Optimizing the Performance of Silica Nanoparticles Functionalized with a Near-Infrared Fluorescent Dye for Bioimaging Applications

Abstract

Modified fluorescent nanoparticles continue to emerge as promising candidates for drug delivery, bioimaging, and labeling tools for various biomedical applications. The ability of nanomaterials to fluorescently label cells allow for the enhanced detection and understanding of diseases. Silica nanoparticles have a variety of unique properties that can be harnessed for many different applications, causing their increased popularity. In combination with an organic dye, fluorescent nanoparticles demonstrate a vast range of advantageous properties including long photostability, surface modification, and signal amplification, thus allowing ease of manipulation to best suit bioimaging purposes. In this study, the Stöber method with tetraethyl orthosilicate (TEOS) and a fluorescent dye sulfo-Cy5-amine was used to synthesize fluorescent silica nanoparticles. The fluorescence spectra, zeta potential, quantum yield, cytotoxicity, and photostability were evaluated. The increased intracellular uptake and photostability of the dye-silica nanoparticles show their potential for bioimaging.

1. Introduction

Cancer, is the second leading cause of death in the United States, and is recognized as a worldwide health issue ¹. Currently, diagnostic screening is of the essence to ensure early intervention for patients. Over the past few decades, diagnostic screening has continued to develop, producing a range of diagnostic tools such as screening through the detection of specific biomarkers, examining blood for circulating DNA from tumor cells, or assessment through imaging techniques ^2–5. Yet, as diagnostic methods develop, cancer becomes more elusive due to its heterogeneous nature. Specifically, this is observed in screening methods as most tests only demonstrate a sensitivity of 70–80%, whereas their specificity is even lower around 60–70% ². In validation of analytical methods, sensitivity is often defined as the number of true positive responses versus the sum of true positive and false negative responses, whereas specificity is defined as the number of true negative responses versus the sum of true negative and false positive reseponses ⁶. In addition to the lack in specificity and selectivity, there is often the concern for overdiagnosis causing unnecessary treatment for patients ⁷. Thus, it is with precedence that diagnostic testing utilizing better specificity and sensitivity be established for all cancer types.

The role of imaging as an approach to diagnostic screening for cancer has proved most useful over the years, particularly for breast, colorectal, and lung cancer ^2,8,9,. Molecular imaging requires both relevant biomarkers as well as targeted contrast agents to provide the most accurate information regarding the disease state ¹⁰. Near-infrared (NIR) fluorescent dyes provide increasing tissue penetration and higher image clarity ^{11, 12}. In particular, cyanine dyes such as Cy5 demonstrate high molar absorption coefficients and relatively high quantum yields, proving useful for bioimaging purposes ¹³. In addition to contrast agents, diagnostic screening requires applicable biomarkers, which may involve a nanomedicine approach to load increased amounts of reporter probe or serve as a scaffold for multiple ligands ¹⁴.

Nanotechnology has advanced translationally, especially in terms of therapeutic applications. Particularly, nanoparticle development can be observed in clinical trials through chemotherapy, immunotherapy, radiotherapy, or gene therapy ¹⁵. Nanoparticles also demonstrate great characteristics to be used in a diagnostic setting ¹⁶. Nanoparticles can be synthesized in many shapes and sizes that influence their intrinsic properties. In particular, silica nanoparticles have garnered much attention due to their easily-tuned chemistry ¹⁷. Silica nanoparticles can be referred to as a workbench, in which their size, shape and surface properties can be modified to meet the requirements of the application they are employed for. In addition to their modifiable properties, silica nanoparticles demonstrate intrinsic stability and large surface areas with which to apply conjugation strategies ¹⁸. Furthermore, silica nanoparticles possess silanol groups present on the exterior of the nanoparticle’s surface that can aid in targeting cell types of interest ¹⁹. Thus, it is pertinent to utilize the robustness of nanoparticle properties to develop improved cancer therapeutics.

