Facile synthesis of Gd/Ru-doped fluorescent carbon dots for fluorescent/MR bimodal imaging and tumor therapy | Journal of Nanobiotechnology

Metal doping was performed via a one-step microwave-assisted synthesis procedure using BPEI, citric acid, Ru(dcbpy)3Cl2, and GdCl3 as precursors (Fig. 1a). In this formulation, citric acid plays a multifaceted role: serving as a carbon source, a complexing agent for the gadolinium ion, and promoting condensation with the BPEI amino group through its carboxyl component. Concurrently, BPEI functioned as a nitrogen dopant while forming complexes with Ru(dcbpy)3Cl2, engendering red fluorescence. To optimize the experimental conditions, we attempted to synthesize carbon dots with different metal doping ratios, as shown in Table S1. Gd/Ru CDs with an input mass ratio of 10:1 not only exhibits good fluorescence properties, but also has considerable potential for magnetic resonance imaging (Figures S1&S2). Therefore, it was selected for the following study. The ensuing Gd/Ru-CDs exhibited favorable water dispersion, characterized by a nearly uniform 4.2 nm diameter as evidenced by TEM imaging (Fig. 1b). Compared with undoped CDs (Figure S3), the average particle size of Gd/Ru CDs is almost close, and there is no significant change in the particle shape. High-resolution TEM analysis further discerned an individual Gd/Ru-CDs lattice spacing of 0.21 nm, aligning with graphitic structural properties (Fig. 1c) [36]. DLS assessments substantiated these findings, denoting an average hydrodynamic diameter of approximately12.4 nm, harmonizing with TEM observations (Figure S4). The zeta potentials of Gd/Ru-CDs were measured to be 20.3 mV, similar with undoped CDs (18.6 eV), attributed to the abundant presence of amino groups on their surface (Table S2). ICP-OES measurements quantified the Gd and Ru constituent at roughly 200 μg and 70 μg per mg of Gd/Ru-CDs, respectively. XRD analyses elucidated the crystalline nature of the CDs and Gd/Ru-CDs, both revealing predominant reflections around 22.1° attributable to a turbostatic layer of graphitic carbon (Fig. 1d) [37]. This, juxtaposed with a broad and somewhat noisy spectrum, hinted at the material’s underlying amorphous carbon composition.

Fig. 1
figure 1

(a) The synthesis route of Gd/Ru-CDs by facile microwave methods in 5 min; (b) TEM image of Gd/Ru-CDs (insert image represent the size distribution); (c) HRTEM image of Gd/Ru-CDs, Lattice spacing: 0.21 nm; (d) XRD of CDs and Gd/Ru-CDs; (e) XPS survey of CDs and Gd/Ru-CDs; High-resolution XPS spectra of Gd/Ru-CDs, (f) C1s, (g) N1s, (h) O1s, (i) Ru3p, and (j) Gd4d

FTIR was performed to illustrate the surface functional groups of Gd/Ru-CDs. As can be seen from Figure S5, the broad band around 3000–3500 cm− 1 clearly confirmed the existence of O-H and N-H bonds, which are mainly caused by its stretching vibration. Meanwhile, the peak at 2768 cm− 1 corresponding to -NH vibrations, indicating the presence of -NH2 in Gd/Ru-CDs. In addition, the existence of C = O and C=N bonds is evidenced by two peaks located at 1740 and 1350 cm− 1, respectively. The peak at 1180 cm− 1 belonged to the stretching vibration of the C-C and C-N bonds. All the above evidence shows the surface of Gd/Ru-CDs contains a variety of functional groups of hydroxyl, carboxyl and amino. Subsequent XPS evaluations mapped the chemical and elemental landscape of the Gd/Ru-CDs, ascertaining a composition rich in C, O, N, Ru, and Gd elements (Fig. 1e). In stark contrast, there are no characteristic peaks of Gd and Ru in the survey XPS spectrum of undoped CDs. Delving deeper, the C1s, N1s, and O1s spectra of both CDs and Gd/Ru-CDs demystified the presence of varied bonding environments, encompassing C = C, C-C, C-O, C-N, C-OH, and O-C-O configurations, highlighting the roles of oxygen-bearing groups and nitrogen doping agents (Fig. 1f-h&S6). Notably, the high-resolution spectrums of Gd/Ru-CDs delineated peaks pertinent to Ru 3p at 464.7 and 485.3 eV, and Gd 4d at 142.7 and 147.7 eV, corroborating the incorporation of Gd and Ru within the Gd/Ru-CDs framework (Fig. 1i-j).

