High biocompatible FITC-conjugated silica nanoparticles for cell labeling in both in vitro and in vivo models – Scientific Reports

Synthesize and characterization of FITC-SiO2-COOH nanoparticles

The morphology and structure of FITC-SiO2-NH2 and FITC-SiO2-COOH NPs were visualized using transmission electron microscopy (TEM) and scanning electron microscopy (SEM) as depicted in Fig. 1a. Both types of nanoparticles predominantly exhibited a spherical shape, a smooth surface, and a consistent size. The inset images in Fig. 1a are the size distribution of FITC-SiO2-NH2 NPs and FITC-SiO2-COOH NPs from corresponding TEM and SEM images. The diameter of the particles was determined using the open-source Image-J software. From there, it is possible to decide on the nanoparticles’ average size and corresponding standard deviation. Specifically, the average diameters for FITC-SiO2-NH2 and FITC-SiO2-COOH NPs were 93 ± 12 nm and 107 ± 17 nm, respectively. Dynamic Light Scattering (DLS) and Zeta potentiometry analyses, presented in Table 1, were utilized to assess their colloidal stability in water. The DLS results revealed a monomodal hydrodynamic size distribution with relatively low polydispersity index (PdI) values. The hydrodynamic diameter of FITC-SiO2-COOH NPs was registered at 115 nm, a slight increment from the FITC-SiO2-NH2 NPs which ranged between 109 to 115 nm. Moreover, the measured Zeta potentials were − 11 mV for FITC-SiO2-NH2 and − 42.4 mV for FITC-SiO2-COOH. This shift towards a more negative value for the latter can be attributed to the presence of hydroxyl and amino groups on the surface of FITC-SiO2-NH2 NPs. These groups result in a minor negative zeta potential and consequentially higher aggregation tendencies. The significantly negative zeta potential of FITC-SiO2-COOH NPs (− 42.4 mV) is indicative of carboxylate COO groups present on their surface, suggesting that they remain highly dispersed7.

Figure 1
figure 1

Nanoparticle characterization. (a) TEM and SEM images depicting FITC-SiO2-NH2 and FITC-SiO2-COOH NPs. (b) Raman and (c) FTIR spectra of SiO2-COOH NPs (magenta line), FITC (green line), FITC-SiO2-COOH NPs (orange line), and FITC-SiO2-NH2 NPs (violet line). UV–Vis spectra (d) and their corresponding normalized fluorescent spectra (excitation 480 nm) (e) of free FITC (green line), FITC conjugated with APTES in ethanol (olive line), FITC-conjugated IgG (magenta line) and FITC-SiO2-COOH NPs (orange line), and FITC-SiO2-NH2 NPs (violet line) in water. Zeta potential values (f) and poly index dispersion values (g) of FITC-SiO2-COOH NPs at different pH values. HD hydrodynamic diameter.

Table 1 Characterization data of FITC-SiO2-NH2 and FITC-SiO2-COOH NPs: average particle size determined by TEM (dTEM), standard deviation (σTEM), average hydrodynamic diameter inferred from DLS (dDLS), PdI, and zeta potential.

For Raman and Fourier Transform Infrared Spectroscopy (FTIR) measurements, SiO2-COOH and FITC-SiO2-COOH NPs were centrifugally separated and subsequently dried to produce anhydrous powders. The Raman and FTIR spectra of SiO2-COOH NPs, FITC-SiO2-COOH NPs, and pure FITC molecules are presented in Fig. 1b, c. Each Raman spectrum was captured under uniform conditions as following: 60 s duration, 785 nm laser excitation, 600 line/mm grating, and utilizing a 100× optical microscope objective. The findings indicate pronounced Raman effects (Fig. 1b). The Raman signal of SiO2-COOH NPs (magenta line) was weak compared to those of FITC-SiO2-COOH NPs (orange line), FITC-SiO2-NH2 NPs (violet line), and the FITC powder sample (green line, divided by 20, shown in Fig. 1c), which indicated that the Raman activity of SiO2-COOH NPs was considerably muted. In contrast, the Raman signature of FITC-SiO2-COOH and FITC-SiO2-NH2 NPs were robust and echoed characteristic peaks of the FITC molecules (orange line). Notably, the peaks at 1170 cm−1 and 1320 cm−1 are ascribed to the CCH bend of the xanthene ring and the C-O (phenoxide ion stretch conjugated with xanthene ring stretch, respectively), which is a characteristic of the dianion forms of FITC molecules and became more pronounced8. The disappearance of the peak at 2021 cm−1 assigned to the –N=C=S groups proved the covalent bonding of FITC conjugated with the silica matrix via the APTES molecule9.

