Exosome Classification by Pattern Analysis of ... - ACS Publications

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Exosome Classification by Pattern Analysis of Surfaceenhanced Raman Spectroscopy Data for Lung Cancer Diagnosis Jaena Park, Miyeon Hwang, ByeongHyeon Choi, Hyesun Jeong, Jikhan Jung, Hyun Koo Kim, Sunghoi Hong, Ji Ho Park, and Yeonho Choi Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 25 May 2017 Downloaded from http://pubs.acs.org on May 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Exosome Classification by Pattern Analysis of Surface-enhanced Raman Spectroscopy Data for Lung Cancer Diagnosis Jaena Park1, Miyeon Hwang2, ByeongHyeon Choi3, Hyesun Jeong4, Jik-han Jung5, Hyun Koo Kim3, Sunghoi Hong4, Ji-ho Park5, Yeonho Choi1,2* 1

Department of Bio-convergence Engineering, Korea University, Seoul, 02841, South Korea

2

School of Bio-medical Engineering, Korea University, Seoul, 02841, South Korea

3

Department of Thoracic and Cardiovascular Surgery, College of Medicine, Korea University Guro Hospital, Seoul, 08308, South Korea 4

School of Biosystem and Biomedical Science, Korea University, Seoul, 02841, South Korea

5

Department of Bio and Brain Bioengineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon, 34141, South Korea

ABSTRACT: Owing to the role of exosome as a cargo for intercellular communication, especially in cancer metastasis, the evidence has been consistently accumulated that exosomes can be used as a non-invasive indicator of cancer. Consequently, several studies applying exosome have been proposed for cancer diagnostic methods such as ELISA assay. However, it has been still challenging to get reliable results due to the requirement of a labeling process and high concentration of exosome. Here, we demonstrate a label-free and highly sensitive classification method of exosome by combining surface enhanced Raman scattering (SERS) and statistical pattern analysis. Unlike the conventional method to read different peak positions and amplitudes of spectrum, whole SERS spectra of exosomes were analyzed by principal component analysis (PCA). By employing this pattern analysis, lung cancer cell derived exosomes were clearly distinguished from normal cell derived exosomes by 95.3% sensitivity and 97.3% specificity. Moreover, by analyzing the PCA result, we could suggest that this difference was induced by 11 different points in SERS signals from lung cancer cell derived exosomes. This result paved the way of new real-time diagnosis and classification of lung cancer by using exosome as a cancer marker.

INTRODUCTION Since the potential role of exosomes in cancer metastasis was demonstrated, studies have accrued evidence of exosomes as cancer markers.1-6 Exosomes are 30~150-nmsized vesicles that are shed from every type of cell.7, 8 Exosomes perform functions related to cell-cell interactions, induction of immune regulation and pathogen transmission by carrying proteins and RNAs to the recipient cells.9-11 Because of the abundance of exosomes found in cancer patients, many studies have been conducting on the role of exosomes in cancer.12-14 The molecular characterization of cancer-derived exosomes may offer a strategy for the noninvasive monitoring of cancer genotypes.15 The use of exosomes that are found in urine or saliva as a cancer marker raises the possibility of cancer detection without biopsy, which will be beneficial for cancer diagnosis and further studies.16, 17 For these purposes, exosome tracking and detection are important.18 However, because of its small size and close connection to cell generation, it is challenging to detect exosomes using conventional detection methods.7 The standard method of

detecting exosomes, enzyme-linked immunosorbent assays (ELISA), requires large amounts of highly concentrated sample.19 Some breakthrough studies have investigated several ways to detect and profile exosomes.20-22 Raman spectroscopy is a method that analyzes vibrational modes of the sample by measuring the nonelastic scattering effect caused by a radiating laser.23 However, because Raman scattering is a rare event, the signal intensity is very low and difficult to distinguish. To overcome this problem, surface-enhanced Raman spectroscopy (SERS), which enhances the electric field by a surface treatment (e.g., applying gold nanoparticles to amplify the Raman signals for better analysis), has been developed.24-26 SERS is capable of depicting the vibrational modes of molecules with high sensitivity.27, 28 The labelfree, non-destructive and non-invasive characteristics of SERS enabled its application to biomedical systems for biosensing, therapeutics and metabolites.29, 30 Although SERS analysis has been performed on numerous systems ranging in scale from a molecule to cell lines, few studies

