Magnetic Nanoparticles Decorated with Gold Nanoclusters–Applications in Cancer Theranostics

Nanomedicine presents exciting new opportunities for the detection and treatment of cancer. Current cancer imaging methods and treatment approaches in clinics frequently fall short of entirely curing cancer and can have severe side effects. Theranostic nanoparticles, however, have the potential to revolutionize effective cancer treatment and early cancer detection. The objective of this study is to show how magnetic iron oxide nanoparticles and photoluminescent gold nanoclusters (MN‐AuNCs) may be combined effectively to produce bimodal imaging nanoparticles that possess magnetic and optical properties and can be used for both magnetic resonance imaging and optical biopsy. These findings demonstrate that MN‐AuNCs, when exposed to visible light, also have the capability to produce singlet oxygen, which is necessary for photodynamic therapy of cancer. In addition, it shows that they are non‐toxic, accumulate inside the cells, and cause cell death during exposure to visible light. The creation of these MN‐AuNCs offers a novel remedy for the current shortcomings in cancer diagnosis and treatment. Since they have both therapeutic and imaging capabilities, MN‐AuNCs have the potential to improve patient outcomes while lowering the risk of negative side effects.


Introduction
Cancer is a leading cause of death, taking away millions of lives worldwide annually. [1,2]As of today, individual diagnostic methods used in clinics to detect early-stage cancer are often insufficient.For example, magnetic resonance imaging (MRI) is a great DOI: 10.1002/admi.202300462technique for identifying and classifying cancers.MRI has deep tissue penetration and can provide detailed images of the internal organs of the body.It is non-invasive, can differentiate between various soft tissue types and distinguish them. [3][6][7] SPION are ideal for magnetic targeting [5] and hyperthermia [6] due to their superparamagnetic nature and strong magnetic saturation.
10] Moreover, optical biopsy provides real-time imaging, allowing immediate diagnosis.Various fluorescent probes such as organic dyes [11] and quantum dots (QDs) [12] are usually used in optical biopsy, but they have limitations.Organic dyes are shortlived and photobleachable, [13] while QDs pose a risk of heavymetal poisoning in case of their degradation. [14,15]21] Despite the advantages of mentioned imaging modalities, both MRI and optical biopsy also possess some disadvantages as well.MRI is costly, scans take longer to complete than other imaging techniques, and MRI resolution is limited at the cellular level.On the other hand, optical biopsy could overcome these problems but not without its flaws since this method is limited by weak light penetration into deeper tissues. [9,22,23]Combining both imaging modalities could improve diagnostic capabilities. [24]Bimodal imaging, which combines an MRI contrast agent and a fluorescent probe, is anticipated to show greater diagnostic ability. [25]ultimodal imaging nanoparticles (NPs) that incorporate both magnetic and optical properties are getting increasingly more attention and cultivate the research on novel synthesis methods of dual-imaging nanoparticles. [24,26,27]We previously have studied magnetic cobalt ferrite nanoparticles that were combined with gold nanoclusters and demonstrated their multimodality. [28]nother important aspect is the incorporation of therapeutic properties into such nanoparticles.Current chemotherapeutic medications frequently fail to completely eradicate tumors due to insufficient efficacy and/or drug resistance and have severe side effects. [29]Thus, a shift toward nanotechnological treatment approaches has been increasing in cancer management.
Photodynamic therapy (PDT) could be a more precise direction for cancer treatment. [30,31][33][34] This reaction promotes generation of singlet oxygen ( 1 O 2 ) and/or reactive oxygen species (ROS), which induce oxidation and lead to cell death. [18,35]Conventional PDT uses photosensitizers such as hematoporphyrin or chlorin e6. [31]Porphyrin-based PSs could be considered as theranostic materials because they exhibit both fluorescence (FL) and phototherapeutic effects. [36]However, PSs accumulate in both tumor and healthy tissues, indicating insufficient specificity of accumulation.Furthermore, porphyrin-based PSs have poor water solubility, low optical stability, and can cause dark toxicity. [37]anomaterials can carry a variety of anti-cancer medications, and some nanomaterials themselves have built-in anticancer and imaging characteristics. [26]18] As the number of oncological diseases continues to increase worldwide, [1,2] one of the promising and highly valued fields, nanomedicine, could help to improve cancer diagnostics and treatment. [38,39]Nanomedicine combines nanotechnologies and medicine to offer a range of advantages over conventional cancer treatments.An emerging field in nanomedicine, cancer theranostics, integrates the principles of cancer detection and treatment using nanotechnology. [40]It entails locating cancer at an early stage using diagnostic methods, followed by treating it through delivery of therapeutic drugs to the tumor site. [41]][43] However, more detailed studies on theranostic nanoparticles are necessary to stumble upon the most optimal approaches for cancer eradication.
Thus, the goal of this study was to unlock and show the biomedical potential of multimodal magneto-fluorescent nanoparticles for cancer theranostics.Herein, we report synthesis and characterization of magnetic iron oxide nanoparticles decorated with gold nanoclusters (MN-AuNCs), demonstrate their multimodality, and explore biomedical applications of MN-AuNCs for cancer diagnostics and treatment in vitro.