The pure FITC molecule exhibited distinct spectra in both Raman and FTIR readings (Fig. 1c). In the FTIR spectrum of FITC molecules (green line), a pronounced peak at 2031 cm−1 corresponded to the stretch vibration of the –N=C=S group. Following the covalent attachment of FITC to the silica nanoparticle network, this peak was no longer discernible in the FTIR spectrum of the FITC-SiO2-COOH NPs (orange line) and FITC-SiO2-NH2 (violet line), affirming the conjugation of FITC with the silica matrix via the APTES molecule. Unlike the Raman data, the FTIR spectra for both FITC-SiO2-COOH NPs and SiO2-COOH NPs closely resembled the standard silica gel FTIR spectrum9. There was no discernible peak associated with FITC molecules in the FITC-SiO2-COOH and FITC-SiO2-NH2 NP’s FTIR spectrums. Such findings underscore the enhanced sensitivity of the dianion structure of the FITC molecule in the Raman spectrum compared to the infrared spectrum.

The absorption and fluorescence spectra of free FITC, FITC conjugated with APTES in ethanol, FITC-SiO2-COOH NPs, and antibody-FITC in water are presented in Fig. 1d, e. Figure 1d depicts the UV–Vis spectrum of free FITC in ethanol, representative of the electronic absorption spectrum of the anion form of FITC molecules8. The absorption spectra of FITC conjugated with APTES in ethanol showed a peak at 498 nm (olive line), while FITC conjugated with the IgG antibody in water peaked at 494 nm (magenta line). Notably, conjugated FITC samples exhibited narrow absorption, resembling those of the dianion forms of FITC at high pH. The FITC-SiO2-COOH NPs (orange line) and FITC-SiO2-NH2 NPs (violet line) in water revealed a broad absorption spectrum with a slight blue- shift, peaking at 494 nm and 492 nm corresponding, comparable to the FITC conjugated with IgG antibody. Pure silica exhibited no absorption in its monolithic form; however, as nanoparticles, it enhanced the background spectrum. The variation in absorption spectra is attributable to the xanthene segment of the FITC molecule, which possesses three protonation sites, resulting in multiple forms (cationic, neutral, anionic, and dianionic) that are pH-dependent. Specifically, the fluorescence quantum yield increases from 0.37 in the monoanion form to 0.93 in the dianion form. In Fig. 1e, with an excitation wavelength of 480 nm, typical emission peaks of all samples were identified around 522 nm for both the free FITC molecule and the FITC-SiO2-COOH NPs, at 527 nm for the FITC-APTES conjugate, at 532 nm for the FITC-SiO2-NH2 NPs, and 518 nm for the FITC-IgG conjugate. The fluorescence spectrum of FITC-SiO2-COOH NPs resembled that of FITC conjugated with IgG antibodies. These results suggested that, after conjugation with the amino groups of APTES or antibodies, the fluorescence spectral profiles of FITC molecules become more refined.