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

have focused on its application to exosomes31-34. This is because the Raman spectra of heterogeneous biological particles, such as exosomes, show complex and nonconforming data. Every exosome, even those that are generated from the same cell line, have non-uniform protein and lipid components, making each Raman spectrum different. Therefore, classifying these heterogeneous Raman spectra based on specific Raman peaks is arduous. However, since exosomes from the same mother cell line would exhibit common patterns in their Raman spectra, principal component analysis (PCA) is a convenient way to reduce data and obtain meaningful patterns.35 PCA generates score and loading plots that maximize the covariance of the spectral data. In other words, PCA transforms the coordinate space to maximize the source separation, resulting in exosome classification; thus, essential common patterns can be uncovered by analyzing which components are responsible for each separate spectrum.36 Here, we present a simple and accurate exosome detection method based on SERS and a statistical method. We show that the Raman signals of lung cancer cellderived exosomes and normal alveolar cell-derived exosomes are well distinguished through PCA. Figure 1 shows a schematic diagram of our system. The selective intracellular budding of vesicles in a cell as a result of endocytosis leads to multivesicular endosomes. When these multivesicular endosomes fuse with the cell membrane, exosomes, which formed from the vesicles in the multivesicular endosomes, are secreted outside the cell. Because of their membranous origin, we suppose that the lipids and membrane proteins and inner components will show unique Raman interactions. However, since exosome randomly lies on the substrate, SERS signals, which are amplified more when are closer to substrate, differ in some peaks, even from the exosomes of the same origin. Therefore it is hard to distinguish those SERS signals by conventional method; the existence of unique Raman peak. Stremersch et al tried to solve this problem with multivariate curve resolution alternating least squares, grouping the Raman spectra to several distinctive groups.34 However, samples having similar composition will show comparable SERS signal shape (Figure Figure S1). S1 Therefore, if the Raman signal of cancerous exosome and noncancerous exosome have different shape, using PCA would be an appropriate way of analyzing whole signal. PCA will divide Raman spectra of cell-derived exosomes into lung cancer derived group or normal cell derived group.

Page 2 of 10

alveolar epithelial cell medium supplemented with 2% FBS, epithelial cell growth supplement and penicillin/stretomycin. All cell lines were cultured at 37.8°C and 5% CO2. Exosome Isolation: 50 ml of cell culture medium were centrifuged at 500 × g for 10 minutes at 4°C to remove detached cells. The supernatants were centrifuged at 5000 × g for 30 minutes at 4°C to remove cell debris and then at 10,000 × g for 30 minutes at 4°C to remove further cell debris and microvesicles. The supernatants were collected and concentrated using an Amicon centrifugal filter (Merck Millipore) with a membrane nominal molecular weight limit of 100,000 Da. 500 μl of concentrated media were loaded in a column (Disposable columns, 10ml, Thermoscientific Pierce) made of porous Sepharose beads (Sepharose CL-2B, GE Healthcare Life Science) with a molecular weight separation range of 70 × 103 to 40 × 106 and was distributed in accordance with the commercial protocol.37 Exosomes were isolated by selecting portions that coincide with the 30~150-nm size measured by dynamic light scattering (DLS) and nanoparticle tracking analysis (NTA), and their concentrations were determined (Figure Figure S2). S2 Every post-processing experiment using exosomes was performed immediately after their isolation because of the risk of degradation.36 TEM: All eluted fractions containing exosomes were subject to the following procedures. Briefly, the solution was mixed with an equal volume of 4% paraformaldehyde (PFA). Then, 5 μl of the mixed solution was deposited on formvar-carbon coated electron-microscope grids, and the membranes were left to absorb it for 15 minutes. The grid was transferred to a PBS droplet for washing and a drop of 1% glutaraldehyde was added for 1 min. For washing, the grids were applied to a drop of distilled water for a total of eight times. For contrasting samples, the grids were transferred to a drop of phosphotungstic acid solution, pH 7, for 5 min, and the remaining solution was removed using filter paper. After drying, the grids were observed under a transmission electron microscope at 200 kV. SERS Experiment: We used the SERS substrate previously shown by Lim et al.32 Briefly, cover glasses were cleaned with piranha solution (3:1 of H2SO4 and H2O2), washed with deionized water, and dried with N2 gas. 50 μl of 80 nm spherical gold nanoparticles (GNPs, BBI Solutions) were then mixed with 1 μl of 10 mM CuSO4 and dried over the cover glass. After, exosomes (50 μl, 109 particles/ml) from lung cancer cells (H1299 and H522) and normal (alveolar) cells were placed on the substrate for SERS and dried. The average number of exosomes in a focused laser spot is about 78 per shot (Eq. Eq. S1). S1 10-mW 785-nm CW laser (MDL-III-785, UltraLasers) irradiation was applied to the substrate on the coffee-stained region and was coupled through an inverted microscope (Axiovert, Carl Zeiss) with a 50× objective lens (NA=0.70). The SERS spectrum was collected with a spectrometer (SP2300, PI Ac-ton) at a 10second acquisition time. Our system was connected to a charge-coupled device (CCD) detector cooled to െ70 °C.