The optical properties of Gd/Ru-CDs were scrutinized, with UV-visible absorption spectra portraying a characteristic band at 260 nm, indicative of π-π* transitions and suggestive of the aromatic ring structures inherent to the Gd/Ru-CDs (Fig. 2a). These structures emerge through microwave-assisted carbonization, spearheaded by reactions involving diverse functional groups, such as hydroxyl, carboxyl, and amino entities present in the precursors. Noteworthy are the distinct peak at 352 nm attributable to n-π* electronic transitions facilitated by the energetic apprehension of surface oxygen functional groups in their excited states. Similarly, the UV-Vis absorption spectra of undoped CDs exhibit peaks at the same positions (Figure S7). The absorption peaks at 306 and 464 nm are consistent with that of bipyridine ruthenium dye, indicating the successful introduction of ruthenium dye. (Figure S8). Fluorescence emission spectra of both CDs and Gd/Ru-CDs, acquired under varied excitation wavelengths, exhibited a dependency on the excitation wavelength, a phenomenon possibly rooted in the heterogeneous chemical makeup and differing surface emission traps, or perhaps through mechanisms yet unidentified (Fig. 2a&S7) [38]. It is worth noting that the fluorescence spectrum of Gd/Ru-CDs exhibits a fixed red emission peak near 637 nm, and its behavior remains unchanged compared to undoped CDs, which may be attributed to the structure of Ru complexes. The utilization of longer excitation/emission wavelengths stands to benefit bioimaging applications, enhancing the penetration depth in biological tissues, a fact underscored by a recorded photoluminescence quantum yield (QY) of around 29.57% in water.

Fig. 2
figure 2

(a) UV-vis absorption and fluorescence spectra of Gd/Ru-CDs solutions. (b) T1WI and T1 Map images of Gd/Ru-CDs (Gd3+ concentration: mM); (c) and longitudinal MR relaxation curve of Gd/Ru-CDs. (d) Cell viability of Gd/Ru-CDs towards L929 and 4T1 cells; (e) The hemolytic percentage of Gd/Ru-CDs to human red blood cells (PBS as a negative control and TX-100 as a positive control); (f) The fluorescence stability of Gd/Ru-CDs in 30 days. (g) Fluorescence images of the cellular uptake of Gd/Ru-CDs (200 μg mL− 1) by 4 T1 cells in 6 h, scale bar = 20 μm

The contemplation of relaxivity in nanoprobes is central in magnetic resonance imaging quality assurance. As shown in Figure S9, compared with water, CDs without doping Gd element or only doping Ru element did not show significant T1 shortening. However, when doped with Gd elements, Gd/Ru CDs exhibit significant concentration dependent T1 changes. This indicates the excellent MRI performance of Gd doped carbon dots, and indirectly proves the successful doping of Gd elements. Subsequently, to pursue the T1 relaxation time of Gd/Ru-CDs, a series of measurements were orchestrated on a 3.0 T MR scanner, escalating the Gd concentration from 0 to 1.0 mM, and thereby establishing the vivid augmentation in the brightness of the T1-weighted MR images of the dispersions with increasing Gd concentration (Fig. 2b). Subsequently, a linear fit between 1/T1 and the Gd concentration revealed an r1 relaxation rate of 6.38 mM-1s-1, which can comparable with clinical contrast agents Gd-DTPA with r1 of 4.9 mM-1s-1, pinpointing the pivotal role of proximate interactions between the paramagnetic ions at the surface and the surrounding water molecules in dictating T1 relaxation enhancement Fig. 2c) [39]. The notably diminutive size and hydrophilicity of Gd/Ru-CDs fostered such interactions, granting them a relatively lofty r1 relaxation rate, thus heralding their aptitude for deployment as potent T1 nanoprobes in MRI applications. This showcases a prospective pathway to honing in on diagnostics with heightened precision, potentially paving the way for more insightful and accurate biomedical imaging.