Dye-loaded nanoparticles demonstrate superior efficacy for bioimaging applications due to their increased photostability and enhanced fluorescence ^20,21. Yet, despite the promise, studies reported to date show a number of challenges, such as dye leakage that contributes to decreased fluorescence intensity and lower signal to noise ratios in vivo ²². In many dye-encapsulated nanoparticles, self-quenching is often observed due to the close proximity of the dyes, which causes a decrease in fluorescence ²³. Despite a significant body of research on dye-encapsulated nanoparticles, it is of great importance to apply a systemic approach to develop bright and stable dye-encapsulated nanoparticles demonstrating ideal bioimaging properties.

The present work focuses on the synthesis, characterization and evaluation of uniform fluorescent silica nanoparticles designed for use in diagnostic imaging. The fluorescent nanoparticles containing an organic dye – derivative of Cy5 were synthesized, characterized (size, dispersity, spectral properties) and evaluated (photostability, cell viability and internalization) for future diagnostic potential.

2. Materials and Methods

2.1. Materials and Reagents

Sulfo-Cy5-amine (sCy5-amine) and sulfo-Cy5-NHS (sCy5-NHS) were purchased from Lumiprobe, Inc. (Hunt Valley, MD, USA). 3-Triethoxysiylpropyl succinic anhydride (TSPA) was purchased from Oakwood Chemical (Estill, SC, USA). Tetraethyl orthosilicate (TEOS), Amicon centrifugal filters (10 kDa MWCO), dimethyl sulfoxide (DMSO), ammonium chloride ammonium hydroxide buffer, toluene, and HEPES buffer (N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid), 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid) were purchased from Sigma Aldrich (St. Louis, MO, USA). 3-Aminopropyltriethoxysilane (APTES) and paraformaldehyde were purchased from Alfa Aesar (Ward Hill, MA, USA), and cyclohexane was purchased from Acros Organics (New Jersey, USA). MES buffer, pH 4.7 was purchased from bio-WORLD (Dublin, Ohio, USA). RPMI 1640 media, penicillin-streptomycin, radioimmunoprecipitation assay (RIPA), Halt protease inhibitor and 96-well plates were purchased from Thermo Scientific (Carlsbad, CA, USA). Fetal bovine serum (FBS) was purchased from Mediatech, Inc. (Woodland, CA, USA). CCK-8 kit was purchased from Dojindo (Rockville, MD, USA), 8-well chamber slides were purchased from labTek (Grand Rapids, MI, USA), phosphate buffered saline (PBS) was purchased from Corning (Corning, NY, USA), and DAPI solution was purchased from Invitrogen, Thermo Scientific (Waltham, MA, USA).

2.2. Preparation of bare and sCy5-incorporated silica nanoparticles, SiNP(sCy5)

Bare and fluorescent silica nanoparticles were synthesized by a modified Stöber method. The reaction was carried out in 10 mM ammonium chloride-ammonium hydroxide buffer (pH 9.0) at 50°C without or in the presence of sCy5-APTES conjugate. The amount of sCy5-APTES conjugate added to reaction mixture was varied, producing silica nanoparticles with different content of sCy5 dye. The reaction was allowed to proceed for 30 min, followed by the addition of TEOS and cyclohexane. Presence of cyclohexane as co-solvent ensured better control over silica seed growth. The reaction mixture was vigorously stirred to mix the organic and inorganic layers to form an emulsion, preventing particles from agglomeration and allowing for the production of small sized, monodispersed, silica nanoparticles. The reaction was allowed to proceed for 24 h. After completion, the reaction mixture was cooled to rt and re-separated into an organic (top) and an aqueous (bottom) layer. The aqueous layer that contained the silica nanoparticles was collected, washed with water, and concentrated using Amicon centrifugal filters (10,000 Da MWCO). sCy5-APTES conjugate was prepared by the coupling procedure adapted from literature ^24,25. Briefly, the appropriate amount of sulfo-Cy5-NHS dye was dissolved in anhydrous DMSO at concentration 4 mg/ml and suitable amount of APTES was added. The molar ratio between APTES and sulfo-Cy5-NHS dye was 1.2:1. The reaction was allowed to proceed for 24 h at rt in the dark, which yielded sCy5-APTES as a product. The obtained conjugate was stored at −20°C and used without additional purification.

