Noninvasively Imaging Subcutaneous Tumor Xenograft by a

May 5, 2017 - Noninvasively Imaging Subcutaneous Tumor Xenograft by a Handheld Raman Detector, with the Assistance of an Optical Clearing Agent. Yunfe...
0 downloads 10 Views 2MB Size
Download PDF
Research Article www.acsami.org

Noninvasively Imaging Subcutaneous Tumor Xenograft by a Handheld Raman Detector, with the Assistance of an Optical Clearing Agent Yunfei Zhang,† Haoran Liu,† Jiali Tang,‡ Zhuoyun Li,† Xingyu Zhou,† Ren Zhang,§ Liang Chen,∥ Ying Mao,*,∥ and Cong Li*,† †

Key Laboratory of Smart Drug Delivery, Ministry of Education, School of Pharmacy, Fudan University, Shanghai 201203, China Department of Pharmaceutical Engineering, School of Engineering, China Pharmaceutical University, Nanjing 210009, China § Center of Analysis and Measurement, Fudan University, Shanghai 200433, China ∥ Department of Neurosurgery, Huashan Hospital, Fudan University, 12 Wulumuqi Middle Road, Shanghai 200040, China ‡

S Supporting Information *

ABSTRACT: A handheld Raman detector with operational convenience, high portability, and rapid acquisition rate has been applied in clinics for diagnostic purposes. However, the inherent weakness of Raman scattering and strong scattering of the turbid tissue restricts its utilization to superficial locations. To extend the applications of a handheld Raman detector to deep tissues, a gold nanostar-based surface-enhanced Raman scattering (SERS) nanoprobe with robust colloidal stability, a fingerprint-like spectrum, and extremely high sensitivity (5.0 fM) was developed. With the assistance of FPT, a multicomponent optical clearing agent (OCA) efficiently suppressing light scattering from the turbid dermal tissues, the handheld Raman detector noninvasively visualized the subcutaneous tumor xenograft with a high target-to-background ratio after intravenous injection of the gold nanostar-based SERS nanoprobe. To the best of our knowledge, this work is the first example to introduce the optical clearing technique in assisting SERS imaging in vivo. The combination of optical clearing technology and SERS is a promising strategy for the extension of the clinical applications of the handheld Raman detector from superficial tissues to subcutaneous or even deeper lesions that are usually “concealed” by the turbid dermal tissue. KEYWORDS: Surface-enhanced Raman scattering (SERS), Optical clearing, Gold nanostar, Noninvasive diagnosing, Handheld Raman detector

1. INTRODUCTION Surface-enhanced Raman scattering (SERS) is a plasmonic effect where the molecular reporters are in close proximity to nanoroughened noble metal surfaces, which enhances the Raman signal by 7−10 orders of magnitude.1 The incorporation of gold or silver nanoparticles with a strong SERS response dramatically broadens the medical applications of Raman imaging in diagnosis,2 prognosis,3 and image-guided surgery4 by providing real-time, quantitative, and chemically specific information of the target molecules in the cellular or tissue samples.5 Compared with optical imaging approaches such as fluorescence imaging, SERS possesses higher sensitivity,6 less susceptibility to photobleaching, and better multiplexing capability.7 Near-infrared (NIR) light-excited SERS received wide interest for in vivo studies because of low absorption of the endogenous molecules, reduced phototoxicity, and autofluorescence background from the living tissues. Gambhir et al. noninvasively imaged mouse liver with up to five SERS nanoprobes and obtained distinguishable, fingerprint-like Raman spectra after excitation with a 785 nm laser light.8 © 2017 American Chemical Society

With the assistance of an NIR light-excited SERS probe, Garside et al. visualized inflammatory lesions in animal models using Raman imaging that showed a higher sensitivity than a range of fluorescence-based imaging technologies, including two-photon fluorescence microscopy.9 Currently, most in vivo SERS studies are conducted by confocal Raman spectroscopy because of its high sensitivity and spatial resolution.8 The static Raman microscope, however, is designed for bench-top experiments and is difficult and hardly to be used in clinical settings because of its immobility, insufficient operating space, fixed scanning angle, long acquisition time, and narrow field of view. To overcome the limitations of instrumental apparatus, handheld Raman detectors with high portability, convenient handling, and rapid acquisition rates have been applied in examining cervical tumors,10 imaging malignant skin cancers,11 and delineating Received: March 24, 2017 Accepted: May 5, 2017 Published: May 5, 2017 17769

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces intraoperatively brain tumor margins.12 However, current applications of the handheld Raman detector in clinical settings are restricted to lesions located in superficial tissues. Strong scattering of turbid dermal tissue sharply decreases both the resolution and contrast of Raman imaging when the light propagates into deep tissues. Suppression of light scattering in the dermal tissue could therefore increase detectability and extend applications of the handheld Raman imaging system. The optical clearing technique is promising for reducing the light scattering of turbid tissues through application of osmophilic optical clearing agents (OCAs) that increase optical homogeneity of the tissues through matching the refractive indexes (RIs) of different tissue components. Because of the remarkable increase of light penetration depth in the living tissues, this technique has been applied for improving sensitivity and spatial resolution of multiple optical imaging techniques, including fluorescence imaging,13 photoacoustic imaging,14 optical coherence tomography,15 and laser speckle contrast imaging.16 Recent reports showed that a mixture of FDA-approved poly(propylene glycol) and poly(ethylene glycol) (PEG) increased the optical “transparency” of human skin in vivo.17 Gallwas et al. proposed that OCA composed of a mixture of dimethyl sulfoxide (DMSO) and PEG400 significantly clarified the cervical epithelium specimens.18 The optical clearing technique, however, has not been introduced to SERS imaging. Neither has it assisted the handheld Raman detectors to visualize subcutaneous lesions. We developed a gold nanostar-based SERS nanoprobe that demonstrated robust colloidal stability, spectral fingerprint, and femtomolar sensitivity upon excitation by the NIR light. A homemade apparatus mimicking a vasculature embedded in a simulated turbid “tissue” confirmed that the optical clearing significantly increased the tissue-imaging depth and detectability of the handheld Raman detector. FPT, a multicomponent OCA comprising fructose, PEG400, and thiazone demonstrated the highest efficacy in reduction of light scattering in the dermal tissue. In vivo studies confirmed that FPT could remarkably improve the detectability of the subcutaneous SERS nanoprobe. The subcutaneous tumor xenograft was noninvasively visualized by the handheld Raman detector with high sensitivity shortly after intravenous injection of the SERS nanoprobe followed dermal treatment of FPT (Figure 1). To the best of our knowledge, this work is the first example to introduce the optical clearing technique into SERS imaging. Combining the optical clearing technique and SERS nanotechnology is a promising strategy for extension of clinical applications of handheld Raman detectors from superficial regions to subcutaneous or even deeper tissues.