Characterization of Spatial Properties
Spatial properties of MN-AuNCs were evaluated with various techniques.Images of atomic force microscopy (Figure 1A-F) show spherical nanoparticles having sizes ≈50 nm for MN-AuNCs and ≈25 nm for magnetic nanoparticles (MN) alone (Figure 1G).Results showed that MN-AuNCs solution consisted of two differently sized populations of nanoparticles, smaller and bigger ones (Figure 1B,E,G).Small nanoparticles are ≈10-20 nm in diameter while large ones-≈35-60 nm (Figure 1G).Comparable results were also demonstrated by dynamic light scattering measurements.Smaller nanoparticles were 10-15 nm in diameter, while bigger ones were ≈60-100 nm (Figure 1H).However, it is possible that three types of the nanoparticles are formed during the synthesis and are present in stock solution of MN-AuNCs: individual AuNCs, individual MN, and magnetic iron oxide nanoparticles decorated with gold nanoclusters (MN-AuNCs).The biggest size distribution appears due to actual MN-AuNCs.The smaller ones (at ≈15-20 nm) appearing due to individual MN and the smallest ones, not detected in our experiments -individual AuNCs, which are of the size of bovine serum albumin (BSA) molecules.Magnetic nanoparticles measured alone were ≈70 nm in diameter, which is around three times bigger than AFM measurements.Such an increase in size may have occurred due to the magnetic nanoparticles aggregating in aqueous solution.The exhibition of two populations is also reasoned by polydispersity index, which was 0.312 on average.Zeta potential measurements revealed that MN-AuNCs are stable and have an average surface charge of −25 ± 0.5 mV.

Optical Properties of MN-AuNCs
Magnetic nanoparticles decorated with gold nanoclusters exhibit optical properties.Absorption spectra had no distinct peaks in the visible spectral region.Optical density of MN-AuNCs increased toward the blue wavelengths (Figure 2A).No changes were observed in aqueous solutions (deionized water (diH 2 O) and Dulbecco's Modified Eagle medium supplemented with fetal bovine seru (DMEM+FBS)).MN-AuNCs in aqueous deionized solution, under 488 nm excitation, emit a broad photoluminescence band in the region of 500-900 nm, that peaked at ≈685 nm.Samples prepared in DMEM, supplemented with fetal bovine serum, and excited at 488 nm wavelength had an additional PL band at 540 nm.It could be attributed to the components of DMEM+FBS because the solvent itself had a fluorescence peak in the same spectral region (blue spectrum in Figure 1A).Subtracting a PL spectrum of DMEM+FBS solvent from the one with nanoparticles showed that solvent had no effects on the shape of photoluminescence band of MN-AuNCs (magenta spectrum Figure 2A).
We tracked optical stability of nanoparticles by registering their photoluminescence for two weeks (in diH 2 O, phosphate buffered saline (PBS), Dulbecco's Modified Eagle medium (DMEM), and DMEM+FBS).No significant changes were observed neither in the intensity of PL signal nor in the peak positions of PL bands or full width at half maximum (Figure S1, Supporting Information).Moreover, in our previous study, we have demonstrated that addition of magnetic nanoparticles to AuNCs does not significantly affect their PL properties, and only reduces the intensity of AuNCs photoluminescence. [28]me-resolved fluorescence spectroscopy measurements were made using 405 nm pulsed diode laser and are presented in Figure 2B.Decays of photoluminescence at 685 nm were fitted with a bi-exponential tail-fitting model and two main exponential components were identified in both samples prepared in diH 2 O and in DMEM+FBS solvents.A fast exponential component ( 1 ) had a lifetime of ≈330 ns (≈300 ns for sample in DMEM+FBS) and a long lifetime component ( 2 ) was ≈1450 ns (≈1405 ns for sample in DMEM+FBS).They resulted in an average PL lifetime of ≈1355 ns (≈1325 ns for sample in DMEM+FBS).The resulting long-lived excited PL states of MN-AuNCs create favorable conditions, which could lead to formation of free radicals, such as reactive oxygen species or singlet oxygen.Quality of the fits is represented with residuals and  2 , which for 2exponential fits, were 0.998 and 1.122 for samples made in diH 2 O and DMEM+FBS, respectively (Table 1; Figure S2, Supporting   Information).Mono-exponential tail-fit models for both samples were inaccurate while tri-exponential model for MN-AuNCs in DMEM+FBS shows that no significantly better fit is achieved (Figure S2, Supporting Information), therefore, the approximation by two exponents is enough.Yang et al. demonstrated that noble metal nanoclusters exhibit metal-centered and ligand-centered emissions, which could be attributed to the shorter PL lifetimes in range of nanoseconds and longer lifetimes in range of microseconds, respectively. [44]

Generation of Singlet Oxygen
Literature shows that the therapeutic effect of PDT is largely conditioned by the generation of reactive oxygen species. [33]In our previous works, we demonstrated that fluorescent gold nanoclusters alone can generate singlet oxygen and ROS such as peroxides at high efficiency comparable to photosensitizer chlorin e6. [17,18]n this study, we assessed whether our magnetic nanoparticles decorated with gold nanoclusters generate 1 O 2 .Generation of singlet oxygen upon light irradiation was assessed with fluorescent 1 O 2 indicator "Singlet Oxygen Sensor Green" (SOSG). [45]ts detection principle is summarized in Figure S3 (Supporting Information).Fluorescence intensity of SOSG increased on irradiation dose in all the samples, indicating 1 O 2 generation (Figure 3A), though the rates of increment were different, indicating the efficiency of 1 O 2 generation also varied between irradiation wavelengths.When compared to photosensitizer chlorin e6 (Ce6), the maximum amount of generated 1 O 2 was similar but Ce6 generates 1 O 2 way faster (olive graph, Figure 3A), Nevertheless, irradiation with light resulted in generation of singlet oxygen, however, 1 O 2 generation efficiency under irradiation 630 nm was minimal because optical density was also significantly lower at this wavelength.Registered spectra indicating increased SOSG fluorescence upon irradiation are depicted in Figure S3B (Supporting Information).

