Quantitative Single-Cell Analysis of Isolated Cancer Cells with a

Nov 28, 2018 - The heterogeneous nature of tumor-cell populations suggests that quantitative analysis at the single-cell level may provide better insi...
0 downloads 0 Views 842KB Size
Subscriber access provided by University of Winnipeg Library

Article

Quantitative single-cell-analysis of isolated cancer cells with microwell array Hojun Kim, Sungwook Park, Benedict Jungmin Kang, Youngdo Jung, Hyojin Lee, and Kwan Hyi Lee ACS Comb. Sci., Just Accepted Manuscript • DOI: 10.1021/acscombsci.8b00151 • Publication Date (Web): 28 Nov 2018 Downloaded from http://pubs.acs.org on November 29, 2018

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 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 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.

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 20 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

ACS Combinatorial Science

Quantitative single-cell-analysis of isolated cancer cells with microwell array Hojun Kim1,ǂ, Sungwook Park1,2, ǂ, Benedict J. Kang1, Youngdo Jeong1, Hyojin Lee*,1, and Kwan Hyi Lee*,1,2

1Center

for Biomaterials, Biomedical Research Institute, Korea Institute of Science and Technology

(KIST), 5 Hwarang-ro 14-gil, Seongbuk-gu, Seoul 02792, Republic of Korea 2 Division

of Bio-Medical Science & Technology, KIST School, Korea University of Science and

Technology (UST), Seoul, 02792, Republic of Korea

ǂ These

authors contributed equally for this work.

Correspondence: Kwan Hyi Lee and Hyojin Lee

Tel +82 2 958 6804 Fax +82 2 958 5308 Email [email protected]

1 ACS Paragon Plus Environment

ACS Combinatorial Science 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

Abstract: The heterogeneous nature of tumor cell populations suggests that quantitative analysis at the single-cell level may provide better insights into cancer biology. Specifically, detection of multiple biomarkers from a single cell offers important initial information about cellular behavior. However, conventional approaches have limitations on biomarker detections in a single cell level. Here we fabricated a polymer microwell array to capture single cells from prostate cancer cell lines and quantitatively analyzed three different cancer-related biomarker expressions, CD44, EpCAM, and PSMA without protein extraction step. The resulting information of cell surface biomarker distribution was compared to other standard analytical techniques. Interestingly large variation of CD44 expression levels was observed when cell proliferation cycle is modulated. On the other hand, expression levels of EpCAM in three different cell lines are consistent throughout analytical methods except microarray where it has different substrate material to adhere with. This observation clearly emphasizes that biomarker choice and environmental control are critical to properly understand the single cell state.

Keywords: Prostate cancer, quantitative profiling, laboratory techniques, microarray, single cell analysis, quantum dots

Introduction A great deal of information in fundamental cancer biology and clinical cancer diagnosis depends on the detection of biomarkers, usually molecules displayed on the cell surface.1–5 For

2 ACS Paragon Plus Environment

Page 2 of 20

Page 3 of 20 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

ACS Combinatorial Science

example, angiogenesis of tumors are closely related to cell surface biomarker such as tumor endothelial marker 1, 5, and 8. It was found that these three markers are highly expressed in tumor vessels while they are almost undetectable from normal adult mice.6 In addition, there are several biomarkers closely related to cancer progression, such as prostate-specific membrane antigen (PSMA), epithelial cell adhesion molecule (EpCAM), and cluster of differentiation 44 (CD44) for prostate cancer. PSMA level is directly proportional to Gleason grade and the biomarker has been the subject of intense basic and clinical research for the past 20 years.7–9 EpCAM has been found to be overexpressed in the prostate cancer cells at the primary prostate tissue, circulation, and metastatic sites.10,11 Although CD44 upregulation induced pro-invasive properties in some tumors was reported, prostate cancer has been found to downregulate standard form of CD44 with increased dysregulated splicing leading to increasing other isoforms of CD44.12,13 Increasing levels of the isoforms were found to have pivotal role in prostate cancers. The experimental techniques used to detect biomarkers usually rely on comparisons of experimental groups with control groups.14,15 The qualitative nature of such analysis only reliably determines the presence of the target(s) of interest. The highly heterogeneous nature of tumor cell populations makes it important to perform quantitative analysis at a single-cell level to understand cancer biology better.16 In this study we fabricated microarray system to isolate single cells and monitor protein expressions. Specifically, we chose three different prostate cancer cell lines and three different surface biomarkers, CD44, EpCAM, and PSMA which are associated with prostate cancer, to quantitatively characterize phenotype on individual cells. Comparison of these results with other standard analytical techniques showed significant differences that highlight the virtues of single-cell profiling.