2. MATERIALS AND METHODS 2.1. Materials and Instruments. Unless otherwise specified, all chemical reagents were purchased from Aladdin Chemistry Co. Ltd. (China) and were used without further purification. Intralipid (20%) was obtained from Sichuan Kelun Pharmaceutical Co. Ltd. (China). mPEG-SH (2.0 kDa) was obtained from Tuoyang Co. Ltd. (China). FPT was prepared by mixing aqueous fructose solution (78%, w/w), PEG400, and thiazone at a volume ratio of 11:8:1.19 The RIs of the OCAs, including the saturated aqueous solutions of glucose (52%, w/ w), fructose (78%, w/w), glycerol, DMSO, 1,4-butanediol, 1,2propanediol, and PEG400, were measured using an Abbe refractometer (WAY-2WAJ; Shanghai INESA Physico-Optical Instrument Co. Ltd, China). The gold nanostar-based SERS nanoprobe was characterized by UV−vis spectroscopy (UV-2550; Shimadzu, Japan), transmission electron microscopy (TEM) (Tecnai G2spirit Biotwin, FEI), dynamic light scattering (Zetasizer 3000; Malvern Instruments), and Fourier transform infrared (FT-IR) spectroscopy (Nicolet; Thermo Nicolet Corporation). Molar concentrations of AuNS-Cy7 were determined by a nanoparticle tracking analyzer (Nanosight NS300; Malvern Instruments, U.K.). Ex vivo confocal Raman spectroscopic images were obtained from a Raman microscopic system (XploRA confocal microRaman system; Horiba Jobin Yvon, France). Unless otherwise noted, SERS spectra were recorded with a handheld Raman detector (QE65Pro; Ocean Optics). An optic probe (Inphotonics) equipped with a 785 nm laser (Ocean Optics) was used for both excitation and collection purposes. This handheld Raman detector possesses a 200 μm collection fiber and a 105 μm excitation fiber. The numerical aperture is 0.22. SERS spectra were processed by LabSpec5 software (v2.02, 2010). 2.2. Synthesis. The detailed synthetic route and supplemented spectra including 1H and 13C NMR spectra of Cy7-SH (3) are presented in Supporting Information. Molecular reporter Cy7-SH (3) was synthesized according to the Dinish’s method.20 Gold nanostar (5) was prepared by a previous method with minor modifications.21 In brief, the addition of 15 mL of citrate solution (1%) into 100 mL of boiling chloroauric acid aqueous solution (1.0 mM) yielded gold seeds (4) with an average diameter of 12 nm. The mixture of 4.0 mL of gold seed solution and 800 μL of HCl (0.1 M) was added into 200 mL of chloroauric acid (0.25 mM). 2.0 mL of AgNO3 (3.0 mM) and 1.0 mL of ascorbic acid (100 mM) were then added simultaneously to the previous mixture to yield gold nanostar (6), with vigorous stirring (1100 rpm) for 30 s. Cy7-SH and mPEG-SH were successively added to the gold nanostar solution and stirred gently for 2 h (molar ratio of Cy7-SH/mPEG-SH/Au nanostars: 1500/10 000/1) to give the desired SERS nanoprobe, AuNS-Cy7 (7), after purification in a 10 000 MW cutoff dialysis bag. 2.3. Sensitivity of AuNS-Cy7. Working solutions of AuNS-Cy7 with concentrations of 400, 200, 100, 50, 25, 20, 10, and 5 fM in distilled water were prepared. The Raman signal intensities of AuNSCy7 as a function of concentration were determined by quantifying the characteristic Raman peak assigned to C−S vibration (∼541 cm−1). The incident laser power was 80 mW, with a wavelength of 785 nm. Acquisition time for each measurement was 1.0 s (100 times). The confocal Raman spectroscopic images of the AuNS-Cy7 solutions with concentrations of 100, 50, 20, and 5 fM were collected with a step width of 3 μm and an acquisition time of 1.0 s for each mapping point (100 times). 2.4. Stability of AuNS-Cy7. The colloidal stability of AuNS-Cy7 (1.0 nM) was studied in 0.01 M HCl, 0.01 M NaOH, 1.0 M NaCl, and phosphate-buffered saline (PBS) by measuring the SERS spectra collected at 0, 1, 2, 4, 8, 16, 24, and 48 h post incubation (laser power: 400 mW, acquisition time: 200 ms). 2.5. In Vitro Raman Phantom Imaging. A glass capillary with an inner diameter of 0.3 mm used to mimic a “blood vessel” was immersed at selected depths into an intralipid medium (20%) mimicking turbid skin tissue. Selected OCAs and water were supplemented to keep the final concentration of intralipid (5%, v/v) unchanged. SERS spectra of a flowing AuNS-Cy7 (1.0 nM) solution in a glass capillary, with a flow rate of 2.0 mm/s driven by a syringe

Figure 1. Handheld Raman detector noninvasively visualizes subcutaneous tumor xenograft in vivo, aided by an OCA that suppresses light scattering from turbid skin tissue. 17770

