A multifunctional chitosan-silica hybrid for fluorescent LBP detection and PA2G4-targeted gallbladder cancer therapy

2.1. Chemicals and measurements

All reagents were purchased from the market and utilized as it was. On a Nicolet Impact 750 spectrometer, Fourier transform-infrared (FT‐IR) spectra of KBr‐pellet samples were gathered, ranging from 400 to 4000 cm⁻¹. EA (N, H, C) was conducted on a Perkin‐Elmer CHN analyzer.

2.2. Preparation and characterization for [Cu(biimb)(BPDC)(H₂O)₂]· DMA·10H₂O (CP1) and compound 1

A mixture formed by 24 mg and 0.10 mmol Cu(NO₃)₂·3H₂O and 0.056 g and 0.20 mmol tib was dissolved in Dimethylacetamide (DMA)/H₂O (8 mL, 1:1, v/v), and three drops of concentrated HNO₃ were incorporated. The obtained mixture was placed in a stainless steel vessel lined with Teflon, heated to 130°C for 3 days, and subsequently cooled (at a rate of 5°C·h^-1) to room temperature (RT). Afterward, CP1’s blue block crystals were acquired. Yield of 53% (based on Cu). Anal. (%) calcd. For C₃₀H₂₈CuN₁₄O₈: N, 25.42; H, 3.66; C, 46.70. Found: N, 25.38; H, 3.62; C, 46.81.

In a one‐pot synthesis of compound 1, 1,3-cyclohexanedione (1.0 mmol, 140 mg) and 4-bromobenzaldehyde (1.0 mmol, 185 mg) were dissolved in 10 mL of ethanol, to which piperidine (0.1 mL) was introduced as a base. The mixture was refluxed for 60 min to induce Knoevenagel condensation, yielding the arylidene intermediate. Without isolation, methyl 3-aminocrotonate (1.0 mmol, 115 mg) and 0.05 mL of acetic acid were then incorporated, and reflux continued for 4 h to promote Michael addition and intramolecular cyclization to the 2H-pyran-2-one core. After cooling, the crude material was filtered, cleaned with cold ethanol, and dried after the reaction was concentrated to about 5 mL, and then put into 10 mL of water to precipitate the product. Recrystallization from ethanol afforded pure compound 1 (pale needles, 68% overall yield, 230 mg).

The SuperNova diffractometer was exploited to obtain the X-ray data. The intensity data were investigated through the CrysAlisPro software, which was later converted to HKL files. Initial structural models were formed by the SHELXS program with the direct method, and the SHELXL-2014 program, based on the least-squares method, was employed to modify them. Anisotropic parameters were employed to mix all non-H atoms. AFIX commands were exploited to fix the entire H-atoms geometrically on their matching C atom. The crystallographic parameters together with refinement details of CP1 and compound 1 have been displayed in Table 1.

2.3. Preparation and characterization for 1-CS-CPTMS@CP1@LBP

The 1-CS-CPTMS@CP1@LBP composite was prepared via a three-step protocol. First, chitosan (CS, 500 mg, Mw ≈100–300 kDa, DDA ≥85%) was dissolved in 50 mL Dimethyl sulfoxide (DMSO) at 60°C, then 200 mg of Compound 1, 300 mg of K₃PO₄, 50 mg of 3,4,7,8-tetramethyl-1,10-phenanthroline, and 20 mg of CuI were included, and the mixture was stirred at 60°C under nitrogen for 24 h. After filtration, washing (DMSO, ethanol), and drying, 1-CS (≈550 mg) was obtained. Next, 1-CS (500 mg) was dispersed in 40 mL DMSO, reacted with CPTMS (200 µL) and Et₃N (300 µL) at RT for 12 h, then filtered, washed, and dried to yield 1-CS-CPTMS (≈520 mg). Finally, after sonicating 200 mg of 1-CS-CPTMS in 20 mL of PBS (pH 7.4), 50 mg of CP 1 and 20 mg of LBP were added, and the mixture was gently stirred for 6 h. The product was gathered via centrifugation, cleaned using PBS, and lyophilized to afford 1-CS-CPTMS@CP1@LBP (≈240 mg).