3 ACS Paragon Plus Environment

ACS Combinatorial Science 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

Results and Discussion Cell line characterization via traditional methods To characterize the different prostate cancer cell lines, we compared three known prostate cancer biomarkers (PSMA, CD44, and EPCAM) from three common prostate cancer model cell lines: 22Rv1, LNCaP, and DU145. The first is non-metastatic, while the latter two are derived from metastatic sites, left supraclavicular lymph node and brain, respectively. Each cell line has a distinct morphology which implies the existence of different biomarker profiles. Additionally, RWPE-1, a noncancerous prostate epithelial cell, was used as a normal control.17 First, western blotting (WB) was performed on membrane proteins extracted by commercial kit and the results are shown in Figure 1. Blotting against anti-CD44, PSMA, and EpCAM antibodies showed that PSMA expression signal was only detected in LNCaP cells. In the case of CD44, RWPE-1 showed too strong CD44 signal which leaded to image saturation problem.18 In other words, signals from other cell lines are buried. In the case of EpCAM, however, we found gradual signal decrease in the order of 22Rv1, LNCaP, and DU145. Since western blotting can only yield qualitative results, we also performed enzyme-linked immunosorbent assay (ELISA) to get quantitative results. Figure 2A represents the summary of ELISA experiments with the same samples used in Figure 1. For CD44, DU145 cells gave comparable expression level to RWPE1 cells which was the only cell line that shows significant expression level of CD44 in Western Blot. In the case of PSMA, none of cell lines gave detectable signal. On the other hand, EpCAM tendency was conserved. 22RV1 showed highest expression level and LNCaP and DU145 are following. These difference of CD44 and PSMA expression levels can be understood from the methodological difference between WB and ELISA. Because denaturation step is involved in WB, 4 ACS Paragon Plus Environment

Page 4 of 20

Page 5 of 20 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

ACS Combinatorial Science

antigen can be cleaved into multiple subunits and not all of bands are detected. On the other hand, denaturation step is not necessary in ELISA and thus protein expression level and tendency can be different from each other.18 In addition, ELISA has much lower limit of detection because antigenantibody reaction can occur even with fragment antigens.19 In this regards, WB results should be rather used for confirming the presence of target proteins in the sample. To account for cancer heterogeneity, we arrested cell cycle to G0/G1 and performed ELISA with extracted plasma membrane proteins. As shown in Figure 2B, high EpCAM expressions were found in 22Rv1 and LNCaP, consistent with previous ELISA and WB results. On the other hand, CD44 expression pattern across different cell types were opposite, high at 22Rv1 and LNCaP and low at DU145 and RWPE-1. In the case of PSMA, no detectable signal was found for all cell lines. From this result, it can be inferred that CD44 is highly dependent on proliferation cycle. Indeed, in colorectal cancer, the CD44 is highly expressed in the proportion of G2/M phase while cells in G0/G1 phase show low expression level.20 The previous studies reported that LNCaP had high PSMA expression while our WB and ELISA showed low expression level. (Figure 1 and Figure 2) The reported studies performed WB and ELISA using whole cell lysate and the discrepancies in the results may due to this sample differences.21 This indicates that the expression pattern of biomarker on plasma membrane could be different to total amount of biomarker proteins in a cell. Depending on the cellular function of biomarker, the local protein concentration is a critical factor for controlling the cell behavior.22 For this reason, one should consider the location of target proteins before performing ELISA or WB.