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces pump, were collected by a handheld Raman detector (laser power: 400 mW, acquisition time: 200 ms) as functions of OCA types and the immersed depths below the intralipid medium level (from 0.1 to 1.1 mm, 0.1 mm interval). The supplemented OCAs included glycerol, DMSO, 1,4-butanediol, and 1,2-propanediol. 2.6. Establishment of Subcutaneous Tumor Xenografts. All animal experiments were carried out according to the Guide for the Care and Use of Laboratory Animals, Fudan University (Shanghai, China). ICR mice and athymic nude mice (15−20 g, 3−4 weeks old) were purchased from the Laboratory Animal Center of Fudan University (Shanghai, China). Briefly, murine glioblastoma C6 cancer cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum (FBS), 1% penicillin, and streptomycin (37 °C, 5% CO2). The male nude mice were anaesthetized by intraperitoneal injection of 150 μL of chloral hydrate (5%). The cancer cells (2 × 107 cells in 200 μL of PBS) were subcutaneously injected into the dorsum. The animal models with tumor volumes greater than 10 mm3 were used for imaging studies. 2.7. Ex Vivo Raman Imaging Studies. The hair of the euthanatized ICR mice was shaved and completely removed by depilatory cream. The dorsal skin was excised carefully and cut into pieces with an area of approximately 1.0 cm2 and immersed into solutions of selected OCAs including fructose (78%), glucose (52%), PEG400 (100%), 1,4-butanediol (100%), glycerol (100%), 1,2propanediol (100%), FPT (a multicomponent OCA including fructose aqueous solution (78%, w/w), PEG400, and thiazone, with a volume ratio of 11:8:1) and PBS. After immersion in the selected OCA for a designated time (from 0 to 60 min, with an interval of 10 min), the murine skin pieces were carefully wiped and placed above AuNS-Cy7 (1.0 nM) that was embedded between two glass slides. The SERS spectra of the subcutaneous nanoprobe were collected by the handheld Raman detector, with a laser power of 350 mW and acquisition time of 200 ms. 2.8. In Vivo Raman Imaging Studies. The ICR mice were anaesthetized and their dorsal hair shaved. AuNS-Cy7 (5 nM in 200 μL of distilled water) was subcutaneously injected at the dorsum. FPT (2.0 mL) or fructose solution (2.0 mL) was gently smeared on the dorsal skin of the mice. SERS spectra were collected using the handheld Raman detector at selected time points post-OCAs treatment. For noninvasively visualizing the subcutaneous tumor, AuNS-Cy7 (0.2 nmol in 200 μL of distilled water) was injected intravenously into mouse models bearing the C6 glioblastoma xenograft. At 24 h post the nanoprobe injections, the mice were anaesthetized and the tumor site was locally treated with FPT (2 mL) via gentle smearing. The handheld Raman detector was placed approximately 1.0 cm above the tumor site. The SERS spectra were collected at selected time points post the FPT treatment and the integration time for each measurement was 200 ms. The laser power was 400 mW. The white-light images of the tumor-bearing mouse before and 30 min after the FPT treatment were also captured. 2.9. Ex Vivo Confocal Raman Spectroscopy and TEM Studies. At 24 h after the intravenous injection of AuNS-Cy7, the mouse specimens with subcutaneous C6 glioblastoma xenografts were anaesthetized and perfused with saline and then paraformaldehyde (4%). The tumor was excised and then cyro-sectioned with a thickness of 200 μm. The Raman spectroscopic images of the tumor sections were collected by measuring the intensity of the characteristic Raman peak at 509 cm−1, with 10 μm steps, with the incident laser wavelength of 785 nm, laser power of 80 mW, and acquisition time of 1.0 s. The ultrathin tumor sections (150 nm) were also prepared and used in TEM (Tecnai, G2spirit Biotwin, FEI) imaging studies. 2.10. Side-Effects of Dermal OCA Treatment and AuNS-Cy7. White-light images of the dorsal skin in nude mice before and at selected times post the FPT treatment and injections of AuNS-Cy7 were captured. At 7 and 21 days post the FPT dermal treatment, murine skin, liver, and kidney tissues were harvested and cyrosectioned with a thickness of 5 μm and then stained with hematoxylin and eosin (H&E) for pathological analysis.

3. RESULTS AND DISCUSSION 3.1. Design, Synthesis, and Characterization of AuNSCy7. Surface-enhanced resonance Raman scattering (SERRS) occurs when the electronic excitation energy of the reporter molecules (RMs) attached to a nanoparticle surface is close to that of the incident laser, which results in signal enhancement by more than 3 orders of magnitude compared to that on SERS.22 Because of its good biocompatibility and electronic excitation wavelength similar to that of the incident laser (785 nm) equipped in most handheld Raman detectors, the heptamethine cyanine derivative, Cy7-SH, with maximum absorption at 790 nm, was chosen as the RM. Meanwhile, gold nanostar (AuNS) was selected as the SERRS substrate because it possesses numerous sharp tips. They remarkably enhance the Raman signal via a “hot spot” effect induced by tremendous enhancement of the local electromagnetic field.23 The covalent coating of Cy7-SH on the AuNS surface via Au−S bond kept the RM in proximity to the rough gold surface for maximizing the SERS effect. PEGs (2.0 kDa) were fabricated on the AuNS surface to improve the Au colloidal stability and prolong the circulation lifetime of the nanoprobes.24 Figure 2A illustrates the synthetic procedures of the SERS nanoprobe,

Figure 2. Synthesis and characterization of the SERS nanoprobe, AuNS-Cy7. (A) Synthetic procedure of AuNS-Cy7. (i) n-BuOH/ toluene, 120 °C; (ii) lipoic acid, DCC/TsOH, dichloromethane, 70 °C; (iii) sodium citrate/H2O, 100 °C; (iv) silver nitrate, ascorbic acid, r.t.; (v) Cy7-SH, r.t.; and (vi) mPEG-SH, r.t. (B) TEM images of AuNS-Cy7. Red arrows indicate toward PEG coating. (C) SERS signal increased linearly with AuNS-Cy7 concentration in a range of 5−400 fM (R2 = 0.9995). (D) Confocal Raman spectroscopic images of aqueous AuNS-Cy7 solutions at concentrations of 5, 20, 50, and 100 fM. Scale bar = 8.0 μm. (E) Representative Raman spectrum of AuNSCy7 in aqueous solution (5.0 nM). The characteristic twin peaks are pointed with arrows. r.t.: room temperature, DCC: dicyclohexylcarbodiimide. 17771