The system’s colloidal stability of FITC-SiO2-COOH NPs was determined by examining the zeta potential, the hydrodynamic diameter of nanoparticles, and their polydispersity index values. The obtained results were presented in Fig. 1f, g. The zeta potential values of the NPs in the buffer systems increased compared to that of the nanoparticles dispersed in deionized water. In these high salty concentration buffers, the results showed that the zeta potential was found to be about 0 at pH 5, − 1.4 mV at pH 4, − 12 mV at pH 6, − 16.2 mV at pH 7, and stable in pH 8, pH 9 with zeta potentials of − 18.3 mV and − 19.6 mV, respectively. By following the variation of the zeta potential of NPs as a function of the pH, the isoelectric point was identified at pH 5. At pH 5, pH 4, and pH 6, the hydrodynamic diameter of the particle was enormous, up to 1200 nm, much larger than the size measured by the TEM, indicating that the particle system was aggregated in this pH range. The farther away from the isoelectric point, in the pH range of 7–9, the nanoparticle system became more stable, and the hydrodynamic size of the particles was relatively similar and close to the diameter obtained from TEM images. This result demonstrated that the FITC-SiO2-COOH nanoparticle colloidal system was relatively stable in the pH range greater than 6 and formed aggregates in the low pH range of 4–6. At the same time, it showed that the chemical resistance of this nanoparticle system was excellent. The storage durability of these particles is very long, up to years if stored in deionized water.

The well distribution and selective penetration of FITC-SiO2-COOH NPs in different cell types

In this study, we used two primary cell types, including hFBs and hUVECs, and one cervical cancer cell line HeLa. Using different cell types would inform us if our nanoparticles could be widely applied as living cell trackers in biomedical research. Besides, primary cells, such as hUVECs and hFBS, are derived from tissue and not modified. Thus, these cells could provide a suitable model for studying the normal physiology of the cell response to the tested compounds.

Before adding on the cells, we tested the distribution of the NPs in the cell culture medium. After 30 min, at the two tested doses 50 and 100 µg/mL, the FITC-SiO2-NH2 NPs gradually aggregated as large as clumps with the average size even bigger than the cell nuclei. In the meantime, FITC-SiO2-COOH NPs were well dispersed in the cell culture medium (Fig. 2a). We then incubated these NPs at 100 µg/mL for 24 h with different cell types. Similar results were obtained, that FITC-SiO2-NH2 NPs were aggregated in all cell culture, consequently, stay outside of the cells (red arrows), even though some of them successfully penetrated to the cytoplasm. Interestingly, FITC-SiO2-COOH NPs stayed mainly inside the cells surrounding the cell nuclei, especially in hUVECs and hFBs, and less in HeLa cells (Fig. 2b). These results demonstrated that FITC-SiO2-COOH NPs were more effective in cell penetration compared to that of FITC-SiO2-NH2 NPs in hUVECs, hFBs, and HeLa cells. Therefore, we chose FITC-SiO2-COOH NPs for the next biocompatibility evaluations as a cell labeling in hUVECs, hFBs, and HeLa cells.

Figure 2
figure 2

The cell penetration of FITC-SiO2-COOH NPs in different cell types. (a) The dispersion of FITC-SiO2-COOH and FITC-SiO2-NH2 NPs at two doses after 30 min of incubation with the cell culture medium. (b) The different distributions of two NP types in hUVECs, hFBs, and HeLa cells. The red arrows indicated the localization of NPs outside the cells.

The effect of FITC-SiO2-COOH NPs on the cytotoxicity and cell senescence

Cell viability was evaluated at 24 h, 48 h, and 72 h post-incubation with the NPs. No significant disparities in cell morphology or density were observed between the three cell types (Fig. 3a). Crucially, both the treated and control cells exhibited no significant deviation in viable cell percentages (Fig. 3b). Additionally, cellular senescence markers in the presence of FITC-SiO2-COOH NPs were identified (Fig. 3c). It was evident that hUVECs had a notably higher rate of aging cells (6.4 ± 1.8%) compared to hFBs (0.4 ± 0.15%) and HeLa cells (0.9 ± 0.7%) (Fig. 3d). Yet, the NPs did not expedite the aging process, as similar cellular senescence markers were evident in both the treated and control groups. We also checked the effect of these nanoparticles on the cell population doubling times. The results showed that there was not significant difference in the cell duplication between the control and the treatment at 50 µg/mL and 100 µg/mL NPs in all three cell types (Fig. 3d).