5 ACS Paragon Plus Environment

ACS Combinatorial Science 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

Quantitative single-cell-analysis on the isolated cancer cells with microarray WB and ELISA are the most popular and basic analytical techniques for protein characterization in a given sample. However, WB and ELISA require at least few micrograms of target proteins. In other words, these two techniques are not adequate to measure protein expressions in a single cell level. To understand cancer biology in highly heterogeneous tumor tissues, developing a robust method to measure target protein expression in a single cell level is necessary. Immunofluorescence (IF) with quantum dot (QD) nanoprobe, semiconducting nanoparticles conjugated with secondary antibodies, was used to directly visualize the presence of the biomarker on the cells and for quantitative profiling analysis after imaging. Instead of targeting the cells with conventional fluorescence dye, which signal decreases over time due to photobleaching effect, QD nanoprobes were used to quantify the biomarkers in the prostate cancer cell lines. (Figure 3) The immunofluorescence with QD nanoprobes is useful for the subsequent quantitative analysis because of several advantageous optical properties of QDs for bio-imaging.23–25 The QDs are 20-100 times brighter than traditional fluorescence dye and they show stable optical character.25,26 The number of QDs were quantitated based on the fluorescence intensity per micron square.24 The number of QDs is equivalent to the number of the biomarkers, and thus number density of biomarkers were calculated based on the fluorescence intensity. Isolation and analysis of single-cells, in tandem, has become important as we gain more understanding of cancer heterogeneity. Recently published methods for detecting cancer heterogeneity through single-cell analysis are mostly nucleic acid-based.27–29 Their high specificity and advances in the gene sequencing techniques allowed genomic and proteomic analysis to reveal cancer heterogeneity. (Table. 1) Microfluidics and microarray devices that can variate and chamber

6 ACS Paragon Plus Environment

Page 6 of 20

Page 7 of 20 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

ACS Combinatorial Science

microenvironments are widely incorporated into high throughput single-cell level cancer heterogeneity studies due to their portability, controllability, and cost effectiveness. In our study, microwell array (with 20 µm x 20 µm single well) was fabricated using PDMS to isolate single cell. Subsequently, poly-l-lysine were used to coat the wells to secure the cells within the wells. After 24 hours, wells with single cells are selected to minimize effect of cell proliferation cycle on the protein expressions. (Figure 3A) The isolated cells exhibited spherical morphologies different from what observed in the conventional cell culture dishes or flasks. After targeting the biomarkers, immunofluorescence method with QD nanoprobes was employed to determine the levels of CD44, EpCAM, and PSMA in a single cell level (Figure 3B). The fluorescent signals from the QD nanoprobes were quantified to profile the biomarkers. For PSMA, no signal was detected for all cell lines as it was in ELISA. In the case of CD44, only DU145 shows high expression level. Lastly, EpCAM expressions were high in LNCaP cells, low in 22Rv1, and none in DU145 which are different from ELISA and WB results. (Figure 3B) The protein expression level summary in each method are summarized in the Table 2. In general, CD44 gives consistent pattern across different method when the cell proliferation cycle is arrested. In the case of EpCAM, its expression pattern is conserved regardless of cell cycle. However, both CD44 and EpCAM expression patterns are different in microarray where it has different substrate material to interact with. It is possible that different substrate material and cell isolation cause this dramatic change.

Conclusion In this study, we successfully fabricated and isolated single cancer cells in a micro array coated with poly-l-lysine. Three biomarker expression levels were measured using various

7 ACS Paragon Plus Environment

ACS Combinatorial Science 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 techniques. Overall, LNCaP and 22Rv1 had the similar profile patterns across different methods whereas the DU145 had the opposite pattern. In the standpoint of morphology, 22Rv1 and LNCaP seem to have poor cell-cell interface whereas DU145 forms stable interfaces within cellular contacting area.30 LNCaP had the weakest adhesion strength compare to other prostate cancer cell lines used in this study. All analytical methods (WB, ELISA, IF, IF-single cell) revealed the lowest CD44 expression in LNCaP regardless of proliferation cycle. Since CD44 is involved in cell-cell interaction, cell adhesion, and migration,31 observed low expression of CD44 in LNCaP is in the same line with low adhesion strength. Cancer biomarker expressions can depend on many factors such as location in a cell compartment and proliferation state. When we arrested a cell cycle, we observed same trend for CD44 across different methods.