Before venturing into broader biological applications, the synthesized Gd/Ru-CDs had their biocompatibility rigorously analyzed using a CCK-8 assay on 4T1 and L929 cells. A promising revelation was noted in Fig. 2d, which illustrated that both the 4T1 cells and L929 cells retained over 83% viability after a 24-hour co-incubation period, signaling a negligible cytotoxic effect of Gd/Ru-CDs. This is mainly due to the inert carbon shell that prevents Gd leakage or migration (Figure S10). Hemocompatibility was scrutinized next, through hemolysis observations conducted on red blood cells exposed to an array of substances including TX-100, PBS, and graded solutions of Gd/Ru-CDs. A comforting result emerged, evidenced in Fig. 2e, which exhibited minimal hemolysis across PBS and all Gd/Ru-CDs treated groups, mirroring the control batch. A noteworthy revelation was the nanoprobe’s superb water-dispersible trait and retention of stability, demonstrating no marked aggregation or precipitation even after a span of two months under standard conditions (Fig. 2f). These observations usher in a great signal, corroborating the high biocompatibility credential of Gd/Ru-CDs for biological deployments. Venturing into the realm of fluorescent bioimaging, cellular uptake studies in 4T1 cells unveiled a striking potential of the Gd/Ru-CDs as nanoprobes. As delineated in Fig. 2g &S11, a marked augmentation in intracellular green/red fluorescence over time (1 h, 2 h, 6 h) was visible, affirming their potent imaging efficacy in tumor cells.

Fig. 3
figure 3

(a) UV-vis absorption intensity of ABDA at 380 nm after the addition of Gd/Ru-CDs and then irradiation for different time periods. Inset: the corresponding UV-vis absorption spectra. (b) Intracellular ROS generation of 4T1 cells incubated with Gd/Ru-CDs and probe (DCFH-DA); (c) Viability of 4T1 cells with and without 650 nm excitation for 10 min after incubation with different concentrations of Gd/Ru-CDs for 6 h

Photodynamic therapy (PDT), a process founded on light-induced therapeutic action, stands as a safe pathway to obliterate diseased tissues [40]. Ruthenium metal complexes have been proven to be effective PDT photosensitizers as they can absorb energy under light and release energy to transfer it to the surrounding oxygen, producing highly active singlet oxygen [41, 42]. Leveraging this concept, the Gd/Ru-CDs showcased an impressive ability to generate singlet oxygen (¹O2), a potent oxidizing agent critical in PDT. Through interaction with ABDA, a recognized tool for evaluating ¹O2 production owing to its absorbance reduction at 380 nm when reacting with ¹O2, the high efficiency of ¹O2 generation from Gd/Ru-CDs was corroborated. As shown in Fig. 3a, a remarkable 80% absorption decay observed within a span of 20 min post-irradiation using an LED lamp substantiated this efficiency. The study proceeded with assessing the reactive oxygen species (ROS) production within 4T1 cells utilizing the DCFH-DA probe, revealing a discernible green fluorescence only upon Gd/Ru-CDs exposure and subsequent white light irradiation, confirming the ROS induction capability solely in the presence of light and Gd/Ru-CDs (Fig. 3b). When scrutinizing the PDT efficiency, a drastic decline in the viability of 4T1 cells, spotlighted in Fig. 3c, was evidenced post a 10-minute irradiation session at 650 nm wavelength, underscoring a notable PDT efficacy while retaining a low toxicity profile.