2.4. Cell proliferation

SGC-996 and GBC-SD, human GBC cell lines, were provided by the Cell Bank of the Chinese Academy of Sciences. Cells were cultivated in RPMI-1640 medium (Gibco, USA) with 10% fetal bovine serum (FBS) (Gibco, USA), 100 μg mL^-1 streptomycin, and 100 U mL^-1 penicillin at 37°C and 5% CO₂ humidity. Cells were categorized into four groups (n = 3 for each group). Control group. LBP: 50 μg mL^-1 free LBP dissolved in medium. NPs: Blank Poly(lactic-co-glycolic acid) (PLGA) nanoparticles equivalent to 50 μg mL^-1 LBP (vehicle control). 1-CS-CPTMS@CP1@LBP: 1-CS-CPTMS@CP1@LBP at 50 μg mL^-1 LBP. Treatment solutions were added to serum-free medium. Cells were treated for 48 h, and then the CCK-8 assay (Dojindo, Japan) was carried out as per the manufacturer’s instructions.

2.5. qPCR

After 48 h of treatment, the extraction of total RNA was performed with TRIzol reagent (Invitrogen, USA) based on the manufacturer’s protocol. Via applying TransScript II Green One-Step qRT-PCR SuperMix (Transgen, China), qPCR was implemented. Relative expression of PA2G4 mRNA was measured with the 2⁻^ΔΔCt approach and standardized to GAPDH. Primer sequences, PA2G4-F: 5’-CCTGCTGAAGATGGTGAAGC-3’. PA2G4-R: 5’-GCTGGATGTTCTGGTTGAGG-3’. GAPDH-F: 5’-GGAGCGAGATCCCTCCAAAAT-3’. GAPDH-R: 5’-GGCTGTTGTCATACTTCTCATGG-3’.

3.1. Crystal structures of CP1

Two crystallographically independent Cu(II) atoms, each occupying a distinct position, are present in the asymmetric unit of CP1, which crystallizes in monoclinic space group C2/c (Figure. 1a). Both Cu1 and Cu2 are hexa-coordinated and have twisted octahedral geometries. Cu1 is coordinated with N5 and N5B atoms come from two separate biimb ligands, coordinated O5 and O5B molecules, and O1 and O1B atoms derived from two distinct BPDC^2-. The lengths of Cu1-O bond is between 2.0689(18) and 2.1657(16) Å and that of Cu1-N is 2.141(2) Å. Cu2 is surrounded by N1, N1A, N3, and N3A atoms from four diverse biimb ligands and coordinated O6 and O6A molecules, where the lengths of Cu2-N bond were 2.1247(18) and 2.1423(17) Å, and that of Cu2-O is 2.1391(17) Å. The two imidazole groups of biimb join Cu2 to create the (4, 4) grid 2D layer (Figure 1b), and the coordination between the Cu1 and third imidazole group serves as pillars to complete the 3D pillar-layered framework of CP1 (Figure 1c). Additionally, fully deprotonated BPDC^2- as terminal ligands adorn the axial positions of Cu1, thereby impeding the structural interpenetration of CP1 (Figure 1d). It is noteworthy that the Cu–O and Cu–N bond lengths in CP1 fall within the typical ranges observed for similar Cu(II)-based pillar-layered MOFs reported in the literature [30-32], indicating comparable coordination strength and geometry. Compared with these reported structures, CP1 exhibits slightly elongated Cu–O bonds, which may be attributed to the steric hindrance of the terminal BPDC²⁻ ligands and the twisted nature of the octahedral coordination. Moreover, the prevention of interpenetration by terminal ligands, as observed in CP1, has also been reported in analogous Cu–Cu-biim-dicarboxylate frameworks [33], highlighting a common structural stabilization strategy in this family of compounds. In addition, π–π stacking interactions are observed between adjacent aromatic rings of biimb ligands, with a centroid-to-centroid distance of 3.7940(2) Å, which contributes to the overall supramolecular stability of CP1 (Figure 1e). No significant hydrogen bonding was observed in the crystal structure, as all potential hydrogen donor-acceptor distances exceed 4.0 Å, falling outside the typical hydrogen bond range.

Close

Figure 1.

(a) CP1’s asymmetry unit. (b) CP1’s 2D Cu2-biimb layer. (c) CP1’s 3D Cu-biimb network. (d) CP1’s 3D framework. (e) π–π stacking interactions (3.794 Å); no hydrogen bonding observed.

