Optimized Synthesis of PEG-Encapsulated Gold Nanorods for Improved Stability and Its Application in OCT Imaging with Enhanced Contrast
Gold nanorods (GNRs) are synthesized with a surfactant template, which often poses toxicity issues for biomedical applications. In addition, blue shift of longitudinal surface plasmon resonance (LSPR) peak of GNR is an inherent problem that needs to be addressed for time-course studies. In this work, we resolve these issues by optimizing the encapsulation of GNRs with polyethylene glycol (PEG) where biocompatibility is improved by ∼20 % and blue shift over a period of 8 days is reduced from 20 nm in the case of CTAB-GNR to 2 nm for PEG-encapsulated GNR. The encapsulated GNRs were then bioconjugated for targeted dark-field imaging of cancer cells. As an application, we also demonstrate the contrast-enhancing capability of GNRs in optical coherence tomography (OCT) imaging of tumor xenograft where the LSPR closely matches the OCT excitation wavelength. Our study proves that incorporating GNRs enhances the contrast of tumor tissue interfaces along with a considerable broadening in OCT depth profile by six times.
KeywordsGold nanorods PEG encapsulation Longitudinal surface plasmon resonance Cytotoxicity Contrast enhancement OCT imaging
Numerous studies have been conducted to understand the incorporation of gold nanoparticles into biomedical applications. Recently, there has been avid interest in developing different nanoparticles such as nanospheres [1, 2], nanoshells [3, 4, 5], and nanorods [6, 7] for biomedical applications. Their intrinsic optical properties are manipulated to fit the requirements of those applications. One particular property of nanoparticles is the localized surface plasmon resonance (SPR), which can be adjusted to match the wavelength of the excitation source to enhance the scattering . The SPR of these nanoparticles are determined by their dimensions and in the case of nanospheres, SPR can only be tuned up to 600 nm by adjusting their diameter to 100 nm . For gold nanoshells, their core diameter can go up to 110 nm with a 2-nm gold shell thickness, resulting in an SPR of 820 nm . However, these dimensions are unacceptable for most of the biomedical usages in near-infrared (NIR) imaging, especially for in vivo applications where a mammalian cell can only assimilate particles that are up to ∼100 nm in diameter .
Anisotropic nanoparticles such as nanorods are of great interest as it can overcome the aforementioned limitations. Gold nanorods (GNRs) possess unique optical properties that open the doorway to many possibilities. Firstly, GNRs exhibit two surface plasmon resonances: longitudinal and transverse surface plasmon resonances (LSPR and TSPR, respectively). In a typical GNR synthesis, the TSPR will be about 520 nm while the LSPR can be tuned from 650 to 1,300 nm . This property is particularly important in NIR applications such as optical coherence tomography (OCT), NIR spectroscopy, and infrared thermography, where the excitation wavelength of the source is beyond 700 nm. Secondly, the versatility of GNR synthesis is warranted by the facile and cost-effective synthesis protocol. By using the seed-mediated approach, GNRs of various dimensions can be obtained by adjusting certain parameters such as the amount of the surfactant molecules and the length of the growth time .
The use of GNRs, however, does not come without practical problems. Firstly, in biomedical applications, there is a growing concern regarding the introducing of such nanoparticles into a mammalian system. An important concern associated with GNRs is the use of the cytotoxic cationic surfactant, cetyl trimethylammonium bromide (CTAB), which is a common structure-directing component used in GNR synthesis. It is widely used as an antiseptic agent [14, 15, 16] and lysis buffer for DNA extraction [17, 18, 19]. As such, the presence of a CTAB layer on GNRs poses indiscriminate toxicity to both normal and cancerous cells.
Secondly, surfactant-stabilized GNRs have the inherent issue of shape modification where the aspect ratio (length-to-diameter) shrinks, causing them to lose their anisotropy. This change in aspect ratio in turn causes the LSPR to blue shift. In medical applications, the short shelf-life of GNRs causes problems in time-course in in vivo studies. The phenomenon of LSPR blue shift in GNRs has been well reported . A study has also been performed to understand the effect of varying concentrations of CTAB on the LSPR of GNRs in aqueous environment .
Efforts in protecting the surrounding cells from the CTAB layer have been put in to circumvent the issue of toxicity. Encapsulation of gold nanoparticles has been reported in other works, such as with bovine serum albumin , citrate anions , and silica . In our previous work, Pluronic was used to coat GNRs for a two-prong approach to tackle both toxicity and stability issues . However, Pluronic encapsulation does not warrant stable bioconjugation for targeted biosensing applications.