EXPERIMENTAL SECTION

Cell Culture: Fetal bovine serum(FBS, HyClone) used in this experiment was previously depleted of exosomes by centrifugation with 10,000 x g. H1299 and H522 cell line (Human lung carcinoma) were cultured in RPMI 1640 medium(HyClone) with 5% FBS, 1% penicillin/streptomycin (Gibco). Human pulmonary alveolar epithelial cells (ScienCell) were cultured in

2


Page 3 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


PCA: The baseline of each spectrum was cut by BEADS (Figure Figure S3). S3 Then, the spectra were normalized by setting the total area of the signal to 1. These processes were performed by MATLAB (Figures Figures S4). For each spectrum, 1030 variables were selected from the Raman shifts from 474 to 1800 cm-1. PCA of these variables was then performed using built-in function of MATLAB. The first and second principal components were taken for further data analysis. Ellipse indicating the boundary of lung cancer cell derived exosome is made by making 95% confidence ellipse of H1299 and H522 cell derived exosomes. Ellipse indicating the boundary of normal cell derived exosome is made by making 95% confidence ellipse of normal alveolar cell derived exosomes. Sensitivity and specificity is calculated equation 1 1. Sensitivity =

୘୔ ୘୔ା୊୔

, Specificity =

୘୒ ୘୒ା୊୒

spectrum showed only partial combinations of the peaks. This is because only the part of exosome which is amplified by the substrate is majorly shown in the Raman spectrum. Therefore, to filter out influential Raman peaks from the massive SERS data sets, adequate data reduction before analysis or analysis of the whole spectra are necessary.30 PCA, in this case, is an appropriate tool for analyzing SERS data.38-41 Therefore, we introduced this system to analyze this Raman spectrum in detail. To further analyze the Raman spectra of each set of exosomes, we used PCA for classification (Figure Figure 4). 4 We tried to distinguish exosome groups into 2 distinctive group (non-small cell lung cancer (NSCLC) to normal cell derived exosomes) by PCA. Because all of the variables have the same unit, no further standardization was performed. Loading and score plots were generated according to the derived principal components as PCA was applied to the preprocessed data. The correlation coefficients between the scores of the principal components and the original data are shown as the loadings of the principal components. The loadings of the principal components characterize the contribution of each variable. Thus, the loading plot shows the Raman wavenumbers related to the variation that differentiates signals into various groups.30 Therefore, PCA allows us to reduce our data to fewer dimensions and to distinguish between the two exosome groups (NSCLC and normal alveolar cell-derived exosomes). Figure 4a shows the PCA score plot. We chose the first and second principal components, which account for 26% and 21% of the variance, respectively. Each dot in the score plot represents one Raman spectrum. The red and yellow stars refer to the H1299- and H522derived exosomes, respectively. The blue circles show alveolar cell-derived exosomes, and the square dots indicate control, which is the base buffer. As shown in the graph, the H1299-derived exosomes lie in the fourth quadrant and the H522-derived exosomes lie in the first quadrant, and the alveolar cell-derived exosomes lie in the second quadrant. The red, blue and black ellipses indicate the 95% confidence ellipses of non-small cell lung cancer (NSCLC)-derived exosomes, alveolar cell-derived exosomes and control, respectively. When PCA was used for the Raman spectrographs, the NSCLC cell-derived exosomes were well distinguished from the normal cellderived exosomes. We took 5 samples for each study group and performed 100-fold cross validation, resulting in 95.3% sensitivity and 97.3% specificity for the 95% confidence interval. We also have done PCA with Raman data after 2nd derivative transformation. (Figure Figure S6) S6 New score plot showed similar result compared to the original data that lung cancer cell derived exosomes and normal cell derived exosomes are differentiated by their positions on the plot. However, in this case, complete differentiation between lung cancer cells and normal cells were difficult because the confidence ellipse of two groups overlap each other. Also, because of the smoothing process, loss of data may have occurred. Therefore, for two reasons; the result of