Export to PPT

3.2. Structure description of compound 1

The single-crystal X-ray diffraction (SCXRD) data analysis of compound 1 displays that compound 1 is crystallized in the monoclinic crystal system with the space group of P2(1)/c. Figure 2(a) presents that the compound 1’s molecule structure is composed of two six-membered rings (ring A: C6-C11 and ring B: C3, C4, C6, C7, C12, and N1) along with a benzene ring C (ring C: C13-C18). Through sharing the double bond between C6 and C7, rings A and B are linked to one another. The C=O bond connects the C8 atom of ring A to one oxygen atom, O3; the C4 atom of ring B is linked to a methyl group (-C5H₃) through the C4-C5 single bond, and its C3 atom is linked to a methyl formate group (-C₂O_1O2C₁H₃) through the C2-C3 single bond. Through the C12 atom, the benzene ring C joins the ring B. One Br atom is connected to the C16 atom of ring C. A dense-packed supramolecular structure was also produced by compound 1’s surrounding molecule structures being further joined by the H-bonds (O3...N1-H1D: 2.1580 Å) (Figure 2b) and van der Waals force (Figure 2c).

Close

Figure 2.

(a) Compound 1’s molecule structure; (b) H-bonds between the adjacent molecules; (c) Dense stacked supramolecular structure of compound 1.

Export to PPT

3.3. Characterizations of 1-CS-CPTMS@CP1@LBP

Figure 3 presents a comprehensive characterization of the 1-CS-CPTMS@CP1@LBP composite. In the FT-IR spectra (Figure 3a), 1-CS-CPTMS displays O–H/N–H stretching at 3423 cm⁻¹, C–H stretches at 2921/2853 cm⁻¹, and Si–O–Si asymmetric vibration at 1105 cm⁻¹; after CP1 grafting, new amide bands emerge at 1655 cm⁻¹ (C=O) and 1540 cm⁻¹ (N–H bend), with the Si–O–Si peak shifting to 1098 cm⁻¹. Subsequent LBP loading introduces ester C=O absorption at 1740 cm⁻¹ and conjugated C=C signals at 1620 cm⁻¹, alongside broadening of the hydroxyl band, confirming successful stepwise functionalization. XRD patterns (Figure 3b) show a broad amorphous halo centered at 2θ ≈ 20° for both 1-CS-CPTMS@CP1 and its LBP-loaded counterpart, while weak reflections at 11.3° and 24.7° in the latter correspond to ordered polysaccharide domains. Thermogravimetric analysis (TGA) (Figure 3c) reveals an initial ∼8% mass loss below 150°C (adsorbed moisture), followed by 55% decomposition between 150-500°C attributable to organic components, and a residual 35% inorganic framework at 800°C. Nitrogen sorption-derived pore size distribution (Figure 3d) features a dominant mesopore peak at 3.2 nm with a pore volume of 0.21 cm³/g STP, and a secondary distribution between 5-15 nm (0.04 cm³/g). Collectively, these data validate the successful construction, functionalization, and mesoporous architecture of the 1-CS-CPTMS@CP1@LBP nanocomposite.

Close

Figure 3.

Structural characterization of 1-CS-CPTMS@CP1@LBP: (a) FT-IR spectra, (b) XRD patterns, (c) TGA, and (d) pore size distribution.

Export to PPT

3.4. Fluorescence detection

The optical properties of 1-CS-CPTMS@CP1@LBP were systematically evaluated by fluorescence spectroscopy (Figure 4). Upon excitation at 360 nm, 1-CS-CPTMS@CP1@LBP exhibits four well‐resolved emission peaks between 590 and 710 nm, corresponding to the characteristic transitions of CP1, whereas free LBP alone shows only a weak band centered at 615 nm (Figure 4a). Importantly, loading of LBP into the 1-CS-CPTMS@CP1 scaffold induces a greater than tenfold increase in emission intensity at 620 nm (Figure 4b), demonstrating that 1-CS-CPTMS@CP1 serves as an efficient fluorescent “turn-on” probe for LBP detection. We further investigated the pH dependence of this response by measuring the fluorescence intensity of 1-CS-CPTMS@CP1@LBP (I) relative to the scaffold alone (I₀) across pH 3-11 (Figure 4c). The greatest signal enhancement (I/I₀≈700 a.u.) occurs at pH 9, indicating that mildly alkaline conditions optimize the LBP-probe interaction. Finally, the kinetics of the fluorescence response were assessed at pH 9 by monitoring I/I₀ over 1-8 min after LBP addition (Figure 4d). The signal rises sharply within the first minute and plateaus by 5 min, confirming rapid complex formation. Thus, the ideal sensing conditions for LBP are pH 9 with a 5-min incubation.