2.3. Nanoparticle characterization

Nanoparticle hydrodynamic diameter, polydispersity, and zeta-potential were measured at a concentration of 2 mg/ml using a Malvern Panalytical Zetasizer Nano ZS90 instrument (Malvern, United Kingdom). The absorption and emission properties of the obtained nanomaterials were studied using the Thermo Fisher Scientific Evolution 220 UV-Visible Spectrophotometer (Waltham, MA, USA) and the Horiba Scientific FluoroMax 4 fluorometer (Edison, NJ, USA), respectively. Nanoformulation solutions were excited at 640 nm, and emission spectra were recorded in the range from 650 nm to 800 nm using the bandpass of 5 nm. Relative quantum yields of the nanoparticles were determined using the A|e-UV-Vis-IR Spectral Sofware (Version 2.2, 2007 Free Software Foundation, Inc.) and utilizing the following formula: $\text{QY}={\text{QY}}_{\text{ref}}\frac{\text{η}2}{\text{η}2\text{ref}}\frac{I}{A}\frac{Aref}{Iref}\cdot \text{η}$is the refractive index of the solvent, $I$ is the integrated fluorescence intensity of the sample or reference, and $A$ is the absorbance at its excitation wavelength The reference values for absorbance and fluorescence intensity were based on the data for the 3.49 μg/mL sample. The morphology of the silica nanoparticles was studied using transmission electron microscopy (TEM). Samples were negatively stained with NanoVan and examined on a Tecnai G2 Spirit TWIN (FEI, Hillsboro, OR) operating at an accelerating voltage of 80 kV. Images were acquired digitally with an AMT (Woburn, MA) digital imaging system.

2.4. Nanoparticle photostability

50 μL of each nanomaterial sample (0.3, or 1.7 μg/mg of silica ) and 50 μL of 1 μg/mL Cy5 free dye were first analyzed using the Thermo Fisher Scientific Evolution 220 UV-Visible Spectrophotometer (Waltham, MA, USA). After the initial absorbance reading, using an IX73 Inverted Microscope,, samples were continuously exposed to a Cy5.5 filter cube at 650 nm (Olympus, Japan). Samples were exposed for a total period of 60 min, with absorbance recorded every 10 min.

2.5. Nanoparticle pH stability

Solutions with differing pH were prepared based on the MES buffer through addition of appropriate volumes of 1 M NaOH or 1 M HCl dropwise. Three different buffers were prepared representing acidic (4.38), neutral (7.55), or basic (8.92) conditions. Each nanoparticle sample was then added to each buffer solution. The fluorescence intensity was measured after 30 minutes, 4 hours, and 24 hours of incubation using the Horiba Scientific FluoroMax 4 fluorometer (Edison, NJ, USA). Samples were excited at 640 nm, and the emission spectra were recorded in the range from 650 nm to 800 nm.

2.6. Cell viability, photostability, and uptake

MDA-MB-231 cells (5×10³ cells per 100 μL) were seeded into individual wells of a 96 well-plate and left to incubate for 24 hours at 37°C. Nanoparticle samples (25 μg/mL) were added to each well and left to incubate at 37°C for 24 h. 10 μL of the cell count CCK-8 solution was added to each well and left to incubate for 4 h. Absorbance was recorded at 450 nm every 30 min until 4 h had elapsed using an Agilent microplate reader (Santa Clara, CA, USA).

