (c) I-V properties of the MCS/AAO-aptamer hybrid membranes treated with different MVs concentrations, n = 5. (d) Recetification ratio of the MCS/AAO-aptamer toward increasing MVs concentrations. Inset: The linear relationship between the ion rectification ratio and MVs concentration, n = 5. (e) I-V curves of the MCS/AAO-aptamer membranes under different secretion times, n = 5. (f) The concentrations of MVs secreted from HepG2 cells was calculated by calibration equation, n = 5.

  DownLoad: Full-Size Img PowerPoint

  Above all, the prepared MCS/AAO hybrid membrane exhibited excellent ability for quantitative in situ detection of cell secretions with a wide size range from micromolecules to micro vesicles.

  The prepared MCS (mesoporous carbon-silicon) membrane contained a large number of hydroxyl, phenolic hydroxyl, and carboxyl groups, which endow the MCS with super-assembly interactions with AAO (anodic aluminum oxide), enhancing the stability of the MCS/AAO hybrid membrane and providing conditions for its reuse. Recyclable detection platforms can not only reduce detection costs but also offer more dependable data support for drug development and biomarker screening. For green chemistry, we tested the renewal performance of the MCS/AAO-aptamer membrane. The captured detection targets on the outer surface of the MCS/AAO hybrid membrane were cleaned by concentrated sulfuric acid and then re-modified by new aptamer. The rectification ratio could be restored to over 96% of the initial value after clearing the surface, and it remained unchanged after modifying the new aptamer DNA strands. (Figs. 5a and b). X-ray photon-electron spectrometer (XPS) analysis results presented that the phosphorus element was meaningfully reduced, confirming the cleaned surface of MCS/AAO (Fig. 5c). According to the results of multiple measurements, the MCS/AAO hybrid membrane could be reused at least 5 times while maintaining its original ion transport performance (Fig. 5d).

  Figure 5

  

  Figure 5.  The regeneration of the MCS/AAO-aptamer. (a) I-V properties of the MCS/AAO, MCS/AAO-aptamer, hybridization, dehybridization, and remodified aptamer. (b) The rectification ratio during the regeneration of the MCS/AAO-aptamer interface. (c) XPS analysis of the surface of MCS/AAO before and after cleaning the DNA aptamer. (d) Stability and reversibility of the MCS/AAO-aptamer. Changes in ICR before and after several cycles of MCS/AAO membrane remodification with aptamer.

  DownLoad: Full-Size Img PowerPoint

  In summary, we have successfully constructed an ordered mesoporous carbon-silicon/AAO hybrid membrane (MCS/AAO) with outer surface probes for in situ detecting living cells released secretions with a wide size range (from nano-scale to micron-scale). The hybrid membrane exhibited cation selectivity and high ionic current rectification, ascribing to asymmetric nanochannel structure and charge distribution. Leveraging the unique mass transfer properties of the MCS/AAO hybrid nanochannels and the specific binding ability between aptamers and corresponding cell secretions, this platform successfully captured and quantitatively analyzed cell secretions of varying sizes, including micromolecules, macromolecules, and micro vesicles. The experimental results indicated that this platform can achieve in situ quantitative detection of ATP, VEGF, and HepG2-MVs in the range of 1 fmol/L ~ 100 pmol/L, 10 fg/mL ~ 1 ng/mL, 1 × 10⁵ ~ 5 × 10⁷ particles/mL, respectively. The detection limits of ATP, VEGF, and HepG2-MVs were 0.64 fmol/L, 3.31 fg/mL, 5.37 × 10⁴ particles/mL. Moreover, the prepared hybrid membrane had exceptional mechanical stability, the detection interface could be regenerated at least 5 times, which significantly reduced the detection cost. This strategy offered a novel and promising platform for the multi-target quantitative detection of diverse cell secretions, presenting good practicality and enormous potential in early clinical diagnosis and biomarker screening.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Zixuan Chen: Methodology. Yafeng Wu: Project administration. Zhaoyan Tian: Data curation. Zhaohan Wang: Formal analysis. Weiwei Liu: Investigation. Songqin Liu: Funding acquisition.

  Acknowledgments

  This work is supported by National Natural Science Foundation of China (Nos. 22374019, 22174016), Start-up Research Fund of Southeast University (No. RF1028624037).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.110917 .