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces

determined by measuring the time-dependent Raman intensity under neutral (pH 7.4), acidic (pH 2.0), alkaline (pH 12), and highly saline (1.0 M NaCl) conditions (Figure S8). Less than 15% Raman signal attenuation was observed after 48 h of incubation under all of the above conditions, which indicated good colloidal stability. The femtomolar sensitivity, robust colloidal stability, and spectral fingerprint were the advantages of using AuNS-Cy7 to visualize subcutaneous lesions using a handheld Raman detector. 3.2. Detection of AuNS-Cy7 in a Simulated Turbid Tissue. The use of OCAs to assist in the identification of AuNS-Cy7 immersed in turbid tissue using a handheld Raman detector was investigated using a homemade apparatus comprising a glass capillary mimicking a “vasculature” and a container holding a 5% intralipid medium mimicking the turbid tissue (Figure 3A). The glass capillary, with syringe pump-

AuNS-Cy7. Briefly, the reaction of 1-(2-hydroxyethyl)-2,3,3trimethyl-3H-indol (1) and 1-cyclohexene-1-carboxaldehyde in a mixture of 1-butanol/toluene (7:3, v/v) yielded heptamethine cyanine derivative (2). The conjugation between compound 2 and lipoic acid, with dicyclohexylcarbodiimide (DCC) as the dehydrating agent and p-toluenesulfonic acid as the catalyst, gave the sulfydryl group-modified molecular reporter Cy7-SH (3). The gold nanosphere acted as a gold seed (4) and gold nanostar (AuNS) (5) were prepared according to a previously reported seed-mediated growth method.21 The modifications of Cy7-SH and PEG-SH on the AuNS surface gave the aiming nanoprobe, AuNS-Cy7. Both naked AuNS and AuNS-Cy7 showed the maximum absorptions centered at ∼850 nm that was assigned to the longitudinal plasmon band of the monodispersed AuNS and indicated a localized surface plasmon resonance in the NIR wavelength range (Figure S1). The molar concentration of the nanoprobe was determined by nanoparticle tracking analysis that directly counted the nanoparticle number in a unit volume by an optical microscope equipped with a high-speed charge-coupled device camera. A good linear regression between the maximum absorption and molar concentration of AuNS-Cy7 in aqueous solutions was observed in Figure S2 (from 8 to 157 pM, R2 = 0.9997). The extinction coefficient of the AuNS-Cy7 was determined as 8.2 × 109 M−1 cm−1, according to the Lambert−Beer equation. The covalent conjugation of Cy7-SH to the AuNS surface was verified by FTIR spectra (Figure S3). AuNS-Cy7 demonstrated characteristic FT-IR peaks of Cy7-SH, including 2926 cm−1 (υCH2), 1735 cm−1 (υCO), 1550 cm−1 (υCC), 1393 cm−1 (δCH), 1141 cm−1 (υC−O), and 712 cm−1 (υC−Cl), which indicated the successful conjugation of Cy7-SH on AuNS. TEM images (Figures 2B and S4) showed both monodispersed AuNS-Cy7 and AuNS possessing a highly branched morphology and nearly the same particle size distribution. A semitransparent coating with a thickness of about 5.0 nm was observed in the TEM images of AuNS-Cy7 but not in those of AuNS, which indicated that the PEG layer was modified on the nanoprobe (Figure S5). The diameters of AuNS-Cy7 and AuNS were found to be between 55 and 70 nm, with an average of 60 nm. Each nanoprobe contained 6−16 prolate tips, with an average of 10. The tip lengths were between 15 and 25 nm, with an average of 20 nm. The number of the “protruding tips” has been reported to correlate proportionally with the excitation cross-section of the nanostar plasmons. The multiple tips enhance the local electromagnetic field and hence SERS signal of the nanoprobes.21,25 The hydrodynamic diameter and Zeta potential of AuNS-Cy7 were determined to be ∼80 nm and +25.4 mV in aqueous solution, respectively (Figures S6 and S7). The sensitivity of AuNS-Cy7 was determined by measuring its Raman signal intensity as a function of concentration (Figure 2C). The Raman signal of AuNS-Cy7 in aqueous solutions increased linearly within the femtomolar concentration range (from 5 to 400 fM, R2 = 0.9995). Confocal Raman spectroscopic images showed that the distinguishable Raman signal of AuNS-Cy7 could be detected even at concentrations as low as 5 fM (Figure 2D). Compared to that of the nonresonant SERS nanoprobes, the sensitivity of AuNS-Cy7 was enhanced by 2 orders of magnitude.26 The Raman spectrum of AuNS-Cy7 showed characteristic twin peaks located in 509 and 541 cm−1 that might be assigned to the S−S vibration and C−S vibration (Figure 2E). Colloidal stability of the AuNS-Cy7 under physiological conditions was

Figure 3. OCA increases detectability of the handheld Raman detector. (A) Illustration of the homemade apparatus, in which a glass capillary mimicking a subcutaneous vasculature was immersed into an intralipid medium (5%) mimicking the turbid skin tissue. The flowing AuNS-Cy7 aqueous solution in the glass capillary was driven by a syringe pump. (B) Raman signal intensities of AuNS-Cy7 as functions of immersion depth and OCAs added to the intralipid medium. Raman spectra (C) and Raman signal intensities (D) of AuNS-Cy7 in intralipid medium, in the absence or presence of OCAs. The capillary was placed 0.4 mm below the medium surface. The characteristic twin peaks are highlighted by a red rectangle. The peak at 509 cm−1 was used for signal-intensity quantification purposes. PG: 1,2-propanediol, BDO: 1,4-butanediol.

driven AuNS-Cy7, was immersed into the intralipid medium that has been widely used as a light-scattering medium due to its scattering characteristics being similar to those of the living tissue.27 Although no Raman signal of AuNS-Cy7 could be detected when the immersion depth of the glass capillary exceeded 0.1 mm, this nanoprobe could be visualized as deep as 0.9 mm after the addition of OCAs (Figures 3B and S9). The efficacy of the OCA to suppress light scattering of the turbid media was further evaluated at a glass capillary immersion depth of 0.4 mm. Although only background noise was detected before the addition of OCAs, the fingerprint spectrum of AuNS-Cy7 was evident after addition of various OCAs, including DMSO, glycerol, 1,2-propanediol, or 1,4-butanediol into the intralipid medium (Figure 3C). Quantification of the characteristic peak (509 cm−1) of the nanoprobe showed that DMSO offered the best optical clearing effect. The intensity of the Raman signal for the DMSO group was 1.2, 1.6, and 2.0 times that of glycerol, 1,4-butanediol, and 1,2-propanediol, 17772

