CED
Experimental dermatology • Original article
Clinical and Experimental Dermatology
Raman spectroscopy as an analytical tool for melanoma research E. Brauchle,1,2,3 S. Noor,4 E. Holtorf,4 C. Garbe,4 K. Schenke-Layland1,2 and C. Busch4 1
Fraunhofer Institute for Interfacial Engineering and Biotechnology (IGB), Stuttgart, Germany; 2University Women’s Hospital Tuebingen, Eberhard Karls University Tuebingen, Tuebingen, Germany; 3University of Stuttgart, Institute for Interfacial Engineering and Plasma Technology (IGVP), Stuttgart, Germany; and 4Section of Dermato-Oncology, Department of Dermatology, University of Tuebingen, Tuebingen, Germany doi:10.1111/ced.12357
Summary
Background. Raman spectroscopy is an optical noninvasive screening technology that generates individual fingerprints of living cells by reflecting their molecular constitution. Aim. To discriminate melanoma cells from melanocytes, to identify drug-induced melanoma cell death stages (apoptosis, necrosis, autophagy) and to assess the susceptibility of melanoma cells to anticancer therapy. Methods. We used Raman spectroscopy on normal and melanoma cells, and on wild-type (WT) and mutant melanoma cells, to investigate whether the technique could distinguish between different types of cells, identify mutations and evaluate response to anticancer therapy. Results. Using the multivariate principal component analysis of the Raman spectra, melanocytes could be distinguished from melanoma cells, and WT melanoma cells could be distinguished from melanoma cells with BRAF or NRAS mutations. When we used the apoptosis inducer staurosporine, the necrosis inducer 3-bromopyruvate and the autophagy inducer resveratrol to induce cell death in SKMEL28 melanoma cells, Raman spectroscopy clearly distinguished between these three types of cell death, as confirmed by immunoblotting. Finally, the technique could discriminate between different melanoma cell lines according to their susceptibility to high-dose ascorbate. Conclusions. Raman spectroscopy is a powerful noninvasive tool to distinguish between melanocytes and melanoma cells, to analyze the specific type of cell death in melanoma cells, and to predict the susceptibility of melanoma cells to anticancer drugs.
Introduction Altered cell death signalling is a hallmark of melanoma development and progression. Activation of intracellular signalling molecules such as BRAF and NRAS drives proliferation and migration in melanoma Correspondence: Dr Christian Busch, Section of Dermato-Oncology, Department of Dermatology, University of Tuebingen, Liebermeisterstr.25, 72076 Tuebingen, Germany E-mail: [email protected] Conflict of interest: the authors declare that they have no conflicts of interest. The first two authors contributed equally to this work and should be considered joint first authors. Accepted for publication 26 January 2014
636
Clinical and Experimental Dermatology (2014) 39, pp636–645
cells, and inhibits cell death cascades.1 Therefore, specific targeting of melanoma death receptors and initiation of death signalling in melanoma cells are promising strategies for therapy. Such strategies focus predominantly on the induction of the cellular apoptosis machinery, or alternatively, on the induction of autophagy, which has a tumour-suppressing function.2 Consequently, toxicological studies aiming to identify new anti-melanoma substances need to consider the apoptotic and autophagic pathways in melanoma cells, and more importantly, to discriminate these pathways from that of necrosis,3 which does not exhibit any similarities to the apoptotic cell death cascade.4 Necrosis is an undesirable process that leads to adverse effects in vivo; cell swelling and a rapidly
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
collapsing plasma membrane affect large cell populations, and cause an inflammatory and toxicological response.5 Cyclophilin A (CypA), which was described as a necrosis-specific marker, is a cytosolic peptidylprolyl cis-trans isomerase that is upregulated by cellular stress and released early in necrosis, when the integrity of the plasma membrane is compromised.6 When the apoptotic machinery is started, the effector caspase-3 is cleaved, and chromatin condensation and the formation of apoptotic bodies is initiated.7 Autophagy, a lysosome-dependent process that occurs at basal levels in almost all cells, leads to the digestion of intracellular protein aggregates and damaged organelles.8,9 Formation of autophagosomes and their fusion with endosomes and lysosomes is a complex mechanism involving autophagy-related (Atg) proteins, including Beclin 1 and LC3.10 The cellular balance that is maintained by autophagy is dysregulated in many malignant cells. The effects of disturbed autophagy may alter cell proliferation and survival, and extensive autophagy causes cell death.11 Monitoring drug-induced cellular effects is an important step in understanding melanoma cell-specific death pathways. A number of cellular assays have been used to identify the molecular mechanisms of cell death.12 Such assays usually require large cell populations, and cannot detect the dynamic state of health of single cells. Kanduc et al.13 compared a number of the conventional cytotoxicity assays, and found that the reported viability of treated cells differed greatly depending on the assay used. In the present study, we used Raman spectroscopy as an optical technique for analysis of melanoma cells treated with different substances. Raman spectroscopy is an emerging tool in life sciences, with potential for noninvasive discrimination between healthy and malignantly transformed cells in vitro.14 It generates individual fingerprints representing the molecular constitution of the measured cells. Monochromatic laser light is used to excite vibration modes in molecules. Dependent on the excited mode, the photons from the incident beam can experience a frequency shift.15 The scattered light resulting from this frequency shift is then detected as a spectrum, in which the resulting peaks display the specific molecular bonds of proteins, nucleic acids, lipids and carbohydrates. Raman spectroscopy thus permits the nondestructive measurement of living cells and native tissues without the need for staining or processing steps prior to measurement.14 In cancer research, Raman spectroscopy has been used as a diagnostic tool in various in vitro and in vivo studies;16 for example, a microfluidic device was developed to detect tumour cells within blood samples at the single
ª 2014 British Association of Dermatologists
cell level.17 Wang et al.18 reported the discrimination of HaCaT cells and melanocytes from their malignant counterparts by Raman spectroscopy. Although there are also radiological, pharmacological, histological, cellular and molecular techniques available to distinguish nontransformed from malignantly transformed tissues and cells, most of these standard techniques are destructive and seem to be less specific than Raman spectroscopy.19 Raman spectroscopy has also been used to assess the potential toxicity of chemical compounds. Pyrgiotakis et al.20 combined the power of spectral resolution with one of the most widely used machine learning techniques. Support vector machines (SVMs) use a wellestablished database for cell classification. This database was constructed using three different stages of cells: healthy, necrotic or apoptotic. Based on these data, cells in the respective stages could be identified and discriminated. Recently, the novel small molecule BRAF inhibitor PLX4032 (vemurafenib) was clinically approved for the treatment of BRAF-mutated metastatic melanoma, and showed extraordinary response rates.21 Melanoma cell lines carrying a BRAF mutation show different sensitivities to vemurafenib.22 The strong clinical efficacy of this new targeted drug highlights the importance of analyzing the mutational status of the primary melanoma, which can then be clinically applied to stratify patients for subsequent therapies. In this study, we show that Raman spectroscopy can be used to distinguish melanocytes from melanoma cells, to distinguish wild-type (WT) from melanoma cells with NRAS or BRAF mutations, and even to distinguish NRAS-mutated from BRAF-mutated cells. In addition, Raman spectroscopy used on melanoma cell lines detected spectral shifts that were specific for apoptosis, necrosis or autophagy. Consequently, we were able to predict the type of cell death induced by 6-prenylnaringinin (6-PN) and 8-PN, two novel natural compounds, and these data were confirmed by western blot analysis. Finally, in terms of translational application, we distinguished melanoma cell lines that are highly sensitive to ascorbate-induced cell death from cell lines that are less sensitive, thus predicting the anticancer efficacy of this pharmacological treatment.
Methods Cell culture
The human metastatic melanoma cell lines SKMEL28, MALME-3M, SKMEL5, UACC-62, 451LU, SKMEL2, SKMEL30, WM852, 1205LU, LOX-IMVI (all from
Clinical and Experimental Dermatology (2014) 39, pp636–645
637
Raman spectroscopy in melanoma research E. Brauchle et al.
National Cancer Institute, Bethesda, MD, USA), MEWO and BLM (kind gift of Meenhard Herlyn, Wistar Institute, Philadelphia, PA, USA) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 1% penicillin and streptomycin, and 1% L-glutamine. All cell cultures were maintained at 37 °C in a 95% air/5% CO2 atmosphere at 100% humidity. Cytotoxicity treatment
Melanoma cells were seeded in triplicate in 96-well plates at a density of 2500 cells per well in 50 lL medium (5 9 105 cells/mL). After 24 h, the medium was replaced by fresh medium containing a concentration of ascorbate (32, 16, 8, 4, 2, 1, 0.5, 0.25, 0.125, 0.0625 or 0.03125 mmol/L), and incubated for 24 h. The medium was then discarded, and each well was washed twice with phosphate-buffered saline (PBS) without Ca2+ and Mg2+, and 100 lL of a solution containing 100 mg 4-methylumbelliferyl heptanoate per mL PBS was added. Plates were incubated at 37 °C for 1 h, then fluorescence was measured by a fluorescent microplate reader (Fluoroskan II; Labsystems, Helsinki, Finland) at an emission wavelength of 355 nm and an excitation wavelength of 460 nm. The intensity of fluorescence indicates the number of viable cells in the wells.
gated anti-rabbit antibody (1:10 000; Santa Cruz Biotechnology, Heidelberg, Germany) for 1 h. Membranes were washed three times in TBS-T, and further detection was performed by an ECL western blotting detection system on Hyperfilm-ECL (Amersham Biosciences). Raman spectroscopy and data analysis
A custom-built Raman spectrometer was used for all measurements, as described previously.23 Raman spectra were plotted using the Andor Solis software package (Andor iDus, Belfast, UK). The wave number range from 0 to 2000/cm was recorded. Measurements were performed after passaging on single cells suspended in medium. Every spectrum was accumulated for 100 s using a laser power of 85 mW. The background signal from the glass surface was subtracted using OPUS Software (version 4.2; Bruker Optik GmbH, Ettlingen, Germany), as previously described in detail.24 Principal component analysis (PCA) was used on vectornormalized spectra to identify significant peak shifts between the sample groups.24 SKMEL28 and BLM cells were incubated for 6 h with either 1 lmol/L staurosporine, 250 lmol/L 3-bromopyruvate or 100 lmol/L resveratrol, then photographed under light microscopy at 9 40 magnification.