(Eq. 1)

where TP is the number of dots from cancer cell derived exosomes plotted inside the ellipse boundary of lung cancer cell derived exosome, FP is the number of dots from normal cell derived exosomes plotted inside the ellipse boundary of lung cancer cell derived exosome. TN is the number of dots from normal cell derived exosomes plotted inside the ellipse boundary of normal cell derived exosomes. FN is the number of dots from cancer cell derived exosomes plotted inside the ellipse boundary of normal cell derived exosome. RESULTS AND DISCUSSION Our group used column chromatography to isolate exosomes from lung cancer cells and alveolar cells. Nanoparticle tracking analysis (NTA) showed that the exosomes isolated by the above protocol have a mode size of 138 ± 6.5 nm (Figure Figure 2a). 2a Figure 2b shows a transmission electron microscopy (TEM) image of an H522 cell-derived exosome. According to TEM, the exosome size was about 100 nm, which is comparable to the size measured by NTA but slightly smaller because the exosome shrinks and forms a cup shape upon drying. CD63 and CD9 are known general exosome markers. Figure 2c shows that every exosome solution isolated by column chromatography contains CD63 and CD9. For each sample, 8 μg were used for the western blotting of CD63 and CD9, and the alveolar cell-derived exosomes showed lower expression of CD63 and CD9 compared with the H1299 and H522 lung cancer cell-derived exosomes. Thirty-seven samples of H1299-derived exosomes, 34 samples of H522-derived exosomes, 23 samples of alveolar cell-derived exosomes and 15 samples of control (PBS on substrate) were used for the experiment. Figure 3 shows a typical Raman spectra of H522-, H1299- and alveolar cellderived exosomes and of PBS dried on a gold aggregate substrate. Each set of Raman signals showed different patterns throughout the Raman shifts from 470 to 1800 Figure S5). cm-1 (Figure S5 Although we identified the most commonly repeated peaks of each group, not every sample contained all of these specific peaks at the same time; each

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

PCA data of raw data is reasonable and data loss caused by fitting data, we preferred to use PCA with raw data in this experiment. Figure 4b shows the variable loading plot of PCA. Each dot of the loading plot indicates one variable that is a unique Raman wavenumber. There is a clear difference between the PC1 (principal component 1) and PC2 (principal component 2) loading spectra (Figure Figure S7). The score plot and loading plot are interrelated; that is, the dots of the loading plot that are in the direction of a dot on a score plot are strongly correlated to that dot on the score plot. According to the score plot of Figure 4a, the cancer cell derived exosomes locate on the positive side of x-axis which is PC1 and non-cancer cell derived exosomes locate on the negative side. Therefore, high value in PC1 indicates NSCLC derived exosomes. On the same principle, low value in PC1 with high value in PC2 designates alveolar cell derived exosomes. Figure 4c shows the points that are related to the NSCLC- and alveolar cell-derived exosomes; this plot was prepared by assigning the variables where PC1 is greater than 0.02 to NSCLC and assigning the points where PC1 is smaller than -0.03 and PC2 is greater than 0.02 to the normal group (Figure Figure 4b). 4b The points related to NSCLC are colored as red and orange, and the points related to the alveolar cell-derived exosomes are colored in blue (Figure Figure 4c). The 4c wavenumbers of the peaks of PC1 and PC2 are processed to indicate the specific peaks of each group of exosome, as shown below. The NSCLC cell-derived exosomes present specific peaks at 570, 602, 732, 781, 838, 843, 869, 911, 1072, 1445, and 1539 cm-1. The alveolar cell-derived exosomes have their specific Raman shifts at 487, 537, 700, 1150, 1214, and 1630 cm-1 (Figure Figure 4c). 4c The points associated with each type of exosome are organized in the table of Figure S8. For example, PCA indicated that 732 cm-1 is highly related to the NSCLC-derived exosomes, and this correlation was confirmed by analyzing the intensity of all the Raman spectra. The mean and standard deviation of the Raman intensity at 732 cm-1 were analyzed by 37, 34, 23, and 15 normalized spectra of H1299, H522, Alveolar derived exosomes and control, respectively (same samples used for PCA). Intensity of 732 cm-1 was determined to be high for the NSCLC-derived exosomes and relatively low for the alveolar cell-derived exosomes. The same procedure was performed for 1150 cm-1, which was shown to be associated with normal alveolar cell-derived exosomes; specifically, the alveolar cell-derived exosomes showed a higher Raman intensity compared to that of the NSCLC cell-derived exosomes and the control (Figure Figure S9). According to the band assignments of acquired Raman shifts (Table Table S1 and S2) S2), it is expected that this difference of Raman intensity between 732 and 1150 cm-1 comes from the different composition of nucleic acids. This has a thread of connection with the conventional report that exosome contains DNA which could work as a biomarker in cancer detection.42 On the same principle, as we see in 1445 and 1630 cm-1, membrane proteins seems to differ between those two groups and we are under investigating