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces

assisting the handheld Raman detector to visualize AuNS-Cy7 covered by murine dorsal skin was conducted, as shown in Figure 4A. After 1.0 h of immersion in selected OCA solutions,

respectively (Figure 3D). The above studies revealed that OCAs effectively suppressed light scattering and increased the tissue-imaging depth of the handheld Raman detector. No matter whether exogenous probes or endogenous biomolecules are visualized in subcutaneous lesions, the photons from the excited laser first need to penetrate the dermal tissue, undergo complete inelastic collision with the target molecule, and travel through the skin again to reach the detector. However, the scattering particles, such as collagens, elastic fibers, and intracellular organelles, with high RIs, and their surrounding interstitial fluid, cytoplasm media, with low RI, make the incident light travel with heterogeneous speeds and angles in turbid tissue, which leads to strong light scattering. The penetration of OCAs with high RI into turbid tissue helps to match the RIs between the scatters and the surrounding media, and therefore reduces the scattering coefficient. To further elucidate the mechanism of the OCA to enhance the imaging sensitivity of the nanoprobe immersed into the turbid intralipid medium, a simple approximation of Mie calculation was used as presented below27

Figure 4. OCA helps to detect AuNS-Cy7 covered by excised murine skin. (A) Illustration of a handheld Raman detector visualizing AuNSCy7 placed beneath a piece of murine dorsal skin, (B) Raman spectra, and (C) Raman signal intensities of AuNS-Cy7 beneath a piece of murine dorsal skin immersed in PBS or selected OCAs solution for 60 min. The characteristic twin peaks are highlighted by a red rectangle. The signal intensity of the Raman peak at 509 cm−1 was used for quantification purposes. BDO: 1,4-butanediol, PG: 1,2-propanediol, Gly: glycerol, Glu: glucose, FPT: a multicomponent OCA comprising 78% fructose solution, PEG400, and thiazone at a volume ratio of 11:8:1.

⎛ 2πa ⎞0.37 ⎟ [Q sca(1 − g )]pred = 3.28⎜ (m − 1)2.09 , ⎝ λ ⎠ n 2πa < 50, m = sca 1 < m ≤ 1.1, 5 < λ nbkg

where Qsca is the scattering coefficient, g is the average cosine of the scattering angle, a is the diameter of spherical scatters, λ is the wavelength of incident light, nsca is the refractive index of scatters, such as collagen, and nbkg is the refractive index of scattering medium, such as cytoplasm. As g is usually constant in living tissues,28,29 Qsca is mainly determined by 2πa/λ and m. Considering that most of the scattering objects have similar diameters and λ is usually constant for the handheld Raman detector, m is the key factor determining Qsca. Additionally, due to the larger power exponent of the m-containing term (2.09) than that of 2πa/λ (0.37), a slight increase of m will lead to a significant enhancement of the scattering coefficient. Therefore, the optical clearing technique focuses on reducing m to suppress scattering of the turbid tissue. Major scatters in tissues, including cytoplasmic organelles, connective-tissue fiber, and pigments, have RIs in the range of 1.39−1.47,30 which are much higher than those of extracellular fluid and cytoplasm, which are in the range of 1.35−1.37.31 The RI discrepancy between the scatters and scattering medium is a major reason leading to strong scattering of the turbid tissue. Addition of OCAs with high RI values into the intralipid solution assisted in matching the RI of the scatters (RI = 1.475 for lipid bilayer in intralipid) and the surrounding medium (RI = 1.333 for water) by increasing nbkg and decreasing m (ratio of nsca to nbkg). This led to the intralipid solution to be “transparent” with the concomitant SERS signal enhancement. The RI values of the OCAs were measured by the Abbe refractometer and decreased as follows: DMSO (1.4746) > glycerol (1.4701) > 1,4butanediol (1.4443) > 1,2-propanediol (1.4313) (Table S1). Notably, the RI decreasing sequence of the OCAs correlated to their optical clearing capabilities (Figure 3D), which validated the interpretation of the OCA-mediated Raman signal enhancement in the turbid intralipid media by the Mie equation. 3.3. Visualization of AuNS-Cy7 Covered by Murine Dorsal Skin. Evaluation of the effectiveness of OCA in

including fructose (78%), FPT (a multicomponent OCA comprising 78% fructose, PEG400 and thiazone with volume ratio of 11:8:1), glucose (52%), PEG400, 1,4-butanedial, glycerol, and 1,2-propanediol, the excised murine dorsal skin pieces were covered on AuNS-Cy7 (1.0 nM in aqueous solution) embedded between the glass slides. A handheld Raman detector was used to record the transcutaneous SERS signal above the murine skin. In contrast to the background noise recorded after the treatment of phosphate buffer saline (PBS) as a negative control, all of the OCAs significantly increased the SERS signal, as evidenced by the highly resolved fingerprint-like Raman spectra of AuNS-Cy7 (Figure 4B). Quantitative studies showed that fructose offered the best optical clearing efficiency. The Raman signal of the fructose group was 408 ± 10 counts per second, which was 1.1, 1.2, 1.8, 2.2, 3.0, 3.4, and 5.4 times the intensities of that of skin pieces treated with FPT, glycerol, glucose, 1,4-butanedial, 1,2propanediol, PEG400, and PBS, respectively (Figure 4C). The highest optical clearing efficiency of fructose could be explained by its capability to decrease the RI of the collagen fibers and other scatters by breaking the intermolecular hydrogen bonds via its multiple hydroxyl groups.19 SERS signal intensity variations as functions of the OCA types as well as immersion time are shown in Figure S10. SERS signals reached their maximum values after 30−50 min of immersion and then remained stable, indicating that 1.0 h of immersion time was sufficient to maximize the optical clearing effect of the OCAs. 3.4. FPT Showing the Best Optical Clearing Effect in Vivo. The skin not only provides a multifunctional interface between organisms and the external surroundings but also actively engages in forming a homeostatic barrier that limits molecular transport both from and into the body.32 Delivery of the OCA by overcoming the epidermal barrier to achieve satisfactory optical clearing effects in vivo has been challenging. 17773