  
  
• 1. [1]

     M.G. Vander Heiden, R.J. DeBerardinis, Cell 168 (2017) 657–669. doi: 10.1016/j.cell.2016.12.039

  2. [2]

     F. Soldner, R. Jaenisch, Cell 175 (2018) 615–632. doi: 10.1016/j.cell.2018.09.010

  3. [3]

     S.S. Cheng, Q.Q. Kong, X.Y. Hu, C.L. Zhang, Y.Z. Xian, Anal. Chem. 94 (2022) 1085–1091. doi: 10.1021/acs.analchem.1c04122

  4. [4]

     Z.Y. Tian, X. Qin, F.Y. Shao, et al., Chin. Chem. Lett. 34 (2023) 107656. doi: 10.1016/j.cclet.2022.06.079

  5. [5]

     Y.L. Ying, Y.T. Long, J. Am. Chem. Soc. 141 (2019) 15720–15729. doi: 10.1021/jacs.8b11970

  6. [6]

     J.L. Hao, W.J. Wang, J.W. Zhao, et al., Chin. Chem. Lett. 33 (2022) 2291–2300. doi: 10.1016/j.cclet.2021.10.011

  7. [7]

     T.Y. Xiong, C.W. Li, X.L. He, et al., Science 379 (2023) 156–161. doi: 10.1126/science.adc9150

  8. [8]

     N.N. Liu, Y.N. Jiang, Y.H. Zhou, et al., Angew. Chem. Int. Ed. 52 (2013) 2007–2011. doi: 10.1002/anie.201209162

  9. [9]

     J.W. Shen, Y.B. Li, H.S. Gu, et al., Chem. Rev. 114 (2014) 7631–7677. doi: 10.1021/cr300248x

  10. [10]

      Z.M. Zhang, Y.L. Zhu, X. Ma, et al., Chem. Eng. J. 485 (2024) 149912. doi: 10.1016/j.cej.2024.149912

  11. [11]

      Y.N. Jiang, N.N. Liu, W. Guo, et al., J. Am. Chem. Soc. 134 (2012) 15395–15401. doi: 10.1021/ja3053333

  12. [12]

      M. Ali, S. Nasir, P. Ramirez, et al., J. Phys. Chem. C 117 (2013) 18234–18242. doi: 10.1021/jp4054555

  13. [13]

      H. Zeng, S. Zhou, L. Xie, et al., Anal. Chem. 94 (2022) 2589–2596. doi: 10.1021/acs.analchem.1c04825

  14. [14]

      Y.A. Huang, H. Zeng, L. Xie, et al., J. Am. Chem. Soc. 144 (2022) 13794–13805. doi: 10.1021/jacs.2c04862

  15. [15]

      W.Q. Zhang, Y.D. Tu, H. Liu, et al., Angew. Chem. Int. Ed. 63 (2024) e20231643.

  16. [16]

      P.C. Gao, Q. Ma, D.F. Ding, et al., Nat. Commun. 9 (2018) 4557–4567. doi: 10.1038/s41467-018-06873-z

  17. [17]

      L. Dai, W.Q. Zhang, D.F. Ding, et al., ACS Nano 18 (2024) 7677–7687. doi: 10.1021/acsnano.3c12270

  18. [18]

      F.F. Liu, X.P. Zhao, X.W. Liao, et al., Anal. Chem. 92 (2020) 5509–5516. doi: 10.1021/acs.analchem.0c00330

  19. [19]

      P.C. Gao, D.G. Wang, C. Che, et al., Nat. Protoc. 16 (2021) 4201–4226. doi: 10.1038/s41596-021-00574-6

  20. [20]

      X.C. Li, T.Y. Zhai, P.C. Gao, et al., Nat. Commun. 9 (2018) 40–49. doi: 10.1038/s41467-017-02447-7

  21. [21]

      W.W. Zhang, M.Y. Chen, Q. Ma, et al., Anal. Chem. 96 (2024) 2445–2454. doi: 10.1021/acs.analchem.3c04413

  22. [22]

      Q. Ma, Y. Li, R.S. Wang, et al., Nat. Commun. 12 (2021) 1573–1582. doi: 10.1038/s41467-021-21507-7

  23. [23]

      Y.J. Qiao, J.J. Hu, Y.X. Hu, et al., Angew. Chem. Int. Ed. 62 (2023) e20239671.