In this work, we propose an approach to encapsulate GNRs with optimized polyethylene glycol (PEG) concentration for improved stability where blue shift over a period of 8 days is reduced from 20 nm (in the case of CTAB-GNR, CGNR) to 2 nm for PEG-encapsulated GNR (PGNR). To the best of our knowledge, this is the first demonstration of stabilizing LSPR of GNR by PEG encapsulation. GNRs are characterized by UV–vis absorption spectra analysis and transmission electron microscopy (TEM). Improved biocompatibility of PGNR is demonstrated by cell viability study. Also, we demonstrate the bioconjugation of epithelial growth factor receptor (EGFR) antibodies onto PGNR to target epidermoid carcinoma (A431) cancer cells. Specific binding of GNRs onto the cells was examined with dark-field imaging. We also report the use of GNRs as contrast-enhancement agents in OCT imaging in a tumor xenograft tumor. Our study proves that the coupled application of GNRs enhances the contrast of tissue layers, allowing one to closely monitor the epithelial thickening in the course of cancer development.
Chloroauric acid (HAuCl4), sodium borohydride (NaBH4), benzyldimethylammoniumchloride hydrate (BDAC), cetyl trimethylammonium bromide (CTAB), silver nitrate (AgNO3), l-ascorbic acid, 1-ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), and cell counting kit (CCK8) were purchased from Sigma Aldrich was diluted with HPLC-grade water before use. Thiolated and carboxylated polyethylene glycol (PEG, molecular weight 5,000 and 3,000 Da, respectively) were purchased from Rapp Polymere. EGFR antibody (100 μg/ml) was purchased from Santa Cruz. Dulbecco’s Modified Eagle Medium (DMEM), fetal bovine serum (FBS), and penicillin streptomycin (Pen Strep) were purchased from Gibco, and 0.05 % (v/v) Trypsin–EDTA (1×) were obtained from PAA. Clean-mount solution to fix the glass cover slip over the eight-chamber slides was purchased from Electron Microscopy Sciences
Optimized Synthesis of GNRs
The synthesis of GNR with LSPR matching the excitation wavelength of the OCT was achieved by varying the BDAC-to-CTAB mass ratio. Seed and growth solutions were prepared using modified protocol by Nikoobakht et al. . In the seed solution, 1 ml of 0.5 mM HAuCl4 was added to 1 ml of 200 mM CTAB to obtain an amber-colored solution. One hundred twenty microliters of 10 mM NaBH4 was then added to the solution. In the growth solution, 5 ml of 150 mM BDAC was added to CTAB (50, 100, and 200 mg) and mixed into a clear solution at 40 °C. Two hundred microliters of 4 mM AgNO3 was added to the solution, followed by the addition of 5 ml of 1 mM HAuCl4 to obtain an amber solution. Seventy microliters of 80 mM l-ascorbic acid was added into the amber solution, yielding a colorless solution. Twelve microliters of the seed solution was injected into the growth solution and was allowed to incubate at 40 °C over a period of 24 to 48 h. The resultant GNR solution was then centrifuged and redispersed in water a few times to remove unbound excess surfactant. In the final round, the GNR solution was centrifuged at 5,000 rpm and the supernatant containing the GNRs was obtained and resuspended in 4 ml PBS.
Optimized PEG Encapsulation of GNRs
GNR solution was mixed with various concentrations of thiolated PEG (10, 20, 40, and 60 μM) in 4-to-1 volume ratio for optimized encapsulation and was allowed to incubate for 2 h before removing excess thiolated PEG by centrifugation. The solution was later resuspended in 2 ml water. LSPR shifts were measured using a UV–vis absorption spectrometer over a period of 8 days.
Bioconjugation of GNRs
Bioconjugation protocol was adapted and modified from Lee at al. . Briefly, 1 ml of 10 μM carboxylated PEG was mixed with 4 ml of GNR solution and the mixture was allowed to mix for 15 min. Subsequently, 2 ml of thiolated PEG of optimized concentration was added to the mixture and was allowed to stir for further 3 h. Excess PEG solutions were removed by centrifugation and resuspended in 2 ml of PBS. Later, 5 μl of 25 mM EDC and NHS were added to PEG-encapsulated GNR solution and thoroughly mixed for 20 min. Excess cross-linkers were removed by centrifugation and 200 μl of purified EGFR antibody solution was added to conjugate the GNRs. The solution was left to incubate for 2 h at room temperature before storing at 4 °C.
A431 (epidermoid carcinoma, which overexpresses EGFR) and SKOV3 (ovarian carcinoma, which do not overexpress EGFR) cells was cultured in DMEM containing 10 % FBS with Pen Strep. All cultures were maintained at 37 °C with 5 % CO2.