In the case of EpCAM, its expression level is not correlated

with proliferation cycle. However, epithelial cell adhesion molecule shows completely opposite expression pattern in microarray system. Considering EpCAM is epithelial cell adhesion molecule, it is plausible that different substrate material (poly-l-lysine) affect adhesion of cells and that alter EpCAM expression. We summarize the comparison of each method as below and also table 2: Conventional methods (WB, ELISA) vs. single cell analysis in IF with QD. A drawback of WB and ELISA is that it cannot display biomarker spatial distributions across cell surfaces. For the surface profiling, additional steps including membrane protein extraction are required. The ability to show spatial distributions of biomarkers is one strong advantage that IF can offer. With the use of QD nanoprobe, it allows quantitative analysis that can give information on relative level of biomarker

8 ACS Paragon Plus Environment

Page 8 of 20

Page 9 of 20 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

ACS Combinatorial Science

per micron square of area. (Figure 3) However, IF cannot detect biomarker that is secretory while ELISA can measure even fragment antigens having binding sites. To do quantitative analysis at a single cell level, a cancer cell must be isolated from a population of heterogeneous cells. However, there would be trade-offs when it comes to cell isolation. Isolating a single cell from a tumor tissue, which comprises numerous heterogeneous cells, and seeding on to microwell could lead to large changes in protein expression levels. Indeed, Figure S1 indicates multi cell analysis with IF showed similar, but not perfectly agreed results. This difference can be arouse from the different environments in the microwells or heterogeneity population of each cell line. In summary, cleavage of target protein, presence of binding sites in each subunit, cell proliferation cycle, image saturation problems (for WB) and many more can affect protein expressions. Thus, it is important to know if target protein is particularly sensitive to these parameters. For example, we found EpCAM is sensitive with substrate material, CD44 is sensitive to cell proliferation cycle. By doing so, one can qualitatively compare protein expression levels across different cell lines in a given environment.

Experimental Procedures Cell lines and cell culture. DU145 and 22Rv1 were cultivated in RPMI 1640 (Gibco, Carlsbad, California, USA, Cat. # A10491-01) supplemented with 10 % FBS (Gibco Cat. # 10082-147) and 1 % antibioticantimycotic (Gibco Cat. # 15240-062). Cells were seeded onto wells and microwells with seeding densities according to the Physical Sciences-Oncology Center Network Bio-resource Core Facility

9 ACS Paragon Plus Environment

ACS Combinatorial Science 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 10 of 20

protocols. The cells were cultivated for 2 days under sterile condition to ensure firm adhesion onto the surface of multiwell and microarray plates at 37 ℃ and 5 % CO2. To arrest cell cycle, we starved cell by incubating cells without FBS for two days. Then we changed with complete growth media for two hours right before performing ELISA assay. Microwells were coated with poly-l-lysine for 10 min. at room temperature and then aspirated. The coated microwell were left to dry up all the residual poly-l-lysine prior to cultivation of the cell lines on the device. Western blotting. 5 x 106 cells/ml were harvested from 100 π dish using a scraper. 1.5 ml tube contained the cells. The cells were washed three times with 1X PBS by pulse centrifugation. To extract cellular membrane proteins from the cells’ lysate, Mem-PER™ Plus Membrane Protein Extraction Kit was used (ThermoFisher, Massachusetts, USA, Cat.# 89842). The protein amount was estimated by BCA assay (ThermoFisher, Cat.# 23225). Next, an equal volume of sample buffer (125 mM Tris pH 6.8, 4 % SDS, 10 % glycerol, 0.006 % bromophenol blue, and 1.8 % mercaptoethanol) was added to all samples, and the resulting solution was boiled for 5 min. A 15 μg amount of total proteins from cells was loaded in each well of a protein precast gel (Bio-Rad, Hercules, CA, USA). After electrophoresis at 120 V and 60 min, the proteins were transferred from the gel to a PVDF membrane at 1 A constant current for 1 hour in transfer buffer (Thermo Fisher). The blot from the transfer apparatus was removed and immediately placed into blocking buffer (5 % nonfat dry milk, 10 mM Tris pH 7.5, 100mM NaCl, and 0.1 % Tween 20). After blocking for 1 hour at room temperature, the membrane was incubated with primary antibodies. After incubation with primary antibody solution, the membrane was twice washed (10 mM Tris pH 7.5, 100 mM NaCl, and 0.1 % Tween 20). Then, the membrane was incubated with horseradish peroxidase (HRP)-conjugate anti10 ACS Paragon Plus Environment

