Integrated Analysis of Exosomal Protein Biomarkers on Alternating

Subscriber access provided by - Access paid by the | UCSB Libraries

Integrated Analysis of Exosomal Protein Biomarkers on Alternating Current Electrokinetic Chips Enables Rapid Detection of Pancreatic Cancer in Patient Blood Jean M. Lewis, Ankit D. Vyas, Yuqi Qiu, Karen S. Messer, Rebekah White, and Michael J. Heller ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.7b08199 • Publication Date (Web): 23 Mar 2018 Downloaded from http://pubs.acs.org on March 23, 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 26 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 Nano

Integrated Analysis of Exosomal Protein Biomarkers on Alternating Current Electrokinetic Chips Enables Rapid Detection of Pancreatic Cancer in Patient Blood

Jean M. Lewis1*, Ankit D. Vyas1, Yuqi Qiu3, Karen S. Messer2,3, Rebekah White4, and Michael J. Heller1*

1

Department of NanoEngineering, University of California San Diego, La Jolla, CA 92093, USA. 2Moores Cancer Center, University of California San Diego, La Jolla, CA 92093, USA. 3 Division of Biostatistics and Bioinformatics; Dept. of Family Medicine and Public Health, UC San Diego, La Jolla, CA 92093, USA. 4Department of Surgery, University of California San Diego, La Jolla, CA 92093, USA. KEYWORDS: exosome; biomarkers; pancreatic cancer; cancer diagnostics; colon cancer; whole blood; glypican-1

ABSTRACT Pancreatic ductal adenocarcinoma (PDAC) typically has non-specific symptoms and is often found too late to treat. Because diagnosis of PDAC involves complex, invasive, and expensive procedures, screening populations at increased risk will depend on developing rapid, sensitive, specific, and cost-effective tests. Exosomes, which are nanoscale vesicles shed into blood from tumors, have come into focus as valuable entities for non-invasive liquid biopsy diagnostics. However, rapid capture and analysis of exosomes with their protein and other biomarkers have proven difficult. Here we present a simple method integrating capture and analysis of exosomes and other extracellular vesicles (EVs) directly from whole blood, plasma, or serum onto an AC Electrokinetic microarray chip.

In this process, no pre-treatment or dilution of sample is

ACS Paragon Plus Environment

1

ACS Nano 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 2 of 26

required, nor is it necessary to use capture antibodies or other affinity techniques. Subsequent on-chip immunofluorescence analysis permits specific identification and quantification of target biomarkers within as little as 30 minutes total time. In this initial validation study, the biomarkers glypican-1 and CD63 were found to reflect the presence of PDAC, and thus used to develop a bivariate model for detecting PDAC.

Twenty PDAC patient samples could be

distinguished from eleven healthy subjects with 99% sensitivity and 82% specificity. In a smaller group of colon cancer patient samples, elevated glypican-1 was observed for metastatic, but not for non-metastatic disease. The speed and simplicity of ACE exosome capture and onchip biomarker detection, combined with the ability to use whole blood, will enable seamless “sample-to-answer” liquid biopsy screening and improve early-stage cancer diagnostics.

Pancreatic ductal adenocarcinoma (PDAC) is projected to be the second-highest cause of cancer death in the U.S. by the year 2020.

1

Common presenting symptoms of PDAC, such as weight

loss and epigastric pain, are non-specific, and patients usually have advanced disease by the time they are diagnosed. Diagnosis of PDAC currently relies on expensive imaging studies—such as contrast-enhanced computed tomography (CT), magnetic resonance imaging (MRI), and endoscopic ultrasound —with subsequent confirmation by tumor biopsy. Because PDAC is typically not diagnosed until the disease is advanced, new technology and methods for liquid biopsy that may facilitate earlier diagnosis could greatly increase the proportion of patients able to undergo potentially curative surgical resection. Biomarker-based screening strategies have promise for early diagnosis, but existing tumor biomarkers, such as carbohydrate antigen 19-9 (CA19-9) and carcinoembryonic antigen (CEA) are neither sensitive nor specific enough to serve as useful early diagnostic tests. 2,3 Glypican-1 has recently been identified as an exosome-based biomarker of interest for pancreatic cancer. 4–6

Exosomes are small, nanoscale entities of about 70-130 nm in size, that are shed from healthy and cancer cells alike into the blood and other body fluids, and carry both protein and nucleic acid (RNA) biomarkers.