Results Immunoblotting
All cell lines were seeded into T25 flasks and grown overnight. The following day, the cells were treated with either 1 lmol/L staurosporine, 250 lmol/L 3-bromopyruvate or 100 lmol/L resveratrol. At 24 h after adding the treatment substances, cells were harvested, washed with PBS and resuspended in 100 lL of lysing buffer (1% Nonident P40, 500 mmol/L Tris-Base pH 7.6, 150 mmol/L NaCl). The cell lysates were incubated on ice for 30 min, then 30 lg aliquots of each lysate were separated in 12% SDS-polyacrylamide gels and transferred to polyvinylidenedifluoride membranes (Hybond-P; Amersham Biosciences, Piscataway, NJ, USA). After blotting, the membranes were blocked for 1 h in Tris-buffered saline (150 mmol/L NaCl, 13 mmol/L Tris, pH 7.5) containing 5% nonfat dry milk. The membranes were treated with antibodies directed against caspase-3, cleaved caspase-3 (1 : 1000), cyclophilin A (1 : 1000) (all Cell Signaling, Frankfurt, Germany) or LC3 (1 : 1000; Sigma-Aldrich, Hamburg, Germany), incubated overnight at 4 °C, then washed three times with TBS-T (TBS containing 0.05% Triton X-100), and incubated with peroxidase-conju-
638
Clinical and Experimental Dermatology (2014) 39, pp636–645
Raman spectroscopy discriminates primary melanocytes from melanoma cells
Primary human melanocytes and five human metastatic melanoma cell lines (SKMEL28, MALME-3M, SKMEL5, UACC-62 and 451LU) were measured. The resulting mean Raman spectra are displayed in Fig. 1a. Using PCA on the Raman data, we identified two distinct clusters, each clearly representing one cell type (Fig. 1b). When analyzing the PCA loading, differences in three prominent peaks were visualized: phenylalanine (1001/cm), DNA (1370/cm) and amide I (1650/cm). In addition to the band at 1370/cm, the spectra of melanoma cell lines versus melanocytes showed differences at several DNA-related positions (789 and 1094/cm; Table 1). Raman spectroscopy distinguishes melanoma cell lines with BRAF or NRAS mutations from wild-type cells
Raman spectra of three NRAS-mutated (SKMEL2, SKMEL30, WM852), three BRAF-mutated (1205LU, LOX-IMVI and 451LU) and three WT metastatic
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
Figure 1 Raman spectroscopy discriminates primary melanocytes
from melanoma cells. (a) Raman spectra of melanoma cell lines (SKMEL28, MALME-3M, SKMEL5, UACC-62, 451LU) and melanocytes. (b) Principal component analysis (PCA) plot comparing data obtained from primary melanocytes and melanoma cell lines. Five different melanoma cell lines were measured (n = 30 each). Principal component (PC)4 and PC6 accounted for 3% and 2%, respectively, of the total variance between the obtained Raman spectra. Melanoma cells and melanocytes clustered in one group each within the diagram.
melanoma (MEMO, UACC-257 and MDA-MB-435) cell lines were measured, and PCA was performed to compare the cell lines based on their mutation status. The resulting score plot (Fig. 2a) displayed a slight separation for the principal component (PC)1 and PC3 scores obtained from spectra of WT and BRAF-mutated cell lines. PC1 and PC3 accounted for 41% of the total explained variance. When the spectra of WT and NRAS-mutated cell lines were compared (Fig. 2b), PC3 and PC7 depicted slight differences, and explained 11% of the total spectral variance. When the BRAFmutated and NRAS-mutated cells were compared, the PC1 and PC3 scores displayed a smaller overlapping area (Fig. 2c). This score plot included a total explained variance of 51%, and could successfully distinguish BRAF-mutated from NRAS-mutated cells. Different mechanisms of cell death show specific clusters after principal component analysis
We used the apoptosis inducer staurosporine, the necrosis inducer 3-bromopyruvate (3-BrPv)25 or the
ª 2014 British Association of Dermatologists
autophagy inducer resveratrol to induce cell death in SKMEL28 cells, which was further analyzed by immunoblotting (induction of cleaved caspase-3 led to apoptosis; secretion of cyclophilin A led to necrosis; shift from LC3-I to LC3-II led to autophagy). In staurosporine-treated cells, cleaved caspase-3 was detected 24 h after treatment. Analysis after 24 h of cell culture supernatants of untreated, DMSO-treated or 3-BrPv-treated cells detected expression of cyclophilin A in 3-BrPv-treated cells only. The shift from LC3-I to LC3-II could only be detected 48 h after resveratrol stimulation (Fig. 3a). Raman spectra of the cells treated with the three different substances showed alterations in their spectral patterns (Fig. 3b). The staurosporine-treated SKMEL28 cells exhibited decreased relative intensities at the wave numbers of 938, 1004 and 1032/cm, and dramatically increased Raman peaks at 1065, 1157, 1303 and 1658/cm (Table 1) compared with untreated cells. The 3-BrPv treatment group exhibited a decrease in all these peaks. In contrast to cells treated with staurosporine or 3-BrPv, the resveratrol-treated cells had less pronounced spectral changes; however, the spectra exhibited a clearly decreased signal at 718/cm (Table 1), which is associated with phospholipids and was shown be a useful spectral marker of autophagic response.26 PCA resulted in a score plot separating apoptotic, necrotic and autophagic cells for PC1 and PC2. The clearest separation without overlapping areas was obtained for vital and necrotic cells. Apoptotic cells were also distinguished from vital cells, although there were some areas of overlap. The spectra of autophagic cells were almost identical to those of vital cells, which is in line with the slight changes detected in the mean spectra (Fig. 3c). In addition to using Raman spectroscopy, we photographed untreated and treated (staurosporine, 3-bromopyruvate or resveratrol) SKMEL28 and BLM cells after 6 h of incubation to analyze the morphological changes induced by the drugs (Figure S1). The morphological assessment of the melanoma cells corroborated the previous results; in both cell lines, staurosporine induced the typical apoptotic morphology, with shrinkage of the cells and fragmentation into membrane-bound apoptotic bodies. 3-BrPv induced a morphological phenotype resembling necrosis, with rounding of the cells, swelling of the cytoplasm and initiation of lysis in some cells. Upon incubation with resveratrol, the cells had an increased number of cytoplasmic vesicles, indicative of autophagosome formation (Figure S1).
Clinical and Experimental Dermatology (2014) 39, pp636–645
639
Raman spectroscopy in melanoma research E. Brauchle et al.
Table 1 Peak assignments.
Peak position, per cm Peak assignment
Reference(s)
623 645 669 684 718 729 751 781 789 830 833 853 900 934
Naumann44; Puppels et al.45 Naumann44; Puppels et al.45 Puppels et al.45; Overman et al.46 Puppels et al.45; Overman et al.46 Gniadecka et al.33 Puppels et al.45; Overman et al.46 Puppels et al.45; Overman et al.46 Notingher et al.47; Stone et al.48; Stuart et al.49 Puppels et al.45; Overman et al.46 Puppels et al.45 Puppels et al.45 Puppels et al.45 Puppels et al.45 Stone et al.48
937 1001 1032 1092 1094 1155 1176 1209 1240 1252 1298 1305 1336 1339 1375 1422 1446 1449 1460 1487 1520–1538 1578 1607 1615 1654 1659
Phe Tyr T, G (DNA) G (ring breathing) Phospholipid A (ring breathing) T (ring breathing) Cytocine/uracil (ring breathing) C, T backbone Tyr Ribose–phosphate Tyr Backbone C-C stretch of proline ring and valine and protein backbone/glycogen Skel. C-C stretch of proteins Phe Phe O-P-O stretch Backbone PO2 C-C (and C-N) stretch of proteins Tyr, Phe Tyr, Phe T, Amide III T, Amide III CH2 deformation mode of lipids A, CH deformation DNA purine bases A, CH deformation T, A, G A, G CH2 bending mode of proteins CH deformation CH deformation G, A C=C carotenoids G, A proteins Tyr, Phe C=C stretching mode for Tyr and Tryoptophan Amide I, C=C lipid stretch Amide I
Carter et al.50 Stone et al.48; Stuart et al.49; Yan et al.51 Puppels et al.45 Notingher et al.47; Stone et al.48 Puppels et al.45 Stone et al.48 Puppels et al.45 Puppels et al.45 Puppels et al.45 Puppels et al.45 Stone et al.48 Carter et al.50 Notingher et al.47 Puppels et al.45; Overman et al.46 Overman et al.46 Puppels et al.45 Stone et al.48 Naumann44 Naumann44 Puppels et al.45 Stone et al.48 Puppels et al.45 Naumann44 Stone et al.48 Stone et al.48; Yan et al.51 Carter et al.50
Raman spectroscopy and principal component analysis are able to predict the cell death mechanism
We next explored the applicability of Raman spectroscopy to detect the mechanism of cell death after treatment with the natural flavonoids 6-PN or 8-PN. SKMEL28 cells were incubated with 6-PN or 8-PN for 24 h and scanned. The resulting spectra were
640
Clinical and Experimental Dermatology (2014) 39, pp636–645
compared with spectra of vital, necrotic, autophagic and apoptotic SKMEL28 cells to screen for possible clustering. The resulting scores plot showed that the spectra of cells incubated with 6-PN or 8-PN were clearly distinct from those of apoptotic and necrotic cells, but they matched the cluster previously determined for autophagic cells (Fig. 4a). This was verified by immunoblot analysis showing the shift from LC3-I
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
(n = 15) yielded different cytotoxic sensitivities towards ascorbate, with half maximal effective concentration (EC50) values varying from 0.19 to 12 mmol/L (not shown). We then chose six different cell lines, three with high ascorbate sensitivity (EC50: MDA-MB-435, 0.19 mmol/L; SKMEL2, 1.4 mmol/L; and WM852, 1.5 mmol/L) and three with low ascorbate sensitivity (EC50: UACC-62, 5.5 mmol/L; MALME-3M, 6.9 mmol/L; and SKMEL28, 12 mmol/L) for Raman spectroscopy (Fig. 5a). The score plot displayed a separation of both groups, although there was some overlap between the spectra (Fig. 4b). PC4 accounted for 4% and PC3 for 10% of the total variance.