Page 11 of 20 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

ACS Combinatorial Science

mouse IgG (secondary antibody) diluted with 5 % nonfat dry milk solution at room temperature. After 1 hour incubation, the antibody solution was removed, and the membrane was washed three times by washing buffer (10 mM Tris pH 7.5, 100 mM NaCl, 0.1 % Tween 20). Finally, the band signal of proteins on the membrane was developed by enhanced chemifluorescence substrate (Atto, Tokyo, Japan). The amount of each protein in the blots was determined by counting the total number of pixels in each band (integrated density value) with ImageJ. Immunofluorescence. DU145 and 22Rv1 are cultivated in RPMI 1640 (Gibco Cat. # A10491-01) supplemented with 10 % FBS (Gibco Cat. # 10082-147) and 1 % antibiotic-antimycotic (Gibco Cat. # 15240-062). Cells are seeded onto wells and microwells with seeding densities according to the Physical SciencesOncology Center Network Bio-resource Core Facility protocols. The cells were cultivated for 2 days under sterile condition to ensure firm adhesion onto the surface of multiwell and microarray plates at 37 ℃ and 5 % CO2. Then, the cells were fixed with 4 % paraformaldehyde in PBS for 10min, permeablized with 0.1 % Triton X-100 in PBS, blocked with 5 % BSA in PBS at room temperature (RT) and incubated with primary antibodies in the blocking buffer. The cells were incubated with secondary antibodies labeled with fluorophore and quantum dot at RT for 1 hour. Images were taken with a confocal laser scanning microscope (LSM700, Carl Zeiss, Oberkochen, Germany). Quantitative profiling and analysis. Standard curve was generated by plotting number of quantum dot (QD) nanoprobes against fluorescence intensities at various concentrations. Equation was generated based on the linear graph. The obtained fluorescence signals are plugged into the equation and a number of QD

11 ACS Paragon Plus Environment

ACS Combinatorial Science 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 12 of 20

nanoprobes that corresponds to the obtained signal intensities. The number of QD nanoprobe equates the number of biomarkers. Microwell fabrication. The PDMS microwell device is composed of arrays of 84,672 microwells with the dimension of 50 μm x 50 μm. This device was manufactured by injecting the Sylgard 184 (Dow Corning Corporation, Midland, USA) into the custom-built mold, which was designed so that the Sylgard mold will adhere directly to a glass slide. The Sylgard 184 was heated at 90°C for 1 hour, and cooled to room temperature until it was removed from the mold. Enzyme-linked Sandwich Assay. 100 μL of coating antibody was added to each well. Which one to be coated on the well surface was empirically tested. The plate was incubated for 2 hours at room temperature and washed three times, 5 min each, afterwards. 300 µL of blocking buffer was used to block the wells and then incubated at room temperature for 1 hour. Blocking buffer was removed before samples or standards were added to the well. The plate with the sample or standard was incubated at room temperature for 1 hour. After incubation, the plate was washed three times, 5 min each. Detection antibodies were added to each well and incubated at room temperature for 1 hour. The excess of the detection antibody was removed and secondary antibody with horseradish peroxidase was added to each well and removed the excess with washing. Substrate solution was added and developed at room temperature for 30 min. Tecan instrument was used to measure the luminescent signal.

