An activatable fluorescent probe for fibroblast activation protein α and its bioimaging applications

Abstract

Fibroblast activation protein α (FAP) is a member of the plasma membrane-bound serine protease family, which exhibits exopeptidase and endopeptidase activity in organisms. Recently, considerable studies have focused on the key role of FAP in various pathological processes, indicating that it may be a therapeutic target for treating human disorders and diseases. In the present study, a fluorescent probe for FAP was constructed by the introduction of a substrate recognition dipeptide that was rationally modified by tert-butoxycarbonyl group to the fluorophore DDAN. The probe DDAF, synthesized from DDAN, exhibited high selectivity and sensitivity for FAP, with a detection limit of 0.71 ng mL⁻¹. DDAF was successfully applied for the rapid screening of potent inhibitors of FAP, and a quinoline derivative was identified that exhibited a strong inhibitory effect on FAP. Subsequently, the in vitro visualization and monitoring of FAP in living cancer cells and tissues were performed using DDAF, indicating the widespread expression of FAP in a variety of tumors. These findings provide a promising tool for discovering potent therapeutic agents targeting FAP and for the real-time visualization and monitoring of FAP in tumors.

Fibroblast activation protein α (FAP) is a promising therapeutic target. We designed fluorescent probe DDAF with high FAP selectivity and sensitivity. DDAF can facilitate the screening of FAP inhibitors and visualizes FAP in tumor cells and tissues.

Introduction

Fibroblast activation protein α (FAP) is a member of the plasma membrane-bound serine protease family, which also includes dipeptidyl peptidase IV (DPP 4), DPP VII, DPP VIII, DPP IX, and prolyl carboxypeptidase.^1,2 The protein is a cell-surface antigen that consists of an intracellular domain, a transmembrane domain, and an extracellular domain.³ FAP and its family members possess a catalytic center composed of Ser-Asp-His, with serine acting as a nucleophile to facilitate the easy cleaving of the N-terminal Pro-X peptide bonds; these enzymes usually exhibit similar exopeptidase and endopeptidase activities in organisms.⁴ Notably, FAP has been found to possess gelatinase activity, indicating its potential role in tissue remodeling through the hydrolysis of type I collagen, which is the major component of the extracellular matrix.^5,6

FAP can mediate the maintenance of the epithelial–mesenchymal state that is expressed at sites of tissue remodeling and repair.⁷ It has been reported that FAP is involved in various pathological processes and may be a therapeutic target for fundamentally treating a multitude of human disorders and diseases.^8–10 Furthermore, a growing proportion of investigations have indicated that FAP is aberrantly expressed in a variety of tumors, including lung cancer, ovarian cancer, pancreatic cancer, and breast cancer.^11,12 It has been reported that the high expression of FAP in patients with rectal cancers was associated with a poor prognosis.¹³ Considering the important biological function of FAP in tumors, the inhibition of FAP has been considered an effective means for tumor treatment and diagnosis. Preclinical and clinical trials have confirmed the potential role of the FAP inhibitor talabostat in anti-neoplastic therapy.^14,15 However, to date, FAP-targeted inhibitors remain severely limited. In general, FAP protein expression and/or activity has been associated with many human disorders and diseases, including fibrosis and cancer, and FAP can therefore be regarded as a highly promising biomarker and therapeutic target.

Activity-based fluorescent probes are powerful molecular tools for analyzing enzyme activity, visualizing target expression in biological research, and facilitating the discovery of small-molecule therapeutic drugs.^16,17 Additionally, fluorescent probe molecules have been widely applied to analyze the activity of target enzymes, including cytochrome P450 enzymes, UDP-glucuronosyltransferases, and various other enzymes.^18–20 Recently, several studies have reported activity-based fluorescent probes for FAP.^21–24 However, the current probes are limited by their low sensitivity, low selectivity, and significant background interference in practical applications. Consequently, there is an urgent need to develop novel probes for FAP that exhibit high sensitivity and selectivity, thereby enhancing their compatibility with efficient screening and diagnostic imaging techniques. This would facilitate the discovery of small-molecule inhibitors and improve the imaging and analysis of FAP.