Close

Figure 4.

(a) Fluorescence emission spectrum of 1-CS-CPTMS@CP1@LBP (λₑₓ = 390 nm); (b) Emission spectra of LBP (blue), 1-CS-CPTMS@CP1 (orange), and 1-CS-CPTMS@CP1@LBP (green) in aqueous solution at pH 9; (c) pH-dependent fluorescence intensity (λₑₘ ≈ 620 nm); (d) Time course of fluorescence response (I/I₀-1) upon addition of LBP. (e) Fluorescence lifetime decay of 1-CS-CPTMS@CP1@LBP (τ = 0.3402 ns).

Export to PPT

Crucially, fluorescence lifetime measurements were conducted at an excitation wavelength of 390 nm to further understand the nature of the emission mechanism. As shown in Figure 4(e), the time-resolved fluorescence decay profile of 1-CS-CPTMS@CP1@LBP reveals a single-exponential behavior with a fitted lifetime (τ) of 0.3402 ns. This short lifetime is consistent with the characteristics of CP1 and supports a ligand-to-metal charge transfer (LMCT) or surface-related emission pathway. The fast decay kinetics further corroborate the proposed “turn-on” mechanism, where LBP adsorption alters the local coordination environment, enhancing the radiative recombination rate. The stability of the lifetime over repeated measurements also confirms the robustness and reproducibility of the sensing signal. Together, these findings validate 1-CS-CPTMS@CP1@LBP as a highly responsive, stable, and rapid-acting fluorescence platform for the detection of LBP under physiological-like conditions.

Figure 5 further demonstrates the exceptional temporal stability of the 1-CS-CPTMS@CP1@LBP probe. As shown in Figure 5(a), when continuously irradiated at room temperature, the fluorescence intensity remains essentially constant over a 2-h period, with less than a 5% fluctuation from its initial value. More strikingly, Figure 5(b) shows that after storing the nanocomposite dispersion at ambient conditions for 30 days, the emission intensity still retains over 95% of its original magnitude. These results confirm that 1-CS-CPTMS@CP1@LBP possesses both excellent photostability and long-term storage stability, making it highly suitable for practical sensing and delivery applications.

Close

Figure 5.

Fluorescence intensity of 1-CS-CPTMS@CP1@LBP (λₑₓ = 360 nm, λₑₘ = 620 nm) over (a) 0–120 min (≤5% change) and (b) 30 days (≥95% retention).

Export to PPT

The fluorescence response of 1-CS-CPTMS@CP1@LBP was assessed across LBP concentrations from 0 to 50 μM (Figure 6a), revealing a progressive increase in emission intensity at 620 nm upon 360 nm excitation. The calibration plot of (I/I₀–1) versus [LBP] (0.05–50 μM) exhibits excellent linearity (I/I₀-1 = 3.935 × 10⁵·[LBP] + 0.886, R² = 0.995) (Figure 6b), the limit of detection (LOD) was thus calculated to be 3 nM (LOD = 3σ/S, where σ represents the standard deviation of the blank and S indicates the slope of the calibration curve). Selectivity tests against a panel of potential interferents, encompassing glucose, amino acids (Glu, Cys, Ala, Lys and His), antibiotics (Amoxicillin/Clavulanic acid (AMC), Erythromycin (E), Florfenicol (FFC), Vancomycin (VA), Azithromycin (AZM), Streptomycin (STR)), and metal ions (Na⁺, Ca²⁺, K⁺, Mg²⁺, Ba²⁺), show that only LBP produces a significant “turn-on” signal (I/I₀–1 ≈ 21), whereas all other species induce negligible fluorescence changes (Figure 6c). Even when interferent concentrations match that of LBP (Figure 6d), the fluorescence response to LBP remains undiminished, confirming excellent anti-interference performance. These findings demonstrate that 1-CS-CPTMS@CP1@LBP offers high sensitivity (LOD = 3 nM), a broad linear detection range, and outstanding selectivity, making it a promising fluorescent probe for LBP detection and removal in complex sample matrices.

