Using a CuAAC reaction, we synthesized a novel probe named NGF by conjugating G0 with N3-PEG4-ALKADK and 2-Azido-2-deoxy-D-glucose. The synthetic route of NGF was showed in Figure S1. The characterization of NGF involved mass spectrometry, HPLC analyses, and 1H/13C NMR spectroscopy (Figures S2-S5). Mass spectrometry and 1H/13C NMR spectroscopy results confirmed that the molecular weight of NGF corresponded to its theoretical value, affirming the synthesis's accuracy. Furthermore, HPLC analysis confirmed the purity of NGF.

We investigated the optical properties of the NGF probe. Figure 2A illustrates that NGF exhibited a absorption peak at 783 nm, whereas G0 showed a peak at 780 nm. Regarding fluorescence emission, NGF peaked at 807 nm compared to 804 nm for G0. These slight wavelength differences indicate that incorporating targeting molecules N3-PEG4-ALKADK and 2-Azido-2-deoxy-D-glucose did not alter the core fluorescence characteristics of G0. Thus, NGF retains the original fluorescence properties of G0 while gaining new targeting capabilities, thereby enhancing its specificity for specific biological applications. Further analysis revealed that a 4 μM solution of NGF in methanol exhibited a maximum absorbance of 1.06 and a fluorescence intensity of 12365 a.u., comparable to those observed for the G0 fluorophore.

Optical properties of NGF. Absorbance spectra of probe NGF at different concentrations (0.5, 1, 2, 3 and 4 μM) in (A) MeOH and (B) H2O. (Inset) Fluorescence spectra of NGF at 4 μM in MeOH and H2O, respectively. The stability of NGF in PBS, 10% mouse serum (C) and laser (D). (E) Photophysical and photochemical properties of NGF

The absorbance and fluorescence intensity ratios measured in water (H2O) were notably lower compared to those in methanol (MeOH), indicating some degree of aggregation of NGF in aqueous solutions. As compared with QS-1 [41], the maximum absorbance of both 4 μM NGF and QS-1 solutions showed consistency between MeOH and H2O conditions (Figs. 2A-B). However, NGF exhibited a slightly higher maximum fluorescence intensity of 5374 a.u. in H2O compared to QS-1. This finding suggests that the inclusion of 2-Azido-2-deoxy-D-glucose improved the solubility of NGF in water, enhancing its hydrophilicity, as evidenced by a Log P value of −0.95 ± 0.07.

Additionally, the photo, serum and aqueous stability of NFG were tested. As can be seen from Figs. 2C-D, NGF demonstrated strong stability in both PBS and 10% mouse serum. After 48 h of incubation in these solutions, the absorbance of NGF was recorded at 97.7% and 95.2% of the initial absorbance (A0), respectively. Furthermore, after being irradiated with a laser for 30 min, the absorbance of NGF decreased to 56.7% of A0. All the results indicate that NGF possesses strong stability and is suitable for subsequent cell and animal experiments.

Meanwhile, the molar absorption coefficient (ε) and fluorescence quantum yield (ΦF) of NGF were investigated. The results (Fig. 2E) showed a significant difference between in MeOH and H2O, which might be attributed to the partial aggregation of NGF in aqueous solution. The photophysical properties of NGF under serum conditions were also investigated, and the results showed that the ε of NGF was significantly lower than that observed in MeOH, this phenomenon might be due to the conjugation of NGF with albumin in serum. In summary, NGF functions as an innovative cyanogen probe with enhanced hydrophilicity and improved optical properties.

Prior to investigating the targeting specificity of the probe, we evaluated the expression of NRP1 and GLUT1 in various cancer cells. Figures 3A-B indicate that MDA-MB-231 cells exhibited higher levels of NRP1 protein compared to the other cell lines studied, and high level of GLUT1 expression was observed in HCT116 cells, whereas NRP1 and GLUT1 were minimally expressed in NCI-H1299 cancer cells, this suggests that the cancer cell lines MDA-MB-231, HCT116 and NCI-H1299 can be utilized to assess the binding specificity of NGF for NRP1 and GLUT1.