In the present study, a characteristic dipeptide structure that enables easy recognition by FAP and a suitable modified terminal group that promotes differential recognition between FAP and its isozymes were introduced into the fluorophore DDAN, thereby developing a novel fluorescent probe for FAP, named DDAF. DDAF exhibited high specificity and sensitivity for FAP, with a detection limit of 0.71 ng mL⁻¹. DDAF was applied for the efficient screening of FAP inhibitors and imaging of FAP in cancer cells and tissues. Using DDAF, a strong inhibitor was screened, and its inhibition behavior was characterized. Additionally, DDAF enabled the visualization of FAP in tumor sections. These findings suggest significant potential for clinical applications in tumor detection, treatment selection, and prognosis assessment.

Experimental section

Chemicals and materials

Recombinant human FAP and DPP 4 were obtained from R & D Systems. Acetylcholinesterase (AChE), butyrylcholinesterase (BChE), β-glucosidase (GLC), leucine aminopeptidase (LAP), galactosyltransferase (GAL), alkaline phosphatase (ALP), β-lactamase (LAC) and human serum albumin (HSA) were obtained from Sigma-Aldrich. Recombinant human carboxylesterase 1c (CE1c) and carboxylesterase 2 (CE2) were obtained from Corning Life Sciences. Recombinant matrix metalloproteinase 9 (MMP9) was obtained from UA BIOSCIENCE. Carbonic anhydrase (CA), β-N-acetylglucosaminidase (NAG), elastase (EA), pepsin and lipase were purchased from Shanghai Yuanye Bio-Technology. Talabostat was purchased from Bidepharm.

Synthesis of DDAF

Compound DDAN was synthesized according to our previous method.²⁵ Compound Boc-Gly-Pro-OH (272.3 mg, 1.0 mmol) was dissolved in 12 mL of anhydrous DMF, then DIPEA (258.6 mg, 2.0 mmol) and HATU (380.2 mg, 1.0 mmol) were added at 0 °C and the mixture was stirred for 40 min. Next, compound DDAN (306.0 mg, 1.0 mmol) dissolved in 6 mL of anhydrous DMF was added dropwise, and the reaction mixture was stirred for a further 5 h at room temperature. Then, the reaction solution was diluted with CH₂Cl₂ and washed with water (300 mL × 3) and brine (300 mL). Next, the organic phase was dried over anhydrous magnesium sulfate and evaporated under reduced pressure, and the resulting crude product was further purified by silica gel chromatography, using CH₂Cl₂/MeOH (40/1 v/v) as the eluent, to afford DDAF as an orange solid (102.5 mg, yield: 18.3%). ¹H NMR (400 MHz, CDCl₃): δ 9.82 (s, 1H), 7.98 (s, 1H), 7.65 (d, J = 3.0 Hz, 1H), 7.58 (d, J = 8.5 Hz, 1H), 7.38 (dd, J = 8.5, 2.2 Hz, 1H), 5.37 (d, J = 4.3 Hz, 1H), 4.84 (d, J = 7.6 Hz, 1H), 4.03 (qd, J = 17.7, 4.8 Hz, 2H), 3.62 (t, J = 7.8 Hz, 1H), 3.48 (dd, J = 17.0, 9.8 Hz, 1H), 2.64 (dd, J = 12.5, 6.0 Hz, 1H), 2.17 (ddd, J = 60.3, 34.1, 6.4 Hz, 3H), 1.91 (d, J = 2.7 Hz, 6H), 1.48 (d, J = 11.3 Hz, 9H). ¹³C NMR (125 MHz, CDCl₃): δ 173.05, 170.13, 168.93, 155.84, 148.30, 142.13, 140.90, 139.91, 138.57, 137.26, 136.78, 135.35, 132.46, 119.27, 117.46, 80.20, 61.26, 46.72, 43.21, 39.23, 28.34, 26.76, 26.58, 25.00. HRMS (ESI negative) calculated for [M–H]⁻ (C₂₇H₂₉Cl₂N₄O₅⁻): 559.1520, found 559.1521.