Close

Figure 6.

(a) Fluorescence spectra of 1-CS-CPTMS@CP1@LBP (λₑₓ = 360 nm) with 0-50 μM LBP; (b) Linear calibration; (c) Selectivity; (d) Anti-interference.

Export to PPT

3.5. LBP loading performance

The loading performance of 1-CS-CPTMS@CP1 toward LBP was further evaluated. Figure 7(a) illustrates that following loading, the UV-Vis absorption peaks of LBP remarkably diminish. The pH range of 4–12 was utilized to evaluate the impact of pH on LBP loading (Figure 7b), revealing robust loading capacity across this interval and a maximum loading at pH 9. The loading kinetics (Figure 8a) indicate a rapid uptake during the first 90 min, attributable to abundant available sites, followed by a gradual approach to equilibrium by 500 min as available sites are consumed, after which the loading rate further diminishes. To elucidate the loading kinetics, the data were fitted to pseudo-first-order (PFO), pseudo-second-order (PSO), and intraparticle diffusion models. The intraparticle diffusion model exposes diffusion control mechanisms, but the PSO model better represents chemisorption behavior. The PFO model usually describes physisorption. Their empirical equations are (Eqs. 1, 2):

Close

Figure 7.

(a) UV–vis spectra of LBP before and after adsorption; (b) The optimist of adsorption pH conditions of 1-CS-CPTMS@CP1.

Export to PPT

Close

Figure 8.

LBP adsorption on 1-CS-CPTMS@CP1: (a) kinetics at 25°C; (b) isotherms at 25/30/35°C; (c) Langmuir linear plots; (d) Freundlich linear plots.

Export to PPT

In which, qe (mg·g⁻¹) and qt (mg·g⁻¹) are the loading at equilibrium and at time t (min), separately; k₁ and k₂ represent the rate constants for the PFO and PSO models; and Ci and ki (mg·g⁻¹·h⁻⁰^․⁵) denote the boundary layer thickness and intraparticle diffusion constant, separately. In summary, the highest correlation coefficient (R2 = 0.997) was obtained for the PSO model, suggesting that the loading of LBP is mainly caused by chemisorption. Furthermore, the fact that the fit does not pass through the origin affirms that intraparticle diffusion is not the only rate-limiting step.

For further examination of the loading mechanism, the equilibrium data were fitted to the Langmuir and the Freundlich isotherm models (Figure 8b), expressed by (Eqs. 3, 4):

Here, qe (mg·g⁻¹) and Ce (mg mL^-1) are the loading and equilibrium concentration; q_max (mg·g⁻¹) is the maximal loading capacity; KL (L·mg⁻¹) represents the Langmuir constant; and kF and nnn denote the Freundlich constants. Comparison of the correlation coefficients in Figures 8(c and d) shows that LBP loading conforms better to the Freundlich model, indicating a multilayer loading process. Under ambient conditions, the maximum loading capacity reaches 384.71 mg·g⁻¹.

3.6. 1-CS-CPTMS@CP1@LBP significantly inhibited gallbladder cancer cell proliferation

The proliferation inhibition rate of GBC cells in each group at 48 h was identified via CCK-8 assay. Figure 9a and 9b suggested that blank NPs did not inhibit the proliferation of SGC-996 and GBC-SD cells. The inhibition rates of free LBP on SGC-996 and GBC-SD cells were approximately 21.5% and 28.4%, respectively. 1-CS-CPTMS@CP1@LBP group exhibited obviously better inhibition effect in contrast to the free LBP group, and the inhibition rate of SGC-996 and GBC-SD cells achieved 52.1% and 48.8%, respectively. Subsequently, we explored the regulatory effects of 1-CS-CPTMS@CP1@LBP on PA2G4. The qPCR assay showed that there was no clear difference in the levels of PA2G4 mRNA between the NPs and control groups (p>0.05) (Figure 9c and 9d). Both free LBP and 1-CS-CPTMS@CP1@LBP groups significantly inhibited the relative expression of PA2G4 mRNA, but the inhibitory effect of 1-CS-CPTMS@CP1@LBP was significantly stronger than that of the free LBP group. Our results suggested that 1-CS-CPTMS@CP1@LBP remarkably suppresses GBC cell proliferation through nanocarrier potentiation, and the mechanism is closely related to the down-regulation of PA2G4 gene expression.