Figure 3
figure 3

The effect of FTIC-SiO2-COOH NPs on cytotoxicity and cell senescence. (a) The morphology and cell density of hFBs, hUVECs, and HeLa cells in the presence of the NPs. (b) The cell viability (%) of hFBs, hUVECs, and HeLa cells with different doses assessed at 24, 48, and 72 h. (c) The cellular senescence signals detected by β-galactosidase staining; black arrows indicate aging cells. Cell nuclei were stained with Hoechst (blue). (d) The percentage of senescence cells (%) and the population doubling time among hFBs, hUVECs, and HeLa cells. Data was collected from three biological trials (n = 3) and presented as mean ± SD.

These data underscores that FITC-SiO2-COOH NPs did not influence cell viability or aging in either hFBs, hUVECs, and HeLa cells.

The effect of FITC-SiO2-COOH NPs on cell migration and in vitro angiogenesis

To discern the nanoparticles’ influence on cell functionality, a wound healing assay was conducted using FITC-SiO2-COOH NPs at concentrations of 50 and 100 µg/mL. Results indicated that the NPs did not hinder the migration of hFBs or hUVECs, irrespective of the concentration (Fig. 4a). Moreover, the wound closure rate was consistent across all assessment times (Fig. 4b). At the elevated concentration of 100 µg/mL, a delay in cell migration was observed compared to controls in hUVECs (p < 0.0001) at 20 h, but this delay was not statistically significant (p > 0.05) at 24 h (Fig. 4b), suggesting that the NPs did not impede cellular migration in 2D cultures.

Figure 4
figure 4

The effect of FTIC-SiO2-COOH NPs on cell migration and in vitro angiogenesis. (a) The images of wounded sites on the cell monolayers at 0, 12, 20 and 24 h. (b) The proportion of wound closure in the presence of NPs at different concentrations with time. (c) The effect of FITC-SiO2-COOH NPs at 100 µg/mL on in vitro angiogenesis of hUVECs. Data was collected from three biological trials (n = 3) and presented as mean ± SD. *p < 0.05; ****p < 0.0001.

Further analysis showed that the NPs had no discernible impact on the tube formation ability of hUVECs. Even at the higher concentration of 100 µg/mL of FITC-SiO2-COOH NPs, metrics such as total tube length, branching length, and segment length remained consistent with controls across time periods of 2, 4, and 8 h post-seeding on Matrigel, though there was a minor reduction in branching length in treated cells at 6 h (p < 0.05) (Fig. 4c). Intriguingly, the NPs exhibited cell-labeling activity that persisted beyond 8 h of observation, suggesting potential utility as a labeling agent for cell tracking.

FITC-SiO2-COOH NPs function as a cell labeling agent in 2D culture

Based on the in vitro angiogenesis assay results, the potential of FITC-SiO2-COOH NPs was assessed for their efficacy as a fluorescence labeling agent in live cell tracking. Two concentrations of the NPs were examined. Notably, in hUVECs, following a 24-h incubation, fluorescence was discernible in nearly all cells. The NPs appeared to localize predominantly in the cytoplasm, surrounding the cell nuclei (Fig. 5a). We also compared the fluorescence signal of our NPs with the commercial fluorescence labeling agent provided with the angiogenesis kit at different times of incubation. Remarkably, these NPs maintained their fluorescence within the cells for a longer duration than the commercial dye. As shown in Fig. 5b, hUVECs pre-labeled with the commercial kit exhibited a robust signal after a 30-min incubation, but this rapidly diminished after 16 h, and vanished by the 24-h mark. In contrast, the fluorescence from FITC-SiO2-COOH NPs persisted up to 48 h post-incubation. These findings suggest that these NPs could serve as effective long-term live cell labeling agents for hUVECs.