Cancer Xenograft Model Development
Before inoculation, A431 cells were washed with PBS, trypsinized, and counted using a hemocytometer. Male Balb/c nude mice, 6–8 weeks of age, weighing an average of 24–25 g were obtained from the Animal Resource Centre (ARC), Western Australia. Approximately 3.0 × 106 A431 cells suspended in 150 μl of Hanks’ balanced salt solution (Gibco, USA) was injected subcutaneously into the lower flanks of the mice. The tumors were allowed to grow to 6–7 mm in diameter for approximately 14 days after inoculation. All handling procedures for mice were approved by the Institutional Animal Care and Use Committee at SingHealth, Singapore, in accordance with international standards.
The mouse was anesthetized before 150 μl of bioconjugated GNRs suspended in PBS at an optical density of 1.7 was injected intratumorally with minimal leakage. After incubating the GNRs for 5 h, the tumor xenografts were imaged with OCT.
Five thousand OSCC cells were seeded in a 96-well plate for 24 h before loading each well with 10 μl of synthesized GNRs. After further incubating the cells for 24 h, 10 μl of CCK8 was added to each well and allowed to incubate for 4 h in dark at 37 °C with 5 % (v/v) CO2. Cell population absorbance was performed with the SpectraMax 384 Plus spectral analyzer. The absorbance from the tetrazolium dye in CCK8 was measured at 450-nm excitation.
Five thousand cells of each cell line (A431 and SKOV3) suspended in media were seeded in each well of the eight-well chamber glass slide. When cells were confluent, media was removed and cells were rinsed with PBS. Media was replenished in the wells. The GNRs were then loaded and incubated with the cells in dark for 4 h at room temperature. The media and synthesized substances were removed and the cells were washed with PBS again, and fixed with 4 % paraformaldehyde for 10 min. Thereafter, 4 % paraformaldehyde was removed and the cells were washed twice with PBS. The slides were then mounted with clean-mount. Imaging was performed with Nikon Eclipse 80i at ×100 magnification.
OCT Setup and Image Acquisition
Characterizations of GNRs
TEM images to characterize the morphologies of the GNRs were taken using JEOL JEM-1010 operating at accelerating voltage of 80 kV. The specimens were prepared on 200-mesh nickel-coated grids. The absorption spectra of the GNRs were obtained with the Hitachi U-2900 UV–vis spectrophotometer at a spectral bandpass of 1.5 nm.
Results and Discussion
Optimized Synthesis of GNRs
It was observed that the LSPR of GNRs increases as the BDAC-to-CTAB ratio increases. This trend is consistent with other reported works [12, 27]. In general, the micelle concentration of CTAB has to be above the critical micelle concentration (CMC) to grow anisotropic nanoparticles [28, 29, 30]. This theory holds only for solutions containing a single surfactant that is CTAB. When BDAC is introduced into the system, the CMC falls, thus allowing the elongated micelles to grow further unilaterally .
Optimized PEG Encapsulation of GNRs for Improved Stability
Stability of encapsulated GNRs for storage over a period of time has not been extensively studied. This result is more promising than in the previous report of Pluronic-encapsulated GNR by our group where the LSPR peak of GNRs experienced a minimal blue shift of 7 nm, as compared to that of CTAB-stabilized GNRs with a shift of 20 nm .
Cell Viability Studies with PGNRs
CTAB has been known to be toxic to cells and tissues. Thus, the presence of the surfactant limits the use of GNRs in medical applications [33, 34, 35, 36]. GNRs are generally washed by centrifugation to remove unbound excess surfactant present in the solution. However, the presence of an appropriate amount of surfactant serves as a stabilizing agent for GNRs in aqueous solutions [37, 38, 39]. This implies that total removal of surfactant, if possible, by centrifugation would render a major isotropic change to the GNRs. In light of this, the advantage and disadvantage of the presence of surfactant can be tackled by means of encapsulation.
Dark-Field Imaging of Cancer Cells with PGNRs
The ability of nanoparticles to label cancer cells or sites is particularly important for many biosensing and imaging applications. In this study, the particle uptake of bionconjugated
GNRs were investigated using dark-field imaging. The presence of 260-nm absorption peak  in UV–vis spectra confirmed that stable bioconjugation of EGFR antibodies onto GNRs was achieved.
Study of GNRs as Contrast-Enhancement Agents in OCT Imaging
The optimization, encapsulation for improved stability, biocompatibility, and antibody-conjugation for targeted homing of GNRs were discussed earlier. To realize the practical application of GNRs, its utilization as a contrast agent in OCT imaging of tumor xenograft in mice was demonstrated.