MDA-MB-231 cells (10⁴ cells per 50 μL) were seeded into individual wells of an 8-chamber well slides and left to incubate for 24 h at 37°C. Cells were washed three times with PBS and nanoparticle sample (25 μg/mL in serum-free RPMI) were added to each well and left to incubate at 37°C for 4 h. Cells were then washed three times with PBS and fixed with 4% paraformaldehyde for 15 min at 37°C. Cells were washed three times again with PBS and then permeabilized at room temperature for 10 min using 0.1% triton-x-100. Cells were washed three times with PBS and mounting media containing DAPI was added to each well. After 5 min, a coverslip was added, and cells were imaged with an Olympus IX73 Inverted Microscope, captured with DP80 Digital Camera and displayed by CellSens Dimension software in the 345 nm (DAPI) and 650 nm (Cy5) channels. After the initial imaging of uptake, the slides were then exposed to a 40 W powered lamp (wavelength approximately 600 nm) over a period of 1 hour. Every 10 min, the slides were then imaged on the Odyssey M (LI-COR Biosciences) in the 700 nm channel to avoid the detection of fluorescence produced from the DAPI staining. Fluorescence was measured with the ImageJ software, in which entire wells (n=3) were selected as regions of interest to analyze for photostability in vitro.

2.7. Image Analysis

Olympus DP80 Digital Camera and CellSens Dimension software were used to take cell internalization images. Empiria Studio Software (Version 2.3.0.154, 2022 LI-COR Biosciences) was used to analyze the signal intensity during the photostability measurements in vitro. ROIs were made by creating a square around the area of interest. The same square was used for each measurement.

3. Results and discussion

3.1. Synthesis of bare and dye encapsulated silica nanoparticles (sCy5-SiNPs)

Bare or dye-encapsulated silica nanoparticles were prepared by using the Stöber method with some modifications ^26–28. Synthesis started by using TEOS - a routinely used silica precursor. Through alkali hydrolysis (pH 9.0), TEOS ethyl groups were converted to hydroxyl groups first, followed by a series of condensation reactions (Scheme 1).

Scheme 1:

Synthesis of bare nanoparticles via Stöber method: (a) hydrolysis; (b) condensation. Reaction conditions: 10 mM chloride-ammonium hydroxide buffer, pH 9.0, cyclohexane, 50°C, 24h.

In order to create dye-encapsulated nanoparticles, differing amounts of cyanine dye molecules were conjugated to APTES to produce sCy5-APTES conjugates (Scheme 2a). The silane groups present on the APTES molecule were able to go through their first set of hydrolysis and condensation reactions to encapsulate the cyanine dye within the silica matrix to form a fluorescent core (Scheme 2b), followed by addition of TEOS that led to a second set of hydrolysis and condensation reactions as described above (Scheme 2c). Thus, generating sCy5 encapsulated nanoparticles with varying dye loading ratios (μg sCy5 dye per 7.5 mg of silica: 0, 0.625, 1.25, 2.5, 5, 10, 20, 30, 40). Differential loading ratios were established to optimize the amount of dye that can be encapsulated without leading to quenching of fluorescence.

Scheme 2.

Synthesis of cyanine dye-incorporated silica nanoparticles. Conditions: (a) conjugation: sCy5-NHS ester was reacted with APTES at a molar ratio of 1.2:1 (APTES:dye). Reaction proceeded at RT for 24 h in dark; (b) first reaction of hydrolysis and condensation: 10 mM chloride-ammonium hydroxide buffer, pH 9.0, 50^oC, 30 min; (c) second reaction of hydrolysis/condensation and stabilization reactions: TEOS, cyclohexane, 50^oC, 24 h.