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces Even though fructose showed the highest optical clearing efficiency in ex vivo experiments, the in vivo performance was far from satisfactory due to the low epidermal permeability.33 To overcome this problem, FPT, a multicomponent OCA composed of thiazone as a skin penetration promoter, fructose as an OCA, and PEG400 functionalized as both an OCA and as a dispersive menstruum, was studied. Figure 5A shows the

Figure 6. Handheld Raman detector noninvasively visualizes subcutaneous tumor xenograft aided by FPT. (A) Illustration of imaging the subcutaneous tumor xenograft after intravenous injection of AuNS-Cy7 and local FPT treatment. (B) In vivo Raman spectra of a tumor area at selected time points post FPT treatment. The characteristic twin peaks are highlighted by a red rectangle. The peak at 509 cm−1 was used for quantification purpose. (C) Timedependent Raman signal intensities of the tumor area post FPT treatment. (D) TEM images of the tumor section at 24 h PI of AuNSCy7. The yellow arrows highlight monodispersed AuNS-Cy7 with diameters of ∼60 nm. (E) Representative white-light images of a nude mouse before and at 30 min post local FPT treatment. AuNS-Cy7 was injected 24 h before the local FPT treatment. Insert: enlarged photographic images of the tumor area. (F) Confocal Raman spectroscopic images of the tumor section. Scale bar = 80 μm.

Figure 5. FPT exhibits the most efficient optical clearing effect in vivo. (A) Illustration of a handheld Raman detector imaging the subcutaneously injected AuNS-Cy7 in mouse dorsum after FPT or fructose treatment. In vivo Raman spectra of subcutaneous AuNS-Cy7 at selected time points post local treatment with FPT (B) or fructose (C). AuNS-Cy7 (5 nM, 200 μL) was injected subcutaneously before the OCA treatment. The characteristic twin peaks are highlighted by a red rectangle. The Raman peak at 509 cm−1 was chosen for quantification. (D) Time-dependent Raman signal intensities of the subcutaneous AuNS-Cy7 after local treatment with FPT or fructose. Raman signal intensities were normalized to their initial values before the optical clearing.

imaging approach, in which FPT or fructose was locally smeared on the murine dorsal skin where the AuNS-Cy7 was injected subcutaneously. Time-dependent Raman spectra demonstrated that while the distinguishable Raman spectrum was not clearly observed after 20 min of fructose treatment, a discernable Raman signal was visualized as soon as 5 min after the FPT treatment (Figure 5B−D). The maximum Raman signal of the subcutaneous AuNS-Cy7 after FPT treatment was observed at 15 min but not until 50 min after treatment with fructose (Figure 5D). Additionally, FPT showed much higher optical clearing efficiency in vivo than fructose. The signal-tobackground ratio of the subcutaneous nanoprobe after FPT mediation was 3.5 times that of fructose at 15 min post skin optical clearance (Figure 5D). The fructose and PEG400 in FPT decreased the RI of the dermal scatters (nsca) by dissociating intermolecular hydrogen bonds and increasing the RI of cytoplasmic background (nbkg) via a dehydration effect, and thiazone played an important role in enhancing dermal uptake of both fructose and PEG400 by increasing permeability of the epidermal layer. Because of its rapid and efficient optical clearing effect in vivo, FPT was chosen for assisting handheld Raman detector to visualize subcutaneous tumor xenograft. 3.5. AuNS-Cy7 Imaging of Subcutaneous Tumor via Intravenous Injection. In vivo visualization of subcutaneous tumor xenograft via the handheld Raman detector was conducted in mice bearing subcutaneous murine glioblastoma C6 xenograft at 24 h post intravenous injection of AuNS-Cy7 (Figure 6A). When the AuNS-Cy7 of ∼60 nm was injected via

the tail vein and then circulated around the tumor tissue, it could accumulate much more effectively in the tumor tissue due to better vascular permeability resulting from a larger vascular endothelial gap and weaker lymphatic return of the tumor tissue compared to those of normal tissue, which is the socalled Enhanced Permeability and Retention (EPR) effect. Meanwhile, the PEG coating could significantly prolong the circulation time of AuNS-Cy7 by decreasing the nonspecific uptake of AuNS-Cy7 via reticuloendothelial system.24 Timedependent Raman spectra of the subcutaneous tumor showed that although there was no signal detected before the FPT treatment via local smearing, the characteristic twin peaks of AuNS-Cy7 appeared as early as 10 min post FPT intervention (Figure 6B). The Raman signal intensity of the tumor increased consistently, and its maximum value (3.6-fold enhancement) was obtained at 25 min (Figure 6C). TEM images clearly demonstrated the dispersive distribution of AuNS-Cy7 in the tumor section (Figure 6D). The partial loss of the sharp tips on the nanoprobe can be explained by the ultrathin slide thickness of 150 nm. White-light photograph studies showed the obvious darkening of the tumor region after the cutaneous FPT treatment (Figure 6E). Considering the intratumoral delivery of the inky AuNS-Cy7 solution via the EPR effect, the FPTinduced dermal transparency led to the visualization of the subcutaneous tumor stained by inky nanoprobe with the naked eye. Figure 6F shows the Raman spectroscopic image of the 17774

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

ACS Applied Materials & Interfaces tumor slides at 24 h post the intravenous injection of AuNSCy7. Although the SERS signal of AuNS-Cy7 was distributed in whole tumor section, it was not detected in the tumorassociated skin tissue (Figure S11). The above studies indicated that the noninvasively detected Raman signal was from the tumor parenchyma but not the tumor attached skin. Figure S12 shows the white-light photos of the murine dorsal skin tissue before and post FPT treatment. Skin injuries, including incrustation, erythema, shrinking, and hemorrhage, did not occur in the 3 weeks following FPT treatment. Additionally, H&E staining studies showed that no necrosis, inflammatory infiltration, or other pathologic changes were found in the skin tissue or main organs including kidney and liver, which indicate the tolerable systemic toxicity of the OCA (FPT) applied via local smearing (Figures S12 and S13).