Binding specificity of NGF in vitro. (A) Western blot analysis of NRP1 and GLUT1 expression in different cancer cell lines HCT116, NCI-H1299 and MDA-MB-231. (B) Quantification analysis of western blot data by ImageJ software, *P < 0.05, **P < 0.01, ***P < 0.001 (vs. NCI-H1299 cells). (C) Flow cytometry analysis of NGF (2 μM) incubated with different cancer cell lines MDA-MB-231, HCT116 and NCI-H1299. (D) Flow cytometry analysis of NGF (2 μM) in cancer cells MDA-MB-231 with different treatment. (E) Flow cytometry analysis of NGF (2 μM) in cancer cells HCT116 with different treatment. (F) The quantification analysis of mean fluorescence intensity by FlowJo software. (G) Confocal fluorescence microscopy imaging of MDA-MB-231, HCT116 and NCI-H1299 cells incubated with NGF (2 μM) or blocking (scale bar = 50 μm). BG, BN, and BG+N indicated the cells were blocked by D-glucose (50 μM), NRP1 peptide (50 μM), and D-glucose + NRP1 peptide (50 μM), respectively. *P < 0.05, **P < 0.01, ***P < 0.001

We investigated the specific targeting capability of the NGF probe towards NRP1 and GLUT1 using flow cytometry and microscopy imaging experiments in MDA-MB-231, HCT116, and NCI-H1299 cells. The flow cytometry illustrated in Fig. 3C showed that NGF was accumulated significantly in HCT116 and MDA-MB-231 cells than that in NCI-H1299 cells, indicating the probe NGF can specifically target cancer cells with high NRP1 and GLUT1 expression. Meanwhile, the Fig. 3G shows that upon incubation with MDA-MB-231 cells, NGF was significantly internalized, displaying intense fluorescence at non-nuclear locations within the cellular structure. This result highlights NGF's strong targeting affinity for NRP1-positive MDA-MB-231 cells. Similarly, substantial internalization was observed in HCT116 cells, albeit with a weaker fluorescence signal compared to MDA-MB-231 cells, indicating NGF's specificity towards GLUT1, and its correlation with GLUT1 expression levels. In contrast, minimal fluorescence was detected in NRP1- and GLUT1-negative NCI-H1299 cells, underscoring NGF's selective targeting of NRP1-positive tumor cells. Flow cytometry and fluorescence imaging experiments confirmed that NGF effectively targets tumor cells expressing high levels of NRP1 and GLUT1, aligning with their respective expression profiles.

In addition, a blocking study of NGF was conducted by pre-treating MDA-MB-231 and HCT116 cells with the D-glucose (50 μM), NRP1 peptide (50 μM), and D-glucose + NRP1 peptide (50 μM) for 30 min. The results illustrated in Figs. 3E-G showed that the fluorescence signals can be blocked by the D-glucose, NRP1 peptide, and D-glucose + NRP1 peptide. According to the quantitative analysis of flow cytometry, the uptake of dual cancer receptor targeting probe in MDA-MB-231 and HCT116 was 1.7–2.5 and 1.4–1.5 times higher, respectively, compared to mono-receptor block conditions. These results indicating that probe NGF exhibited high specificity for both NRP1 and GLUT1 in vitro.

Furthermore, we used flow cytometry to explore the relationship between cellular uptake in MDA-MB-231 cells and the concentration of NGF probe. Figure 4 demonstrates a clear and direct correlation between the fluorescence intensity of MDA-MB-231 cells and increasing concentrations of NGF, exhibiting a strong linear relationship (R2 = 0.98). These findings provide robust confirmation of NGF's specific binding capability to tumor cells expressing high levels of NRP1, while also establishing a quantitative association between fluorescence intensity and NGF concentration.

The cellular uptake of NGF in MDA-MB-231 cancer cells. (A) Flow cytometry analysis of different concentrations of NGF incubated with MDA-MB-231 cancer cells for 1 h. (B) The quantification analysis of mean fluorescence intensity by FlowJo software

To evaluate the probe's targeting efficacy towards NRP1 in vivo, we selected MDA-MB-231, HCT116, and NCI-H1299 tumor-bearing mouse models, each representing different levels of NRP1 and GLUT1 expression. As shown in Fig. 5, the tumor of MDA-MB-231 and HCT116 mice could be clearly identified at 1 h post-injection, indicating the probe's tumor-targeting capability. In contrast, fluorescence intensity was comparatively weak in NCI-H1299 tumors. Importantly, while the fluorescent signal cleared from non-target organs within 12 h, it persisted at the tumor site for up to 72 h. The long half-life of NGF in blood and tumor was found to be associated with serum binding by SDS-PAGE analysis (Figure S7). This sustained and specific fluorescence signal underscores the NGF probe's potential for tumor diagnosis. Significantly, after 24 h, only the tumor site exhibited a fluorescent signal, further confirming the probe's excellent tumor-specific imaging capability.

NIR fluorescence images of NGF (25 μM, 200 μL) in MDA-MB-231, HCT116 and NCI-H1299 tumor-bearing mice (n = 3–5)