Photostability of MN-AuNCs
In photodynamic therapy the use of light is essential, therefore, photostability of MN-AuNCs was tested.Photostability was evaluated by irradiating various samples with different wavelengths (Figure 3B).To collect the same final dose of irradiation (40 J cm −2 ) for all wavelengths, irradiation duration was set according to the fluorescent lamp's power of each wavelength.The results show that the PL intensity of MN-AuNCs decreased by more than half after irradiation with 405 nm (violet graph) and 470 nm (cyan graph).On the other hand, 630 nm light (red graph) did not cause such a strong photobleaching.Interestingly, we noticed that after irradiation, if kept in the dark, PL signal intensity of MN-AuNCs would fully recover.This effect is called reversible photobleaching and was previously introduced by Hemmateenejad et al. [46]

Determination of Fluorescence Quantum Yield
Fluorescent markers with a high quantum yield (QY) are usually used for optical diagnostics.For this reason, novel contrast agents should exhibit as high QY as possible.Relative quantum yield of photoluminescent part of MN-AuNCs was evaluated and determined with comparative method using rhodamine 6G, which QY is ≈95%. [47]QY was measured under 513 nm wavelength excitation, because at this wavelength rhodamine 6G and MN-AuNCs would guarantee the same number of absorbed photons (Figure S4, Supporting Information).The calculated photoluminescence quantum yield of gold nanoclusters was 6.11%.

Magnetic Properties of MN-AuNCs
Besides optical properties, MN-AuNCs also exhibit magnetic properties.We tested how well MN-AuNCs perform as MRI contrast agent.The synthesized MN-AuNCs have been marked 1, No. 2 is pure Fe 3 O 4 nanoparticles and No. 3 is commercially available gadopentetic acid "Magnegita" contrast.MRI magnetic characterization is presented in Figure 4.
To demonstrate visually the contrast possibilities of MN-AuNCs, T2-weighted and T1-weighted magnetic resonance (MR) images of different samples are presented in Figure 4B.Pure Fe 3 O 4 nanoparticles (MN only) and Magnegita were used as references.The tubes with the samples were immersed in water and the MR intensity values of water were measured from the surrounding area around the tubes.
As it seen from T2-weighted images (Figure 4B), MN-AuNCs (1) and MN only (2) result in the highest T2 contrast.The contrast for magnetic nanoparticles alone and for MN-AuNCs in T2-weighted images is higher than for the Magnegita solution.This is as expected, since Fe 3 O 4 -based nanoparticles have shorter T2 relaxation times and are widely applied as negative probing agents in MRI, leading to the contrast in reconstructed images of water protons by an inhomogeneous magnetic field around the outer sphere region of NPs.On the other hand, Gd-based contrast agents (such as Magnegita) can also be used as a T2 contrast agents, but at significantly shorter TE values.Figure 4B   contrast characteristics and can be further analyzed as a potential T2-weighted imaging contrast agent.Graphs present the intensities of T1-(Figure 4C) and T2weighted (Figure 4D) images depending on the different TI and TE times after RF excitation.Under very short TE, Gd-based solution shows the highest intensity differences from the water (Figure 4D).Under longer TE, the intensity drops and becomes similar to the intensity of water.However, the intensities of MN-AuNCs or MN only solutions, under longer TE, have significant differences in signal intensity and show a clear potential to serve as MRI contrast agents.Analyzing Figure 4C it is clear, that only Gd-based solution has apparent T1 intensity differences from the water, while -Fe 3 O 4 -based NPs, as expected, show no significant intensity difference from the water.T1 and T2 values of all analyzed substances are presented in Table 2.