Discussion
(c)
Figure 2 Raman spectroscopy distinguishes melanoma cell lines
with BRAF or NRAS mutations from wild-type (WT) cells. (a) WT cells compared with BRAF-mutated cells (LOX-IMVI, 1205LU, 451LU). Principal component (PC)3 accounted for 9% and PC7 for 2% of the total variance in the Raman spectra. (b) WT cells (UACC-257, MEWO, MDA-MB-435) compared with NRAS-mutated cells (SKMEL2, BLM, WM852). PC1 accounted for 34% and PC3 for 7% of the total explained variance. (c) NRAS-mutated cells and BRAF-mutated cells clustered in distinct groups in the score plot; PC1 accounted for 38% and PC3 for 13% of the total explained variance.
to LC3-II after incubation with 6-PN and 8-PN for 24 h (Fig. 4b). Raman spectroscopy detects differences in sensitivity towards drug-induced melanoma cell death
Finally, we tested the efficacy of different pharmacological doses of ascorbate27,28 in killing melanoma cells. The human metastatic melanoma cell lines tested
ª 2014 British Association of Dermatologists
In this study, we were able to discriminate melanoma cell lines from melanocytes using Raman spectroscopy, a diagnostic tool used for cancer monitoring,29,30 screening of the molecular and structural composition of skin,31 and identifying differences between pathological and healthy skin tissue.32–34 In our study, specific Raman peaks and differences in Raman signal intensities identified biochemical differences between melanoma cell lines and melanocytes. We applied multivariate analyses to the spectral data to distinguish melanocytes from melanoma cells. Our data confirm recently published data describing a precise separation between HaCaT cells and melanocytes from their malignant counterparts.18 This technique also serves as diagnostic tool for skin cancer.35 In the latter study, malignant and pre-malignant lesions were distinguished from benign disorders, melanomas from benign pigmented skin lesions, and melanomas from pigmented seborrhoeic keratosis. A high percentage of melanocytic tumours harbour activating BRAF mutations. Approximately 70–80% of acquired melanocytic naevi and 40–50% of primary melanomas contain a BRAF mutation, the vast majority of which result in a single amino acid change at codon 600 (BRAFV600E).36 Although the RAS signalling pathway plays an important role in the homeostasis of normal cell turnover, death and survival, studies hae identified activating mutations in RAS family members in various malignancies.37 In melanoma, NRAS mutations, found in 10–25% of tumour samples, are an important oncogenic driver.38 In the current study, we successfully differentiated WT cells from NRAS-mutated melanoma cells, WT cells from BRAFmutated melanoma cells, and NRAS-mutated from BRAF-mutated melanoma cells. Therefore, we suggest
Clinical and Experimental Dermatology (2014) 39, pp636–645
641
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
(c)
Figure 3 Different mechanisms of cell death showed specific clusters after principal component (PCA) analysis. (a) SKMEL28 cells were
incubated with staurosporine, 3-bromopyruvate (3-BrPv) or resveratrol, and cell lysates were subjected to SDS-PAGE followed by immunoblotting. Apoptosis was verified by induction of cleaved caspase-3, necrosis was detected by cyclophilin A secretion into the medium, and autophagy was verified by the shift from LC3-I to LC3-II. (b) Melanoma cells (BLM, MEWO, SKMEL28) incubated with staurosporine, 3-BrPv or resveratrol were lysed with trypsin and used for Raman spectroscopy. Untreated cells served as controls. (c) PCA comparing these four groups showed that principal component (PC)1 accounted for 35% and PC2 for 20% of the total variance in the Raman spectra.
Raman spectroscopy as a possible alternative to the current methods of melanoma mutation analysis by sequencing or PCR. This also has clinical relevance, as the US Food and Drugs Authority and other authorities have approved the small molecule vemurafenib as targeted therapy for BRAF-mutated metastatic melanoma. In addition to targeted therapy with vemurafenib or immune therapy, with an anti-CTLA4 antibody (ipilimumab), (poly-) chemotherapy is used as second-line or third-line treatment options in metastatic melanoma. All clinically applied chemotherapeutics kill cancer cells by different modes of action. We therefore investigated whether Raman spectroscopy could distinguish the different types of cell death in melanoma cells. After apoptosis, necrosis or autophagy were induced by defined substances,39 we were able to identify unambiguously specific clusters for each type of cell death. The spectra of each group showed alterations in their peaks, which were highly specific for apoptosis, necrosis or autophagy.26 Morphological changes in the cells after treatment with the various drugs supported the induction of the relevant type of cell death (apoptosis, necrosis or autophagy). We then used the obtained data to test 6-PN and 8PN, which induce cell death in different cancer cell lines.40–42 Raman spectroscopy indicated that 6-PN and 8-PN did not induce either apoptosis nor necrosis, but did induce autophagy, as the PN-incubated cells
642
Clinical and Experimental Dermatology (2014) 39, pp636–645
(a)
(b)
Figure 4 Raman spectroscopy and principal components analysis
(PCA) were able to predict the cell death mechanism in melanoma cells. (a) SKMEL28 cells incubated with the cell death inducers staurosporine, 3-bromopyruvate (3-BrPv) and resveratrol or the flavonoids 6-prenylnaringenin (6-PN) or 8-PN were analysed by Raman spectroscopy. The spectra of untreated, staurosporine-treated and 3-BrPv-treated cells clustered in distinct groups. One additional cluster was detected for resveratrol and the PNs; PC1 accounted for 36% and PC2 for 25% of the total explained variance. (b) Immunoblot analysis demonstrated the shift from LC3-I to LC3-II, an indicator of autophagy.
clustered in the same area as that of autophagic cells. Immunoblot analysis confirmed that the shift from LC3-I to LC3-II was induced by 6-PN or 8-PN.
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
melanocytes and melanoma cells and that it can be used to determine the cellular mutation status, to analyze the specific type of cell death in melanoma cells, to predict the susceptibility of melanoma cells to anticancer drugs.
Acknowledgements
(b)
This project was supported by grants from the Deutsche Forschungsgemeinschaft SFB 773: ‘Understanding and overcoming drug resistance of solid tumors’ (to CG and CB), the Wissenschaftsf€ orderung der Deutschen Brauwirtschaft eV (to CB), the Fraunhofer-Gesellschaft Internal programm (to KSL), and the Ministry of Science, Research and the Arts of Baden-W€ urttemberg (33-729.55-3/214, to KSL).
What’s already known about this topic? ● Raman spectroscopy is a noninvasive diagnos-
tic tool that can distinguish between healthy and malignant cells.
Figure 5 Raman spectroscopy detect differences in sensitivity
towards drug-induced melanoma cell death. (a) Three untreated melanoma cell lines with high ascorbate sensitivity (EC50: MDAMB-435, 0.19 mmol/L; SKMEL2, 1.4 mmol/L; and WM852, 1.5 mmol/L) and three with low ascorbate sensitivity (EC50: UACC-62, 5.5 mmol/L; MALME-3M, 6.9 mmol/L SKMEL28, 12 mmol/L) were analyzed by Raman spectroscopy. (b) The score plot displayed a separation of both groups, although the spectra of both groups did overlap. PC4 accounted for 4% and PC3 for 10% of the total variance.
Increasing evidence supports the hypothesis of Pauling and Cameron43 that ascorbate selectively kills cancer cells by a pro-oxidant mode of action.27,28 Indeed, we were able to induce cell death in several melanoma cell lines with pharmacological doses of ascorbate; however, we detected large differences in ascorbate sensitivity. Using Raman spectroscopy, we generated different clusters for highly sensitive and less sensitive melanoma cells, suggesting that the applied technique can allow prediction of the cytotoxic susceptibility of melanoma cells towards a particular drug.
Conclusion Altogether, our results show that Raman spectroscopy is a powerful noninvasive tool for distinction between
ª 2014 British Association of Dermatologists
What does this study add? ● The technique is an analytical approach for
the determination of the type of cell death in melanoma cells. ● Raman spectroscopy provides an optical readout for the susceptibility of melanoma cells to anticancer drugs.
