Open Access
Add to BibSonomy
Abstract
Intrinsic radiosensitivity is a biological parameter known to influence the response to radiation therapy in cancer treatment. In this study, Raman spectroscopy and surface enhanced Raman spectroscopy (SERS) were successfully used in conjunction with principal component analysis (PCA) to discriminate between radioresistant (LY-R) and radiosensitive (LY-S) murine lymphoma sublines (L5178Y). PCA results for normal Raman analysis showed a differentiation between the radioresistant and radiosensitive cell lines based on their specific spectral fingerprint. In the case of SERS with gold nanoparticles (AuNPs), greater spectral enhancements were observed in the radioresistant subline in comparison to its radiosensitive counterpart, suggesting that each subline displays different interaction with AuNPs. Our results indicate that spectroscopic and chemometric techniques could be used as complementary tools for the prediction of intrinsic radiosensitivity of lymphoma samples.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Around 40-50% of diagnosed cancer patients are treated with radiation therapy [1]. Radiation response varies widely amongst individual cases, and the course of action is regularly determined based on the histopathological and anatomical characteristics of the tumor. Nonetheless, several biological mechanisms like repopulation capacity, tumor microenvironment [2] and intrinsic radiation sensitivity [3] have been shown to affect the therapeutic response of a tumor [4]. According to Begg et al., studying the differences between normal and malignant tissues between individual patients is key to improve radiotherapy outcomes [3].
Predictive assays could potentially be used for the rapid evaluation of intrinsic radiosensitivity of some tumors [5,6]. Intrinsic radiosensitivity can be hereditary, and also depend on the tissue origin and acquired mutations [7]. Intrinsic radiation sensitivity has been predominantly determined by a clonogenic survival assay (SF2) [4,8]. One of the main drawbacks of this technique is the time consuming and difficult nature of the method [9]. Some efforts have focused on genetic and molecular biology techniques to detect differences in gene expression [7] but are still underway. It has been noted that the failure to predict radiation outcomes can be related to the minimal manifestation of quantitative differences between tumors and normal tissue in addition to high tissue heterogeneity [5].
Raman spectroscopy can be used as a tool for the fast analysis and classification of tissues and cells. A high-throughput and label-free vibrational spectroscopy technique like Raman spectroscopy provides information about the chemical composition of cells, tissues and fluids, resulting in a specific “fingerprint” constituted by the spectra of different molecular components (i.e. proteins, lipids and nucleic acids) [10]. Biochemical changes caused by a disease can translate to changes in the Raman spectra [11]. Spontaneous or normal Raman scattering in analysis of biological samples is limited by low analyte concentration and instrumental parameters. Additionally, the use of high laser powers can induce sample damage. Surface Enhanced Raman Spectroscopy (SERS) is a technique that allows the amplification of Raman signals by means of localized surface plasmon excitations in metallic nanostructures [12].
Raman spectroscopy and SERS are routinely combined with multivariate analysis algorithms like Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Support Vector Machine (SVM) and random forest [10,13–16], where the ability to differentiate between cell samples based on small changes has been demonstrated. Considerable effort has been made in the analysis of cancerous cells, cell lines and tumors [17–20]. Analysis of cell chemoresistance [21], normal and neoplastic hematopoietic cells [22], and changes caused by exposure to ionizing radiation have been reported [23–25]. In one of the few studies analyzing radiosensitivity by Raman spectroscopy, Yasser et al. studied the acquired radioresistance of oral cancer sublines finding changes in the bands associated to DNA and proteins [26].
In this work, cells from two established lymphoma sublines were selected as a model for the analysis of both radioresistance (LY-R) and radiosensitivity (LY-S), and analyzed by Raman spectroscopy. The L5178Y (LY) Murine lymphoma cell line is derived from an induced thymic tumor; the parent line, LY-R, is resistant to ionizing radiation and highly tumorigenic in DBA/2 mice, whilst the LY-S subline is sensitive to ionizing radiation [27,28]. In addition to these characteristics, the LY-S and LY-R cells present different responses to external physicochemical agents [27].
To the best of our knowledge, this is the first work characterizing the spectral signature of the LY-S (ATCC CRL-1723) and LY-R (ATCC CRL-1722) lymphoma sublines by vibrational spectroscopy and exploring the discrimination of this lymphoma model by Raman spectroscopy and SERS. Comprehensive normal Raman characterization of both sublines was performed to find the ideal conditions for probing these cells, and PCA of these spectra was successfully used for the initial differentiation between sublines. Moreover, an advanced technique (SERS using gold nanoparticles) was successfully employed to achieve further differentiation.
In line with other recent work [26,29], this study represents an initial approach at identifying spectral differences of radiosensitive and radioresistant cells, in order to contribute to the development of fast radiosensitivity predictive assays.
2. Materials and methods
2.1 Chemicals and reagents
Hydrogen tetrachloroaurate trihydrate (HAuCl4 3H2O, 99.9%, Sigma Aldrich), cetyltrimethylammonium bromide (CTAB, BioUltra for molecular biology, ≥ 99.0%, Sigma Aldrich), sodium borohydride (NaBH4, 98%, Aldrich), ascorbic acid (reagent grade Sigma Aldrich), sodium bromide (NaBr, Anhydrous ≥ 99% Sigma Aldrich), and Trypan Blue Solution (0.4%, Gibco) were used without further purification. Distilled water was used for washing material and preparation of aqueous solutions.
2.2 Nanoparticle synthesis and characterization
AuNPs synthesis was adapted from Wu et al [30]. Two solutions were freshly prepared; an aqueous solution containing 0.001 M HAuCl4 and 0.025 M sodium citrate, and 10 mL of 0.01 M ice-cold NaBH4. Afterwards, 0.60 mL of the NaBH4 solution were added to the HAuCl4 under stirring. The solution turned brown immediately, indicating the formation of gold nanoparticles, and was aged for 2 h at room temperature.
Separately, four vials were labeled A to D. A growth solution was prepared in each vial with a total volume of 10 mL of HAuCl4 (0.001 M), CTAB (0.1M) and NaBr (0.01 M). To begin the growth reaction, 1 mL of the aged solution was added to vial A, then a 1 mL aliquot was taken from vial A to B, this procedure was repeated from B to C and from C to D. The vial D turned blue, and was left at 30°C for 24 hours with stirring. The resulting solution had a brown color and strong blue backlight.
Microscopic characterization was carried out by a Nova Nano SEM 200 (FEI, Germany) field emission scanning electron microscope (FE-SEM) with a STEM1 XT detector coupled with an INCA X-Sight (Oxford Instruments, UK) energy dispersive X-ray (EDX) microanalysis system. Samples were prepared by diluting 100 µL of AuNPs in 1 ml of deionized water. The dilution was sonicated for 5 min. Afterwards, a drop of the solution was deposited on a formvar/carbon 300 mesh copper grid (Ted Pella Inc., USA) and dried at room temperature.