ABBREVIATIONS

■

REFERENCES

(1) Nie, S. M.; Emery, S. R. Probing Single Molecules and Single Nanoparticles by Surface-Enhanced Raman Scattering. Science 1997, 275, 1102−1106. (2) Sha, M. Y.; Xu, H.; Penn, S. G.; Cromer, R. SERS Nanoparticles: A New Optical Detection Modality for Cancer Diagnosis. Nanomedicine 2007, 2, 725−734. (3) Perumal, J.; Balasundaram, G.; Mahyuddin, A. P.; Choolani, M.; Olivo, M. SERS-Based Quantitative Detection of Ovarian Cancer Prognostic Factor Haptoglobin. Int. J. Nanomed. 2015, 10, 1831−1840. (4) Karabeber, H.; Huang, R.; Iacono, P.; Samii, J. M.; Pitter, K.; Holland, E. C.; Kircher, M. F. Guiding Brain Tumor Resection Using Surface-Enhanced Raman Scattering Nanoparticles and a Hand-Held Raman Scanner. ACS Nano 2014, 8, 9755−9766. (5) Kong, K.; Kendall, C.; Stone, N.; Notingher, I. Raman Spectroscopy for Medical Diagnostics - from in-Vitro Biofluid Assays to in-Vivo Cancer Detection. Adv. Drug Delivery Rev. 2015, 89, 121− 134. (6) Yang, S.; Dai, X.; Stogin, B. B.; Wong, T.-S. Ultrasensitive Surface-Enhanced Raman Scattering Detection in Common Fluids. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, 268−273. (7) Ando, J.; Fujita, K.; Smith, N. I.; Kawata, S. Dynamic SERS Imaging of Cellular Transport Pathways with Endocytosed Gold Nanoparticles. Nano Lett. 2011, 11, 5344−5348. (8) Zavaleta, C. L.; Smith, B. R.; Walton, I.; Doering, W.; Davis, G.; Shojaei, B.; Natan, M. J.; Gambhir, S. S. Multiplexed Imaging of Surface Enhanced Raman Scattering Nanotags in Living Mice Using Noninvasive Raman Spectroscopy. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 13511−13516. (9) McQueenie, R.; Stevenson, R.; Benson, R.; MacRitchie, N.; McInnes, I.; Maffia, P.; Faulds, K.; Graham, D.; Brewer, J.; Garside, P. Detection of Inflammation in Vivo by Surface-Enhanced Raman Scattering Provides Higher Sensitivity Than Conventional Fluorescence Imaging. Anal. Chem. 2012, 84, 5968−5975. (10) Duraipandian, S.; Zheng, W.; Ng, J.; Low, J. J.; Ilancheran, A.; Huang, Z. Simultaneous Fingerprint and High-Wavenumber Confocal Raman Spectroscopy Enhances Early Detection of Cervical Precancer in Vivo. Anal. Chem. 2012, 84, 5913−5919. (11) Lui, H.; Zhao, J.; McLean, D.; Zeng, H. Real-Time Raman Spectroscopy for in Vivo Skin Cancer Diagnosis. Cancer Res. 2012, 72, 2491−500. (12) Jermyn, M.; Mok, K.; Mercier, J.; Desroches, J.; Pichette, J.; Saint-Arnaud, K.; Bernstein, L.; Guiot, M.-C.; Petrecca, K.; Leblond, F. Intraoperative Brain Cancer Detection with Raman Spectroscopy in Humans. Sci. Transl. Med. 2015, 7, No. 274ra19. (13) Matsui, A.; Lomnes, S. J.; Frangioni, J. V. Optical Clearing of the Skin for near-Infrared Fluorescence Image-Guided Surgery. J. Biomed. Opt. 2009, 14, 024019. (14) Yang, X.; Zhang, Y.; Zhao, K.; Zhao, Y.; Liu, Y.; Gong, H.; Luo, Q.; Zhu, D. Skull Optical Clearing Solution for Enhancing Ultrasonic and Photoacoustic Imaging. IEEE. Trans. Med. Imaging 2016, 35, 1903−1906. (15) Guo, L.; Shi, R.; Zhang, C.; Zhu, D.; Ding, Z.; Li, P. Optical Coherence Tomography Angiography Offers Comprehensive Evalua-

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b04205. Detailed synthetic procedure of AuNS-Cy7; supplemented spectra including 1H, 13C NMR, and high resolution mass spectra of Cy7-SH; hydrodynamic diameter, Zeta potential, and UV−vis absorption of AuNS-Cy7; morphological comparison of AuNS and AuNS-Cy7 by TEM; SERS and FT-IR spectra of AuNSCy7 and its derivatives; stability evaluation of AuNSCy7; refractive index of OCAs; evaluation of potential side-effects after local treatment with FPT and the systemic injection of AuNS-Cy7 (PDF)

■

■

SERS, surface-enhanced Raman scattering; OCA(s), optical clearing agent(s); EPR, enhanced permeability and retention; FPT, a multicomponent OCA comprising fructose, PEG400 and thiazone; PEG, poly(ethylene glycol); i.p., intraperitoneal; TEM, transmission electron microscopy; H&E, hematoxylin and eosin; SERRS, surface-enhanced resonance Raman scattering; RMs, reporter molecules; AuNS, gold nanostar; r.t., room temperature; NTA, nanoparticle tracking analysis; UV−vis, ultraviolet visible; PG, 1,2-propanediol; BDO, 1,4butanediol; RI, refractive index; PBS, phosphate buffer saline; Gly, glycerol; Glu, glucose; PI, post injection

4. CONCLUSIONS To the best of our knowledge, this work is the first example to introduce the optical clearing technique into SERS imaging and demonstrates the feasibility of assisting the handheld Raman detector in imaging, in vivo. FPT, a multicomponent OCA showed the best optical clearing effect in vivo by suppressing light scattering in dermal tissue, thus enhancing the intratumoral Raman signal of an intravenously injected SERS nanoprobe. Considering the high optical clearing efficiency of FPT and ultrahigh sensitivity of AuNS-Cy7, this strategy could accelerate the clinical translation of the handheld Raman detector and extend its application into subcutaneous or even deeper lesions that are usually “concealed” by turbid tissue.