Page 4 of 10

these differences by matching proteomics data with Raman peaks. We showed the experiment with the control to verify that the achieved Raman spectra are from exosomes. Because every exosome is depleted in PBS, the spots where the exosome concentrations are low would be similar to the PBS Raman spectra. In that situation, the point of the score plot in Figure 4a will be located inside the black ellipse which indicates Raman spectrum of PBS on substrate. However, our experiment was capable of differentiating PBS from exosomes. Since Raman spectrum shows the data of exosome and most of them are in the boundary made by PCA, most of the position that exosome make seems not to have hindering effects on cancer detection. This means that the part of exosome on SERS active area possess proteins or nucleic acid which distinguish lung cancer from normal cell derived exosome. Therefore, we could conclude that exosome is composed of distinguishable protein or nucleic acid which differs from their cancerous origin and is distinguishable by SERS. Our detection method was capable of detecting exosomes at small concentrations of 109 particles/ml. Benchtop chemiluminescnece ELISA needs about 100-fold higher concentration than our method.20, 43 Also, since SERS is a label-free detection method, it doesn’t need specific sample preparation leading to a shorter experimental time. We also tried to test our PCA ellipse as an indicator of cancer detection to unknown samples (Figure Figure S10). 10 We projected newly achieved Raman spectra of H1299 and H522 derived exosomes to the PCA score plot in Figure 4a by multiplying loadings to each variables and saw whether the projected dots locate inside the ellipse which indicates the NSCLC derived exosomes. The result showed 85.7% and 90.0% sensitivity for H522 and H1299 cell line derived exosomes with 21 and 20 Raman spectra respectively. Therefore, we insist to use this projection as a potential cancer detection method. Moreover, we applied this method to clinical samples (Figure Figure S11 S11). 2 people with lung cancer and 2 people of healthy control group were tested. Since we only have a small number of samples, we did this experiment as a preliminary experiment. Unlike the experiment done with cell line derived exosomes (Figure Figure S10) 10), most of the projected dots of clinical samples (Figure Figure S11 S11) did not lie inside the ellipse. On the contrary, they lied in between the ellipse indicating lung cancer and normal cell. We suppose this phenomenon as a result of the diverse origin of exosomes and different proportion of exosomes from cancer and healthy cells. Since human blood derived exosomes are comprised of various origins, unlike pure exosomes made from one cell line, exosomes derived by blood mean they are made up of hundreds of cell line origin including cells which are not from lung cells. Exosomes derived from lung cancer patient would possess both exosomes from cancer cells and from normal cells. However, lung cancer patient would possess more cancer cell derived exosomes compared to the healthy controls.

4


Page 5 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry proteins result in specific Raman spectra; thus, the NSCLCderived exosomes and alveolar cell-derived exosomes vary in their Raman spectral patterns. We examined whether NSCLC-derived exosomes are distinguishable from normal cell-derived exosomes by using two NSCLC cell lines (H1299 and H522) and one normal cell line (alveolar). The original cell lines were distinguished and specific Raman patterns were identified using PCA. This method led to 95.3% sensitivity and 97.3% specificity for the 95% confidence interval for NSCLC. Our experiment showed the successful segregation of NSCLC-derived exosomes from normal alveolar cell-derived exosomes using the noninvasive method of SERS accompanied with PCA. Moreover, we prepared a library of H1299-, H522- and alveolar cell-derived exosome Raman spectra. Because exosomes originate from their mother cells, exosome classification could lead to cell classification. We are currently classifying the Raman spectra of exosomes derived from different cell lines, and we are also comparing proteomics data with our Raman spectral modes. The goal is to diagnose and stage cancer by obtaining exosome profiles. By looking to loading, we could point out wavenumbers related to differentiating exosomes of cancerous cell and normal cells.