The UV-Vis spectrum of the as-prepared colloidal suspension was recorder in the range of 300-700 nm using an Agilent 8453 spectrometer (Agilent Technologies, Germany.
Dynamic light scattering was performed using a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK). The gold solution in aqueous medium was analyzed with a gold absorption value of 3.32, a refractive index value of 1.33 for the dispersant (water) and a viscosity of 0.8872 cP.
2.3 Cell culture
L5178Y-R (ATCC CRL-1722) and L5178Y-S (ATCC CRL-1723) cell lines were cultured in RPMI 1640 medium containing L-Glutamine (Sigma-Aldrich, Missouri, USA) supplemented with 10% heat-inactivated Fetal Bovine Serum (Gibco, Massachusetts, USA), 10 units/ml penicillin and 100 µg/ml streptomycin (Sigma-Aldrich, Missouri, USA). Cells were incubated at 37°C in 5% CO2 until a density about 1 × 106 cells/ml was achieved.
2.4 Normal Raman characterization of the L5178Y sublines
As there are no previous studies addressing the spectral signature of the L5178Y cell line by normal Raman spectroscopy, cells from both sublines were probed with 457, 514, 633 and 830 nm excitation wavelengths. 30 individual cells were probed for each subline and wavelength and cells were obtained from different subcultures (cell passages) to consider possible reproducibility variations between days.
For sample preparation, 2 mL of either LY-R or LY-S cells in RPMI media were centrifuged at 1000 rpm for 3 minutes, the supernatant was removed, and cells were resuspended in PBS buffer. Then, 10 μL of cell suspension were deposited on a clean quartz substrate (Ted Pella Inc., USA) and analyzed using an InVia Raman microscope (Renishaw, UK) with a 50x microscope objective (NA 0.75). Each sample was analyzed in mapping mode, measuring at least 40 points covering different locations of the cell (nucleus and cytoplasm). Point measurements were centered at 1500 cm−1, with 1 s acquisition time and 3 accumulations per mapping point. Laser excitation was ∼9 mW for 457 nm, ∼17 mW for 514 nm, ∼17 mW for 633 nm and ∼10 mW for 830 nm.
Laser power was determined by testing different attenuation values until the best signal to noise ratio was obtained whilst avoiding negatively affecting the sample during spectral collection, as total acquisition time was ∼3 min. These effects were mainly observed as evaporation of the buffer solution that subsequently caused the formation of salt crystals and cell collapse. In the case of shorter wavelengths (i.e. 457 and 514 nm), high laser powers combined with long exposure times induced changes in the spectra, attributed to photodamage of the sample [16].
A separate study was performed with the purpose of analyzing the cell spectra by multivariate analysis, where cell pellets were analyzed instead of single cells in order to obtain more homogeneous spectra. Raman analysis was carried out as follows; 2 mL of either LY-R or LY-S cells in RPMI media were centrifuged at 1000 rpm for 3 minutes, the supernatant was removed and the concentrated cell pellet was deposited onto a quartz substrate. For each subline, Raman spectra of 30 cell pellets were obtained using the 514 nm excitation wavelength attenuated to ∼17 mW laser power. Spectra were acquired in mapping mode, with at least 40 mapping points per pellet.
2.5 SERS analysis
SERS of the L5178Y sublines was employed to enhance the Raman signal and provide further characterization and differentiation. Briefly, 100 μL of colloidal AuNPs were added to 500 μL of cells in PBS buffer. The cells were gently mixed and incubated for 1 hour. After incubation, cells were centrifuged for 3 min at 1000 rpm and the supernatant was removed. The pellet was resuspended in PBS buffer to remove unbound AuNPs and centrifuged again; afterwards the supernatant was carefully removed, leaving a concentrated cell pellet. 10 μL of cell pellet were placed on a quartz substrate, and SERS spectra of 30 pellets per subline were collected using the 830 nm excitation wavelength attenuated to ∼1 mW laser power. Since SERS enhancement can be highly variable [31], SERS spectra were acquired in mapping mode, where at least 40 points per pellet were measured and averaged. Cell samples from different days were probed to ensure data reproducibility. Cell viability after incubation with AuNPs was determined by the Trypan blue assay.
The following control samples were measured under the same instrumental conditions used for SERS; cells without nanoparticles (cell suspension in PBS buffer) and a nanoparticle blank (100 μL of AuNPs diluted in 500μL of PBS buffer).
2.6 Principal component analysis
All spectra were baseline corrected by 5th order polynomial fitting using WiRE 4 (Renishaw, UK) and averaged with OriginPro 8.5 (OriginLab, USA). Principal component analysis was carried out with the Unscrambler X version 10.1 (CAMO Software A.S., Norway). For normal Raman analysis, the mean spectra of 30 cell pellets per subline, obtained with 514 nm excitation were input into the software. Spectra were pre-processed by Savitzky-Golay Smoothing and normalized by maximum normalization prior to PCA. For SERS samples, the mapping points from each cell pellet were averaged and pre-processed by Savitzky-Golay Smoothing. Confidence ellipses (95%) were calculated with XLSTAT (Addinsoft, France).
2.7 Electron microscopy of cells incubated with AuNPs
Scanning electron microscopy was used to investigate the presence of metallic nanoparticles in the cell surface after incubation with colloidal AuNPs. Briefly, cells incubated with nanoparticles were centrifuged (1000 rpm, 3 min) and the supernatant was removed. Cells were fixed with a 3% glutaraldehyde 3% formaldehyde solution for 20 min at room temperature [32]. Afterwards, the solution was carefully removed to avoid resuspending the cell pellet. Cells were dehydrated following a graded series of ethanol solutions in deionized water. Finally, cells in 100% ethanol were critical point dried in a Supercritical Automegasamdri critical point drier (Tousimis, USA). Cell morphology and the presence of nanoparticles in the cell surface were studied with an analytical High Resolution SEM (JEOL 7600F) operating at a low accelerating voltage (1 kV) in Gentle Beam (GB) mode coupled with a low-angle backscattered electron (LABE) detector.
3. Results and discussions
3.1 Normal Raman characterization
To the best of our knowledge, no prior spectroscopic characterization of the LY-R or LY-S pair has been conducted. Therefore, a range of excitation wavelengths (457, 514, 633 and 830 nm) were used to find the experimental conditions that generate the best signal to noise ratio [33]. Figure 1 shows mean Raman spectra of cells from the LY-R and LY-S sublines obtained with different excitation wavelengths. Standard deviation is presented as the shaded area.
Fig. 1. Mean Normal Raman spectra of LY-R and LY-S cells obtained with 514 nm, 633 nm, 457 nm and 830 nm excitation. Shaded area represents the standard deviation. Cyt c: cytochrome c, Phe: phenylalanine, Trp: tryptophan, NA: nucleic acids.