12 ACS Paragon Plus Environment

Page 13 of 20 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

ACS Combinatorial Science

Associated Content Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: xx.xxxx/acscombsci.xxxxxxx. Immunofluorescence-Quantum Dot Assay of Multi Cell Analysis (PDF)

Acknowledgments This study was supported by the Bio & Medical Technology Development Program of the NRF funded by the Korean government MSIP (2015M3A9E2029265).

References (1) (2) (3) (4) (5) (6) (7) (8) (9)

Weigelt, B.; Pusztai, L.; Ashworth, A.; Reis-Filho, J. S. Challenges translating breast cancer gene signatures into the clinic. Nat. Rev. Clin. Oncol. 2012, 9, 58–64. Kang, B.J.; Jeun, M.; Jang, G. H.; Song, S.H.; Jeong, I.G.; Kim, C.S.; Searson, P.C.; Lee, K. H. Diagnosis of prostate cancer via nanotechnological approach. Int. J. Nanomedicine 2015, 10, 6555. Jeun, M.; Park, S.; Jang, G. H.; Lee, K. H. Tailoring Mg x Mn1–x Fe2O4 Superparamagnetic Nanoferrites for Magnetic Fluid Hyperthermia Applications. ACS Appl. Mater. Interfaces 2014, 6, 16487-16492. Lee, H.; Lee, D.; Park, J. H.; Song, S. H.; Jeong, I. G.; Kim, C.-S.; Searson, P. C.; Lee, K. H. High throughput differential identification of TMPRSS2-ERG fusion genes in prostate cancer patient urine. Biomaterials 2017, 135, 23–29. Park, J. S.; Park, W.; Park, S.; Larson, A. C.; Kim, D.-H.; Park, K.-H. Multimodal Magnetic Nanoclusters for Gene Delivery, Directed Migration, and Tracking of Stem Cells. Adv. Funct. Mater. 2017, 27, 1700396. Deer, E. L.; González-Hernández, J.; Coursen, J. D.; Shea, J. E.; Ngatia, J.; Scaife, C. L.; Firpo, M. A.; Mulvihill, S. J. Phenotype and Genotype of Pancreatic Cancer Cell Lines. Pancreas 2010, 39, 425–435. Bostwick, D. G.; Pacelli, A.; Blute, M.; Roche, P.; Murphy, G. P. Prostate specific membrane antigen expression in prostatic intraepithelial neoplasia and adenocarcinoma: a study of 184 cases. Cancer 2000, 82, 2256–2261. Ristau, B. T.; O’Keefe, D. S.; Bacich, D. J. The prostate-specific membrane antigen: Lessons and current clinical implications from 20 years of research. Urol. Oncol. Semin. Orig. Investig. 2014, 32, 272–279. Wright Jr, G. L.; Haley, C.; Beckett, M. Lou; Schellhammer, P. F. Expression of prostate-specific membrane antigen in normal, benign, and malignant prostate tissues. Urol. Oncol. Semin. Orig. 13 ACS Paragon Plus Environment

ACS Combinatorial Science 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

(10)

(11) (12) (13) (14) (15)

(16) (17) (18) (19) (20) (21)

(22) (23) (24) (25) (26)