2,7–12

Because they reflect both the genotype and phenotype of the

originating cell, exosomes have the appealing potential to be screened for cancer or other


2

Page 3 of 26 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 Nano

diseases without tumor biopsy. However, given their very small size and low buoyant density, exosomes are especially difficult to isolate.

Conventional protocols rely on differential

centrifugation, which is time- and labor-intensive, or on various types of affinity capture that by definition select for only one subtype of exosome. Some progress towards developing devices for the rapid isolation and/or analysis of exosomes has recently been made. In many of these systems, antibodies are used for both capture and analysis of exosomes, such as the tetraspanin membrane proteins CD63 and CD9 that are enriched on exosomes but may also be present on whole cells.

13–19

The methods of detection employed include electrochemical sensors, surface

plasmon resonance, and measurement of mechanical forces of flow on tethered beads. Several of these technologies employ microfluidics to restrict flow rates, or in one example, to isolate exosomes from body fluids by extruding them through nano-pores, followed by antibodies for detection. 19 Patient samples applied to this device were necessarily subjected to a filtration step. Indeed, such devices based on microfluidic separation must overcome obstacles such as clogging, low flow rates, sheer stress, and mixing issues before they can meet the criteria required to be considered a rapid diagnostic device. Furthermore, the use of antibodies as the initial exosome isolation introduces a selection step that may cause sub-populations of exosomes, crucial to diagnosis, to be overlooked. Another recent approach uses addition of tagged lipids to tether bilayer membrane-containing moieties in a buffer sample to magnetic beads. 20 When the tagged magnetic beads were used to isolate EVs from human plasma, the efficiency was less than 50%, suggesting inhibition of lipid binding by plasma elements, as well as differential lipid association/composition for different EV populations; cancer exosomes are likely to differ in this regard. Although several strategies have been devised for exosome-based diagnostics, to date these techniques cannot be used directly with whole blood, require sample dilution and are relatively time-consuming and involved procedures that are useful chiefly in research settings. 13,14,21,22

Thus, methods and devices suitable for viable liquid biopsy diagnostics remain elusive.

We previously demonstrated a technique using an AC electrokinetic (alternating current; “ACE”) microarray chip device to isolate glioblastoma exosomes and other nanoscale entities directly from whole blood, plasma or serum within 15-20 minutes.

23

Here we present results

demonstrating exosome isolation from pancreatic and colon cancer patient samples with subsequent immunofluorescent detection of specific exosome-associated protein biomarkers, all


3

ACS Nano 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 4 of 26

carried out directly on the ACE microarray chip (in situ). Using this method, glypican-1 and CD63 could be detected in undiluted pancreatic cancer patient whole blood, plasma or serum within 90 minutes total time; where fluorophore-conjugated antibodies are available, the entire procedure can be done in under 45 minutes from receipt of the patient sample (Figure 1). Unlike other exosome isolation strategies, this procedure requires no dilution or additions to the sample, such as antibodies, affinity beads, or lipids, that might be differentially bound to exosome populations or adsorbed by other contaminants. 15,20 The assay is quantitative, sensitive, and can readily be adapted for use with biomarkers for different cancers or other disease states. In this context, we asked if this rapid format could be used to assess the relative amounts of glypican-1 and the exosome-associated biomarker CD63. Using the ACE microarray chip, we isolated exosomes and other nanoscale material (EVs and cf-DNA) from 25 µL of undiluted pancreatic cancer patient whole blood, serum or plasma in just 20 minutes. Integrated immunofluorescent labeling and analysis of the biomarkers glypican-1 and CD63 were then performed directly on the chip. Quantification of the fluorescently-labeled biomarkers permitted comparison of a small cohort (n=20) of PDAC patient samples both to those from healthy individuals, and to those from patients with benign pancreatic disease or colon cancer. The assay was found to readily discriminate PDAC from healthy samples. Furthermore, when extended to a small set of colon cancer patient samples (n=10), the assay identified the three with metastatic disease. Thus, the assay is rapid and accurate, suggesting that it should be developed further as a powerful diagnostic tool for pancreatic or other cancers.