Biocompatibility of MN-AuNCs In Vitro
Biocompatibility studies of cell-nanoparticle interactions must be performed and their consequences on cell viability and functionality must be thoroughly evaluated.We assessed the cytotoxicity of MN-AuNCs on cancerous MCF-7, MDA-MB-231, and fibroblast NIH-3T3 cell lines.The cells were incubated with MN-AuNCs for 24 h using different concentrations of nanoparticles.The viability of the control (without MN-AuNCs) was also studied for comparison.The cell viability assay is presented in Figure 5 and shows that none of the MN-AuNCs concentrations (0.255, 0.51, 1.27, and 2.55 mg mL −1 ) had a significant cytotoxic effect on the viability of any of the cell lines (Figure 5).
To study intracellular accumulation, MN-AuNCs were incubated with cancerous human breast cancer cells MCF-7 (poorlyaggressive and non-invasive cell line) [48] and MDA-MB-231 (highly aggressive and poorly differentiated triple-negative cancer cell line), [49] and embryonic mouse fibroblasts NIH-3T3, representing healthy cell line.Cells were incubated with nanoparticles for 24 h using 2.55 mg mL −1 concentration.Live cells were imaged using laser scanning confocal microscopy.In Figure 6 red photoluminescence of MN-AuNCs upon accumulation inside the cells is visible.Blue color represents cell nuclei stained with Hoechst 33 258.
After 24 h of incubation, nanoparticles accumulated in all tested cell lines, although uptake of nanoparticles was not equal.All MDA-MB-231 and NIH3T3 cells have homogenously accumulated nanoparticles, while only few of MCF-7 cells have taken up MN-AuNCs.Photoluminescence of the nanoparticles comes from the intracellular region, suggesting that MN-AuNCs showed no specificity for cancerous or healthy cell lines.MN-AuNCs internalized into the cells, concentrated within the cytoplasm, and localized in the perinuclear region (Figure 6 right panel).For better clarification of nanoparticles localization inside the cells, we also performed Z-scanning of MCF-7 cells incubated with MN-AuNCs and showed that nanoparticles localized around the nuclei but not inside the nuclei itself (Figure 6).

Photodynamic Effect of MN-AuNCs In Vitro
As previously shown in Figure 3, MN-AuNCs in an aqueous solution can generate reactive oxygen species, therefore the potential of these nanoparticles in photodynamic therapy was tested.The PDT effect caused by MN-AuNCs was studied in monolayers of MCF-7 and MDA-MB-231 cells, irradiating them with 402 nm light (irradiation dose 50 J cm −2 ) absorbed by nanoparticles.The photodynamic effect was studied with fluorescent cell viability dyes: calcein AM (green) stains living cells and propidium iodide (red) was used for visualization of dead cells after irradiation (Figure 7).
The photodynamic effect on MCF-7 cells incubated with MN-AuNCs became visible when irradiating cells with 50 J cm −2 dose of 402 nm light.Irradiated nanoparticles caused cell death (stained red, Figure 7F).Even though not all cells were affected, some of the viable (stained green) cells gained round morphology, indicating possible apoptotic processes happening [50] (Figure 7F).In comparison, less of MDA-MB-231 cells were affected by photodynamic effect and remained viable (Figure 7H).However, round or apoptotic morphology of cells was also present, thus, besides the fact that MDA-MB-231 are more resistant, cells still get affected during incubation with MN-AuNCs and irradiation with blue light (Figure 7H).The untreated groups of cells, which had no incubation with nanoparticles, were not affected by anything, and remained viable in all cases, even when irradiated (Figure 7A,C,E,G).Also, MN-AuNCs had no dark toxicity for either cell lines (Figure 7B,D).