The in vitro photodynamic therapeutic effect was firstly evaluated on 4 T1 tumor cells by the live/dead cell staining assay. 4T1 cells were incubated with (i) Saline, (ii) Saline + L, (iii) Gd/Ru-CDs, (v) Gd/Ru-CDs + L and co-stained with calceinAM and PI. As depicted in Fig. 3d, 4T1 cells in the control (Saline) revealed vivid green fluorescence, indicating the good living condition of the cells. Moreover, cells in saline + L, Gd/Ru-CDs groups also exhibited strong green fluorescence, demonstrating that they could not induce any strong injury to the tumor cells. However, when combined with laser irradiation, Gd/Ru-CDs efficiently brought about most of cell death, which were stained to be red in the field. Next, an Annexin V-FITC/PI fluorescence staining was also performed on the cells and analyzed by a flow cytometry system. As shown in Fig. 3e, more than 90% of cells were found to be distributed in the left lower quadrant for the saline, saline + L, Gd/Ru-CDs without laser irradiation group, indicating the good living condition of the cells after the above treatments. However, after laser irradiation, the apoptosis rate achieved 94.4% for the Gd/Ru-CDs incubated cells, respectively. The flow cytometry data suggested that cells treated by various Gd/Ru-CDs displayed irreversible damage upon photodynamic treatment. The photodynamic antitumor effect of the various groups was further quantitatively assessed by the CCK-8 assay. As shown in Figure S12, without laser irradiation, Gd/Ru-CDs did not induced evident damage to the 4T1 cells compared to the control group. However, in contrast, the cell viabilities decreased to about 58% for the Gd/Ru-CDs exposed to 650 nm laser, respectively, attributed to their photodynamic effect. Thus, these results indicated that the Gd/Ru-CDs would be a great potential PDT application in cancer treatment.

Fig. 4
figure 4

In vivo fluorescence and MR imaging (three mice per group). (a) In vivo fluorescence imaging of 4T1 tumor-bearing mice after intravenous injection of Gd/Ru-CDs; (b) FL signal intensities within tumor regions at corresponding time points after intravenous injection of Gd/Ru-CDs; (c) T1-weighted MR imaging of mice model pre- and post-injection of Gd/Ru-CDs at various time intervals; (d) T1-weighted MRI signals of the tumors at determined time points after administration of Gd/Ru-CDs

An empirical venture was undertaken to demonstrate the tumor imaging and imaging-guided therapy capabilities of Gd/Ru-CDs in a 4T1 xenograft mouse model. Due to enhanced permeability and retention (EPR) effects, prominent fluorescent signals of Gd/Ru-CDs accumulation in tumor tissues were witnessed 3 h post-injection, as illustrated in Fig. 4a. Tumors are easily distinguishable from surrounding tissues, with fluorescence intensity maintained for 12 h (Fig. 4b). Moreover, we use a fluorescence imaging system to measure its biological distribution in major organs and tumors. After the injection of Gd/Ru-CDs in 1 h, strong fluorescence signals were found in the bladder, indicating that the Gd/Ru-CDs may have been cleared through the kidneys (Figure S13). Dissect the mice 4 h after injection, respectively. As expected, strong fluorescence signals appeared in the tumor area. More importantly, a large amount of Gd/Ru-CDs remained in the kidneys, which further confirms that Gd/Ru-CDs can be cleared through the renal pathway (Figure S14). Subsequently, we collected urine samples after 4 h and observed strong fluorescence characteristics of Gd/Ru-CDs (Figure S15). This result was also confirmed by TEM, confirming renal clearance and high stability of carbon dot metabolism. In addition, most of the injected carbon dots were detected in urine and their content in feces was negligible, indicating that the clearance of Gd/Ru-CDs is mainly carried out through the kidneys (Figure S16). The above results indicate that Gd/Ru-CDs can effectively accumulate at the tumor site and be excreted from the body through the kidneys.

T1-weighted MR imaging conducted serially at distinct intervals pre and post intravenous Gd/Ru-CDs administration depicted a peak in tumor T1 signal intensity at the 3-hour mark, post which it gradually declined, normalizing in a 12-hour window (Fig. 4c). The nanoparticles facilitated crisp delineation of tumor peripheries, enhancing edge imaging and offering a precise demarcation between normal and tumor tissues. A commendable attribute was the prolonged retention of a high T1 signal in tumors, upheld till 5 h subsequent to the administration, showcasing potential for long-term tracking (Fig. 4d). The enhanced visualization of the tumor border in the critical 1–5 h post-injection timeframe underscored a substantial window for effective MR imaging, thereby promising stellar capabilities in tumor diagnostic imaging. This pathway illustrates not just the promising potential of Gd/Ru-CDs in crafting detailed fluorescence and MR images but potentially revolutionizes tumor diagnostics, affording a larger temporal window to carry out precise and informative MRI sessions, thereby possibly steering towards better prognostic outcomes.