3.2. Characterization of bare and dye encapsulated silica nanoparticles (sCy5-SiNPs)

The optical properties of the dye-encapsulated nanoparticles were evaluated to optimize the amount of sCy5 that could be added without leading to self quenching (Fig. 1). Encapsulation of sCy5 dye was confirmed with absorption spectroscopy that demonstrated a prevalent peak at 650 nm (Fig. 1A). Absorbance peaks were noted to increase as the amount of dye loaded into the nanoparticle increased (Fig. 1B). After confirming sCy5 encapsulation, a calibration curve was generated in order to determine the actual amount of sCy5 present in the nanoparticles as a result of the synthesis (Fig. S1A). Absorbance values for each SiNP sample were placed into the equation generated from the calibration curve, and multiplied by their dilution factor to obtain the μg sCy5 dye per 7.5 mg of silica. Our values indicate that relatively the theoretical amount was equal to that of the actual encapsulated amount of dye, thus demonstrating potential quality loading efficiency (Fig. S1B).

Figure 1.

Absorbance spectra of different dye-loaded sCy5-SiNPs (μg of dye per mg of silica) (A) and the absorbance of samples with varying dye content at 650 nm (the dotted line represents the best linear fit) (B).

To identify the optimal content of the dye that will provide maximum brightness per amount of the dye loaded, fluorescence spectra of the dye-incorporated nanoparticles were also studied. In fluorescence spectra, an intensive emission peak was observed at approximately 660 nm (Fig. 2A). Looking at the fluorescence intensity dependence upon the amount of the dye incorporated into the NPs, a decrease in fluorescence is notable for loading ratios greater than 3.49 μg per mg of silica (Fig. 2B). Self-quenching, likely due to the aggregation of the dye molecules loaded into the silica core, significantly contributes to the loss of excitation energy ²⁹. Thus, the decrease in fluorescence intensity that is observed at higher dye concentrations may be explained by this phenomenon. ³⁰ Indeed, normalized absorption spectra (Fig. S2) show a distinct shoulder at approximately 590 nm that increases in intensity with increase of the dye content. In earlier studies, this spectral feature was attributed to the formation of Cy5 dimers or H-aggregates. ^31,32 At the same time, this plot does not account for the higher absorbance of the dye-loaded NPs with higher dye content. To take this important factor in consideration, the relative quantum yields of the nanoparticles with different dye content were determined (Fig. 3). The sample with the lowest dye content (0.10 μg per mg of silica) was used as a reference for the relative quantum yield. Our results revealed that nanoparticles with dye content up to 1.73 μg per mg of silica maintained high relative quantum yields greater than 0.5. Therefore, two leading nanoparticles with the highest relative quantum yield were chosen: 0.32 μg and 1.73 μg dye per mg of silica, as both demonstrated optimal spectral qualities, but produced different fluorescence intensities. For future reference, these will be referred to as NP-low (0.32 μg per mg of silica) and NP-high (1.73 μg per mg of silica).

Figure 2.

Fluorescence spectra of dye-loaded sCy5-SiNPs (μg of dye per mg of silica) (A) and their fluorescence intensity at 660 nm (B).

Figure 3.

Relative quantum yield values of sCy5 encapsulated nanoparticles.

The size and polydispersity of the nanoparticles were examined using TEM (Fig. 4). The obtained microscopy images showed spherical nanoparticles with sizes of roughly 30 nm, which was also independently confirmed using DLS (Table 1). It was noted that the nanoparticles slightly increased in size as more dye was encapsulated (Table 1). The nanoparticles demonstrated zeta potentials on average of –15 mV (Table 1). This negative surface potential is due to hydroxyl groups present on the exterior of the nanoparticle’s surface, and suggests that the nanoparticles will have good colloidal stability in solution ³³.

Figure 4.

Transmission electron microscopy (TEM) images of bare and dye-incorporated silica nanoparticles prepared with different content of sCy5 dye: (A) bare; (B) low-NP; (C) high-NP.

Table 1.

Colloidal characteristics of synthesized silica nanoparticles

Formulation	DLSa			TEMb
	Deff (nm)	PDI	$\zeta$-potential (mV)	Dav (nm)
Bare SiNPs	26.4 ± 0.6	0.09	−16.3 ± 1.8	26 ± 1
sCy5-SiNPs, low-NPs	28.1 ± 0.9	0.12	−14.8 ± 2.0	29 ± 2
sCy5-SiNPs, high-NPs	32.5 ± 1.4	0.13	−15.1 ± 2.4	32 ± 2

^aParticle size (D_eff), particle size distribution (PDI) and ζ-potential were determined by DLS (2.5 mg/mL, PBS, pH 7.4, 25°C). Data presented as mean ± SD (n = 3–5).