Discussion
For decades conventional cancer treatment strategies have included surgical resection of tumors, radiotherapy, and/or chemotherapy. [51]However, neither radiotherapy nor chemotherapy are selective treatment strategies that also come with significant side effects.Therefore, more advanced approaches for cancer diagnosis and treatment are needed, and the emerging field of nanomedicine could help personalize and improve eradication of cancer.In our study, we demonstrated that our synthesized magnetic nanoparticles decorated with gold nanoclusters have both magnetic and optical properties, enabling their applications in bimodal imaging.MN-AuNCs also exhibit intrinsic therapeutic characteristics, generating singlet oxygen that is suitable for photodynamic therapy.In pair, diagnostic and therapeutic properties of MN-AuNCs make them promising theranostic nanoparticles.
It is important that MN-AuNCs exhibit magnetic properties from the magnetic core and red photoluminescence from gold nanoclusters (see Figures 2-4).Thus, MN-AuNCs could serve as magnetic contrast agent for MRI and a fluorescent probe for optical biopsy.Combination of various imaging modalities has the advantage over individual imaging techniques. [25,52]The fusion of photoluminescent and magnetic properties of nanoparticles is highly desirable due to their potential exploitation in biomedical applications. [53,54]he spatial characterization of nanoparticles revealed that MN-AuNCs had increased diameter compared to magnetic nanoparticles alone (as shown in Figure 1).[57] However, in our case, magnetic nanoparticles did not have a layer of gold, as there was no plasmon resonance absorption band characteristic to gold nanoparticles or layers of gold. [58]Instead, MN-AuNCs emitted red photoluminescence, which leads to a conclusion that magnetic iron oxide nanoparticles were decorated with photoluminescent gold nanoclusters (see model in Figure 8).
The addition of Au nanoclusters and BSA to Fe 3 O 4 nanoparticles increases the diameter of the NPs and the distance between Fe ions and water protons.This results in a weaker interaction between Fe and water protons, which is demonstrated by slightly longer T2 values for MN-AuNCs in comparison to MN alone.This further can be demonstrated in less intense MR images of  MN-AuNCs.However, MN-AuNCs still exhibit significant contrast from the water and have the potential to be employed as multimodal contrast agents or multifunctional NPs.
To use MN-AuNCs in biomedical applications, they must exhibit good colloidal stability that should not change during experiments.Thus, we demonstrated that MN-AuNCs are highly soluble and stable.Although the components of cell growing media could affect the stability of MN-AuNCs, our results showed that no changes appeared in either colloidal or optical properties.
Another important property of nanoparticles is photostability.We observed that the optical properties of MN-AuNCs were not affected by changes of the medium nanoparticles were dispersed in (Figure 2A).However, when continuously irradiated with visible light (402, 470, and 630 nm), photobleaching occurred, and intensity of PL started decreasing (Figure 3B).Nevertheless, after stopping the irradiation, the PL intensity almost fully recovered, indicating that MN-AuNCs could serve as photoluminescent markers in optical biopsy.Hemmateenejad et al. previously observed a reversible effect with bovine serum albumincapped gold nanoclusters, [46] where the authors claimed that photobleaching was caused by light-induced ratio changes of two separate oxidative states of Au 0 and Au + .They also hypothesized that photobleaching or spectral changes arose due to the formation of oxygen-associated photoproducts. [46]ccording to our observations, MN-AuNCs have similar properties to nanoparticles used by Hemmateenejad et al.MN-AuNCs exhibit a long-lived excited photoluminescence state (Figure 2B), which creates favorable conditions for the formation of reactive oxygen species.Singlet oxygen is essential in photodynamic therapy. [33]In our previous studies, we revealed that proteintemplated gold nanoclusters can generate singlet oxygen and other reactive oxygen species, such as peroxides and peroxynitrites, upon visible light irradiation. [17,18]MN-AuNCs also generate singlet oxygen during exposure to visible light, which we demonstrated using SOSG indicator (Figure 3A).Therefore, our synthesized magnetic nanoparticles decorated with gold nanoclusters could be used in photodynamic therapy.
All the characteristics listed above make MN-AuNCs promising theranostic nanoparticles, which incorporate bimodal diagnostic as well as therapeutic properties for PDT.However, their potential had to be proven during biocompatibility studies in cells.We showed that our tested concentrations of MN-AuNCs had no dark cytotoxicity to MCF-7, MDA-MB-231, and NIH3T3 cell lines (Figure 5).These results are also supported by our previous publications, in which blood plasma stabilized gold nanoclusters [17] and technetium-99m labeled bovine serum albumin-gold nanoclusters, [59] at nearly ten and six times higher concentrations, respectively, had almost no cytotoxicity after 24 h of incubation. [17]In another study, which included magnetic nanoparticles combined with gold nanoclusters, slight cytotoxicity to 293T cells appeared using a 0.1 mg mL −1 concentration (which is 25 times lower than the concentration we used), but cell viability still remained above 80%. [60]Similar results were demonstrated on HL-60 and HepG2 cancer cells using up to 4 nm concentration of aptamer/magnetic nanoparticles conjugated with fluorescent gold nanoclusters, as cells exhibited no significant decrease in viability. [61]uring PDT studies in vitro, we demonstrated that MN-AuNCs can kill MCF-7 and MDA-MB-231 breast cancer cells when irradiated with visible light (Figure 7).Although, MN-AuNCs are not as effective in generating singlet oxygen as conventional photosensitizers, [31,34,62] but we still managed to display therapeutic effect.For even better results, MN-AuNCs could be combined with photoactive drug that would increase singlet oxygen generation and cancerous cells killing efficiency. [50,63]ll in all, given the fact that MN-AuNCs have dual-imaging capabilities and exhibit therapeutic properties, it makes them a promising theranostic nanoplatform, which is the topic of many researchers around the world who are currently working with various types of theranostic nanoparticles. [42,64,65]Emerging theranostic NPs, administered systematically and specifically accumulated in the cancerous tissue, could help to detect tumors at early stages of its formation by noninvasive imaging modalities, such as MRI.Combining anatomical and contrast images, spread of cancer, that is otherwise indiscernible, could be detected.Moreover -therapeutic agents linked to NPs provide targeted tumor therapy and ensure that the same sites that are imaged and recognized as malignant, will be affected by anticancer treatment.Otherwise, there is no guarantee, that using separate agents for diagnostics and anticancer drugs for treatment would be specifically targeting the exact same location (tumor).Moreover, combined theranostic nanoparticles might reduce the number of invasive procedures needed for the patients (e.g.intravenous administration).However, lack of information and research on multifunctional theranostic nanoplatforms, especially on the hybrid nanocomposites of magnetic nanoparticles and fluorescent gold nanoclusters, requires additional studies.

Conclusion
This publication demonstrates that our synthesized nanoplatform of magnetic nanoparticles decorated with gold nanoclusters exhibits both optical and magnetic properties, which enables bimodal imaging capabilities in cancer diagnostics.In vitro experiments revealed that MN-AuNCs accumulate in cancerous cells, are non-toxic in the dark, and therefore are biocompatible.In addition, we showed that photoluminescent gold nanoclusters, under visible light irradiation, can generate a significant amount of singlet oxygen, which in turn can cause oxidative damage or even death to targeted cells.This property makes MN-AuNCs not only good bimodal contrast agents for optical biopsy and/or MRI but also a therapeutic nanodrug.Thus, the combination of diagnostic and therapeutic properties in one nanoplatform makes MN-AuNCs promising theranostic nanoparticles.However, more detailed and extensive studies are still needed to unlock the full potential of our nanoparticles.