Page 14 of 20

Investig. 1995, 1, 18–28. Armstrong, A. J.; Marengo, M. S.; Oltean, S.; Kemeny, G.; Bitting, R. L.; Turnbull, J.; Herold, C. I.; Marcom, P. K.; George, D.; Garcia-Blanco, M. A. Circulating Tumor Cells from Patients with Advanced Prostate and Breast Cancer Display Both Epithelial and Mesenchymal Markers. Mol. Cancer Res. 2011, 9, 997–1007. Went, P.; Vasei, M.; Bubendorf, L.; Terracciano, L.; Tornillo, L.; Riede, U.; Kononen, J.; Simon, R.; Sauter, G.; Baeuerle, P. A. Frequent high-level expression of the immunotherapeutic target Ep-CAM in colon, stomach, prostate and lung cancers. Br. J. Cancer 2006, 94, 128–135. Nagabhushan, M.; Pretlow, T. G.; Guo, Y.-J.; Amini, S. B.; Pretlow, T. P.; Sy, M.-S. Altered Expression of CD44 in Human Prostate Cancer During Progression. Am. J. Clin. Pathol. 1996, 106, 647–651. De Marzo, A. M.; Bradshaw, C.; Sauvageot, J.; Epstein, J. I.; Miller, G. J. CD44 and CD44v6 downregulation in clinical prostatic carcinoma: Relation to Gleason grade and cytoarchitecture. Prostate 1998, 34, 162–168. Dhanasekaran, S. M.; Barrette, T. R.; Ghosh, D.; Shah, R.; Varambally, S.; Kurachi, K.; Pienta, K. J.; Rubin, M. A.; Chinnaiyan, A. M. Delineation of prognostic biomarkers in prostate cancer. Nature 2001, 412, 822–826. Shen, J.; Person, M. D.; Zhu, J.; Abbruzzese, J. L.; Li, D. Protein Expression Profiles in Pancreatic Adenocarcinoma Compared with Normal Pancreatic Tissue and Tissue Affected by Pancreatitis as Detected by Two-Dimensional Gel Electrophoresis and Mass Spectrometry. Cancer Res. 2004, 64, 9018–9026. McGranahan, N.; Swanton, C. Clonal Heterogeneity and Tumor Evolution: Past, Present, and the Future. Cell 2017, 168, 613–628. Bennett, N. C.; Hooper, J. D.; Johnson, D. W.; Gobe, G. C. Expression profiles and functional associations of endogenous androgen receptor and caveolin-1 in prostate cancer cell lines. Prostate 2013, 74, 478–487. Bass, J. J.; Wilkinson, D. J.; Rankin, D.; Phillips, B. E.; Szewczyk, N. J.; Smith, K.; Atherton, P. J. An overview of technical considerations for Western blotting applications to physiological research. Scand. J. Med. Sci. Sports 2016, 27, 4–25. He, J.; Schepmoes, A. A.; Shi, T.; Wu, C.; Fillmore, T. L.; Gao, Y.; Smith, R. D.; Qian, W.-J.; Rodland, K. D.; Liu, T.; et al. Analytical platform evaluation for quantification of ERG in prostate cancer using protein and mRNA detection methods. J. Transl. Med. 2015, 13:54. Blacking, T. M.; Waterfall, M.; Argyle, D. J. CD44 is associated with proliferation, rather than a specific cancer stem cell population, in cultured canine cancer cells. Vet. Immunol. Immunopathol. 2011, 141, 46–57. Evans, M. J.; Smith-Jones, P. M.; Wongvipat, J.; Navarro, V.; Kim, S.; Bander, N. H.; Larson, S. M.; Sawyers, C. L. Noninvasive measurement of androgen receptor signaling with a positron-emitting radiopharmaceutical that targets prostate-specific membrane antigen. Proc. Natl. Acad. Sci. 2011, 108, 9578–9582. Sun, Z.; Chen, X.; Wang, G.; Li, L.; Fu, G.; Kuruc, M.; Wang, X. Identification of functional metabolic biomarkers from lung cancer patient serum using PEP technology. Biomark. Res. 2016, 4:11. Lee, H.; Kim, C.; Lee, D.; Park, J. H.; Searson, P.; Lee, K. H. Optical coding of fusion genes using multicolor quantum dots for prostate cancer diagnosis. Int. J. Nanomedicine 2017, Volume 12, 4397– 4407. Kim, C.; Galloway, J. F.; Lee, K. H.; Searson, P. C. Universal Antibody Conjugation to Nanoparticles Using the Fcγ Receptor I (FcγRI): Quantitative Profiling Of Membrane Biomarkers. Bioconjug. Chem. 2014, 25, 1893–1901. Ma, J.; Hou, Z.; Song, Y.; Wang, L.; Guo, E. Visual and Quantitative Screening of α1-Adrenoceptor Antagonists in Living Cells Using Quantum Dots. ACS Comb. Sci. 2014, 16, 155–159. Walling, M. A.; Novak, J. A.; Shepard, J. R. E. Quantum Dots for Live Cell and In Vivo Imaging. 14 ACS Paragon Plus Environment