^bThe TEM measurements were performed at silica nanoparticles concentration of 1 mg/mL.

3.3. Evaluation of bare and dye-encapsulated silica nanoparticles (sCy5-SiNPs) for imaging

The evaluation of the sCy5-SiNPs as materials for imaging began with determining their photostability. Stability of the lead dye-encapsulated NPs were examined in comparison with the free sCy5 dye. Silica nanoparticles, generally very stable particles, were observed to maintain their integrity, size, and polydispersity for at least two weeks upon storage at room temperature or 4°C (Fig.S3). Therefore, we assume these nanoparticles to be intact during the photostability studies. To illustrate their relative photostability, percent of degradation was calculated over time after exposure to light in the Cy5.5 channel (694 nm, Fig. 5A). After exposure for 60 minutes, the free dye sample demonstrated the greatest amount of photodegradation as compared to the lowNP and high-NP samples (Fig. 5 and Fig. S4). These results concur with literature reports that the silica matrix encapsulating the sCy5 dye protects it from photobleaching ³⁴. In addition, the nanoformulations demonstrated good stability in MES buffer at pH of 7.55 and 8.92 over a period of 24 hours (Fig. S5). Thus, these results portray a more robust and stable material for bioimaging purposes.

Figure 5.

Percent degradation for each SiNP sample type over a period of 60 minutes of irradiation at 694 nm.

After confirming a stability profile superior to the free dye, the nanoparticles were tested in vitro. It is essential for these nanoparticles to exhibit low cytotoxicity in order to be used for cellular targeting in the future. Thus, the nanoparticle samples were investigated in a breast cancer cell line, MDA-MB-231. Our findings demonstrated a lack of toxicity for all samples (Fig. 6E).

Figure 6.

Cellular internalization of NP-high (A), NP-low (B), bare SiNP (C), and media-only control (D) after 4 hours of incubation. Cytotoxicity of SiNP samples on MDA-MB-231 breast cancer cell line (E). Photostability of internalized SiNP samples (n=3, F). Blue color corresponds to the DAPI channel, and red color indicates fluorescence in the Cy-5 channel.

On the next step, the SiNP samples were subjected to a photostability investigation in vitro. After 4 hours, internalization of the sCy5-SiNPs into cells was observed (Fig. 6A–D), and their fluorescence intensity were observed every 10 minutes. The NP-high sample demonstrated higher fluorescence intensity (Fig. 6A) in comparison to the NP-low sample (Fig. 6B–D). After a 1-hour exposure to a 40 W daylight LED lamp (wavelength of approximately 600 nm), the sCy5-SiNPs remained relatively stable in vitro, thus further indicating the ability of the silica matrix to protect the sCy5 dye from photobleaching (Fig. 6F).

4. Conclusions

In this study, we developed novel fluorescent silica nanoparticles as a potential tool for diagnostic imaging. The synthesis, characterization, and evaluation of these nanoparticles was discussed and characteristics such as uniform size, high colloidal stability, efficient fluorescence and photostability were observed. These characteristics are of great importance for a material to be effectively used in imaging applications and study a wide range of cancer types. In addition, the combination of these vital properties with a non-toxic nanoparticle provides a compelling evidence for the perspective of future clinical use of these fluorescent nanoparticles. Importantly, silica nanoparticles provide a large surface area, thus offering a multitude of modification opportunities including conjugations with different targeting ligands. Such modification can significantly increase their ability to target cancer cells with greater affinity and specficity. Thus, the synthesized composite Cy5 dye-encapsulated silica nanoparticles is a compelling candidate for application in fluorescence-guided surgery or diagnostic imaging.