Experimental Section
Synthesis of MN-AuNCs-Materials: For synthesis of magnetite (Fe 3 O 4 ) nanoparticles and their conjugation with gold nanoclusters, the following reagents were used: FeSO 4 • 7H 2 O, Fe 2 (SO 4 ) • 5H 2 O, phosphate buffered saline (PBS), d,l cysteine (C 3 H 7 NO 2 S, CYS) 96% and d,l methionine (MET) amino acids, bovine serum albumin (BSA), and HAuCl 4 .All respective reagents were purchased from Sigma-Aldrich.NaOH was obtained from Posh SA (Poland) and purified by oversaturation of its aqueous solution.Dialysis tubes SnakeSkin were purchased from Thermo Fisher Scientific (MWCO, 10 kDa, Rockford, IL, USA).Deionized/distilled water was used throughout all the steps of synthesis.
Synthesis of Magnetite NPs: Fe 3 O 4 nanoparticles were synthesized from an alkaline solution of FeSO 4 , Fe 2 (SO 4 ) 3 , and 0.2 mol L −1 amino acid as chelating agent at pH 12.1.The synthesis was conducted at 70 °C for 3 h in N 2 atmosphere and mild stirring of the medium.The synthesized NPs were collected by centrifugation, rinsed, and dried at 60 °C.XRD pattern of the obtained product confirmed the formation of pure magnetite NPs (Figure S5, Supporting Information).Hydrodynamic Diameter and Zeta Potential Measurements: The hydrodynamic diameter of particles was measured using the dynamic light scattering technique.Hydrodynamic diameter and zeta potential of the samples were examined using particle size and zeta potential analyzer Zeta Plus PALS (Brookhaven Inc., Suffolk County, NY, USA).