■

Research Article

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (Y.M.). *E-mail: [email protected] (C.L.). ORCID

Cong Li: 0000-0001-7731-8031 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National Basic Research Program of China (973 Program 2013CB932500, 2015CB755500), the National Natural Science Foundation of China (Nos. 81371624 and 81571741), and the Nanotechnology Program of Shanghai Science and Technology Committee (15140901300). 17775

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776

Research Article

ACS Applied Materials & Interfaces tion of Skin Optical Clearing in Vivo by Quantifying Optical Properties and Blood Flow Imaging Simultaneously. J. Biomed. Opt. 2016, 21, No. 081202. (16) Wang, J.; Shi, R.; Zhu, D. Switchable Skin Window Induced by Optical Clearing Method for Dermal Blood Flow Imaging. J. Biomed. Opt. 2013, 18, No. 061209. (17) Khan, M. H.; Choi, B.; Chess, S.; Kelly, M. K.; McCullough, J.; Nelson, J. S. Optical Clearing of in Vivo Human Skin: Implications for Light-Based Diagnostic Imaging and Therapeutics. Lasers Surg. Med. 2004, 34, 83−85. (18) Gallwas, J.; Stanchi, A.; Ditsch, N.; Schwarz, T.; Dannecker, C.; Mueller, S.; Stepp, H.; Mortensen, U. Effect of Optical Clearing Agents on Optical Coherence Tomography Images of Cervical Epithelium. Lasers Surg. Med. 2015, 30, 517−525. (19) Wang, J.; Ma, N.; Shi, R.; Zhang, Y.; Yu, T.; Zhu, D. SugarInduced Skin Optical Clearing: From Molecular Dynamics Simulation to Experimental Demonstration. IEEE J. Sel. Top. Quantum Electron. 2014, 20, 256−262. (20) Dinish, U.; Song, Z.; Ho, C. J. H.; Balasundaram, G.; Attia, A. B. E.; Lu, X.; Tang, B. Z.; Liu, B.; Olivo, M. Single Molecule with Dual Function on Nanogold: Biofunctionalized Construct for in Vivo Photoacoustic Imaging and SERS Biosensing. Adv. Funct. Mater. 2015, 25, 2316−2325. (21) Yuan, H.; Khoury, C. G.; Hwang, H.; Wilson, C. M.; Grant, G. A.; Vo-Dinh, T. Gold Nanostars: Surfactant-Free Synthesis, 3d Modelling, and Two-Photon Photoluminescence Imaging. Nanotechnology 2012, 23, No. 075102. (22) McNay, G.; Eustace, D.; Smith, W. E.; Faulds, K.; Graham, D. Surface-Enhanced Raman Scattering (SERS) and Surface-Enhanced Resonance Raman Scattering (SERRS): A Review of Applications. Appl. Spectrosc. 2011, 65, 825−837. (23) Hrelescu, C.; Sau, T. K.; Rogach, A. L.; Jäckel, F.; Laurent, G.; Douillard, L.; Charra, F. Selective Excitation of Individual Plasmonic Hotspots at the Tips of Single Gold Nanostars. Nano Lett. 2011, 11, 402−407. (24) Papahadjopoulos, D.; Allen, T. M.; Gabizon, A.; Mayhew, E.; Matthay, K.; Huang, S. K.; Lee, K. D.; Woodle, M. C.; Lasic, D. D.; Redemann, C. Sterically Stabilized Liposomes: Improvements in Pharmacokinetics and Antitumor Therapeutic Efficacy. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 11460−11464. (25) Hao, F.; Nehl, C. L.; Hafner, J. H.; Nordlander, P. Plasmon Resonances of a Gold Nanostar. Nano Lett. 2007, 7, 729−732. (26) Kircher, M. F.; De La Zerda, A.; Jokerst, J. V.; Zavaleta, C. L.; Kempen, P. J.; Mittra, E.; Pitter, K.; Huang, R.; Campos, C.; Habte, F.; et al. A Brain Tumor Molecular Imaging Strategy Using a New TripleModality Mri-Photoacoustic-Raman Nanoparticle. Nat. Med. 2012, 18, 829−834. (27) Flock, S. T.; Jacques, S. L.; Wilson, B. C.; Star, W. M.; van Gemert, M. J. Optical Properties of Intralipid: A Phantom Medium for Light Propagation Studies. Lasers Surg. Med. 1992, 12, 510−519. (28) Bolin, F. P.; Preuss, L. E.; Taylor, R. C.; Ference, R. J. Refractive Index of Some Mammalian Tissues Using a Fiber Optic Cladding Method. Appl. Opt. 1989, 28, 2297−2303. (29) Saidi, I. S.; Jacques, S. L.; Tittel, F. K. Mie and Rayleigh Modeling of Visible-Light Scattering in Neonatal Skin. Appl. Opt. 1995, 34, 7410−7418. (30) Schmitt, J. M.; Kumar, G. Optical Scattering Properties of Soft Tissue: A Discrete Particle Model. Appl. Opt. 1998, 37, 2788−2797. (31) Liang, X.; Liu, A.; Lim, C.; Ayi, T.; Yap, P. Determining Refractive Index of Single Living Cell Using an Integrated Microchip. Sens. Actuators, A 2007, 133, 349−354. (32) Naik, A.; Kalia, Y. N.; Guy, R. H. Transdermal Drug Delivery: Overcoming the Skin’s Barrier Function. Pharm. Sci. Technol. Today 2000, 3, 318−326. (33) Ding, Y.; Wang, J.; Fan, Z.; Wei, D.; Shi, R.; Luo, Q.; Zhu, D.; Wei, X. Signal and Depth Enhancement for in Vivo Flow Cytometer Measurement of Ear Skin by Optical Clearing Agents. Biomed. Opt. Express 2013, 4, 2518−2526.

17776

DOI: 10.1021/acsami.7b04205 ACS Appl. Mater. Interfaces 2017, 9, 17769−17776