References 1 Bennett DC. How to make a melanoma: what do we know of the primary clonal events? Pigment Cell Melanoma Res 2008; 21: 27–38. 2 Shintani T, Klionsky DJ. Autophagy in health and disease: a double-edged sword. Science 2004; 306: 990–5. 3 Ocker M, Hopfner M. Apoptosis-modulating drugs for improved cancer therapy. Eur Surg Res 2012; 48: 111–20. 4 Degterev A, Huang Z, Boyce M et al. Chemical inhibitor of nonapoptotic cell death with therapeutic potential for ischemic brain injury. Nat Chem Biol 2005; 1: 112–19. 5 Proskuryakov SY, Gabai VL. Mechanisms of tumor cell necrosis. Curr Pharm Des 2010; 16: 56–68.
Clinical and Experimental Dermatology (2014) 39, pp636–645
643
Raman spectroscopy in melanoma research E. Brauchle et al.
6 Christofferson DE, Yuan J. Cyclophilin A release as a biomarker of necrotic cell death. Cell Death Differ 2010; 17: 1942–3. 7 Hersey P, Zhang XD. How melanoma cells evade trail-induced apoptosis. Nat Rev Cancer 2001; 1: 142–50. 8 Kim I, Rodriguez-Enriquez S, Lemasters JJ. Selective degradation of mitochondria by mitophagy. Arch Biochem Biophys 2007; 462: 245–53. 9 Levine B, Klionsky DJ. Development by self-digestion: molecular mechanisms and biological functions of autophagy. Dev Cell 2004; 6: 463–77. 10 Eskelinen EL. Maturation of autophagic vacuoles in mammalian cells. Autophagy 2005; 1: 1–10. 11 Levine B. Cell biology: autophagy and cancer. Nature 2007; 446: 745–7. 12 Jaeschke H, Gujral JS, Bajt ML. Apoptosis and necrosis in liver disease. Liver Int 2004; 24: 85–9. 13 Kanduc D, Mittelman A, Serpico R et al. Cell death: apoptosis versus necrosis. Int J Oncol 2002; 21: 165–70. 14 Brauchle E, Schenke-Layland K. Raman spectroscopy in biomedicine – non-invasive in vitro analysis of cells and extracellular matrix components in tissues. Biotechnol J 2013; 8: 288–97. 15 Raman C, Krishna CM. The Raman effect – inelastic insight. Nature 1928; 121: 501–2. 16 Tu Q, Chang C. Diagnostic applications of Raman spectroscopy. Nanomedicine 2012; 8: 545–58. 17 Dochow S, Krafft C, Neugebauer U et al. Tumour cell identification by means of Raman spectroscopy in combination with optical traps and microfluidic environments. Lab Chip 2011; 11: 1484–90. 18 Wang H, Tsai TH, Zhao J et al. Differentiation of HaCaT cell and melanocyte from their malignant counterparts using micro-Raman spectroscopy guided by confocal imaging. Photodermatol Photoimmunol Photomed 2012; 28: 147–52. 19 Chan JW, Taylor DS, Zwerdling T et al. Micro-Raman spectroscopy detects individual neoplastic and normal hematopoietic cells. Biophys J 2006; 90: 648–56. 20 Pyrgiotakis G, Kundakcioglu OE, Finton K et al. Cell death discrimination with Raman spectroscopy and support vector machines. Ann Biomed Eng 2009; 37: 1464–73. 21 Flaherty KT, Puzanov I, Kim KB et al. Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med 2010; 363: 809–19. 22 Sondergaard JN, Nazarian R, Wang Q et al. Differential sensitivity of melanoma cell lines with BRAFV600E mutation to the specific Raf inhibitor PLX4032. J Transl Med 2010; 8: 39. 23 Pudlas M, Berrio DAC, Votteler M et al. Non-contact discrimination of human bone marrow-derived mesenchymal stem cels and fibroblasts using Raman spectroscopy. Med Laser Appl 2011; 26: 119–25. 24 Pudlas M, Brauchle E, Klein TJ et al. Non-invasive identification of proteoglycans and chondrocyte
644
Clinical and Experimental Dermatology (2014) 39, pp636–645
25
26
27
28
29
30
31
32
33
34
35
36 37 38
39
differentiation state by Raman microspectroscopy. J Biophotonics 2013; 6: 205–11. Qin JZ, Xin H, Nickoloff BJ. 3-Bromopyruvate induces necrotic cell death in sensitive melanoma cell lines. Biochem Biophys Res Commun 2010; 396: 495–500. Konorov SO, Jardon MA, Piret JM et al. Raman microspectroscopy of live cells under autophagy-inducing conditions. Analyst 2012; 137: 4662–8. Chen Q, Espey MG, Krishna MC et al. Pharmacologic ascorbic acid concentrations selectively kill cancer cells: action as a pro-drug to deliver hydrogen peroxide to tissues. Proc Natl Acad Sci USA 2005; 102: 13604–9. Sinnberg T, Noor S, Venturelli S et al. The ROS-induced cytotoxicity of ascorbate is attenuated by hypoxia and HIF-1alpha in the NCI60 cancer cell lines. J Cell Mol Med 2014; 18: 530–41. Larraona-Puy M, Ghita A, Zoladek A et al. Development of Raman microspectroscopy for automated detection and imaging of basal cell carcinoma. J Biomed Opt 2009; 14: 054031. Lyng FM, Faolain EO, Conroy J et al. Vibrational spectroscopy for cervical cancer pathology, from biochemical analysis to diagnostic tool. Exp Mol Pathol 2007; 82: 121–9. Bernard G, Auger M, Soucy J et al. Physical characterization of the stratum corneum of an in vitro psoriatic skin model by ATR-FTIR and Raman spectroscopies. Biochim Biophys Acta 2007; 1770: 1317–23. El GA, Commandeur S, Rietveld MH et al. Replacement of animal-derived collagen matrix by human fibroblast-derived dermal matrix for human skin equivalent products. Biomaterials 2009; 30: 71–8. Gniadecka M, Wulf HC, Nielsen OF et al. Distinctive molecular abnormalities in benign and malignant skin lesions: studies by Raman spectroscopy. Photochem Photobiol 1997; 66: 418–23. Nijssen A, Bakker Schut TC, Heule F et al. Discriminating basal cell carcinoma from its surrounding tissue by Raman spectroscopy. J Invest Dermatol 2002; 119: 64–9. Lui H, Zhao J, McLean D et al. Real-time Raman spectroscopy for in vivo skin cancer diagnosis. Cancer Res 2012; 72: 2491–500. Pollock PM, Harper UL, Hansen KS et al. High frequency of BRAF mutations in nevi. Nat Genet 2003; 33: 19–20. Bos JL. ras oncogenes in human cancer: a review. Cancer Res 1989; 49: 4682–9. Platz A, Ringborg U, Brahme EM et al. Melanoma metastases from patients with hereditary cutaneous malignant melanoma contain a high frequency of N-ras activating mutations. Melanoma Res 1994; 4: 169–77. Morselli E, Marino G, Bennetzen MV et al. Spemidine and resveratrol induce autophagy by distinct pathways converging on the acetylproteome. J Cell Biol 2011; 192: 615–29.
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
40 Brunelli E, Pinton G, Bellini P et al. Flavonoid-induced autophagy in hormone sensitive breast cancer cells. Fitoterapia 2009; 80: 327–32. 41 Brunelli E, Pinton G, Chianale F et al. 8-Prenylnaringenin inhibits epidermal growth factor-induced MCF-7 breast cancer cell proliferation by targeting phosphatidylinositol-3-OH kinase activity. J Steroid Biochem Mol Biol 2009; 113: 163–70. 42 Delmulle L, Vanden Berghe T, Keukeleire DD et al. Treatment of PC-3 and DU145 prostate cancer cells by prenylflavonoids from hop (Humulus lupulus L.) induces a caspase-independent form of cell death. Phytother Res 2008; 22: 197–203. 43 Cameron E, Pauling L. Supplemental ascorbate in the supportive treatment of cancer: prolongation of survival times in terminal human cancer. Proc Natl Acad Sci USA 1976; 73: 3685–9. 44 Naumann D. FT-Infrared and FT-Raman spectroscopy in biomedical research. Appl Spectrosc Rev 2001; 36: 239–98. 45 Puppels GJ, Garritsen HS, Segers-Nolten GM et al. Raman microspectroscopic approach to the study of human granulocytes. Biophys J 1991; 60: 1046–56. 46 Overman SA, Aubrey KL, Reilly KE et al. Conformation and interactions of the packaged double-stranded DNA genome of bacteriophage T7. Biospectroscopy 1998; 4: S47–56.
ª 2014 British Association of Dermatologists
47 Notingher I, Verrier S, Hague S et al. Spectroscopic study of human lung epithelial cells (A549) in culture, living cells versus dead cells. Biopolymers 2003; 72: 230–40. 48 Stone N, Kendall C, Smith J et al. Raman spectroscopy for identification of epithelial cancers. Faraday Discuss 2004; 126: 141–57. 49 Stuart NS, McIIImurray MB, Bishop JL et al. Vinorelbine and infusional 5-fluorouracil in the anthracycline and taxane pre-treated metastatic breast cancer. Clin Oncol (R Coll Radiol) 2008; 20: 152–6. 50 Carter EA, Edwards HGM. Infrared and Raman spectroscopy of biological materials. J Raman Spectrosc 2004; 35: 754–60. 51 Yan XL, Dong RX, Zhang L et al. Raman spectra of single cell from gastrointestinal cancer patients. World J Gastroenterol 2005; 11: 3290–2.
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1 (a) SKMEL28 and (b) BLM cells were either left untreated or were subjected to treatment with either staurosporine, 3-bromopyruvate or resveratrol, and photographed after 6 h of treatment to analyze the morphological changes.