  24. [24]

      J.J. Hu, W.L. Jiang, Y.J. Qiao, et al., ACS Nano 17 (2023) 11935–11945. doi: 10.1021/acsnano.3c04017

  25. [25]

      W.X. Lv, H.C. Ye, Z.Q. Yuan, et al., TrAC Trends Anal. Chem. 123 (2020) 115767. doi: 10.1016/j.trac.2019.115767

  26. [26]

      Y.J. Tang, K. Cheng, L.F. Tan, et al., Chem. Eng. J. 463 (2023) 142438. doi: 10.1016/j.cej.2023.142438

  27. [27]

      Y.C. Liu, L.H. Yeh, M.J. Zheng, et al., Sci. Adv. 7 (2021) eabe9924. doi: 10.1126/sciadv.abe9924

  28. [28]

      S. Zhou, L. Xie, X.F. Li, et al., Angew. Chem. Int. Ed. 60 (2021) 26167–26176. doi: 10.1002/anie.202110731

  29. [29]

      T.L. Xiao, J. Ma, Z.Y. Liu, et al., J. Mater. Chem. A 8 (2020) 11275–11281. doi: 10.1039/d0ta02162h

  30. [30]

      J. Gao, W. Guo, D. Feng, et al., J. Am. Chem. Soc. 136 (2014) 12265–12272. doi: 10.1021/ja503692z

  31. [31]

      D. Feng, Y.Y. Lv, Z.X. Wu, et al., J. Am. Chem. Soc. 133 (2011) 15148–15156. doi: 10.1021/ja2056227

  32. [32]

      X. Zhang, S. Zhou, Y.J. He, et al., Anal. Chem. 96 (2024) 13971–13979. doi: 10.1007/s11071-024-09784-5

  33. [33]

      W.W. Xin, Z. Zhang, X.D. Huang, et al., Nat. Commun. 10 (2019) 3876–3885. doi: 10.1038/s41467-019-11792-8

  34. [34]

      X.B. Zhu, J.R. Hao, B. Bao, et al., Sci. Adv. 4 (2018) eaau1665. doi: 10.1126/sciadv.aau1665

  35. [35]

      H. Zeng, S. Zhou, L. Xie, et al., Biosens. Bioelectron. 222 (2023) 114985. doi: 10.1016/j.bios.2022.114985

  36. [36]

      X. Sui, Z. Zhang, C. Li, et al., ACS Appl. Mater. Interfaces 11 (2019) 23815–23821. doi: 10.1021/acsami.8b02578

  37. [37]

      L. Cao, I.C. Chen, X.W. Liu, et al., ACS Nano 16 (2022) 18910–18920. doi: 10.1021/acsnano.2c07813

  38. [38]

      S.F. Pan, P. Liu, Q. Li, et al., Angew. Chem. Int. Ed. 62 (2023) e202218129. doi: 10.1002/anie.202218129

  39. [39]

      Z.J. Yang, L.H. Yeh, Y.H. Peng, et al., Angew. Chem. Int. Ed. 63 (2024) e202408375. doi: 10.1002/anie.202408375

  40. [40]

      A.R. Fauziah, C.W. Chu, L.H. Yeh, Chem. Eng. J. 452 (2023) 139244. doi: 10.1016/j.cej.2022.139244

  41. [41]

      C. Wang, X.P. Zhao, F.F. Liu, et al., Nano Lett. 20 (2020) 1846–1854. doi: 10.1021/acs.nanolett.9b05066

  42. [42]

      R. Mo, T.Y. Jiang, R. DiSanto, et al., Nat. Commun. 5 (2014) 3364. doi: 10.1038/ncomms4364

  43. [43]

      J. Zhao, J.H. Gao, W.T. Xue, et al., J. Am. Chem. Soc. 140 (2018) 578–581. doi: 10.1021/jacs.7b11161

  44. [44]

      S. Dehghani, R. Nosrati, M. Yousefi, et al., Biosens. Bioelectron. 110 (2018) 23–37. doi: 10.1016/j.bios.2018.03.037

  45. [45]

      J.L. Li, K.X. Sun, Z.P. Chen, et al., Biosens. Bioelectron. 89 (2017) 964–969. doi: 10.1016/j.bios.2016.09.078

  46. [46]

      D. Wu, Y. Yu, D. Jin, et al., Anal. Chem. 92 (2020) 4006–4015. doi: 10.1021/acs.analchem.9b05531

  47. [47]

      D. Zhang, Y.S. Zheng, Z.G. Lin, et al., Angew. Chem. Int. Ed. 59 (2020) 12022–12028. doi: 10.1002/anie.202002145

  48. [48]

      H.X. Zhang, Z.H. Wang, Q.X. Zhang, et al., Biosens. Bioelectron. 124 (2019) 184–190.