This tendency is shown in Figure S11. Truly, projected dots of cancer patient derived exosomes lie closer to the red ellipse indicating lung cancer origin than the projected dots of healthy control derived exosomes. This position would be related to the sample proportion between lung cancer cell derived exosome and normal cell derived exosome. And further study of finding out the proportion of them by SERS is under investigation. Our objective of this article is to find Raman signals of exosomes to distinguish the cancerous cell lines from the non-cancerous cell lines. However, future research is necessary to test this technique with heterogeneous exosomes at once and this subsequent study is under investigation; aside with this experiment, we are trying to distinguish properties of exosomes by machine learning. Also because we had done this with only few clinical samples, we are also trying to study with more number of samples. CONCLUSION In summary, we established a method for detecting cancer cell-derived exosomes using SERS, with the aim of real-time diagnosis of lung cancer. Lipid and membrane

5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 10

Figure 1. Schematic diagram of lung cancer diagnosis by SERS classification of exosome. (a, b) Lung cancer cell and normal cell releases exosomes to the extracellular environment having their own profiles by fusing multivesicular endosomes to plasma membrane respectively. (c, d) Raman spectrum of lung cancer cell and normal cell derived exosomes was achieved by SERS respectively. (e) SERS spectra achieved by (c) and (d) are shown. Red lines indicate specific peaks of lung cancer derived exosomes. (f) Exosome classification is done by PCA of SERS spectra. Identifying the origin of exosome is done by checking the location of the plotted x-shaped dot. Dot plotted inside of the red circle indicates lung cancer cell derived exosome and dot plotted inside of the blue circle indicates normal cell derived exosome.

6


Page 7 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry Figure 2. Characterization of purified exosomes by column chromatography. Exosomes isolated by column chromatography were characterized by its size and expression of CD9 and CD63. (a) NTA result of alveolar cell derived exosome showed 138 ± 6.5 nm size of exosomes in mode. (b) TEM displays morphology and size of exosomes which are negatively stained. (c) CD9 and CD63 expression of Alveolar, H1299 and H522 cell derived exosomes are shown.

Figure 3. Types of SERS for examining lung cancer cell derived exosomes. Four different types of SERS were acquired from control, Alveolar, H522 and H1299. Control was measured by SERS of dried PBS on GNPs.

7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 10

Figure 4. Principal component analysis (PCA) of exosome data sets. (a) Principal component scatter plot with colored clusters of control, Alveolar, H522 and H1299 derived exosome. Black square, blue circle, orange star and red star indicates control, Alveolar, H522 and H1299 derived exosomes respectively. Lung cancer cell derived exosomes were distinguishable with 95.3% sensitivity and 97.3% specificity in the confidence interval of 95%. 37 samples of H1299, 34 samples of H522, 23 samples of Alveolar and 13 samples of control Raman spectra were took for PCA. (b) Principal components of PCA result in Figure 4a. Red area indicates the Raman shifts related with NSCLC derived exosomes and blue region indicates the Raman shifts related with alveolar cell derived exosomes. (c) Raman peaks related with each cell achieved by PCA are shown on the graph with colored lines.