Page 15 of 20 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

ACS Combinatorial Science

(27) (28) (29) (30) (31)

Int. J. Mol. Sci. 2009, 10, 441–491. Park, S. H.; Hong, A.; Kim, J.-H.; Yang, H.; Lee, K.; Jang, H. S. Highly Bright Yellow-GreenEmitting CuInS2 Colloidal Quantum Dots with Core/Shell/Shell Architecture for White LightEmitting Diodes. ACS Appl. Mater. Interfaces 2015, 7, 6764–6771. Khoo, B. L.; Chaudhuri, P. K.; Ramalingam, N.; Tan, D. S. W.; Lim, C. T.; Warkiani, M. E. Singlecell profiling approaches to probing tumor heterogeneity. Int. J. Cancer 2016, 139, 243–255. Saadatpour, A.; Lai, S.; Guo, G.; Yuan, G.-C. Single-Cell Analysis in Cancer Genomics. Trends Genet. 2015, 31, 576–586. Chang, L.; Graham, P. H.; Hao, J.; Bucci, J.; Cozzi, P. J.; Kearsley, J. H.; Li, Y. Emerging roles of radioresistance in prostate cancer metastasis and radiation therapy. Cancer Metastasis Rev. 2014, 33, 469–496. Goodison, S.; Urquidi, V.; Tarin, D. CD44 cell adhesion molecules. Mol. Pathol. 1999, 52, 189–196.

15 ACS Paragon Plus Environment

ACS Combinatorial Science 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

Table 1 Single-cell analysis in literatures.

Abbreviations: AUC, area under the curve; LS, least squares; NE, not estimable.

16 ACS Paragon Plus Environment

Page 16 of 20

EpCAM CD44 PSMA

120 100 80 60 40 20 0

RWPE1

22Rv1

LNCaP

DU145

Cell Line

Figure 1 Western Blotting and semi-quantitative data based on the band intensities Respective antibodies to detect the protein of interest were used and generated a semiquantitative graph. The band signal intensity of the bands were normalized against GAPDH, a housekeeping gene, band signal. Abbreviations: αPSMA; αCD44; αEpCAM; αGAPDH; 100

EpCAM CD44 PSMA

80 60 40 20 0

RWPE1

22Rv1

LNCaP

DU145

Amount of Biomarker (μg/ml)

B

A Amount of Biomarker (μg/ml)

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

ACS Combinatorial Science

Relative Intensity [arb. u.]

Page 17 of 20

100 80

EpCAM CD44 PSMA

60 40 20 0

RWPE1

Cell Line

22Rv1

LNCaP

DU145

Cell Line

Figure 2 ELISA quantitative analysis. Enzyme-linked immunosorbent assay of (Left) Non-starved cells and (Right) Starved cells that quantified CD44, EpCAM, and PSMA of prostate cancer cells. Assays were

ACS Paragon Plus Environment

ACS Combinatorial Science 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

carried out three times. The error bars were derived based standard deviations for each set of three trials.

Figure 3 Microarray and single cell quantitative profiling.

18 ACS Paragon Plus Environment

Page 18 of 20

Page 19 of 20 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

ACS Combinatorial Science

(A) Schematic diagram of single cell assay using microwell array (B) The microarray device isolated the single cells of 22Rv1, LNCaP, and DU145. A standard curve was generated based on the QD nanoprobes in serial dilution. Isolated prostate cancer cells photographed under the light microscope. Quantitative single cell analysis performed on each cancer cell. The error bars were derived based standard deviations for each set of three trials. Abbreviations: QD, quantum dots;

Table 2 Comparison of different methods on three protein expressions from four different cell lines.

Abbreviations: S, starved; NS, Non-starved; 22Rv1, 22; LNCaP, LN; DU145, DU; RWPE1, RE;

Notes: Immunofluorescence assay result is in the figure S1. Table 3 Assays for cellular molecular characterization.

Notes: $ represents relative degree of cost

19 ACS Paragon Plus Environment

ACS Combinatorial Science 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

TOC

20 ACS Paragon Plus Environment

Page 20 of 20