General procedures for FAP activity assay

Incubation systems consisted of 100 mM phosphate buffer (pH 7.4) and recombinant human FAP, or human tissue preparations, with a total reaction volume of 200 µL. DDAF or inhibitor was dissolved in DMSO and the final concentration of DMSO was maintained at less than 1% (v/v) in the mixtures. After incubation for a certain time at 37 °C, 100 µL of acetonitrile was added to terminate the hydrolysis reaction. The samples were centrifuged at 20 000g for 10 min, and the supernatant was analyzed.

Cytotoxicity assays

LoVo cells were seeded in 96-well plates at approximately 1 × 10⁴ cells per well in a medium supplemented with 10% FBS and incubated for 24 h (37 °C, 5% CO₂). The cells were then exposed to DDAF and DDAN for 12 h. After rinsing once with PBS prewarmed to 37 °C, CCK-8 (0.5 mg mL⁻¹) was introduced into the wells. Following 2 h of further incubation, the absorbance at 450 nm was recorded. Cell viability was calculated by normalizing the absorbance of the treated groups to that of the vehicle control group (set as 100%).

Cell culture and cell bioimaging experiment

MCF-7, LoVo or U87 cells were seeded in dishes and cultured overnight. The adherent cells were then washed with FBS-free medium. DDAF was prepared as a 50 mM stock solution in DMSO and then diluted to a final concentration of 10 µM for cell incubation. After 60 min of incubation, fluorescence images were acquired using a confocal microscope (Leica SP8, Germany). For the probe DDAF, the excitation wavelength was set at 633 nm; for DAPI, the excitation wavelength was set at 405 nm.

Tissue bioimaging experiment

Slices were cut using a cryostat (Leica CM3050S). Tissue slices were incubated with DDAF for 60 min at 37 °C and imaged using either a Leica SP8 (Germany) or an Olympus FV1000 (Japan). For the probe DDAF, the excitation wavelength was set at 633 nm; for DAPI, the excitation wavelength was set at 405 nm.

Results and discussion

Synthesis and design of DDAF

The N-terminal Gly-Pro dipeptide protected with tert-butoxycarbonyl was used as the recognition site of FAP and introduced into the fluorophore DDAN, resulting in the design and development of the fluorescent probe DDAF for the detection of FAP. The synthetic procedure for DDAF is provided in Scheme S1, and its structure was confirmed by ¹H NMR, ¹³C NMR and HRMS (Fig. S1–S3, respectively).

Spectral and kinetic characterization of DDAF

The amide bond of DDAF was cleaved by FAP-mediated hydrolysis, releasing DDAN, which triggered a fluorescence peak at around 666 nm (Scheme 1 and Fig. 1). Additionally, the fluorescence intensity at 666 nm increased linearly when the concentration of FAP gradually increased from 0 to 0.4 µg mL⁻¹ (R² = 0.9928). DDAF is weakly fluorescent over a wide pH range, whereas DDAN exhibits intense fluorescence under the same conditions (λ_ex = 600 nm, Fig. S4). Thus, DDAF could be used as an “OFF–ON” fluorescent probe for the detection of FAP. The detection limit of DDAF for sensing FAP was measured as 0.71 ng mL⁻¹, indicating the ultra-sensitivity of DDAF. To further study the hydrolysis kinetics of the probe, kinetic analysis of the hydrolysis of DDAF by FAP was performed (Fig. S5). The apparent kinetic parameters including K_m and V_max values were calculated to be 3.87 ± 0.44 µM and 0.068 ± 0.0020 µmol per min per mg protein, respectively. The high sensitivity observed for DDAF in FAP activity analysis was attributed to its high binding affinity for FAP.

Scheme 1. Proposed mechanism of the sensitive “turn-on” detection mode of DDAF.

Fig. 1. Absorption (A) and fluorescence spectra (B) of DDAF and DDAN. (C) Fluorescence spectra of DDAF after reacting with a range of different concentrations of FAP (0–0.4 µg mL⁻¹). (D) Fluorescence intensity of DDAF after reacting with a range of different concentrations of FAP (0–0.4 µg mL⁻¹).