Conjugation of Magnetic NPs with Gold
Steady-State Absorption and Fluorescence Spectroscopy: Optical densities of synthesized MN-AuNCs solutions dispersed in deionized water (diH 2 O), PBS, Dulbecco's Modified Eagle medium (DMEM), and DMEM supplemented with fetal bovine serum (DMEM+FBS), were registered in the wavelength range from 350 to 800 nm.Measurements were performed with UV/Vis spectrophotometer Varian Cary 50 (Varian Inc, Australia).Photoluminescence spectra were measured under 488 nm excitation with fluorescence spectrometer Edinburgh Instruments FLS 920 (Edinburgh Instruments Ltd., Livingston, UK).Colloidal stability of MN-AuNCs was evaluated by monitoring the intensity of PL for the period of 2 weeks.All spectral measurements were performed in 1 cm pathlength polystyrene cuvettes.
Time-Resolved Fluorescence Spectroscopy: Fluorescence decay measurements were acquired by a time-correlated single-photon counting technique using FLS 920 spectrometer (Edinburgh Instruments Ltd.) equipped with a single-photon photomultiplier detector (S900-R) (Hamamatsu, Japan) and diode laser ( exc = 405 nm, pulse length -<200 ps, pulse repetition rate -100 kHz) for excitation of samples.Fluorescence decays were measured by collecting 1000 counts at the peak emission wavelength (685 nm) of the MN-AuNCs fluorescence band after preparation and 1 as well as 2 weeks after.Average fluorescence lifetime was calculated using Equation (1): where B 1 is the amplitude of the first exponential term, B 2 is that of the second term, and so forth and  1 is the FL lifetime of the first exponential term,  2 is that of the second term, and so forth.
Evaluating Generation of Singlet Oxygen: Singlet oxygen sensor green (SOSG) (100 μg, Invitrogen, Waltham, MA, USA) was used to assess singlet oxygen generation under three different wavelength illuminations (402, 470, and 630 nm).SOSG was dissolved in methanol (33 μL), and then diluted with deionized water to prepare a SOSG stock solution (50 μm).Three separate samples of MN-AuNCs + SOSG were prepared for investigation of singlet oxygen generation.For additional comparison with a well-known photosensitizer, an aqueous solution containing 1 μm of chlorin e6 was prepared.Afterward, the samples were irradiated by a respective wavelength from a MAX-302 xenon light (Asahi Spectra Co., Ltd, Japan).A total irradiation dose of 40 J cm −2 was achieved via five irradiation sessions, collecting 1, 3, 6, 10, and 20 J cm −2 irradiation doses, respectively.Irradiation time was calculated according to the light source power.Fluorescence spectra of SOSG were measured with a spectrophotometer before and after irradiations.The difference in SOSG FL intensity at its peak ( em = 525 nm) before and after the irradiation was calculated and plotted as a graph (Figure 3A).
Photostability: The photostability of MN-AuNCs was measured by irradiating 2 mL of the sample solutions in plastic cuvettes (exposed area was 1 cm 2 ) with a Xenon light source Max-302 (Asahi Spectra Co., Ltd, Japan).Various wavelengths (402, 470, and 630 nm) were used for irradiation.A total irradiation dose of 40 J cm −2 was achieved via five irradiation sessions, collecting 1, 3, 6, 10, and 20 J cm −2 irradiation doses, respectively.Irradiation time was calculated according to the light source power.After collecting 40 J cm −2 , samples were placed in the dark and kept for 24 h.Results were represented as the variation of PL intensity at the maximum of MN-AuNCs PL spectra ( exc = 488 nm).
Evaluation of Photoluminescence Quantum Yield: Quantum yield of MN-AuNCs photoluminescence was evaluated by comparing it with in the literature well-known fluorophore rhodamine 6G (R6G) (Merck Group, Germany), which quantum yield reached 95%. [47]First, absorption spectra of both R6G and MN-AuNCs were measured and a point (wavelength) of spectra intersection was found ( exc = 513 nm).Under excitation of this wavelength, PL spectra were registered, then area under the spectra curve integrated, and quantum yield calculated according to Equation ( 2): where  -quantum yield, I -integrated PL intensity, A -optical density, n -refractive index and f -a known fluorophore.
Since the R6G and MN-AuNCs were both dissolved in deionized water and wavelength, at which optical density was the same for both solutions, was selected, Equation ( 2) can be simplified to Equation (3): Characterization of Magnetic Properties: MR imaging and characterization of MN-AuNCs were performed using a clinical 1.5T MRI scanner (Philips Achieva).A series of 2 mL plastic Eppendorf tubes were filled with different NPs and fixed into non-MRI contrasting holder.The concentration of prepared samples was 0.2 mg mL −1 of Fe 3 O 4 .For comparison, a tube with a solution of a certain concentration (0.12 mg mL −1 Gd) of clinical MRI contrast Magnegita was added to the imaging field.Eppendorf tubes were immersed in water to ensure homogeneous magnetic field distribution.T1-weighted turbo spin-echo inversion recovery sequence was used to provide images to estimate longitudinal relaxation time T1.Transversal relaxation time was calculated using a turbo spin echo sequence that was T2-weighted.
Cell Culturing: The MCF-7 and MDA-MB-231 human breast cancer cell lines and immortalized mouse embryonic fibroblast cell line NIH3T3 were selected as model in vitro systems for cellular experiments (MCF-7 were purchased from the European Collection of Cell Cultures and MDA-MB-231 and NIH3T3 were purchased from American Type Culture Collection).Cells were cultured in a cell growth medium (DMEM), supplemented with 10% (v/v) fetal bovine serum, 100 U mL −1 penicillin, and 100 μg mL −1 streptomycin (all from Gibco, Thermo Fisher Scientific, Waltham, MA, USA).Cells were maintained at 37 °C in a humidified atmosphere containing 5% of CO 2 .In 25 cm 2 cell culture flasks, the cells were ordinarily subcultured two to three times per week.
Biocompatibility Assay: MCF-7 and MDA-MB-231 breast cancer cells as well as NIH-3T3 mouse embryonic fibroblasts were seeded into a 96well plate (BD Falcon, USA) at a density of 20 000 cells per well to conduct the XTT cell viability test.The cell growing medium included 150 μL of suspended cells.To later assess the optical density of the media, one column of the plate was filled entirely with nutritional medium without any cells.Following 24 h, the cell growing medium was removed from the cells and replaced with 150 μL of fresh medium that contained MN-AuNCs which were diluted 10, 20, 50, and 100 times in the previous solution (stock concentration 25.5 mg mL −1 ).Only 150 μL of nutritive media devoid of nanoparticles was used to fill the control wells.Cells were incubated with nanoparticles for 24 h.Following incubation, the nanoparticle-containing media was removed, and Dulbecco's PBS was used to wash the cell cultures.Afterward, 50 μL of the XTT reaction solution and 100 μL of fresh nutritional medium were added.
The XTT reagent (Biological Industries, Israel) and activation reagent (Biological Industries, Israel) were combined in a 50:1 ratio to create enough XTT reaction solution to cover the entire plate (96 wells).The cells were cultured in the incubator for 4 h with the added XTT reaction solution.Using a microplate reader, the medium's absorbance was determined at 490 nm (BioTek, USA).Each group of wells' average optical density was computed, and the Equation ( 4) was then used to translate the results into percentages: Viability of cells = Optical density of a specific well Average optical density of control wells × 100% (4) Accumulation of MN-AuNCs in Cell Monolayers: The accumulation of magnetic nanoparticles decorated with gold nanoclusters in cells was imaged using a Nikon TE-2000U microscope (Nikon, Japan) with a C1si laser scanning confocal system.An immersion (oil) lens of x60 magnification and 1.4 numerical aperture Plan Apo VC lens (Nikon, Japan) was used for imaging.The photoluminescence of MN-AuNCs was excited using a 488 nm wavelength argon ion laser, while the Hoechst 33 258 nuclear stain was excited with a 404 nm diode laser (Roithner, Austria).A three-channel RGB detector with 450/17, 545/45, and 688/67 nm bandpass filters corresponding to blue, green, and red channels, respectively, was used to record photoluminescence.
The Frame Lambda function was used to make sure that only one chosen laser and one desired detector were engaged at a time to prevent signal leakage into nearby detectors.During the measurement, other lasers and detectors were turned off to prevent the excitation of unnecessary fluorophores or the detection of their emissions in the channels.Processing of the images was done with Nikon EZ-C1 software.
Confocal microscopy was also used to view three-dimensional images (Z-stack) of nanoparticle accumulation in cells.Sample photos in various z-axis planes were registered to create a 3D image.Software was used to recreate a 3D image of cells that were treated with nanoparticles from numerous z-axis sections of the sample.
Photodynamic Effect in Cells: The PDT effect using MN-AuNCs was tested in MDA-MB-231 and MCF-7 cell monolayers.Cells were seeded into 8-well glass bottom plates (Lab-Tek, USA) with 30 000 cells well −1 .After seeding (24 h), the medium was replaced with fresh medium supplemented with 2.55 mg mL −1 MN-AuNCs.Control wells received no additional treatment but were subjected to the same manipulations as the experimental wells.Cells were incubated with nanoparticles for 24 h.The following day, cells were irradiated with 402 nm wavelength light or left unirradiated (control wells, cells with nanoparticles).The power density of the radiation was measured, and wells were irradiated until a dose of 50 J cm −2 was accumulated.After irradiation, plates with cells were returned to the incubator and incubated overnight.The following day, cells were stained with LIVE/DEAD viability dyes (ThermoFisher, USA).The set included two dyes: Calcein-AM, which stained only live cells, and Ethidium homodimer-1, which stained dead cells.Stained cells were imaged with a Nikon Eclipse confocal microscope using a 20x/0.5 NA dry objective lens (Nikon, Japan).Calcein was excited with 488 nm laser and its fluorescence was registered in the green channel, Ethidium homodimer-1 was excited using a 543 nm laser and its fluorescence was registered in the red channel.