Fig. 5
figure 5

In vivo anticancer effect in 4T1 tumor bearing mice. (a) Schematic illustration of animal experiments design; (b) body weight (c) Relative tumor size, (d) Tumor weight, and (e) Photographs of 4T1 tumor bearing mice and the images of the tumor at 21 days for various groups: PBS, Laser only, Gd/Ru-CDs, Gd/Ru-CDs + Laser. (h) H&E stained and Tunnel images of the tumors from different groups. Scale bars: 100 μm

In vivo experiments evaluated the imaging-guided photodynamic therapy (PDT) efficacy of Gd/Ru-CDs in 4T1 tumor-bearing mice. This study divided mice into four different groups. The first group served as the control group and received intravenous injection of PBS solution. The second group exposed the tumor site to 650 nm laser for 20 min. Both the third and fourth groups were intravenously injected with Gd/Ru-CDs. One hour after injection, the fourth group received 20 min of 650 nm laser irradiation at the tumor site. This cycle of Gd/Ru-CDs injection followed by light exposure was repeated every three days for 21 days (Fig. 5a). A consistent observation was that all mouse groups maintained normal body weight, indicating the low toxicity of Gd/Ru-CDs and photoirradiation therapy (Fig. 5b). Fluctuations in tumor volume were carefully monitored during this period (Fig. 5c). It can be seen from the size of the mice in each group after treatment, the weight of the resected tumors and the photo records that the fourth group showed the most significant tumor suppression effect, leaving minimal tumor residues at the end of three weeks, or even completely tumor elimination (Fig. 5d-f). After the two-week experimental period, all tumors were carefully dissected to facilitate H&E and TUNEL analysis. As shown in Fig. 5g, the PDT group (Gd/Ru-CDs + laser) had severe histological damage and typical pathological changes, such as severe pyknosis, apoptosis or necrosis of tumor cells. In contrast, there was almost no tumor destruction and necrosis in the control group and other treatment groups, and the treatment effect was limited. Therefore, the emergence of Gd/Ru-CDs depicts a robust, efficient, and safe approach to cancer treatment, which is expected to revolutionize the field of PDT and herald the future of precise, efficient, and targeted cancer treatment with minimized side effects.

Fig. 6
figure 6

In vivo biosafety evaluation of Gd/Ru-CDs. (a) H&E staining of the tissue sections (Heart, Liver, Spleen, Lung, Kindey) of mice. The scale bar is 100 μm. (b) Detection of blood routine and liver and kidney function in mouse serum

The biocompatibility of Gd/Ru-CDs is pivotal for their potential transition into clinical settings. Therefore, to assess the in vivo biocompatibility and prospective toxicity of Gd/Ru-CDs, both serum biochemical and histological analyses were undertaken. As depicted in Fig. 6a, H&E-stained tissue sections from various mouse organs (namely the heart, liver, spleen, lung, and kidney) exhibited no histopathological abnormalities or lesions one day post Gd/Ru-CDs injection. Moreover, blood routine white blood cell (WBC), red blood cell (RBC), liver function markers, specifically alanine aminotransferase (ALT), aspartate aminotransferase (AST), along with crucial renal indicators such as creatinine (CREA) and urea (UREA) remained within the standard range across all treatment groups throughout the one-week evaluation period (Fig. 6b). These findings substantiate that Gd/Ru-CDs do not induce sustained impairment to normal renal and hepatic functions. Preliminary data thus denote the low toxicity and commendable biocompatibility of Gd/Ru-CDs nanoprobes in vivo, fostering their prospective applications in the biomedical domain.