Specificity of DDAF to detect FAP

The specificity of DDAF toward FAP was evaluated using various hydrolases or proteins with hydrolytic activity. As shown in Fig. 2A, a dramatic enhancement in fluorescence was observed in samples incubated with recombinant human FAP, while samples incubated with other proteins generated only weak fluorescence signals, indicating a strong preference of DDAF for FAP. Additionally, the potential influence of various interfering species on the FAP-mediated hydrolysis of DDAF was evaluated. It was found that the fluorescence response of DDAF was barely affected when DDAF was incubated with various interfering species (Fig. S6).

Fig. 2. (A) Fluorescence response of DDAF at 666 nm (λ_ex = 600 nm) after reacting with FAP and various hydrolases. (B) Concentration-dependent inhibitory effects of talabostat on the hydrolysis of DDAF by FAP, human liver microsomes (HLM) and human kidney microsomes (HKM).

The selectivity of DDAF for FAP was further evaluated via a chemical inhibition assay in complicated bio-samples using the FAP-selective inhibitor talabostat. As shown in Fig. 2B, talabostat exhibited an inhibitory effect on the hydrolysis of DDAF mediated by recombinant human FAP, human liver microsomes (HLM) and human kidney microsomes (HKM), respectively. Furthermore, the concentration-dependent inhibitory plots were independent of the enzyme source. These results confirmed the high specificity of DDAF for FAP and its superior anti-interference ability in bio-samples.

Bioimaging FAP in living cells

Three human cell lines were used in the cell imaging assay to investigate the imaging capability of DDAF in living cells. Precisely, the cell viability of LoVo treated with DDAF and DDAN was evaluated using a CCK-8 assay. The results indicated that DDAF and its metabolite, DDAN, had only weak effects on cell viability (Fig. S7). Then, cell imaging was performed after incubation with DDAF for 1 h, and strong fluorescence signals were observed in MCF-7, LoVo and U87 cells (Fig. 3). Furthermore, the fluorescence signals were markedly weakened when the cells were co-incubated with talabostat, an FAP-selective inhibitor, and DDAF (Fig. S8 and S9). These results indicated that the observed fluorescence signal was mainly attributed to FAP in the cells.

Fig. 3. Confocal microscopy images of MCF-7, LoVo and U87 cells. The scale bar represents 50 µm. Channel 1 for DDAF: excitation, 633 nm; emission, 650–690 nm. Channel 2 for DAPI: excitation, 405 nm; emission, 430–490 nm.

The identification of FAP inhibitors based on DDAF

FAP represents a promising biomarker and therapeutic target because it is involved in tumor proliferation and invasion, making FAP inhibitors potential antitumor agents.¹⁵ In the present study, the excellent performance of DDAF indicated that it facilitated the speedy and accurate detection of FAP activity, thereby enabling the discovery and characterization of FAP regulators. Then, potential FAP inhibitors were screened by a speedy and visual assay in a 96-well-microplate. As shown in Fig. 4, the effects of small-molecular ligands on FAP activity could be rapidly and visually characterized based on the fluorescence assay. It was found that compound located in 2D, a quinoline derivative ((S)-N-(2-(2-cyano-4,4-difluoropyrrolidin-1-yl)-2-oxoethyl) quinoline-4-carboxamide, CDQC), exhibited a strong inhibitory effect on FAP. Subsequently, the inhibitory effects of CDQC were evaluated using concentration–effect curves. As shown in Fig. 4, CDQC inhibited the formation of DDAN in a concentration-dependent manner, and the IC₅₀ value of CDQC against FAP was determined as 0.97 nM. Kinetic inhibition analysis revealed that CDQC acted as a mixed inhibitor of FAP, with a K_i value of 0.95 nM. These results indicated that CDQC functioned as a strong inhibitor of FAP and further demonstrated that DDAF, together with its visualization-based assay, enables the rapid, on-site detection of FAP activity, making it highly promising in applications for the screening and characterization of inhibitors of FAP.