• 

  Scheme 1  Schematic illustration of the preparation and modification of MCS/AAO hybrid membrane, which was used for in situ and noninvasive monitoring of cell secretions with different sizes (ATP, VEGF and MVs).

  下载: 全尺寸图片 幻灯片
  

  Figure 1  Characterization of the MCS/AAO hybrid membrane. SEM image of the (a) bare AAO surface. SEM images of (b) the MCS membrane and (c) cross-section of MCS/AAO hybrid membrane. (d) STEM image and EDS mapping of the element distribution on the surface of the MCS membrane. TEM image of MCS membrane film along (e) the [110] zone axis and (f) the [001] zone axis. (g) SAXS pattern of the MCS membrane. (h) TEM image of the MCS membrane of MCS/AAO hybrid membrane with two-dimensional (2D) hexagonal symmetry structure. Inset: the magnified image of the MCS membrane. (i) The FT-IR spectra of MCS membrane.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Electrochemical detection and wettability characterization of the MCS/AAO hybrid membrane. (a) The zeta potential of the bare AAO, AAO—NH₂, MCS/AAO, and MCS/AAO-aptamer. (b) The contact angle of the bare AAO, MCS/AAO and MCS/AAO-aptamer. (c) The ionic current curves of the bare AAO, MCS/AAO and MCS/AAO-aptamer, n = 5. (d) Schematic illustration of charge distribution of the MCS/AAO hybrid membrane. CV measurements of the bare AAO (black curve) and MCS/AAO hybrid membrane (red curve) electrode in 10 mmol/L KCl containing 0.5 mmol/L (e) Ru(NH₃)₆³⁺ and (f) Fe(CN)₆³⁻. The scan rate was 50 mV/s.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The quantitative detection of cell secretion ATP. (a) I-V properties of the MCS/AAO-aptamer treated with different concentrations of ATP solution, n = 5. (b) Recetification ratio of the MCS/AAO-aptamer toward increasing ATP concentrations. Inset: The linear relationship between the ion rectification ratio and ATP concentration, n = 5. (c) Selectivity assessment of this MCS/AAO-aptamer toward ATP, BSA, MUC1, GSH, and VEGF. (d) The ATP release rate ((R₀–R)/R₀) of three cell lines (HepG2 cells, LX2 cells, and A549 cells) within 72 h. R₀ was the ICR of MCS/AAO-aptamer at 0 h, R was the ICR of MCS/AAO-aptamer treated with different cells at different times. (e) The ionic current curves of the MCS/AAO-aptamer after the HepG2 cells treated with PBS, 2, 4-dinitrophenol (DNP, ATP inhibitor), and glucose (ATP inducer), n = 5. (f) The concentrations of ATP secreted from HepG2 cells induced by PBS, DNP, and glucose was calculated by calibration equation, n = 5. ***P < 0.001, **P < 0.01, P < 0.05.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Detection of MVs secreted by HepG2 cells. (a) The TEM images of MVs secreted by HepG2 cells. (b) The MVs concentration and size distribution determined by NTA, mean concentration was 1.2 × 10¹¹ particles/mL and the size distribution was between 40 nm and 750 nm with an average of 137.4 nm. (c) I-V properties of the MCS/AAO-aptamer hybrid membranes treated with different MVs concentrations, n = 5. (d) Recetification ratio of the MCS/AAO-aptamer toward increasing MVs concentrations. Inset: The linear relationship between the ion rectification ratio and MVs concentration, n = 5. (e) I-V curves of the MCS/AAO-aptamer membranes under different secretion times, n = 5. (f) The concentrations of MVs secreted from HepG2 cells was calculated by calibration equation, n = 5.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  The regeneration of the MCS/AAO-aptamer. (a) I-V properties of the MCS/AAO, MCS/AAO-aptamer, hybridization, dehybridization, and remodified aptamer. (b) The rectification ratio during the regeneration of the MCS/AAO-aptamer interface. (c) XPS analysis of the surface of MCS/AAO before and after cleaning the DNA aptamer. (d) Stability and reversibility of the MCS/AAO-aptamer. Changes in ICR before and after several cycles of MCS/AAO membrane remodification with aptamer.

  下载: 全尺寸图片 幻灯片
•