Clinical and Experimental Dermatology (2014) 39, pp636–645
645
Experimental dermatology • Original article
Clinical and Experimental Dermatology
Raman spectroscopy as an analytical tool for melanoma research E. Brauchle,1,2,3 S. Noor,4 E. Holtorf,4 C. Garbe,4 K. Schenke-Layland1,2 and C. Busch4 1
Fraunhofer Institute for Interfacial Engineering and Biotechnology (IGB), Stuttgart, Germany; 2University Women’s Hospital Tuebingen, Eberhard Karls University Tuebingen, Tuebingen, Germany; 3University of Stuttgart, Institute for Interfacial Engineering and Plasma Technology (IGVP), Stuttgart, Germany; and 4Section of Dermato-Oncology, Department of Dermatology, University of Tuebingen, Tuebingen, Germany doi:10.1111/ced.12357
Summary
Background. Raman spectroscopy is an optical noninvasive screening technology that generates individual fingerprints of living cells by reflecting their molecular constitution. Aim. To discriminate melanoma cells from melanocytes, to identify drug-induced melanoma cell death stages (apoptosis, necrosis, autophagy) and to assess the susceptibility of melanoma cells to anticancer therapy. Methods. We used Raman spectroscopy on normal and melanoma cells, and on wild-type (WT) and mutant melanoma cells, to investigate whether the technique could distinguish between different types of cells, identify mutations and evaluate response to anticancer therapy. Results. Using the multivariate principal component analysis of the Raman spectra, melanocytes could be distinguished from melanoma cells, and WT melanoma cells could be distinguished from melanoma cells with BRAF or NRAS mutations. When we used the apoptosis inducer staurosporine, the necrosis inducer 3-bromopyruvate and the autophagy inducer resveratrol to induce cell death in SKMEL28 melanoma cells, Raman spectroscopy clearly distinguished between these three types of cell death, as confirmed by immunoblotting. Finally, the technique could discriminate between different melanoma cell lines according to their susceptibility to high-dose ascorbate. Conclusions. Raman spectroscopy is a powerful noninvasive tool to distinguish between melanocytes and melanoma cells, to analyze the specific type of cell death in melanoma cells, and to predict the susceptibility of melanoma cells to anticancer drugs.
Introduction Altered cell death signalling is a hallmark of melanoma development and progression. Activation of intracellular signalling molecules such as BRAF and NRAS drives proliferation and migration in melanoma Correspondence: Dr Christian Busch, Section of Dermato-Oncology, Department of Dermatology, University of Tuebingen, Liebermeisterstr.25, 72076 Tuebingen, Germany E-mail: [email protected] Conflict of interest: the authors declare that they have no conflicts of interest. The first two authors contributed equally to this work and should be considered joint first authors. Accepted for publication 26 January 2014
636
Clinical and Experimental Dermatology (2014) 39, pp636–645
cells, and inhibits cell death cascades.1 Therefore, specific targeting of melanoma death receptors and initiation of death signalling in melanoma cells are promising strategies for therapy. Such strategies focus predominantly on the induction of the cellular apoptosis machinery, or alternatively, on the induction of autophagy, which has a tumour-suppressing function.2 Consequently, toxicological studies aiming to identify new anti-melanoma substances need to consider the apoptotic and autophagic pathways in melanoma cells, and more importantly, to discriminate these pathways from that of necrosis,3 which does not exhibit any similarities to the apoptotic cell death cascade.4 Necrosis is an undesirable process that leads to adverse effects in vivo; cell swelling and a rapidly
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
collapsing plasma membrane affect large cell populations, and cause an inflammatory and toxicological response.5 Cyclophilin A (CypA), which was described as a necrosis-specific marker, is a cytosolic peptidylprolyl cis-trans isomerase that is upregulated by cellular stress and released early in necrosis, when the integrity of the plasma membrane is compromised.6 When the apoptotic machinery is started, the effector caspase-3 is cleaved, and chromatin condensation and the formation of apoptotic bodies is initiated.7 Autophagy, a lysosome-dependent process that occurs at basal levels in almost all cells, leads to the digestion of intracellular protein aggregates and damaged organelles.8,9 Formation of autophagosomes and their fusion with endosomes and lysosomes is a complex mechanism involving autophagy-related (Atg) proteins, including Beclin 1 and LC3.10 The cellular balance that is maintained by autophagy is dysregulated in many malignant cells. The effects of disturbed autophagy may alter cell proliferation and survival, and extensive autophagy causes cell death.11 Monitoring drug-induced cellular effects is an important step in understanding melanoma cell-specific death pathways. A number of cellular assays have been used to identify the molecular mechanisms of cell death.12 Such assays usually require large cell populations, and cannot detect the dynamic state of health of single cells. Kanduc et al.13 compared a number of the conventional cytotoxicity assays, and found that the reported viability of treated cells differed greatly depending on the assay used. In the present study, we used Raman spectroscopy as an optical technique for analysis of melanoma cells treated with different substances. Raman spectroscopy is an emerging tool in life sciences, with potential for noninvasive discrimination between healthy and malignantly transformed cells in vitro.14 It generates individual fingerprints representing the molecular constitution of the measured cells. Monochromatic laser light is used to excite vibration modes in molecules. Dependent on the excited mode, the photons from the incident beam can experience a frequency shift.15 The scattered light resulting from this frequency shift is then detected as a spectrum, in which the resulting peaks display the specific molecular bonds of proteins, nucleic acids, lipids and carbohydrates. Raman spectroscopy thus permits the nondestructive measurement of living cells and native tissues without the need for staining or processing steps prior to measurement.14 In cancer research, Raman spectroscopy has been used as a diagnostic tool in various in vitro and in vivo studies;16 for example, a microfluidic device was developed to detect tumour cells within blood samples at the single
ª 2014 British Association of Dermatologists
cell level.17 Wang et al.18 reported the discrimination of HaCaT cells and melanocytes from their malignant counterparts by Raman spectroscopy. Although there are also radiological, pharmacological, histological, cellular and molecular techniques available to distinguish nontransformed from malignantly transformed tissues and cells, most of these standard techniques are destructive and seem to be less specific than Raman spectroscopy.19 Raman spectroscopy has also been used to assess the potential toxicity of chemical compounds. Pyrgiotakis et al.20 combined the power of spectral resolution with one of the most widely used machine learning techniques. Support vector machines (SVMs) use a wellestablished database for cell classification. This database was constructed using three different stages of cells: healthy, necrotic or apoptotic. Based on these data, cells in the respective stages could be identified and discriminated. Recently, the novel small molecule BRAF inhibitor PLX4032 (vemurafenib) was clinically approved for the treatment of BRAF-mutated metastatic melanoma, and showed extraordinary response rates.21 Melanoma cell lines carrying a BRAF mutation show different sensitivities to vemurafenib.22 The strong clinical efficacy of this new targeted drug highlights the importance of analyzing the mutational status of the primary melanoma, which can then be clinically applied to stratify patients for subsequent therapies. In this study, we show that Raman spectroscopy can be used to distinguish melanocytes from melanoma cells, to distinguish wild-type (WT) from melanoma cells with NRAS or BRAF mutations, and even to distinguish NRAS-mutated from BRAF-mutated cells. In addition, Raman spectroscopy used on melanoma cell lines detected spectral shifts that were specific for apoptosis, necrosis or autophagy. Consequently, we were able to predict the type of cell death induced by 6-prenylnaringinin (6-PN) and 8-PN, two novel natural compounds, and these data were confirmed by western blot analysis. Finally, in terms of translational application, we distinguished melanoma cell lines that are highly sensitive to ascorbate-induced cell death from cell lines that are less sensitive, thus predicting the anticancer efficacy of this pharmacological treatment.
Methods Cell culture
The human metastatic melanoma cell lines SKMEL28, MALME-3M, SKMEL5, UACC-62, 451LU, SKMEL2, SKMEL30, WM852, 1205LU, LOX-IMVI (all from
Clinical and Experimental Dermatology (2014) 39, pp636–645
637
Raman spectroscopy in melanoma research E. Brauchle et al.
National Cancer Institute, Bethesda, MD, USA), MEWO and BLM (kind gift of Meenhard Herlyn, Wistar Institute, Philadelphia, PA, USA) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS), 1% penicillin and streptomycin, and 1% L-glutamine. All cell cultures were maintained at 37 °C in a 95% air/5% CO2 atmosphere at 100% humidity. Cytotoxicity treatment
Melanoma cells were seeded in triplicate in 96-well plates at a density of 2500 cells per well in 50 lL medium (5 9 105 cells/mL). After 24 h, the medium was replaced by fresh medium containing a concentration of ascorbate (32, 16, 8, 4, 2, 1, 0.5, 0.25, 0.125, 0.0625 or 0.03125 mmol/L), and incubated for 24 h. The medium was then discarded, and each well was washed twice with phosphate-buffered saline (PBS) without Ca2+ and Mg2+, and 100 lL of a solution containing 100 mg 4-methylumbelliferyl heptanoate per mL PBS was added. Plates were incubated at 37 °C for 1 h, then fluorescence was measured by a fluorescent microplate reader (Fluoroskan II; Labsystems, Helsinki, Finland) at an emission wavelength of 355 nm and an excitation wavelength of 460 nm. The intensity of fluorescence indicates the number of viable cells in the wells.
gated anti-rabbit antibody (1:10 000; Santa Cruz Biotechnology, Heidelberg, Germany) for 1 h. Membranes were washed three times in TBS-T, and further detection was performed by an ECL western blotting detection system on Hyperfilm-ECL (Amersham Biosciences). Raman spectroscopy and data analysis
A custom-built Raman spectrometer was used for all measurements, as described previously.23 Raman spectra were plotted using the Andor Solis software package (Andor iDus, Belfast, UK). The wave number range from 0 to 2000/cm was recorded. Measurements were performed after passaging on single cells suspended in medium. Every spectrum was accumulated for 100 s using a laser power of 85 mW. The background signal from the glass surface was subtracted using OPUS Software (version 4.2; Bruker Optik GmbH, Ettlingen, Germany), as previously described in detail.24 Principal component analysis (PCA) was used on vectornormalized spectra to identify significant peak shifts between the sample groups.24 SKMEL28 and BLM cells were incubated for 6 h with either 1 lmol/L staurosporine, 250 lmol/L 3-bromopyruvate or 100 lmol/L resveratrol, then photographed under light microscopy at 9 40 magnification.