Close

Figure 9.

(a) Comparison of 48-h proliferation inhibition rates in GBC-SD cells. (b)Relative PA2G4 mRNA expression levels in GBC-SD cells following the indicated treatments. (c) Comparison of 48-h proliferation inhibition rates in SGC-996 cells. (d) Relative PA2G4 mRNA expression levels in SGC-996 cells following the indicated treatments. **, p < 0.01; ***, p < 0.001; ****, p < 0.0001; ns, not significant.

Export to PPT

3.7. Discussion

This study presents a multifunctional nanoplatform, 1-CS-CPTMS@CP1@LBP, engineered for simultaneous fluorescent detection and efficient delivery of LBP in GBC therapy. The rational design leverages the structural and optical advantages of Cu-based MOFs, the biocompatibility of chitosan, and the covalent stability provided by CPTMS functionalization to overcome the typical limitations of traditional LBP delivery, such as poor solubility, instability, and inefficient cellular uptake.

The successful synthesis of CP1, a Cu(II)-based MOF with a pillar-layered 3D framework, was confirmed via SCXRD. Its robust architecture, built from biimb and BPDC linkers, provides a high surface area and multiple coordination sites ideal for molecular interaction and loading. The structural rigidity and spatial arrangement of CP1 contribute significantly to its fluorescence properties, exhibiting well-defined emission bands with long lifetimes and large Stokes shifts, which are essential for precise and interference-free detection. Compound 1, incorporated into the chitosan matrix through CPTMS coupling, further enhances the fluorescence signal via supramolecular interactions and introduces additional hydrogen bonding and van der Waals stabilization. This functionalization not only reinforces the mechanical stability of the carrier but also introduces sites for enhanced LBP recognition. Characterization data from FTIR, XRD, TGA, and nitrogen adsorption confirm the successful stepwise assembly of the hybrid nanostructure and its mesoporous nature, which is favorable for high-capacity loading and controlled release. The fluorescence sensing studies demonstrate that the nanoplatform offers an excellent “turn-on” response to LBP under mildly alkaline conditions (optimal at pH 9), with a rapid plateau reached within 5 min. This indicates efficient interaction between LBP and the probe surface, enabling real-time detection. Notably, the probe shows ultra-sensitive detection limits (3 nM) and high selectivity even in the presence of a wide range of biological interferents, which is critical for real-world diagnostic applications. The LBP loading behavior follows pseudo-second-order kinetics and the Freundlich adsorption isotherm, suggesting chemisorption and multilayer uptake mechanisms. This is attributed to strong electrostatic and hydrogen bonding interactions between LBP and the hybrid matrix. A maximum loading capacity of 384.7 mg/g confirms the efficient encapsulation ability of the nanoplatform.

In vitro cytotoxicity assays against SGC-996 and GBC-SD cells reveal that 1-CS-CPTMS@CP1@LBP exerts significantly higher antiproliferative effects than free LBP, validating the enhanced delivery and therapeutic efficacy of the nanocomposite. Mechanistically, this is supported by the marked downregulation of the PA2G4 oncogene, a proliferation-associated factor implicated in gallbladder carcinoma progression. The improved gene suppression further underlines the contribution of nanocarrier-facilitated uptake and intracellular delivery to therapeutic enhancement. Compared with previously reported systems, our platform achieves superior LBP loading, fluorescence sensitivity, and biological effect with a simpler synthetic route and improved structural integration. The use of a single-component MOF core circumvents the fluorescence self-quenching and synthetic complexity often encountered in multi-component systems. Overall, this study establishes a dual-functional nanocomposite with high diagnostic sensitivity, drug-loading efficiency, and therapeutic relevance. By addressing critical challenges in LBP delivery and detection, it offers a promising paradigm for multifunctional nanomedicine in GBC management and potentially other inflammation-driven malignancies.