Figure 5
figure 5

Analysis of fluorescence signals of FITC-SiO2-COOH NPs. (a) The uptake of FITC-SiO2-COOH NPs and a commercial fluorescence kit in the hUVECs after different times of incubation. (b) The quantitative comparison of fluorescence signals at different time points of hUVECs and a commercial kit. (c) hFBs and HeLa cells incubated with FITC-SiO2-COOH NPs at various times. (d) The intracellular distribution of FITC-SiO2-COOH NPs in hFBs. The relative localization of FITC-SiO2-COOH NPs (green) in the cytoplasm and CD63 membrane protein (red, left), or alpha-tubulin (red, right). (e) Quantitative comparison of fluorescence signals at different time points at 50 µg/mL and 100 µg/mL NPs in hFBs and HeLa cells. Data was collected from three biological trials (n = 3) and presented as mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001.

A similar intracellular distribution of FITC-SiO2-COOH NPs was observed in fibroblasts and HeLa cells, with the fluorescence signal observed from 16 h and last for 72 h after incubation. The signal was more pronounced in cells treated with 100 µg/mL NPs compared to those treated with 50 µg/mL NPs across all assessment periods (Fig. 5c). To validate the cellular uptake of these NPs, cells were stained with CD63 antibodies to identify cell membrane proteins and alpha-tubulin antibodies to delineate cytoskeletal components. Observations revealed a uniform distribution of NPs within fibroblasts after 24 h of incubation (Fig. 5d).

In hFBs, the peak fluorescence intensity was achieved after 48 h of incubation with the lower dose of NPs, and 24 h with the higher dose. In HeLa cells, the peak fluorescence intensity was achieved after 48 h of incubation with both doses of NPs (Fig. 5e). The signal was still observed after 72 h with the fluorescence intensity decreased of about 45.3%. The fluorescence exhibited a noticeable decline by 72 h, especially for the higher dose in both cell types. Such findings affirm that the NPs successfully penetrated fibroblasts and HeLa cells, serving as cellular trackers for durations exceeding 48 h.

FITC-SiO2-COOH NPs function as a cell labeling agent in 3D cultures

While FITC-SiO2-COOH NPs have established their labeling efficacy in 2D cultures, their performance in 3D cell cultures and their penetrative capabilities into deeper spheroid layers warranted investigation. In the first group, we incubated the NPs with spheroids containing both hFBs and hUVECs. After a 24-h incubation, the NPs permeated several layers (spanning layers 9–55), reaching from the apex nearly to the base of the 3D structure (Fig. 6a). In the second group, NP-incubated fibroblasts were used to create multicellular spheroids with a mean diameter of 500 µm. Results showed that the NPs emitted a strong signal across multiple cell layers even 24 h post-treatment (Fig. 6b). The FITC fluorescence mirrored the intensity of the nuclear-staining dye Hoechst, across all cellular layers (Fig. 6b). Notably, when spheroids were formulated using a mix of hFBs and NP-incubated hUVECs, a distinct cellular distribution was discerned: hUVECs populated the spheroid core while hFBs were peripherally located (Fig. 6c, d). These observations suggest that FITC-SiO2-COOH NPs neither inhibit cell spheroid formation nor compromise their structural integrity, validating their applicability as labeling agents in 3D cultures.

Figure 6
figure 6

The distribution of FITC-SiO2 labeled cells in 3D culture. (a) The z-stack images of a mixture of spheroids incubated with FITC-SiO2-COOH NPs for 24 h. (b) The z-stack images of FITC-SiO2-COOH NPs-hFBs spheroid. The cell nuclei were visualized by Hoechst staining. The above number indicates the depth of different cell layers in the spheroid. (c) The mixed spheroid consisted of hFBs and FITC-SiO2-COOH NPs-hUVECs. Xy, yz, and xz slices of the confocal z-stack of FITC-SiO2-COOH (green) and Hoechst-stained cellular DNA (blue). (d) Confocal microscopy 3D reconstruction of the nucleus and distribution of FITC-SiO2-COOH NPs.