Excitation with 830 nm is generally recommended for biological samples as it can be less damaging to tissues [34] and help reduce cell autofluorescence [35]. Nevertheless, as observed in Fig. 1, less representative bands were obtained with this wavelength. It is known that the Raman scattering efficiency is inversely proportional to λ4, where λ is the wavelength, assuming non-resonant excitation [35]. Therefore, shorter wavelengths (i.e. 457 nm, 514 nm) generate better scattering efficiencies than wavelengths approaching the infrared region (i.e. 830 nm). As a consequence, higher laser output is required to compensate for the low efficiency, but high laser power can also affect the sample; this was observed during preliminary measurements working with the 830 nm wavelength at higher laser powers. The ∼3 min acquisition time required to complete a point mapping combined with instrumental limitations concerning laser attenuation meant exposing the sample to high irradiance. Under these conditions, buffer evaporation, formation of salt crystals and cell collapse were observed.
In addition, although autofluorescence arising from the sample is reduced with longer wavelengths, fluorescence from the substrate and the objective may cause interference when working with near infrared wavelengths [33]. The 457 nm wavelength is expected to exhibit better scattering efficiencies, but shorter wavelengths can cause photochemical damage to biological tissues and DNA affecting the spectral fingerprint [36,37]. In addition to sample autofluorescence, lasers with high scattering efficiencies can be used advantageously to induce resonance of certain molecules such as heme moieties and cytochrome c, that are excited with 488, 514 and 568 nm [38] . The Raman spectra acquired with 514 nm presented the bands with the highest intensity. These bands are not as defined with 633 nm, probably due to the lack of a resonance effect of molecules with aromatic rings in their structure.
The balance between high excitation efficiencies and minimizing sample damage has been demonstrated to lie between 532 and 785 nm [39]. Therefore, due to good scattering efficiency, and despite slight cell autofluorescence, 514 and 633 nm are recommended for the analysis of LY-R and LY-S cells.
The most prominent feature observed with all excitation wavelengths emerges from the phenylalanine symmetric ring breathing at ∼1002 cm−1. Notably, this band was consistently more intense for LY-S cells with all excitation lasers. Another phenylalanine band is observed at ∼1030 cm−1 and is assigned to C-H stretching vibrations. The band ∼1097 cm−1 corresponds symmetric stretching of the phosphate DNA backbone [40]. Other bands that exhibited different intensities in LY-R and LY-S cells are those assigned to cytochrome c (∼1127 cm−1), nucleic acids (∼1150 and 1360 cm−1), and amide III (∼ 1312 cm−1). Tentative assignment of main Raman bands can be found in Table 1. Bands at 751, 1131 and 1312 cm−1, assigned to different vibrational modes of cytochrome c have been observed when excited with a 532 nm laser [37,41,42], under resonance scattering conditions. This is in agreement with our observations for LY-R and LY-S cells, since the band at ∼ 751 cm−1 only arises with 457 and 514 nm excitation. Porphyrins and pyrimidine rings also exhibit resonance at 514 nm.
Table 1. Raman band assignment for the average spectra of LY-R and LY-S cells obtained with different excitation wavelengths.
Hamada et al. analyzed HeLa cells with 488, 514, 532 and 633 nm excitation, and found that excitation with 633 nm generated a much lower scattering signal, and the bands at 1451 and 1660 cm−1 were the most prominent [41]. These prominent and broad bands, that correspond to CH2CH3 deformations and the amide I band, respectively, have been also reported for lymphocytes by Bankapur et al. [43]. In the present work, the CH2CH3 deformation band was observed between 1441 and 1448 cm−1. The band at 1448 is also acknowledged as a protein marker band, and is representative of the protein concentration in the cell [44]. This band has a slight higher intensity in the LY-S subline. The amide III bands are also associated to changes in the overall protein content [45], and this band (∼1312 cm−1) is also more intense in LY-S cells. The band at 1543 cm−1 is attributed to amide II vibrations (60% NH bending, 40% CN str). The amide I band is assigned to a combination of C = O stretching and N-H in plane bending [46].
Information concerning the spectral signature of lymphoblasts is scarce. The normal Raman spectra of LY-R and LY-S cells is dominated by bands corresponding to proteins (Amide I, II, III), lipids and nucleic acids (Table 1). This is expected since LY cells have a lymphoblastic morphology, characterized by a high nuclear/cytoplasmic ratio [47]. This is similar to observation made for lymphocytes, where the nucleus can constitute up to 80% of the total cell volume [22]. Furthermore, these cells are able to express multiple surface proteins [27,48,49], and this can contribute to the overall Raman signature.
Moradi et al. [29] studied the differentiation between carcinoma cell lines resistant and sensitive to chemotherapy (A2780s and cisplatin-resistant variant A2780cp). The authors found that based on spectral information, proteins and glutathione could be found at higher concentrations in the resistant cell line.
Other authors have observed that lung cancer cells have higher signals associated to proteins compared to normal cells [50]. For example, oral cancer cells present lower spectral contributions from proteins, lipids and glycogen in comparison to normal cells [51]. Studies focused on B-leukemia cells have found spectral differences associated with nucleic acid and proteins when compared to normal B-lymphocytes [52].
Despite the observations mentioned above, there are no general criteria regarding spectral differences between normal and cancerous cells, or between different classes of cancerous cells, since these variations will depend on the type of malignant cell.
3.2 Principal component analysis of normal Raman spectra
Biochemical variations in biological samples can manifest in minor changes of the overall Raman signature [14,60,61]. Thus, PCA was employed to reduce data dimensionality and facilitate the comparison between similar complex spectra [15]. By measuring cell pellets, several cells were probed at once at each mapping point, which generates more consistent spectra [26] and improves reproducibility between samples [21]. Based on the observations from section 3.1, the 514 nm laser was used for probing cell pellets. PCA was applied to the full Raman spectral range (800-1700cm−1), as well as the range limiting the peak at 1001 cm−1 (997-1003 cm−1). In order to contribute to the development of accessible methods using commercial software, data processing was minimal and consisted on normalization and Savitzky-Golay smoothing.
The PCA scores plot (Fig. 2(A)), shows two clearly defined groups outlined by the 95% confidence ellipses, indicating a differentiation between the radiosensitive and radioresistant sublines. There is a degree of intragroup variability, though most of the samples fall within the confidence ellipses. The groups are not fully separated, but some overlapping is expected in biological samples due to their complex chemical composition. This behavior has been reported in other studies examining normal Raman spectra of cells [40,62,63]. Differentiation between LY-R and LY-S samples occurs predominantly along the PC-1 axis (contribution to the variance of 53%), with most LY-S samples situated on the negative side of the axis and LY-R samples on the positive side. A degree of intragroup variability along the PC-2 axis accounts to its 23% contribution to the variance. Upon inspection of the loadings plots (Fig. 2(B)), it is observed that the peak at ∼1090 cm−1 associated to O-P-O stretching in the DNA phosphate backbone [22,40,55] is prominent and a positive feature in PC-1. Also, the peaks at ∼1001 and 1032 cm−1 assigned to phenylalanine, are present in the loadings plots for PC-1 and PC-2. This is in agreement with the mean spectra presented in Fig. 1, where these peaks exhibit different intensities in LY-S and LY-R cells. It has been reported that LY-R and LY-S cells present similar average DNA and protein content [28]. Nevertheless, PCA based on the spectral signature suggest there are subtle differences in biochemical composition, since most of the peaks identified in the loadings plots are associated to nucleic acids and proteins.