4

Page 5 of 26 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 Nano

Fig. 1. Schematic diagram illustrates the ACE (AC electrokinetic) direct immunoassay procedure. Undiluted whole blood, plasma, or serum sample is added directly to the chip. A 10minute application of AC current to the chip effects dielectrophoretic separation and isolation of target nanoscale extracellular vesicles (EVs) and other particulates onto the chip electrodes. Unbound material is washed off the chip with 0.5 X PBS during an additional 10 minutes of AC current. After the current is stopped, EVs and nucleic acid targets can be analyzed by different methods, including addition of fluorescent reporter antibodies and nucleic acid dyes that are applied to the chip, then incubated for the appropriate times. Following the final wash step, the chip is ready for direct imaging and analysis of the fluorescent signal.

RESULTS AND DISCUSSION Capturing exosomal biomarkers from PDAC patient blood, serum, or plasma with AC Electrokinetic device We have previously shown that exosomes purified from cultured U87 glioblastoma cells can be selectively immobilized onto the microelectrode chip from either buffer or spiked human plasma, then identified on-chip by labeling with the exosome-selective antibodies CD63 and TSG101 (TSG101

labeling,

as

expected

for

a

lumenal

exosome

protein,

required

prior

permeabilization).23,24 For reasons of simplicity, we will refer to the biomarker-rich material that is captured by the microchips as “exosomes/EVs”, noting that there are also likely to be other types of 50-200 nm particles present, such as low-density lipoproteins.25 We now asked if these


5

ACS Nano 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 26

techniques could be used to isolate, immobilize, and identify exosomal biomarkers directly from whole blood of a patient with pancreatic cancer. To this end, a 25 µl sample of whole blood, taken just two hours earlier from a PDAC patient, was loaded directly onto a microelectrode chip, subjected to dielectrophoretic isolation, and incubated with antibodies. Figure 2 (panels AC) illustrates the visible collection of large blood cells and debris during the dielectrophoresis step into low-field regions between the microelectrodes, followed by their removal during the wash. The exosomes, EVs and smaller nano-particulates remaining around the electrodes were found to be strongly positive for the PDAC exosome-selective biomarkers glypican-1 and CD63 (panels D and E), whereas control samples with primary antibodies omitted were not strongly labeled (panels G and H).

Fig. 2. Whole blood from pancreatic cancer patient is glypican-1 and CD63 positive in ACE Immunoassay. Panels A-C: Whole blood drawn from a patient with PDAC was applied to an ACE chip (A); application of current caused larger whole blood cells and particulates to migrate to low-field regions between electrodes (B). A buffer wash removed most c ells and particulates from the chip (C). Panels D-F: Antibodies against glypican-1 and CD63 were incubated together on the chip, followed by fluorophore-conjugated secondary antibodies for detection (D, CD63, green; E, glypican-1, red; F, brightfield image). Panels G-I: As negative controls, primary antibodies were omitted, and the fluorophore-conjugated secondary antibodies alone did not label the electrodes (G, green channel; H, red channel; I, bright field image).