FITC-SiO2-COOH NPs as a labeling agent in the Medaka fish model

After a 4-h exposure to a medium infused with 200 µg/ml of the NPs, FITC fluorescence was discernible in the developing gastrointestinal tract of fish larvae, a phenomenon absent in control specimens (Fig. 7a). This FITC fluorescence persisted within the small intestinal tubules of 7 dpf larvae, spreading progressively within the structure over subsequent days (spanning beyond 3 days) post-exposure. At this developmental stage, larvae predominantly rely on their yolk sac for nourishment, and minimal intestinal motility was noted (Video S1 supplemental data). Given the passive ingestion mechanisms, as larvae intake water and medium fluids, it is conceivable that the NPs gained access to the intestinal tract. These observations underscore that the NPs maintain their fluorescence for a minimum of three days in the relatively inactive intestines of the larvae. Any subsequent signal attenuation could likely be attributed to the NPs’ excretion from the fish’s alimentary canal rather than intrinsic fluorescence diminution. This explanation seems reasonable considering the 11 dpf larvae. Those exposed to the 200 µg/ml NP solution but deprived of feed exhibited a robust FITC fluorescence immediately after a 4-h exposure in their fully operational gastrointestinal tract (Fig. 7b) (Video S2 supplemental data). This fluorescence exhibited a marked reduction 24 h post-exposure and was virtually undetectable 48 h post-exposure (Fig. 7b, c).

Figure 7
figure 7

FITC-SiO2-COOH NPs as a labeling agent in the Medaka fish model. (a) The FITC signal in the small intestinal tubes of 7 dpf larvae. (b) The FITC signal in the small intestinal tubes of 11 dpf larvae. (c) Mean gray value of the FITC signal in the larvae 0, 4, 24, and 48 h after NP exposure through a 4-h feeding window. (d) The FITC signal in the intestinal tubes of a newly hatched live brine shrimp. (e) The FITC signal in the intestine of the 11-dpf fish group, which ingested the NP-ingested brine shrimp. (f) The morphology of the 11-dpf fishes in control and NPs treated groups.

Upon immersion in a medium with a concentration of 200 µg/ml for 4 h, newly hatched live brine shrimps exhibited pronounced FITC signals within their intestine. These signals persisted for several days post-exposure until the organisms succumbed (Fig. 7d). For the 11-dpf fish cohort that consumed NP-laden brine shrimp, the NP-associated fluorescence was evident in their intestines immediately after a 4-h feeding window. Yet, this fluorescence was imperceptible in the fish 24 h post-ingestion. This diminished FITC signal retention in the intestines of these fish, relative to their counterparts directly exposed to NPs, can likely be attributed to heightened digestive activity associated with brine shrimp metabolism (Fig. 7e). Beyond the gastrointestinal tract, no FITC signals were detected in vital detoxifying and excretory organs, namely the liver and kidneys, regardless of whether fish were directly exposed to NPs or consumed NP-laden brine shrimp. This suggests that the pathway for NP uptake in these fish predominantly ensues from passive water drinking or through trophic transfer; NPs likely remain unmetabolized and are subsequently expelled by the digestive system. Consequently, the retention span of NPs within the intestinal tract depends on the efficiency of the digestive system. More importantly, no developmental abnormalities or adverse or lethal effects of the NPs in the fish exposed to NPs were recorded. All tested fish remained alive and showed normal behaviors and physical activities during and for weeks following exposure (Fig. 7f).

Crucially, no developmental abnormalities or detrimental effects, lethal or otherwise, were ascribed to the NPs in the examined fish. All evaluated specimens manifested normal behavior and physiological activity both during the exposure and in the ensuing weeks.