Results Immunoblotting
All cell lines were seeded into T25 flasks and grown overnight. The following day, the cells were treated with either 1 lmol/L staurosporine, 250 lmol/L 3-bromopyruvate or 100 lmol/L resveratrol. At 24 h after adding the treatment substances, cells were harvested, washed with PBS and resuspended in 100 lL of lysing buffer (1% Nonident P40, 500 mmol/L Tris-Base pH 7.6, 150 mmol/L NaCl). The cell lysates were incubated on ice for 30 min, then 30 lg aliquots of each lysate were separated in 12% SDS-polyacrylamide gels and transferred to polyvinylidenedifluoride membranes (Hybond-P; Amersham Biosciences, Piscataway, NJ, USA). After blotting, the membranes were blocked for 1 h in Tris-buffered saline (150 mmol/L NaCl, 13 mmol/L Tris, pH 7.5) containing 5% nonfat dry milk. The membranes were treated with antibodies directed against caspase-3, cleaved caspase-3 (1 : 1000), cyclophilin A (1 : 1000) (all Cell Signaling, Frankfurt, Germany) or LC3 (1 : 1000; Sigma-Aldrich, Hamburg, Germany), incubated overnight at 4 °C, then washed three times with TBS-T (TBS containing 0.05% Triton X-100), and incubated with peroxidase-conju-
638
Clinical and Experimental Dermatology (2014) 39, pp636–645
Raman spectroscopy discriminates primary melanocytes from melanoma cells
Primary human melanocytes and five human metastatic melanoma cell lines (SKMEL28, MALME-3M, SKMEL5, UACC-62 and 451LU) were measured. The resulting mean Raman spectra are displayed in Fig. 1a. Using PCA on the Raman data, we identified two distinct clusters, each clearly representing one cell type (Fig. 1b). When analyzing the PCA loading, differences in three prominent peaks were visualized: phenylalanine (1001/cm), DNA (1370/cm) and amide I (1650/cm). In addition to the band at 1370/cm, the spectra of melanoma cell lines versus melanocytes showed differences at several DNA-related positions (789 and 1094/cm; Table 1). Raman spectroscopy distinguishes melanoma cell lines with BRAF or NRAS mutations from wild-type cells
Raman spectra of three NRAS-mutated (SKMEL2, SKMEL30, WM852), three BRAF-mutated (1205LU, LOX-IMVI and 451LU) and three WT metastatic
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
Figure 1 Raman spectroscopy discriminates primary melanocytes
from melanoma cells. (a) Raman spectra of melanoma cell lines (SKMEL28, MALME-3M, SKMEL5, UACC-62, 451LU) and melanocytes. (b) Principal component analysis (PCA) plot comparing data obtained from primary melanocytes and melanoma cell lines. Five different melanoma cell lines were measured (n = 30 each). Principal component (PC)4 and PC6 accounted for 3% and 2%, respectively, of the total variance between the obtained Raman spectra. Melanoma cells and melanocytes clustered in one group each within the diagram.
melanoma (MEMO, UACC-257 and MDA-MB-435) cell lines were measured, and PCA was performed to compare the cell lines based on their mutation status. The resulting score plot (Fig. 2a) displayed a slight separation for the principal component (PC)1 and PC3 scores obtained from spectra of WT and BRAF-mutated cell lines. PC1 and PC3 accounted for 41% of the total explained variance. When the spectra of WT and NRAS-mutated cell lines were compared (Fig. 2b), PC3 and PC7 depicted slight differences, and explained 11% of the total spectral variance. When the BRAFmutated and NRAS-mutated cells were compared, the PC1 and PC3 scores displayed a smaller overlapping area (Fig. 2c). This score plot included a total explained variance of 51%, and could successfully distinguish BRAF-mutated from NRAS-mutated cells. Different mechanisms of cell death show specific clusters after principal component analysis
We used the apoptosis inducer staurosporine, the necrosis inducer 3-bromopyruvate (3-BrPv)25 or the
ª 2014 British Association of Dermatologists
autophagy inducer resveratrol to induce cell death in SKMEL28 cells, which was further analyzed by immunoblotting (induction of cleaved caspase-3 led to apoptosis; secretion of cyclophilin A led to necrosis; shift from LC3-I to LC3-II led to autophagy). In staurosporine-treated cells, cleaved caspase-3 was detected 24 h after treatment. Analysis after 24 h of cell culture supernatants of untreated, DMSO-treated or 3-BrPv-treated cells detected expression of cyclophilin A in 3-BrPv-treated cells only. The shift from LC3-I to LC3-II could only be detected 48 h after resveratrol stimulation (Fig. 3a). Raman spectra of the cells treated with the three different substances showed alterations in their spectral patterns (Fig. 3b). The staurosporine-treated SKMEL28 cells exhibited decreased relative intensities at the wave numbers of 938, 1004 and 1032/cm, and dramatically increased Raman peaks at 1065, 1157, 1303 and 1658/cm (Table 1) compared with untreated cells. The 3-BrPv treatment group exhibited a decrease in all these peaks. In contrast to cells treated with staurosporine or 3-BrPv, the resveratrol-treated cells had less pronounced spectral changes; however, the spectra exhibited a clearly decreased signal at 718/cm (Table 1), which is associated with phospholipids and was shown be a useful spectral marker of autophagic response.26 PCA resulted in a score plot separating apoptotic, necrotic and autophagic cells for PC1 and PC2. The clearest separation without overlapping areas was obtained for vital and necrotic cells. Apoptotic cells were also distinguished from vital cells, although there were some areas of overlap. The spectra of autophagic cells were almost identical to those of vital cells, which is in line with the slight changes detected in the mean spectra (Fig. 3c). In addition to using Raman spectroscopy, we photographed untreated and treated (staurosporine, 3-bromopyruvate or resveratrol) SKMEL28 and BLM cells after 6 h of incubation to analyze the morphological changes induced by the drugs (Figure S1). The morphological assessment of the melanoma cells corroborated the previous results; in both cell lines, staurosporine induced the typical apoptotic morphology, with shrinkage of the cells and fragmentation into membrane-bound apoptotic bodies. 3-BrPv induced a morphological phenotype resembling necrosis, with rounding of the cells, swelling of the cytoplasm and initiation of lysis in some cells. Upon incubation with resveratrol, the cells had an increased number of cytoplasmic vesicles, indicative of autophagosome formation (Figure S1).
Clinical and Experimental Dermatology (2014) 39, pp636–645
639
Raman spectroscopy in melanoma research E. Brauchle et al.
Table 1 Peak assignments.
Peak position, per cm Peak assignment
Reference(s)
623 645 669 684 718 729 751 781 789 830 833 853 900 934
Naumann44; Puppels et al.45 Naumann44; Puppels et al.45 Puppels et al.45; Overman et al.46 Puppels et al.45; Overman et al.46 Gniadecka et al.33 Puppels et al.45; Overman et al.46 Puppels et al.45; Overman et al.46 Notingher et al.47; Stone et al.48; Stuart et al.49 Puppels et al.45; Overman et al.46 Puppels et al.45 Puppels et al.45 Puppels et al.45 Puppels et al.45 Stone et al.48
937 1001 1032 1092 1094 1155 1176 1209 1240 1252 1298 1305 1336 1339 1375 1422 1446 1449 1460 1487 1520–1538 1578 1607 1615 1654 1659
Phe Tyr T, G (DNA) G (ring breathing) Phospholipid A (ring breathing) T (ring breathing) Cytocine/uracil (ring breathing) C, T backbone Tyr Ribose–phosphate Tyr Backbone C-C stretch of proline ring and valine and protein backbone/glycogen Skel. C-C stretch of proteins Phe Phe O-P-O stretch Backbone PO2 C-C (and C-N) stretch of proteins Tyr, Phe Tyr, Phe T, Amide III T, Amide III CH2 deformation mode of lipids A, CH deformation DNA purine bases A, CH deformation T, A, G A, G CH2 bending mode of proteins CH deformation CH deformation G, A C=C carotenoids G, A proteins Tyr, Phe C=C stretching mode for Tyr and Tryoptophan Amide I, C=C lipid stretch Amide I
Carter et al.50 Stone et al.48; Stuart et al.49; Yan et al.51 Puppels et al.45 Notingher et al.47; Stone et al.48 Puppels et al.45 Stone et al.48 Puppels et al.45 Puppels et al.45 Puppels et al.45 Puppels et al.45 Stone et al.48 Carter et al.50 Notingher et al.47 Puppels et al.45; Overman et al.46 Overman et al.46 Puppels et al.45 Stone et al.48 Naumann44 Naumann44 Puppels et al.45 Stone et al.48 Puppels et al.45 Naumann44 Stone et al.48 Stone et al.48; Yan et al.51 Carter et al.50
Raman spectroscopy and principal component analysis are able to predict the cell death mechanism
We next explored the applicability of Raman spectroscopy to detect the mechanism of cell death after treatment with the natural flavonoids 6-PN or 8-PN. SKMEL28 cells were incubated with 6-PN or 8-PN for 24 h and scanned. The resulting spectra were
640
Clinical and Experimental Dermatology (2014) 39, pp636–645
compared with spectra of vital, necrotic, autophagic and apoptotic SKMEL28 cells to screen for possible clustering. The resulting scores plot showed that the spectra of cells incubated with 6-PN or 8-PN were clearly distinct from those of apoptotic and necrotic cells, but they matched the cluster previously determined for autophagic cells (Fig. 4a). This was verified by immunoblot analysis showing the shift from LC3-I
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
(n = 15) yielded different cytotoxic sensitivities towards ascorbate, with half maximal effective concentration (EC50) values varying from 0.19 to 12 mmol/L (not shown). We then chose six different cell lines, three with high ascorbate sensitivity (EC50: MDA-MB-435, 0.19 mmol/L; SKMEL2, 1.4 mmol/L; and WM852, 1.5 mmol/L) and three with low ascorbate sensitivity (EC50: UACC-62, 5.5 mmol/L; MALME-3M, 6.9 mmol/L; and SKMEL28, 12 mmol/L) for Raman spectroscopy (Fig. 5a). The score plot displayed a separation of both groups, although there was some overlap between the spectra (Fig. 4b). PC4 accounted for 4% and PC3 for 10% of the total variance.