Fig. 4. (A) Visualization of the fluorescence response of DDAF toward potential FAP inhibitors using a fluorescence image analyzer (λ_ex: 635 nm, λ_em: 670 nm). F12, G12 and H12 were parallel control samples that were not treated with the test compound, whereas E12 was a positive control containing an FAP inhibitor. (B) Concentration-dependent inhibitory effects of CDQC toward FAP. (C) Lineweaver–Burk and Dixon plots for CDQC.

Bioimaging FAP in tumor sections

It has been proven by a growing number of studies that there is a close connection between the abnormal expression of FAP and the occurrence and development of cancer. FAP is a very interesting biomarker that has been applied in the diagnosis, prognosis and treatment evaluation of tumors in recent years. Herein, encouraged by the above-mentioned results, we explored whether DDAF could be used as a fluorescent tool to image FAP in tumors. As shown in Fig. 5, stronger fluorescence signals were observed on sections from a variety of tumors including breast tumor, malignant melanoma, thyroid cancer, ovarian cancer, lung cancer, and gastrointestinal stromal tumor. These results suggest that FAP was expressed in a variety of tumors, which enables DDAF to be used as a crucial tool for understanding the potential role of FAP in tumors and assisting clinicians in detecting tumors, making treatment choices and assessing prognosis.

Fig. 5. Confocal laser scanning microscopy images of tissue sections. (A) Breast cancer, (B) melanomas, (C) thyroid carcinoma, (D) ovarian cancer, (E) lung adenocarcinoma, and (F) gastrointestinal tumor. The scale bar represents 50 µm. Channel 1 for DDAF: excitation, 633 nm; emission, 650–690 nm. Channel 2 for DAPI: excitation, 405 nm; emission, 430–490 nm.

3D imaging of tumor tissues

Inspired by the distinct “light-up” fluorescence response of DDAF toward FAP, we further examined the fluorescence imaging capability of DDAF for the 3D visualization of FAP in tumor tissues. Herein, a slice of colorectal cancer patient-derived xenograft tumor was incubated with DDAF. As shown in Fig. 6A, strong fluorescence responses were observed at different depths within the tumor tissue. The 3D image was reconstructed, displaying the spatial distribution of FAP within the tumor tissue, which indicated the feasibility of visualizing the biological activity of FAP within its multidimensional network of interactions. Meanwhile, the multi-layer and 3D images of breast cancer tissues are shown in Fig. 6B. All these results demonstrated that DDAF displayed good tissue penetration capability and was suitable for visualizing the 3D morphological characteristics of FAP, laying a foundation for the further investigation of the role of FAP in tumors.

Fig. 6. Confocal laser scanning microscopy images of tumor tissues obtained by multi-layer scanning and 3D reconstruction. (A) Colorectal cancer patient-derived xenograft tumors. (B) Breast cancer tissue. Excitation, 633 nm; emission, 650–690 nm.

Conclusion

In the present study, a fluorescent probe named DDAF for FAP was constructed by the introduction of a substrate recognition dipeptide that was rationally modified by the Boc group to the fluorophore DDAN. The probe DDAF exhibited high specificity and sensitivity for FAP, with a detection limit of 0.71 ng mL⁻¹. Subsequently, DDAF was successfully applied for the rapid screening of FAP inhibitors and for imaging of FAP in cancer cells and tissues. Based on DDAF, a strong inhibitor was screened, and its inhibitory activity was characterized. Collectively, DDAF realized the visualization and monitoring of FAP in tumors, indicating its promising clinical applicability in detecting tumors, guiding treatment decisions and assessing prognosis.

Author contributions

Lei Feng: investigation, data curation, writing – original draft. Hengjing Chen, Zhenhao Tian, Wenpeng Li & Yifei Dang: investigation. Chengjuan Xing, Xiaohong Shu & Fangyu Yang: resources. Guobiao Liang, Jing Ning & Xiaochi Ma: writing – review & editing, funding acquisition.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: NMR, HRMS. See DOI: https://doi.org/10.1039/d6ra05595h.