Fig. 2. PCA score plots (Confidence interval of 95%) of normal Raman spectra from radioresistant (LY-R) and radiosensitive (LY-S) cell pellets. Pellets were analyzed with 514 nm excitation and 17 mW laser power. (A) Score plot obtained by analyzing the full (800-1700cm−1) spectral range; (B) Loadings plot for PC-1 and PC-2. (C) Score plot obtained when the range of peak at ∼1001 cm−1 (997-1003 cm−1) was input into the software; (D) Loadings plot for PC-1 and PC-2.
Group separation was improved, particularly along the PC-1 axis, by analyzing the spectral range adjacent to the peak at ∼1001 cm−1 (Fig. 2(C)). In this case, PC-1 contribution to the variance was 76% and PC-2 17%. Additional spectral ranges were analyzed by PCA, such as 570-1700cm−1, 950-1170 cm−1, 1560-1814cm−1, 1000-1814cm−1 and 1017-1150 cm−1, but separation between sublines was not improved (data not shown).
3.3 Nanoparticle synthesis and characterization
Figure 3 shows a STEM micrograph and size distribution histogram with the mean diameter of the AuNPs used for SERS analysis. Mean nanoparticle size was 48.6 nm with a standard deviation of 14.9. Mean size obtained by DLS was 70.3 nm, which was higher than the geometric size calculated from STEM images. This result is expected, as the hydrodynamic size of colloidal solutions is influenced by the substances absorbed onto the surface (CTAB) and the thickness of the electrical double layer. According to DLS measurements, 93% of the AuNPs are over 43.38 nm in size, 86% over 50 nm, 12.7% over 68 nm and 13.3% over 78 nm. Therefore, although there is polydispersion, only 7% of the AuNPs are below 50 nm and 13.3% above 78 nm. This is in agreement with the polydispersity index, which was 0.18.
Assuming a nearly spherical particle shape, our results fall within the optimal size for SERS enhancements via AuNPs. As it has been reported that nanoparticles of sizes up to 80 nm generate significant enhancements. While NPs with smaller sizes may not generate the highest enhancements, they also contribute to the overall SERS signal amplification [64–66].
3.4 SERS analysis using AuNPs
Preliminary studies were carried out with different laser excitation wavelengths to generate SERS enhancement, however significant enhancement was only achieved using 830 nm, even tough absorption maxima of the as-prepared colloidal solution was 537 nm. This is attributable to the aggregation of AuNPs under our experimental conditions, causing a red-shit of the localized surface plasmon absorption maxima to the near infrared [67]. Nanoparticles in a biological system do not exhibit the same characteristics as the original colloidal solution due to several reasons; first, the ionic strength of buffer solutions change the AuNPs dielectric environment and consequently affect their stability in suspension [68]. Also, membrane proteins can generate microdomains that facilitate nanoparticle aggregation upon contact with the cell [69,70]. High electrostatic attraction between the cationic capping agent (CTAB) and molecules present on the cell surface can also induce aggregation of the AuNPs due to loss of colloidal stability [71,72].
The average SERS spectra of 30 cell pellets per subline are shown in Fig. 4(A). These measurements were acquired during different days and experimental runs to ensure reproducibility of the results. There is an evident difference between the radioresistant and radiosensitive lines regarding SERS enhancement; LY-S pellets present lower enhancement but more spectral homogeneity, while SERS spectra of LY-R pellets exhibit significantly higher, yet variable enhancements. Mean SERS spectra and standard deviation of each subline are shown in Fig. 4(B). Controls were probed by normal Raman spectroscopy (without AuNPs) and acquired under the same instrumental conditions used for SERS (830 nm, 1 mW). The mean spectrum of AuNPs (nanoparticle blank) does not present significant signals when compared to SERS spectra.
Fig. 4. (A) SERS spectra of radioresistant (LY-R) and radiosensitive (LY-S) cell pellets obtained by incubation with AuNPs. (B) Mean SERS spectra by subline compared to their control (normal Raman spectra). The shaded area represents the standard deviation. All measurements were acquired with 830 nm excitation, 1 mW laser power and acquisition time of 1 s.
The PCA scores plot of SERS spectra (Fig. 5(A)) shows two distinct clusters corresponding to each subline, which are separated by the PC-1 axis (40% contribution to the variance); LY-S samples have negative PC-1 values, whereas the majority of LY-R samples are located along the positive values of PC-1. In the case of LY-R pellets, the high spectral variability observed in Fig. 4(A) translates into high intragroup dispersion along the PC-1 and PC-2 axis. In contrast, dispersion of LY-S samples occurs primarily due to PC-1, whilst variations in the PC-2 axis are minimal. As shown in Fig. 5(B), the loadings plot for PC-1 shows positive peaks at ∼ 1142, 1160, 1260, 1420, 1477, 1553 and 1600 cm−1, which are mostly associated with proteins, nucleic acids and lipids (Table 1). Positive peaks in the loadings plot for PC-2 include peaks at ∼ 1260 (Amide III, 1279 (aromatic C-C stretching), 1335 (nucleic acids, aminoacids and lipids) and 1601 cm−1 (aminoacids). Negative values for the peaks ∼ 1136 (phenylalanine), 1158 (C-N stretch, nucleic acids) and 1203 cm−1 (aromatic aminoacids) are also observed.
Fig. 5. (A) PCA score plots of SERS spectra from cell pellets incubated with AuNPs (confidence interval of 95%). (B) Loadings plots for PC-1 and PC-2.
Although polydispersity of the AuNPs could be related to the high variability of SERS signals, in theory the size distribution is maintained in the colloid. Specifically, the same colloidal suspension was used in order to avoid possible variations caused by nanoparticles from different synthesis batches.
Cell viability was assessed by Trypan blue exclusion to determine if the observed difference in SERS enhancements was associated to a loss of cell viability induced by AuNPs colocalization (Fig. 6). Cell viability decreased for both LY-R and LY-S cells after incubation with AuNPs when compared to their controls (P < 0.05). Nevertheless, there were no significant differences between the viability of neither controls (p = 0.93) nor the LY-R and LY-S cells treated with AuNPs (p = 0.51). Hence, the observed difference in SERS spectra is not arising from differences in cell viability.
Fig. 6. Cell viability assed by Trypan blue exclusion test. Mean ± standard deviation of three independent experiments, p < 0.05 vs. control. Cell viability of both sublines decreased after incubation with AuNPs. In contrast, no significant differences were observed (p > 0.05) between the LY-R and LY-S cells incubated with AuNPs.