Discussion
(c)
Figure 2 Raman spectroscopy distinguishes melanoma cell lines
with BRAF or NRAS mutations from wild-type (WT) cells. (a) WT cells compared with BRAF-mutated cells (LOX-IMVI, 1205LU, 451LU). Principal component (PC)3 accounted for 9% and PC7 for 2% of the total variance in the Raman spectra. (b) WT cells (UACC-257, MEWO, MDA-MB-435) compared with NRAS-mutated cells (SKMEL2, BLM, WM852). PC1 accounted for 34% and PC3 for 7% of the total explained variance. (c) NRAS-mutated cells and BRAF-mutated cells clustered in distinct groups in the score plot; PC1 accounted for 38% and PC3 for 13% of the total explained variance.
to LC3-II after incubation with 6-PN and 8-PN for 24 h (Fig. 4b). Raman spectroscopy detects differences in sensitivity towards drug-induced melanoma cell death
Finally, we tested the efficacy of different pharmacological doses of ascorbate27,28 in killing melanoma cells. The human metastatic melanoma cell lines tested
ª 2014 British Association of Dermatologists
In this study, we were able to discriminate melanoma cell lines from melanocytes using Raman spectroscopy, a diagnostic tool used for cancer monitoring,29,30 screening of the molecular and structural composition of skin,31 and identifying differences between pathological and healthy skin tissue.32–34 In our study, specific Raman peaks and differences in Raman signal intensities identified biochemical differences between melanoma cell lines and melanocytes. We applied multivariate analyses to the spectral data to distinguish melanocytes from melanoma cells. Our data confirm recently published data describing a precise separation between HaCaT cells and melanocytes from their malignant counterparts.18 This technique also serves as diagnostic tool for skin cancer.35 In the latter study, malignant and pre-malignant lesions were distinguished from benign disorders, melanomas from benign pigmented skin lesions, and melanomas from pigmented seborrhoeic keratosis. A high percentage of melanocytic tumours harbour activating BRAF mutations. Approximately 70–80% of acquired melanocytic naevi and 40–50% of primary melanomas contain a BRAF mutation, the vast majority of which result in a single amino acid change at codon 600 (BRAFV600E).36 Although the RAS signalling pathway plays an important role in the homeostasis of normal cell turnover, death and survival, studies hae identified activating mutations in RAS family members in various malignancies.37 In melanoma, NRAS mutations, found in 10–25% of tumour samples, are an important oncogenic driver.38 In the current study, we successfully differentiated WT cells from NRAS-mutated melanoma cells, WT cells from BRAFmutated melanoma cells, and NRAS-mutated from BRAF-mutated melanoma cells. Therefore, we suggest
Clinical and Experimental Dermatology (2014) 39, pp636–645
641
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
(b)
(c)
Figure 3 Different mechanisms of cell death showed specific clusters after principal component (PCA) analysis. (a) SKMEL28 cells were
incubated with staurosporine, 3-bromopyruvate (3-BrPv) or resveratrol, and cell lysates were subjected to SDS-PAGE followed by immunoblotting. Apoptosis was verified by induction of cleaved caspase-3, necrosis was detected by cyclophilin A secretion into the medium, and autophagy was verified by the shift from LC3-I to LC3-II. (b) Melanoma cells (BLM, MEWO, SKMEL28) incubated with staurosporine, 3-BrPv or resveratrol were lysed with trypsin and used for Raman spectroscopy. Untreated cells served as controls. (c) PCA comparing these four groups showed that principal component (PC)1 accounted for 35% and PC2 for 20% of the total variance in the Raman spectra.
Raman spectroscopy as a possible alternative to the current methods of melanoma mutation analysis by sequencing or PCR. This also has clinical relevance, as the US Food and Drugs Authority and other authorities have approved the small molecule vemurafenib as targeted therapy for BRAF-mutated metastatic melanoma. In addition to targeted therapy with vemurafenib or immune therapy, with an anti-CTLA4 antibody (ipilimumab), (poly-) chemotherapy is used as second-line or third-line treatment options in metastatic melanoma. All clinically applied chemotherapeutics kill cancer cells by different modes of action. We therefore investigated whether Raman spectroscopy could distinguish the different types of cell death in melanoma cells. After apoptosis, necrosis or autophagy were induced by defined substances,39 we were able to identify unambiguously specific clusters for each type of cell death. The spectra of each group showed alterations in their peaks, which were highly specific for apoptosis, necrosis or autophagy.26 Morphological changes in the cells after treatment with the various drugs supported the induction of the relevant type of cell death (apoptosis, necrosis or autophagy). We then used the obtained data to test 6-PN and 8PN, which induce cell death in different cancer cell lines.40–42 Raman spectroscopy indicated that 6-PN and 8-PN did not induce either apoptosis nor necrosis, but did induce autophagy, as the PN-incubated cells
642
Clinical and Experimental Dermatology (2014) 39, pp636–645
(a)
(b)
Figure 4 Raman spectroscopy and principal components analysis
(PCA) were able to predict the cell death mechanism in melanoma cells. (a) SKMEL28 cells incubated with the cell death inducers staurosporine, 3-bromopyruvate (3-BrPv) and resveratrol or the flavonoids 6-prenylnaringenin (6-PN) or 8-PN were analysed by Raman spectroscopy. The spectra of untreated, staurosporine-treated and 3-BrPv-treated cells clustered in distinct groups. One additional cluster was detected for resveratrol and the PNs; PC1 accounted for 36% and PC2 for 25% of the total explained variance. (b) Immunoblot analysis demonstrated the shift from LC3-I to LC3-II, an indicator of autophagy.
clustered in the same area as that of autophagic cells. Immunoblot analysis confirmed that the shift from LC3-I to LC3-II was induced by 6-PN or 8-PN.
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
(a)
melanocytes and melanoma cells and that it can be used to determine the cellular mutation status, to analyze the specific type of cell death in melanoma cells, to predict the susceptibility of melanoma cells to anticancer drugs.
Acknowledgements
(b)
This project was supported by grants from the Deutsche Forschungsgemeinschaft SFB 773: ‘Understanding and overcoming drug resistance of solid tumors’ (to CG and CB), the Wissenschaftsf€ orderung der Deutschen Brauwirtschaft eV (to CB), the Fraunhofer-Gesellschaft Internal programm (to KSL), and the Ministry of Science, Research and the Arts of Baden-W€ urttemberg (33-729.55-3/214, to KSL).
What’s already known about this topic? ● Raman spectroscopy is a noninvasive diagnos-
tic tool that can distinguish between healthy and malignant cells.
Figure 5 Raman spectroscopy detect differences in sensitivity
towards drug-induced melanoma cell death. (a) Three untreated melanoma cell lines with high ascorbate sensitivity (EC50: MDAMB-435, 0.19 mmol/L; SKMEL2, 1.4 mmol/L; and WM852, 1.5 mmol/L) and three with low ascorbate sensitivity (EC50: UACC-62, 5.5 mmol/L; MALME-3M, 6.9 mmol/L SKMEL28, 12 mmol/L) were analyzed by Raman spectroscopy. (b) The score plot displayed a separation of both groups, although the spectra of both groups did overlap. PC4 accounted for 4% and PC3 for 10% of the total variance.
Increasing evidence supports the hypothesis of Pauling and Cameron43 that ascorbate selectively kills cancer cells by a pro-oxidant mode of action.27,28 Indeed, we were able to induce cell death in several melanoma cell lines with pharmacological doses of ascorbate; however, we detected large differences in ascorbate sensitivity. Using Raman spectroscopy, we generated different clusters for highly sensitive and less sensitive melanoma cells, suggesting that the applied technique can allow prediction of the cytotoxic susceptibility of melanoma cells towards a particular drug.
Conclusion Altogether, our results show that Raman spectroscopy is a powerful noninvasive tool for distinction between
ª 2014 British Association of Dermatologists
What does this study add? ● The technique is an analytical approach for
the determination of the type of cell death in melanoma cells. ● Raman spectroscopy provides an optical readout for the susceptibility of melanoma cells to anticancer drugs.
References 1 Bennett DC. How to make a melanoma: what do we know of the primary clonal events? Pigment Cell Melanoma Res 2008; 21: 27–38. 2 Shintani T, Klionsky DJ. Autophagy in health and disease: a double-edged sword. Science 2004; 306: 990–5. 3 Ocker M, Hopfner M. Apoptosis-modulating drugs for improved cancer therapy. Eur Surg Res 2012; 48: 111–20. 4 Degterev A, Huang Z, Boyce M et al. Chemical inhibitor of nonapoptotic cell death with therapeutic potential for ischemic brain injury. Nat Chem Biol 2005; 1: 112–19. 5 Proskuryakov SY, Gabai VL. Mechanisms of tumor cell necrosis. Curr Pharm Des 2010; 16: 56–68.