8


Page 9 of 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry (7) Lobb, R. J.; Becker, M.; Wen, S. W.; Wong, C. S.; Wiegmans, A. P.; Leimgruber, A.; Möller, A. J. Extracell. Vesicles 2015, 4. (8) Thery, C.; Ostrowski, M.; Segura, E. Nat. Rev. Immunol 2009, 9, 581-93. (9) Maguire, G., Chapter 7 - Exosomes: Smart nanospheres for drug delivery naturally produced by stem cells A2 - Grumezescu, Alexandru Mihai. In Fabrication and Self-Assembly of Nanobiomaterials, William Andrew Publishing: 2016, 179-209. (10) Hannafon, B. N.; Ding, W. Q. Int. J. Mol. Sci. 2013, 14, 14240-69. (11) Hannafon, B. N.; Carpenter, K. J.; Berry, W. L.; Janknecht, R.; Dooley, W. C.; Ding, W. Q. Mol. Cancer 2015, 14, 133. (12) Park, J.; Choi, Y. Biomedical Engineering Letters 2014, 4, 258-268. (13) Clark, D. J.; Fondrie, W. E.; Liao, Z.; Hanson, P. I.; Fulton, A.; Mao, L.; Yang, A. J. Anal. Chem. 2015, 87, 10462-10469. (14) Sheridan, C. Nat. Biotechnol. 2016, 34, 359-60. (15) Vaksman, O.; Trope, C.; Davidson, B.; Reich, R. Carcinogenesis 2014, 35, 2113-20. (16) Maheswaran, S.; Sequist, L. V.; Nagrath, S.; Ulkus, L.; Brannigan, B.; Collura, C. V.; Inserra, E.; Diederichs, S.; Iafrate, A. J.; Bell, D. W. N Engl. J. Med. 2008, 359, 366-377. (17) Fernandez-Llama, P.; Khositseth, S.; Gonzales, P. A.; Star, R. A.; Pisitkun, T.; Knepper, M. A. Kidney Int. 2010, 77, 736-42. (18) Raimondo, F.; Morosi, L.; Chinello, C.; Magni, F.; Pitto, M. Proteomics 2011, 11, 709-20. (19) Yamashita, T.; Kamada, H.; Kanasaki, S.; Maeda, Y.; Nagano, K.; Abe, Y.; Inoue, M.; Yoshioka, Y.; Tsutsumi, Y.; Katayama, S.; Inoue, M.; Tsunoda, S. Die Pharmazie 2013, 68, 969-73. (20) Im, H.; Shao, H.; Park, Y. I.; Peterson, V. M.; Castro, C. M.; Weissleder, R.; Lee, H. Nat. Biotechnol. 2014, 32, 490-5. (21) Jeong, S.; Park, J.; Pathania, D.; Castro, C. M.; Weissleder, R.; Lee, H. ACS nano 2016, 10, 1802-9. (22) Palazzolo, G.; Albanese, N. N.; G, D. I. C.; Gygax, D.; Vittorelli, M. L.; Pucci-Minafra, I. Anticancer Res. 2012, 32, 84760. (23) Gill, D.; Kilponen, R. G.; Rimai, L. Nature 1970, 227, 743-4. (24) Zhang, R.; Zhang, Y.; Dong, Z. C.; Jiang, S.; Zhang, C.; Chen, L. G.; Zhang, L.; Liao, Y.; Aizpurua, J.; Luo, Y.; Yang, J. L.; Hou, J. G. Nature 2013, 498, 82-6. (25) Premasiri, W. R.; Lee, J. C.; Sauer-Budge, A.; Theberge, R.; Costello, C. E.; Ziegler, L. D. Anal. Bioanal. Chem. 2016. (26) Yoon, S. S.; Lee, W.; Byun, J. H.; Lee, J. U. J. Nanosci. Nanotechnol. 2015, 15, 8996-9001. (27) Nie, S.; Emory, S. R. Science 1997, 275, 1102-6. (28) Schlucker, S. Chemphyschem : a European journal of chemical physics and physical chemistry 2009, 10, 1344-54. (29) Seballos, L.; Zhang, J. Z.; Sutphen, R. Anal. Bioanal. Chem. 2005, 383, 763-7. (30) Huefner, A.; Kuan, W. L.; Barker, R. A.; Mahajan, S. Nano Lett. 2013, 13, 2463-70. (31) Lee, C.; Carney, R. P.; Hazari, S.; Smith, Z. J.; Knudson, A.; Robertson, C. S.; Lam, K. S.; Wachsmann-Hogiu, S. Nanoscale 2015, 7, 9290-7. (32) Lim, J.; Nam, J.; Yang, S.; Shin, H.; Jang, Y.; Bae, G.; Kang, T.; Lim, K.; Choi, Y. Anal. Chem. 2015, 87, 11652-11659. (33) Tirinato, L.; Gentile, F.; Di Mascolo, D.; Coluccio, M.; Das, G.; Liberale, C.; Pullano, S.; Perozziello, G.; Francardi, M.; Accardo, A. Microelectron. Eng. 2012, 97, 337-340. (34) Stremersch, S., Marro, M., Pinchasik, B.E., Baatsen, P., Hendrix, A., De Smedt, S.C., Loza-Alvarez, P., Skirtach, A.G., Raemdonck, K. and Braeckmans, K., 2016. Identification of

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Reproduction of spectra, NTA result of exosomes, calculation of laser spot size and number of exosome on SERS active area, preprocess of Raman spectra, loading plot of principal component 1 and principal component 2, tabulated data of specific patterns of exosomes, and confirmation of Raman pattern, frequencies and band assignments of Raman spectra, random, test with clinical samples. (PDF)

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare that they have no conflict of interest in relation to the work in this article.

ACKNOWLEDGMENT This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (grant numbers : HI14C3477 and HI14C2537).