The presence of AuNPs in the cell surface was confirmed by electron microscopy. Figure 7 shows SEM micrographs of LY-R and LY-S cells incubated with AuNPs. A Low-Angle Backscattered Electron (LABE) detector was used, which offers the advantages of reducing the load effect produced by the cell and improving contrast of nanoparticles [27]. Backscattered electron imaging is sensitive to atomic number, therefore gold appears brighter than the elements constituting the cell (carbon, hydrogen, oxygen, nitrogen, sulfur and phosphorous) [73,74].
Fig. 7. SEM micrograph of radioresistant (LY-R) and radiosensitive cells (LY-S). (A) LY-R control, (B) LY-R incubated with AuNPs, (c) LY-S control and (d) LY-S with AuNPs. Arrows indicate the presence of AuNPs in the cell surface.
It has been reported that SERS enhancement by unlabeled nanoparticles is affected by the occurrence of hotspots and aggregates, as well as the localization of the nanoparticles cell [12,31,75], since SERS is a proximity effect and the enhanced bands are usually correlated to the molecules located in the vicinity of the metallic nanoparticles [76]. As seen in Fig. 8, AuNPs were identified in the cell surface of both sublines even after cell fixation and several washing steps, indicating that a chemical interaction between nanoparticles and cells caused the observed SERS enhancements.
Fig. 8. SEM image of a cell incubated with AuNP showing the presence of nanoparticle aggregates (indicated by arrows) in the cell surface.
LY-R and LY-S cells exhibit differences in their plasmatic membranes, highlighting the electrophoretic pattern of glycoproteins associated to the cell structure and the tendency of the LY-S subline to lose its spherical shape. Additionally, total lipid to protein ratio is 0.213 ± 0.008 for LY-R cells and 0.234 ± 0.009 for LY-S cells, meaning that the LY-S subline has a slightly lower protein content and potentially less thiol binding sites [27, 28, 77]. This is worth noting because surface proteins can bind to metallic nanoparticles by means of the thiol functional group in cysteine residues [78]. Au-thiol binding is confirmed by the peaks at 1335 and 1601 cm−1 observed in the loadings plot (Fig. 5(B)), which have been assigned to the SERS spectra of cysteine molecules adsorbed onto AuNPs [59].
Furthermore, the cationic CTAB-capped AuNPs may bind to phospholipids in the cell membrane by strong electrostatic interaction [79]. This is harder to confirm via SERS results, since several peaks assigned to lipids and proteins overlap, especially near 1440 and 1650 cm−1 [23, 80]. However, enhancement around these regions, which is observed mainly in the SERS spectra of LY-R nanoparticles, indicates an interaction of AuNPs with both lipids and proteins.
There are only a few reports indicating the type of internalization mechanisms for CTAB capped AuNPs, but we can deduce that due to the cationic nature of the nanoparticles, they may be internalized by passive diffusion [79]. However, CTAB is also known to induce non-specific binding to the negatively charged cell surface mediated by electrostatic interactions [72]. In addition, nanoparticles can be incorporated into membranes and bind to proteins [81,82]. For the specific case of lymphoblasts, both non-specific binding and uptake of nanoparticles with sizes up to 100 nm has been observed [83]. Nevertheless, a further study focusing on the internalization mechanisms exhibited by these sublines could help elucidate why the observed SERS enhancements are so contrasting.
4. Conclusion
Normal Raman analysis provided important information about the chemical composition of the radioresistant and radiosensitive sublines, and two distinct but overlapping groups were observed when spectra were analyzed by PCA. SERS was successfully used to enhance Raman signals, and multivariate analysis with PCA showed a clear separation between sublines after incubation with AuNPs. Despite the observed intragroup variability of the LY-R samples, SERS measurements are promising, since they allow for the clear distinction of two sublines with subtle differences in biochemical composition.
Therefore, it is suggested that small chemical differences between radiosensitive and radioresistant lymphoma cells can be identified based on changes in the spectral features (normal Raman), and these changes are enhanced by the interaction between cells and AuNPs (SERS).
Moreover, data processing was minimal in this work, allowing the development of a simple and reproducible method that can be carried out using commercial software.
The spectroscopic differences observed herein can potentially be used in conjunction with established techniques to aid the selection of an adequate therapeutic treatment of cancer patients. Further studies are needed to evaluate possible differences in nanoparticle-cell interaction presented by the LY-R and LY-S sublines.
Funding
Consejo Nacional de Ciencia y Tecnología-CONACYT (grant PN2016-4320, scholarship #400488).
Acknowledgments
The authors thank Nayely Pineda Aguilar for technician support at CIMAV-Monterrey.
This research used resources of the Center for Functional Nanomaterials, which is a U.S. DOE Office of Science Facility, at Brookhaven National Laboratory under Contract No. DE-SC0012704.
Disclosures
The authors declare no competing financial interest.
References
1. A. Yaromina, M. Krause, and M. Baumann, “Individualization of cancer treatment from radiotherapy perspective,” Mol Oncol. 6(2), 211–221 (2012). [CrossRef]
2. N. von Moos and V. I. Slaveykova, “Oxidative stress induced by inorganic nanoparticles in bacteria and aquatic microalgae–state of the art and knowledge gaps,” Nanotoxicology 8(6), 605–630 (2014). [CrossRef]
3. A. C. Begg, F. A. Stewart, and C. Vens, “Strategies to improve radiotherapy with targeted drugs,” Nat. Rev. Cancer 11(4), 239–253 (2011). [CrossRef]
4. L. J. Forker, A. Choudhury, and A. E. Kiltie, “Biomarkers of Tumour Radiosensitivity and Predicting Bene fi t from Radiotherapy Statement of Search Strategies Used and Sources of Information,” Clin. Oncol. 27(10), 561–569 (2015). [CrossRef]
5. D. G. Hirst and T. Robson, “Molecular biology: the key to personalised treatment in radiation oncology?” Br. J. Radiol. 83(993), 723–728 (2010). [CrossRef]
6. K. I. Altman and J. T. Lett, eds., Relative Radiation Sensitivities of Human Organ Systems, Part III. Volume 15 of Advances in Radiation Biology Relative Radiation Sensitivities of Human Organ Systems (Elsevier, 2016).