Despite the similar probing depth of up to ∼600 μm, clear tissue interface along with a two-peak depth profile with a FWHM of 325 μm can only be seen when the tumor was injected with bioconjugated GNRs. We note that this enhanced interfacial contrast is accompanied with a considerable broadening of the depth profile by 155 μm (Fig. 6d). Compared with previous result, this profile broadening increases by about six times. The white arrows in Fig. 6c outline the interface between the upper epidermis and lower dermis layers. The lower dermis layer is the connective tissue whereas the upper epidermis is a cell-rich tissue layer where the epithelial cancer is triggered. Compared with Fig. 6b, clearer interfacial contrast is obtained when using the GNRs, allowing one to examine the process of epithelial thickening in skin cancer. For instance, the A-scan profile in Fig. 6d depicts the peaks and valley in the reflectance profile of tumor injected with GNRs. Accordingly, epithelial thickness is estimated to be 150 μm by measuring the distance between the first peak and the valley.
Our dark-field imaging study has proven the binding specificity of the GNRs to A431 cells used in mouse tumor xenograft model. Thus, the increased image contrast here is attributed to the preferential localization of the GNRs at epidermal tumors, inducing a differential light-scattering contrast between the epidermis and dermis layers . A simulation study has previously confirmed this contrasting effect given by the locally distributed gold nanospheres in tissue . Similar effect was noticed when bare gold nanosphere droplet was left over time on the superficial skin layer for gradual penetration into deeper layers . By contrast, our GNRs were made to specifically bind with the epithelial cancer cells only and thereby improving a gradual accumulation of GNRs in the upper epidermis layer. We hypothesize that the optical contrast is time varying and will reach an optimal level when there is a stark difference in the amount of GNRs between tissue layers while allowing sufficient light to probe the whole epidermis for thickness measurement. Further studies are currently under way to determine the optimal circulation time of GNRs in the mouse model to achieve the best OCT contrast.
We optimized the synthesis of GNRs to closely match the OCT excitation wavelength. PEG encapsulation was successfully carried out to minimize the LSPR blue shift of the GNRs. In our case, PGNRs exhibited a minimal blue shift of only ∼2 nm as compared to that of CGNRs with ∼20 nm shift over a period of 8 days. The morphologies of PGNRs do not alter with encapsulation, thus preserving their characteristic LSPRs. PGNRs exhibited promisingly low cytotoxicity. Dark-field imaging of EGFR-overexpressing A431 cells showed that specific binding was achieved by conjugating the GNRs with antibody. We also demonstrated that these GNRs can be translated as contrast-enhancement agents in OCT imaging of tumor xenograft. Our study showed that incorporation of GNRs into OCT imaging improved the contrast of tissue interfaces that permit a monitoring of epithelial thickening for cancer diagnosis.
The authors would like to thank the Joint Council Office, A*STAR, Singapore, (CCOGA02_002_2008) for the financial support. They would like to thank Dr. Hanhua Feng (Institute of Microelectronics, A*STAR) for her assistance in this project. Authors also thank Mr. Gong Tianxun and Prof. Ken-Tye Yong (Nanyang Technological University, Singapore) for their assistance in dark-field imaging.
- 7.Xu W, Luo T, Li P, Zhou C, Cui D, Pang B, Ren Q, Fu S (2012) RGD-conjugated gold nanorods induce radiosensitization in melanoma cancer cells by downregulating alpha(v)beta(3) expression. Int J Nanomed 7:915–924Google Scholar
- 17.Ye J, Ji A, Parra EJ, Zheng X, Jiang C, Zhao X, Hu L, Tu Z (2004) A simple and efficient method for extracting DNA from old and burned bone. J Forensic Sci 49:754–759Google Scholar
- 19.Hsia AP, Chen HD, Ohtsu K and Schnable PS (2010) DNA extraction from freeze-dried plant tissue with CTAB in a 96-well format. Cold Spring Harb Protoc:pdb prot5516Google Scholar
- 25.Goh D, Gong T, Dinish US, Maiti KK, Fu CY, Yong K-T, Olivo M (2012) Pluronic triblock copolymer encapsulated gold nanorods as biocompatible localized plasmon resonance-enhanced scattering probes for dark-field imaging of cancer cells. Plasmonics. doi:10.1007/s11468-012-9347-3
- 45.Zagaynova EV, Shirmanova MV, Kirillin MY, Khlebtsov BN, Orlova AG, Balalaeva IV, Sirotkina MA, Bugrova ML, Agrba PD, Kamensky VA (2008) Contrasting properties of gold nanoparticles for optical coherence tomography: phantom, in vivo studies and Monte Carlo simulation. Phys Med Biol 53:4995–5009CrossRefGoogle Scholar