REFERENCES (1) Sung, B. H.; Ketova, T.; Hoshino, D.; Zijlstra, A.; Weaver, A. M. Nat. Commun. 2015, 6, 7164. (2) Suetsugu, A.; Honma, K.; Saji, S.; Moriwaki, H.; Ochiya, T.; Hoffman, R. M. Adv. Drug Delivery Rev. 2013, 65, 383-90. (3) Costa-Silva, B.; Aiello, N. M.; Ocean, A. J.; Singh, S.; Zhang, H.; Thakur, B. K.; Becker, A.; Hoshino, A.; Mark, M. T.; Molina, H.; Xiang, J.; Zhang, T.; Theilen, T. M.; Garcia-Santos, G.; Williams, C.; Ararso, Y.; Huang, Y.; Rodrigues, G.; Shen, T. L.; Labori, K. J.; Lothe, I. M.; Kure, E. H.; Hernandez, J.; Doussot, A.; Ebbesen, S. H.; Grandgenett, P. M.; Hollingsworth, M. A.; Jain, M.; Mallya, K.; Batra, S. K.; Jarnagin, W. R.; Schwartz, R. E.; Matei, I.; Peinado, H.; Stanger, B. Z.; Bromberg, J.; Lyden, D. Nat. Cell Biol. 2015, 17, 816-26. (4) Peinado, H.; Aleckovic, M.; Lavotshkin, S.; Matei, I.; CostaSilva, B.; Moreno-Bueno, G.; Hergueta-Redondo, M.; Williams, C.; Garcia-Santos, G.; Ghajar, C.; Nitadori-Hoshino, A.; Hoffman, C.; Badal, K.; Garcia, B. A.; Callahan, M. K.; Yuan, J.; Martins, V. R.; Skog, J.; Kaplan, R. N.; Brady, M. S.; Wolchok, J. D.; Chapman, P. B.; Kang, Y.; Bromberg, J.; Lyden, D. Nat. Med. 2012, 18, 883-91. (5) Joshi, G. K.; Deitz-McElyea, S.; Liyanage, T.; Lawrence, K.; Mali, S.; Sardar, R.; Korc, M. ACS nano 2015, 9, 11075-89. (6) Luga, V.; Zhang, L.; Viloria-Petit, A. M.; Ogunjimi, A. A.; Inanlou, M. R.; Chiu, E.; Buchanan, M.; Hosein, A. N.; Basik, M.; Wrana, J. L. Cell 2012, 151, 1542-56.

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Individual Exosome-Like Vesicles by Surface Enhanced Raman Spectroscopy. Small, 12, 3292-3301. (35) Guicheteau, J.; Argue, L.; Emge, D.; Hyre, A.; Jacobson, M.; Christesen, S. Appl. Spectrosc. 2008, 62, 267-72. (36) Sokolova, V.; Ludwig, A. K.; Hornung, S.; Rotan, O.; Horn, P. A.; Epple, M.; Giebel, B. Colloids and surfaces. B, Biointerfaces 2011, 87, 146-50. (37) Böing, A.N., Van Der Pol, E., Grootemaat, A.E., Coumans, F.A., Sturk, A. and Nieuwland, R., 2014. Single-step isolation of extracellular vesicles by size-exclusion chromatography. J. Extracell. Vesicles, 3. (38) Ong, Y. H.; Lim, M.; Liu, Q. Opt. Express 2012, 20, 2215871.

Page 10 of 10

(39) Feng, S.; Lin, J.; Huang, Z.; Chen, G.; Chen, W.; Wang, Y.; Chen, R.; Zeng, H. Appl. Phys. Lett. 2013, 102, 043702. (40) De Groot, P.; Postma, G.; Melssen, W.; Buydens, L.; Deckert, V.; Zenobi, R. Anal. Chim. Acta 2001, 446, 71-83. (41) Lin, D.; Feng, S.; Pan, J.; Chen, Y.; Lin, J.; Chen, G.; Xie, S.; Zeng, H.; Chen, R. Opt. Express 2011, 19, 13565-77. (42) Thakur, B.K., Zhang, H., Becker, A., Matei, I., Huang, Y., Costa-Silva, B., Zheng, Y., Hoshino, A., Brazier, H., Xiang, J. and Williams, C., 2014. Double-stranded DNA in exosomes: a novel biomarker in cancer detection. Cell Res., 24, 766-769. (43) H. Shao, J. Chung, L. Balaj, A. Charest, D. D. Bigner, B. S. Carter, F. H. Hochberg, X. O. Breakefield, R. Weissleder and H. Lee, Nat. Med., 2012, 18, 1835

Insert Table of Contents artwork here

10


Exosome Classification by Pattern Analysis of ... - ACS Publications

Recommend Documents