6

Page 7 of 26 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 Nano

Because patient whole blood samples are not readily available, we obtained a set of archival, presurgery samples from patients diagnosed with pancreatic cancer to investigate if plasma and serum fractions could be used to optimize the assay (Supplementary Data, Table I). Patient samples were loaded directly onto two separate microelectrode chips and processed as above to verify the presence of the exosome-selective markers CD63 and TSG101. Figure 3 shows the identification of both CD63- and TSG101-positive material immobilized onto the circular electrodes from just 25 µL of sample, and demonstrates that both biofluids can be used in the assay to produce clear results. Use of secondary antibodies alone, in the absence of primary antibodies, was found to produce minimal background. Similar results were obtained using either directly-conjugated mouse anti-CD63 antibody, unconjugated mouse anti-CD63 antibody visualized with a fluorophore-conjugated secondary antibody, or CD63 antibody from a different host animal (rabbit) with a different secondary antibody (not shown). Because processing of plasma samples is likely to be more consistent, we chose to focus on plasma, and using CD63 as a biomarker instead of TSG101 eliminated the need for a permeabilization step.

Fig. 3. TSG101-, CD63, and glypican-1 positive exosomes are captured by ACE chips from either plasma or serum of pancreatic cancer patients. Panels A-F: Twenty-five ul of plasma (upper panels) or serum (lower panels) was applied to two ACE chips, and current applied to isolate and immobilize exosomes onto the chips. Following a wash, samples were permeabilized with saponin. Panels A and D show the presence of TSG101 on each chip, detected using rabbit anti-TSG101 antibody followed by Alexa Fluor 594 (red) anti-rabbit secondary antibody. Panels B and E show the presence of CD63 also on each chip, detected with mouse anti-CD63 antibody followed by Alexa Fluor 488 (green) anti-mouse secondary antibody. Panels C and F, bright field images corresponding to panels A and B, or D and E and showing position of the circular electrodes. Panels G-L: Twenty-five ul of plasma was applied to one ACE chip, and current applied to isolate and immobilize exosomes onto the chip. The chip was then incubated in both mouse anti-CD63 and rabbit anti-Glypican-1 antibody, washed, and incubated again with both Alexa Fluor 488 (green) anti-mouse and Alexa Fluor 594 (red) anti-rabbit secondary antibodies (Panels G and J). Panels H and K, controls with primary antibody omitted, showing minimal labeling due to Alexa Fluor 488 (green) anti-mouse or Alexa


7

ACS Nano 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 26

Fluor 594 (red) anti-rabbit secondary antibodies alone. Panels I and L, bright field images showing position of the circular electrodes in each field.

We next investigated if we could detect the pancreatic cancer-selective marker glypican-1 on exosomes from PDAC patient plasma samples that were isolated by this technique. As shown in Figure 3 G-L, exosomes and other material isolated onto the ACE microelectrode chip from pancreatic cancer patient plasma were shown to be positive for both CD63 and glypican-1 biomarkers. When secondary antibodies were used alone (in the absence of primary antibody), and imaged using the same parameters, minimal fluorescence was observed, showing that the positive fluorescence signal on the microarray chip reflected the presence of the selected antigen targets. Of note, while the current studies were performed chiefly with a polyclonal antiglypican-1 antibody similar to that used for the Melo et al. study, other antibodies (e.g., a monoclonal anti-glypican-1 antibody) produced similar results.5

Quantifying fluorescent exosomes on ACE microarray chips To determine if there is a direct relationship between the amount of sample applied to the chip, and the degree of fluorescence observed for the ACE-immobilized material, we first used exosomes purified from U87-MG cultured human glioblastoma cell supernatants. This cell line has been used extensively as a source of purified exosomes.26,27 The purified exosomes were dyed with the non-specific fluorescent membrane dye PKH, and the concentration of EV particles was determined by NanoSight prior to dielectrophoresis. Purified exosomes were either undiluted or diluted into 0.5 X PBS buffer. Samples with different numbers of particles, ranging from 0 to 130 X 109, were subjected to AC dielectrophoresis and imaged.