Clinical and Experimental Dermatology (2014) 39, pp636–645
643
Raman spectroscopy in melanoma research E. Brauchle et al.
6 Christofferson DE, Yuan J. Cyclophilin A release as a biomarker of necrotic cell death. Cell Death Differ 2010; 17: 1942–3. 7 Hersey P, Zhang XD. How melanoma cells evade trail-induced apoptosis. Nat Rev Cancer 2001; 1: 142–50. 8 Kim I, Rodriguez-Enriquez S, Lemasters JJ. Selective degradation of mitochondria by mitophagy. Arch Biochem Biophys 2007; 462: 245–53. 9 Levine B, Klionsky DJ. Development by self-digestion: molecular mechanisms and biological functions of autophagy. Dev Cell 2004; 6: 463–77. 10 Eskelinen EL. Maturation of autophagic vacuoles in mammalian cells. Autophagy 2005; 1: 1–10. 11 Levine B. Cell biology: autophagy and cancer. Nature 2007; 446: 745–7. 12 Jaeschke H, Gujral JS, Bajt ML. Apoptosis and necrosis in liver disease. Liver Int 2004; 24: 85–9. 13 Kanduc D, Mittelman A, Serpico R et al. Cell death: apoptosis versus necrosis. Int J Oncol 2002; 21: 165–70. 14 Brauchle E, Schenke-Layland K. Raman spectroscopy in biomedicine – non-invasive in vitro analysis of cells and extracellular matrix components in tissues. Biotechnol J 2013; 8: 288–97. 15 Raman C, Krishna CM. The Raman effect – inelastic insight. Nature 1928; 121: 501–2. 16 Tu Q, Chang C. Diagnostic applications of Raman spectroscopy. Nanomedicine 2012; 8: 545–58. 17 Dochow S, Krafft C, Neugebauer U et al. Tumour cell identification by means of Raman spectroscopy in combination with optical traps and microfluidic environments. Lab Chip 2011; 11: 1484–90. 18 Wang H, Tsai TH, Zhao J et al. Differentiation of HaCaT cell and melanocyte from their malignant counterparts using micro-Raman spectroscopy guided by confocal imaging. Photodermatol Photoimmunol Photomed 2012; 28: 147–52. 19 Chan JW, Taylor DS, Zwerdling T et al. Micro-Raman spectroscopy detects individual neoplastic and normal hematopoietic cells. Biophys J 2006; 90: 648–56. 20 Pyrgiotakis G, Kundakcioglu OE, Finton K et al. Cell death discrimination with Raman spectroscopy and support vector machines. Ann Biomed Eng 2009; 37: 1464–73. 21 Flaherty KT, Puzanov I, Kim KB et al. Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med 2010; 363: 809–19. 22 Sondergaard JN, Nazarian R, Wang Q et al. Differential sensitivity of melanoma cell lines with BRAFV600E mutation to the specific Raf inhibitor PLX4032. J Transl Med 2010; 8: 39. 23 Pudlas M, Berrio DAC, Votteler M et al. Non-contact discrimination of human bone marrow-derived mesenchymal stem cels and fibroblasts using Raman spectroscopy. Med Laser Appl 2011; 26: 119–25. 24 Pudlas M, Brauchle E, Klein TJ et al. Non-invasive identification of proteoglycans and chondrocyte
644
Clinical and Experimental Dermatology (2014) 39, pp636–645
25
26
27
28
29
30
31
32
33
34
35
36 37 38
39
differentiation state by Raman microspectroscopy. J Biophotonics 2013; 6: 205–11. Qin JZ, Xin H, Nickoloff BJ. 3-Bromopyruvate induces necrotic cell death in sensitive melanoma cell lines. Biochem Biophys Res Commun 2010; 396: 495–500. Konorov SO, Jardon MA, Piret JM et al. Raman microspectroscopy of live cells under autophagy-inducing conditions. Analyst 2012; 137: 4662–8. Chen Q, Espey MG, Krishna MC et al. Pharmacologic ascorbic acid concentrations selectively kill cancer cells: action as a pro-drug to deliver hydrogen peroxide to tissues. Proc Natl Acad Sci USA 2005; 102: 13604–9. Sinnberg T, Noor S, Venturelli S et al. The ROS-induced cytotoxicity of ascorbate is attenuated by hypoxia and HIF-1alpha in the NCI60 cancer cell lines. J Cell Mol Med 2014; 18: 530–41. Larraona-Puy M, Ghita A, Zoladek A et al. Development of Raman microspectroscopy for automated detection and imaging of basal cell carcinoma. J Biomed Opt 2009; 14: 054031. Lyng FM, Faolain EO, Conroy J et al. Vibrational spectroscopy for cervical cancer pathology, from biochemical analysis to diagnostic tool. Exp Mol Pathol 2007; 82: 121–9. Bernard G, Auger M, Soucy J et al. Physical characterization of the stratum corneum of an in vitro psoriatic skin model by ATR-FTIR and Raman spectroscopies. Biochim Biophys Acta 2007; 1770: 1317–23. El GA, Commandeur S, Rietveld MH et al. Replacement of animal-derived collagen matrix by human fibroblast-derived dermal matrix for human skin equivalent products. Biomaterials 2009; 30: 71–8. Gniadecka M, Wulf HC, Nielsen OF et al. Distinctive molecular abnormalities in benign and malignant skin lesions: studies by Raman spectroscopy. Photochem Photobiol 1997; 66: 418–23. Nijssen A, Bakker Schut TC, Heule F et al. Discriminating basal cell carcinoma from its surrounding tissue by Raman spectroscopy. J Invest Dermatol 2002; 119: 64–9. Lui H, Zhao J, McLean D et al. Real-time Raman spectroscopy for in vivo skin cancer diagnosis. Cancer Res 2012; 72: 2491–500. Pollock PM, Harper UL, Hansen KS et al. High frequency of BRAF mutations in nevi. Nat Genet 2003; 33: 19–20. Bos JL. ras oncogenes in human cancer: a review. Cancer Res 1989; 49: 4682–9. Platz A, Ringborg U, Brahme EM et al. Melanoma metastases from patients with hereditary cutaneous malignant melanoma contain a high frequency of N-ras activating mutations. Melanoma Res 1994; 4: 169–77. Morselli E, Marino G, Bennetzen MV et al. Spemidine and resveratrol induce autophagy by distinct pathways converging on the acetylproteome. J Cell Biol 2011; 192: 615–29.
ª 2014 British Association of Dermatologists
Raman spectroscopy in melanoma research E. Brauchle et al.
40 Brunelli E, Pinton G, Bellini P et al. Flavonoid-induced autophagy in hormone sensitive breast cancer cells. Fitoterapia 2009; 80: 327–32. 41 Brunelli E, Pinton G, Chianale F et al. 8-Prenylnaringenin inhibits epidermal growth factor-induced MCF-7 breast cancer cell proliferation by targeting phosphatidylinositol-3-OH kinase activity. J Steroid Biochem Mol Biol 2009; 113: 163–70. 42 Delmulle L, Vanden Berghe T, Keukeleire DD et al. Treatment of PC-3 and DU145 prostate cancer cells by prenylflavonoids from hop (Humulus lupulus L.) induces a caspase-independent form of cell death. Phytother Res 2008; 22: 197–203. 43 Cameron E, Pauling L. Supplemental ascorbate in the supportive treatment of cancer: prolongation of survival times in terminal human cancer. Proc Natl Acad Sci USA 1976; 73: 3685–9. 44 Naumann D. FT-Infrared and FT-Raman spectroscopy in biomedical research. Appl Spectrosc Rev 2001; 36: 239–98. 45 Puppels GJ, Garritsen HS, Segers-Nolten GM et al. Raman microspectroscopic approach to the study of human granulocytes. Biophys J 1991; 60: 1046–56. 46 Overman SA, Aubrey KL, Reilly KE et al. Conformation and interactions of the packaged double-stranded DNA genome of bacteriophage T7. Biospectroscopy 1998; 4: S47–56.
ª 2014 British Association of Dermatologists
47 Notingher I, Verrier S, Hague S et al. Spectroscopic study of human lung epithelial cells (A549) in culture, living cells versus dead cells. Biopolymers 2003; 72: 230–40. 48 Stone N, Kendall C, Smith J et al. Raman spectroscopy for identification of epithelial cancers. Faraday Discuss 2004; 126: 141–57. 49 Stuart NS, McIIImurray MB, Bishop JL et al. Vinorelbine and infusional 5-fluorouracil in the anthracycline and taxane pre-treated metastatic breast cancer. Clin Oncol (R Coll Radiol) 2008; 20: 152–6. 50 Carter EA, Edwards HGM. Infrared and Raman spectroscopy of biological materials. J Raman Spectrosc 2004; 35: 754–60. 51 Yan XL, Dong RX, Zhang L et al. Raman spectra of single cell from gastrointestinal cancer patients. World J Gastroenterol 2005; 11: 3290–2.
Supporting Information Additional Supporting Information may be found in the online version of this article: Figure S1 (a) SKMEL28 and (b) BLM cells were either left untreated or were subjected to treatment with either staurosporine, 3-bromopyruvate or resveratrol, and photographed after 6 h of treatment to analyze the morphological changes.
Clinical and Experimental Dermatology (2014) 39, pp636–645
645