Figure 1 .
Figure 1.Characterization of spatial properties and chemical composition of MN-AuNCs.A-F) Atomic force microscopy images of magnetic nanoparticles decorated with gold nanoclusters (A,B,D,E) and magnetic nanoparticles (C,F) dried on mica.The height images (A-C) and the phase images (E-G) are presented.Scale bars: 200 nm.G) Nanoparticle size distribution measured from AFM images.H) Nanoparticle size distribution from dynamic light scattering measurements.

Figure 2 .
Figure 2. Optical properties of magnetic nanoparticles decorated with gold nanoclusters.A) Characteristic optical spectra of MN-AuNCs in two different solutions (deionized water (diH 2 O) -presented in black, Dulbecco's Modified Eagle medium supplemented with fetal bovine serum (DMEM+FBS) presented in red).Dashed lines represent absorbance spectra, solid lines -photoluminescence of MN-AuNCs.Additionally, PL spectrum of only DMEM+FBS solvent is given in blue color.PL spectra were registered under 488 nm wavelength excitation.B) Fluorescence decay curves of MN-AuNCs after preparation in diH 2 O (black decay) and in DMEM+FBS (red decay) and their respective bi-exponential fits (light grey and orange lines).Inset represents a zoomed region from 120 to 200 ns.Magenta decay shows instrument response function.
T1weighted image clearly indicates that only Magnegita has significantly different intensity of T1 contrast and none of the Fe 3 O 4based NPs show any apparent intensity difference.Therefore, it can be concluded that solution of MN-AuNCs (1) possesses MRI

Figure 5 .
Figure 5. Cell viability studies of MCF-7, MDA-MB-231, and NIH-3T3 cells after incubation with magnetic nanoparticles decorated with gold nanoclusters.Data is presented by averaging n = 6 measurement results and calculating standard deviation.

Figure 7 .
Figure 7.The PDT effect in cell monolayers using MN-AuNCs.Laser scanning microscopy images with breast cancer MCF-7 and MDA-MB-231 cell lines.Cells were untreated (without nanoparticles) or incubated with MN-AuNCs (2.55 mg mL −1 ).After 24 h of incubation, cells were left non-irradiated or irradiated with 402 nm light (50 J cm −2 irradiation dose).Green color in the images represents live cells and red represents dead cells.Scale bars for all images are 100 μm.

Figure 8 .
Figure 8. Suggested model for MN-AuNCs.Magnetic iron oxide nanoparticles are capped with amino acids (cysteine and/or methionine) and decorated with photoluminescent gold nanoclusters.
Nanoclusters: Red luminescent Fe 3 O 4 NPs were synthesized in a microwave reactor Monowave 300 (Anton Paar, USA) using 160 W radiation for 60 min.Two aqueous solutions composed of Fe 3 O 4 NPs dispersed in PBS buffer and BSA dissolved in 10 mmol L −1 HAuCl 4 together with MET were prepared and mixed.Then, the pH of mixture was adjusted to 12.4 by dropwise addition of NaOH solution and transferred to a microwave reactor.The synthesis of red luminescent Fe 3 O 4 @Au@BSA NPs (MN-AuNCs) was conducted by 160 W radiation at 60 and 45 °C for 60 min.Obtained product was collected by centrifugation and rinsed with water.Dialysis was performed by placing nanoparticles solution in a SnakeSkinTM dialysis bag for 24 h.Atomic Force Microscopy: The Innova Atomic Force Microscope (AFM) (Bruker Corporation, Billerica, Massachusetts, USA) with SPMLab software version 7.11 was used to image MN-AuNCs dried on a mica substrate.Samples were scanned in the tapping mode with a point probe silicon tip of <15 nm radius (model: TESPA) (Bruker Corporation), and an oscillation frequency of 230−410 kHz was applied.Topographic images of 2 μm × 2 μm and 1 μm × 1 μm were acquired at a scan rate of 2 Hz, containing 512 × 512 pixels.

Table 1 .
Fitting results of photoluminescence decay curves of MN-AuNCs.
1 , 2 are corresponding lifetimes of two-exponential fit of photoluminescence decays; b) A 1 ,A 2 are the amplitudes; c)  ave is the intensity-weighted average lifetime; d)  2 represents the goodness of fit parameter.

Table 2 .
T1 and T2 values for different samples.