7. S. J. Mcmahon, K. M. Prise, A. L. Mcnamara, J. Schuemann, and H. Paganetti, “A general mechanistic model enables predictions of the biological effectiveness of different qualities of radiation,” Sci. Rep. 7(1), 10790 (2017). [CrossRef]
8. C. M. L. West, S. E. Davidson, S. A. Roberts, and R. D. Hunter, “Intrinsic radiosensitivity and prediction of patient response to radiotherapy for carcinoma of the cervix,” Br. J. Cancer 68(4), 819–823 (1993). [CrossRef]
9. A. Brahme, ed., Comprehensive Biomedical Physics. Volume 1: Nuclear Medicine and Molecular Imaging (Elsevier, 2014).
10. M. Diem, A. Mazur, K. Lenau, J. Schubert, B. Bird, M. Miljković, C. Krafft, and J. Popp, “Molecular pathology via IR and Raman spectral imaging,” J. Biophotonics 6(11-12), 855–886 (2013). [CrossRef]
11. K. Kong, C. Kendall, N. Stone, and I. Notingher, “Raman spectroscopy for medical diagnostics - From in-vitro biofluid assays to in-vivo cancer detection,” Adv. Drug Deliv. Rev. 89, 121–134 (2015). [CrossRef]
12. R. La Rocca, G. C. Messina, M. Dipalo, V. Shalabaeva, and F. De Angelis, “Out-of-Plane Plasmonic Antennas for Raman Analysis in Living Cells,” Small 11(36), 4632–4637 (2015). [CrossRef]
13. F. M. Lyng, I. R. M. Ramos, O. Ibrahim, and H. J. Byrne, “Vibrational Microspectroscopy for Cancer Screening,” Appl. Sci. 5(1), 23–35 (2015). [CrossRef]
14. A. J. Popp, C. Krafft, M. Schmitt, I. Schie, D. Cialla-may, C. Matthaeus, and T. Bocklitz, “Label-free molecular imaging of biological cells and tissues by linear and non-linear Raman spectroscopic approaches,” Angew. Chem. Int. Ed. Engl. 56(16), 4392–4430 (2017). [CrossRef]
15. A. Maguire, I. Vega-Carrascal, J. Bryant, L. White, O. Howe, F. M. Lyng, and A. D. Meade, “Competitive evaluation of data mining algorithms for use in classification of leukocyte subtypes with Raman microspectroscopy,” Analyst 140(7), 2473–2481 (2015). [CrossRef]
16. H. J. Butler, L. Ashton, B. Bird, G. Cinque, K. Curtis, K. Esmonde-white, N. J. Fullwood, B. Gardner, P. L. Martin-, M. J. Walsh, M. R. Mcainsh, N. Stone, F. L. Martin, H. J. Butler, and P. L. Martin-hirsch, “Using Raman spectroscopy to characterize biological materials,” Nat. Protoc. 11(4), 664–687 (2016). [CrossRef]
17. N. P. Damayanti, Y. Fang, M. R. Parikh, A. P. Craig, J. Kirshner, and J. Irudayaraj, “Differentiation of cancer cells in two-dimensional and three-dimensional breast cancer models by Raman spectroscopy,” J. Biomed. Opt. 18(11), 117008 (2013). [CrossRef]
18. C. Krafft, K. Wilhelm, A. Eremin, S. Nestel, N. von Bubnoff, W. Schultze-Seemann, J. Popp, and I. Nazarenko, “A specific spectral signature of serum and plasma-derived extracellular vesicles for cancer screening,” Nanomedicine Nanotechnology, Biol. Med. 13(3), 835–841 (2017). [CrossRef]
19. T. Tolstik, C. Marquardt, C. Matthäus, N. Bergner, C. Bielecki, C. Krafft, A. Stallmach, and J. Popp, “Discrimination and classification of liver cancer cells and proliferation states by Raman spectroscopic imaging,” Analyst 139(22), 6036–6043 (2014). [CrossRef]
20. C. Krafft and V. Sergo, “Biomedical applications of Raman and infrared spectroscopy to diagnose tissues,” J. Spectrosc. 20(5-6), 195–218 (2006). [CrossRef]
21. D. Franco, S. Trusso, E. Fazio, A. Allegra, C. Musolino, A. Speciale, F. Cimino, A. Saija, F. Neri, M. S. Nicolò, and S. P. P. Guglielmino, “Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy Raman spectroscopy differentiates between sensitive and resistant multiple myeloma cell lines,” Spectrochim. Acta - Part A Mol. Biomol. Spectrosc. 187, 15–22 (2017). [CrossRef]
22. J. W. Chan, D. S. Taylor, T. Zwerdling, S. M. Lane, K. Ihara, and T. Huser, “Micro-Raman Spectroscopy Detects Individual Neoplastic and Normal Hematopoietic Cells,” Biophys. J. 90(2), 648–656 (2006). [CrossRef]
23. S. J. Harder, Q. Matthews, M. Isabelle, A. G. Brolo, J. J. Lum, and A. Jirasek, “A Raman spectroscopic study of cell response to clinical doses of ionizing radiation,” Appl. Spectrosc. 69(2), 193–204 (2015). [CrossRef]
24. S. J. Harder, M. Isabelle, L. Devorkin, J. Smazynski, W. Beckham, A. G. Brolo, J. J. Lum, and A. Jirasek, “Raman spectroscopy identifies radiation response in human non- small cell lung cancer xenografts,” Sci. Rep. 6(1), 21006 (2016). [CrossRef]
25. Q. Matthews, A. Brolo, J. Lum, X. Duan, and A. Jirasek, “Raman spectroscopy of single human tumour cells exposed to ionizing radiation in vitro,” Phys. Med. Biol. 56(1), 19–38 (2011). [CrossRef]
26. M. Yasser, R. Shaikh, M. K. Chilakapati, and T. Teni, “Raman Spectroscopic Study of Radioresistant Oral Cancer Sublines Established by Fractionated Ionizing Radiation,” PLoS One 9(5), e97777 (2014). [CrossRef]
27. J. Z. Beer, E. Budzicka, E. Niepokojczycka, O. Rosiek, I. Szumiel, and M. Walicka, “Loss of Tumorigenicity during In Vitro Growth of L5178Y Murine Lymphoma Cells,” Cancer Res. 43(10), 4736–4742 (1983).
28. I. Szumiel, “L5178Y sublines: a look back from 40 years. Part 1: General characteristics,” Int. J. Radiat. Biol. 81(5), 353–365 (2005). [CrossRef]
29. H. Moradi, A. Ahmad, D. Shepherdson, N. H. Vuong, G. Niedbala, L. Eapen, B. Vanderhyden, B. Nyiri, and S. Murugkar, “Raman micro-spectroscopy applied to treatment resistant and sensitive human ovarian cancer cells,” J. Biophotonics 10(10), 1327–1334 (2017). [CrossRef]
30. H. Wu, C. Kuo, and M. H. Huang, “Seed-Mediated Synthesis of Gold Nanocrystals with Systematic Shape Evolution from Cubic to Trisoctahedral and Rhombic Dodecahedral Structures,” Langmuir 26(14), 12307–12313 (2010). [CrossRef]
31. Q. Zhang, X. Lu, P. Tang, D. Zhang, J. Tian, and L. Zhong, “Gold Nanoparticle (AuNP)-Based Surface-Enhanced Raman Scattering (SERS) Probe of Leukemic Lymphocytes,” Plasmonics 11(5), 1361–1368 (2016). [CrossRef]
32. A. M. Schrand, J. J. Schlager, L. Dai, and S. M. Hussain, “Preparation of cells for assessing ultrastructural localization of nanoparticles with transmission electron microscopy,” Nat. Protoc. 5(4), 744–757 (2010). [CrossRef]
33. L. T. Kerr, J. Byrne, and B. M. Hennelly, “Optimal choice of sample substrate and laser wavelength for Raman spectroscopic analysis of biological specimen,” Anal. Methods 7(12), 5041–5052 (2015). [CrossRef]
34. K. Tsia, Understanding Biophotonics: Fundamentals, Advances and Applications (CRS Press, 2016).
35. F. LaPlant, “Lasers, Spectrographs, and Detectors,” in Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields, P. Matousek and M. Morris, eds. (Springer, 2010), pp. 1–13.
36. I. Notingher, S. Verrier, H. Romanska, A. E. Bishop, J. M. Polak, and L. L. Hench, “In situ characterisation of living cells by Raman spectroscopy,” Spectroscopy 16(2), 43–51 (2002). [CrossRef]
37. B. Kann, H. L. Offerhaus, M. Windbergs, and C. Otto, “Raman microscopy for cellular investigations — From single cell imaging to drug carrier uptake visualization,” Adv. Drug Deliv. Rev. 89, 71–90 (2015). [CrossRef]
38. P. Lasch and J. Kneipp, Biomedical Vibrational Spectroscopy (Wiley, 2008).
39. A. Downes and A. Elfick, “Raman spectroscopy and related techniques in biomedicine,” Sensors 10(3), 1871–1889 (2010). [CrossRef]
40. I. W. Schie, L. Alber, A. L. Gryshuk, and J. W. Chan, “Investigating drug induced changes in single, living lymphocytes based on Raman micro-spectroscopy,” Analyst 139(11), 2726–2733 (2014). [CrossRef]
41. K. Hamada and K. Fujita, “Raman microscopy for dynamic molecular imaging of living cells,” J. Biomed. Opt. 13(4), 044027 (2008). [CrossRef]
42. A. Rygula, K. Majzner, K. M. Marzec, A. Kaczor, M. Pilarczyk, and M. Baranska, “Raman spectroscopy of proteins: a review,” J. Raman Spectrosc. 44(8), 1061–1076 (2013). [CrossRef]
43. A. Bankapur, E. Zachariah, S. Chidangil, M. Valiathan, and D. Mathur, “Raman tweezers spectroscopy of live, single red and white blood cells,” PLoS One 5(4), e10427 (2010). [CrossRef]
44. S. Corsetti, T. Rabl, D. McGloin, and G. Nabi, “Raman spectroscopy for accurately characterising biomolecular changes in androgen-independent prostate cancer cells,” J. Biophotonics 11(3), e201700166 (2018). [CrossRef]
45. Q. Wang, S. D. Grozdanic, M. M. Harper, N. Hamouche, H. Kecova, T. Lazic, and C. Yu, “Exploring Raman spectroscopy for the evaluation of glaucomatous retinal changes,” J. Biomed. Opt. 16(10), 107006 (2011). [CrossRef]
46. I. Rocha-Mendoza, D. R. Yankelevich, M. Wang, K. M. Reiser, C. W. Frank, and A. Knoesen, “Sum frequency vibrational spectroscopy: the molecular origins of the optical second-order nonlinearity of collagen,” Biophys. J. 93(12), 4433–4444 (2007). [CrossRef]
47. P. Greaves, Histopathology of Preclinical Toxicity Studies: Interpretation and Relevance in Drug Safety Evaluation, 3rd edit (Academic Press, 2007).
48. A. J. Hobro, Y. Kumagai, S. Akira, and N. I. Smith, “Raman spectroscopy as a tool for label-free lymphocyte cell line discrimination,” Analyst 141(12), 3756–3764 (2016). [CrossRef]
49. I. Szumiel, “L5178Y sublines: a look back from 40 years. Part 2: response to ionizing radiation,” Int. J. Radiat. Biol. 81(5), 353–365 (2005). [CrossRef]
50. M. Wang, X. Cao, W. Lu, L. Tao, H. Zhao, Y. Wang, M. Guo, J. Dong, and W. Qian, “Surface-enhanced Raman spectroscopic detection and differentiation of lung cancer cell lines (A549, H1229) and normal cell line (AT II) based on gold nanostar substrates,” RSC Adv. 4(109), 64225–64234 (2014). [CrossRef]
51. W. Y. Dai, S. Lee, and Y. C. Hsu, “Discrimination between oral cancer and healthy cells based on the adenine signature detected by using Raman spectroscopy,” J. Raman Spectrosc. 49(2), 336–342 (2018). [CrossRef]
52. S. Managò, C. Valente, P. Mirabelli, D. Circolo, F. Basile, D. Corda, and A. C. De Luca, “A reliable Raman-spectroscopy-based approach for diagnosis, classification and follow-up of B-cell acute lymphoblastic leukemia,” Sci. Rep. 6(1), 24821 (2016). [CrossRef]
53. J. Zhu, J. Zhou, J. Guo, W. Cai, B. Liu, Z. Wang, and Z. Sun, “Surface-enhanced Raman spectroscopy investigation on human breast cancer cells,” Chem. Cent. J. 7(1), 37 (2013). [CrossRef]
54. F. Wei, D. Zhang, N. J. Halas, and J. D. Hartgerink, “Aromatic amino acids providing characteristic motifs in the Raman and SERS spectroscopy of peptides,” J. Phys. Chem. B 112(30), 9158–9164 (2008). [CrossRef]
55. J. M. Benevides, S. A. Overman, and G. J. Thomas, “Raman, polarized Raman and ultraviolet resonance Raman spectroscopy of nucleic acids and their complexes,” J. Raman Spectrosc. 36(4), 279–299 (2005). [CrossRef]
56. Z. Movasaghi, S. Rehman, and I. U. Rehman, “Raman Spectroscopy of Biological Tissues,” Appl. Spectrosc. Rev. 42(5), 493–541 (2007). [CrossRef]
57. S. Casabella, P. Scully, N. Goddard, and P. Gardner, “Automated analysis of single cells using Laser Tweezers Raman Spectroscopy,” Analyst 141(2), 689–696 (2016). [CrossRef]
58. C. Matthäus, B. Bird, M. Miljković, T. Chernenko, and M. Romeo, “Infrared and Raman Microscopy in Cell Biology,” Methods Cell Biol. 89(08), 275–308 (2008). [CrossRef]
59. C. Jing and Y. Fang, “Experimental (SERS) and theoretical (DFT) studies on the adsorption behaviors of l-cysteine on gold/silver nanoparticles,” Chem. Phys. 332(1), 27–32 (2007). [CrossRef]
60. A. Maguire, I. Vegacarrascal, L. White, B. McClean, O. Howe, F. M. Lyng, and A. D. Meade, “Analyses of Ionizing Radiation Effects In Vitro in Peripheral Blood Lymphocytes with Raman Spectroscopy,” Radiat. Res. 183(4), 407–416 (2015). [CrossRef]
61. O. J. Old, L. M. Fullwood, R. Scott, G. R. Lloyd, L. M. Almond, N. A. Shepherd, N. Stone, H. Barr, and C. Kendall, “Vibrational spectroscopy for cancer diagnostics,” Anal. Methods 6(12), 3901–3917 (2014). [CrossRef]
62. A. N. S. Ohail, S. A. K. Han, R. A. U. Llah, S. H. A. Hmad, Q. Ureshi, M. U. B. Ilal, and A. S. K. Han, “Analysis of hepatitis C infection using Raman spectroscopy and proximity based classification in the transformed domain,” Biomed. Opt. Express 9(5), 2041–2055 (2018). [CrossRef]
63. W. Lee, A. Nanou, L. Rikkert, F. A. W. Coumans, C. Otto, L. W. M. M. Terstappen, and H. L. Offerhaus, “Label-Free Prostate Cancer Detection by Characterization of Extracellular Vesicles Using Raman Spectroscopy,” Anal. Chem. 90(19), 11290–11296 (2018). [CrossRef]
64. Z. Starowicz, R. Wojnarowska-Nowak, P. Ozga, and E. M. Sheregii, “The tuning of the plasmon resonance of the metal nanoparticles in terms of the SERS effect,” Colloid Polym. Sci. 296(6), 1029–1037 (2018). [CrossRef]
65. S. Hong and X. Li, “Optimal Size of Gold Nanoparticles for Surface-Enhanced Raman Spectroscopy under Different Conditions,” J. Nanomater. 2013, 1–9 (2013). [CrossRef]
66. S. C. Boca, C. Farcau, and S. Astilean, “The study of Raman enhancement efficiency as function of nanoparticle size and shape,” Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 267(2), 406–410 (2009). [CrossRef]
67. D. Kang, S. Y. Lee, and J. H. Kim, “Highly Sensitive Depolarized Light Scattering to Monitor Aggregation of Spherical Gold Nanoparticles,” J. Phys. Chem. C 123(23), 14625–14631 (2019). [CrossRef]
68. S. Du, K. Kendall, P. Toloueinia, Y. Mehrabadi, G. Gupta, and J. Newton, “Aggregation and adhesion of gold nanoparticles in phosphate buffered saline,” J. Nanoparticle Res. 14(3), 758 (2012). [CrossRef]
69. Q. Hu, L.-L. Tay, M. Noestheden, and J. P. Pezacki, “Mammalian cell surface imaging with nitrile-functionalized nanoprobes: biophysical characterization of aggregation and polarization anisotropy in SERS imaging,” J. Am. Chem. Soc. 129(1), 14–15 (2007). [CrossRef]
70. H. Tang, H. Ye, H. Zhang, and Y. Zheng, “Aggregation of nanoparticles regulated by mechanical properties of nanoparticle-membrane system,” Nanotechnology 29(40), 405102 (2018). [CrossRef]
71. G. Du, L. Wang, D. Zhang, X. Ni, X. Zhou, H. Xu, L. Xu, S. Wu, T. Zhang, and W. Wang, “Colorimetric aptasensor for progesterone detection based on surfactant-induced aggregation of gold nanoparticles,” Anal. Biochem. 514, 2–7 (2016). [CrossRef]
72. E. Yasun, C. Li, I. Barut, D. Janvier, L. Qiu, C. Cui, and W. Tan, “BSA modification to reduce CTAB induced nonspecificity and cytotoxicity of aptamer-conjugated gold nanorods,” Nanoscale 14(3), 758 (2012). [CrossRef]
73. G. Plascencia-Villa, C. R. Starr, L. S. Armstrong, A. Ponce, and M. José-Yacamán, “Imaging interactions of metal oxide nanoparticles with macrophage cells by ultra-high resolution scanning electron microscopy techniques,” Integr. Biol. 4(11), 1358–1366 (2012). [CrossRef]
74. T. Kowoll, E. Müller, S. Fritsch-Decker, S. Hettler, H. Störmer, C. Weiss, and D. Gerthsen, “Contrast of backscattered electron SEM images of nanoparticles on substrates with complex structure,” Scanning 2017, 1–12 (2017). [CrossRef]
75. D. Cialla, A. März, R. Böhme, F. Theil, K. Weber, M. Schmitt, and J. Popp, “Surface-enhanced Raman spectroscopy (SERS): progress and trends,” Anal. Bioanal. Chem. 403(1), 27–54 (2012). [CrossRef]
76. K. Kneipp, A. S. Haka, H. Kneipp, K. Badizadegan, N. Yoshizawa, C. Boone, K. E. Shafer-Peltier, J. T. Motz, R. R. Dasari, and M. S. Feld, “Surface-enhanced raman spectroscopy in single living cells using gold nanoparticles,” Appl. Spectrosc. 56(2), 150–154 (2002). [CrossRef]
77. I. Szumiel, “From radioresistance to radiosensitivity: In vitro evolution of L5178Y lymphoma,” Int. J. Radiat. Biol. 91(1), 1–12 (2015). [CrossRef]
78. M. Matczuk, L. Ruzik, S. S. Aleksenko, B. K. Keppler, M. Jarosz, and A. R. Timerbaev, “Analytical methodology for studying cellular uptake, processing and localization of gold nanoparticles,” Anal. Chim. Acta 1052, 1–9 (2019). [CrossRef]
79. S. Behzadi, V. Serpooshan, W. Tao, M. A. Hamaly, M. Y. Alkawareek, E. C. Dreaden, D. Brown, A. M. Alkilany, O. C. Farokhzad, and M. Mahmoudi, “Cellular uptake of nanoparticles: Journey inside the cell,” Chem. Soc. Rev. 46(14), 4218–4244 (2017). [CrossRef]
80. J. F. Hsu, P. Y. Hsieh, H. Y. Hsu, and S. Shigeto, “When cells divide: Label-free multimodal spectral imaging for exploratory molecular investigation of living cells during cytokinesis,” Sci. Rep. 5(1), 17541 (2015). [CrossRef]
81. T. J. MacCormack, A. M. Rundle, M. Malek, A. Raveendran, and M. V. Meli, “Gold nanoparticles partition to and increase the activity of glucose-6-phosphatase in a synthetic phospholipid membrane system,” PLoS One 12(8), e0183274 (2017). [CrossRef]
82. Z. J. Deng, M. Liang, I. Toth, M. Monteiro, and R. F. Minchin, “Plasma protein binding of positively and negatively charged polymer-coated gold nanoparticles elicits different biological responses,” Nanotoxicology 7(3), 314–322 (2012). [CrossRef]
83. J. E. Jaetao, K. S. Butler, N. L. Adolphi, D. M. Lovato, H. C. Bryant, I. Rabinowitz, S. S. Winter, T. E. Tessier, H. J. Hathaway, C. Bergemann, E. R. Flynn, and R. S. Larson, “Enhanced leukemia cell detection using a novel magnetic needle and nanoparticles,” Cancer Res. 69(21), 8310–8316 (2009). [CrossRef]