Photographic images corresponding to each sample were quantified using ImageJ software (NIH). In brief, the intensity of PKH-related fluorescence, representing numbers of exosomes captured onto the electrode array, was converted to pixel number (relative fluorescence units, RFU) using ImageJ. A representative section of each chip was quantified, and background


8

Page 9 of 26 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 Nano

fluorescence for that region was subtracted. Figure 4A shows that there is a direct correlation between the concentration of sample applied and the resulting fluorescence of the ACEimmobilized exosomes measured. The assay can detect a difference of approximately 300-fold in number of exosomes applied, with a maximum that was limited only by the original concentration of purified exosomes.

These results demonstrate that capture of PKH-dyed

exosomes onto the ACE chip is quantitative and correlates with the number in the original sample.

Fig. 4. Measurement of relative fluorescence of exosome-related material captured onto ACE chips directly corresponds to the amount of sample or exosomes applied. Dilutions of either purified, PKH- (red fluorescence) labeled exosomes (Panel A) or a pancreatic cancer patient plasma sample (Panels B and C) were applied to ACE chips and subjected to DEP. Chips were either viewed directly (Panel A), or labeled with fluorescent antibodies before imaging (Panels B and C). Top right side, images of ACE chips; bottom right side, direct 3D representations of chip images (Image J). Left side, quantitation of relative fluorescence for each sample or dilution.


9

ACS Nano 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 26

Quantifying antibody-labeled PDAC patient exosomes on ACE microarray chips To determine if the quantitation method can distinguish different concentrations of fluorescent antibody-tagged material, as well as for directly dyed, purified exosomes, we next investigated if this relationship extends to pancreatic cancer patient exosomes isolated and immobilized directly from plasma onto the ACE microelectrode chip. Figure 4B shows that when increasing amounts of plasma are applied to the chip, there is a corresponding increase in fluorescently-labeled CD63, an exosome biomarker protein. This linear relationship extends from undiluted plasma to plasma diluted as much as 20-fold. To better visualize the range of fluorescence intensities on the chip, a direct depiction of the fluorescently-labeled ACE chips is shown by 3D graph (Figure 4, right side, lower panels). We next examined whether fluorescent labeling of glypican-1, a cancer-specific exosome biomarker, could also yield quantitative results. Similar to CD63, for glypican-1 we also found a correspondence between the concentration of plasma sample applied and the measurement of fluorescence intensity.

When tested in separate experiments,

measurements of glypican-1 specific fluorescence using the same plasma sample were found to be very similar (sample #P6, 9.47 RFU +/- 0.15; Supplementary Figure 1A). Overall, these results support the idea that the ACE direct immunoassay results are reproducible and can be quantitative. Screening biomarker levels in PDAC patient versus healthy donor samples We next tested the utility of the assay in comparing relative levels of both glypican-1 and CD63 in undiluted plasma from healthy persons (“normal”), pancreatic cancer patients (“PDAC”), and patients with other benign pancreatic diseases (“BPD”; operators were blinded to all patient medical and pathology information; see Supplementary Figure 1). When a set of normal plasma (n=11) was tested for glypican-1, serving as negative controls, levels were found to be clustered in the lower range, with an average of 2.2 RFU (95%CI 1.8, 2.6) (Figure 5A). The assay was next used to screen a series of archival samples, taken from patients prior to surgical resection for PDAC (n=20; Supplementary Table 1). In these pancreatic cancer patient plasma samples, glypican-1 levels were found to be much higher, with an average of 9.1 RFU (95% CI 7.1, 11.1); all but one of the PDAC samples were above 4.0. Comparison of the PDAC with the normal


10

Page 11 of 26 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 Nano

group demonstrated that mean differences in glypican-1 labeling are significant (student’s t-test, p< 0.0001). When glypican-1 for PDAC samples was compared to BPD samples (average of 5.3, 95% CI 3.7, 6.9), however, mean differences were not significant for the small set tested (p = 0.0651). Glypican-1 labeling intensity of the BPD samples was higher than that of the normal, negative controls (p=0.0003). Our BPD set was enriched for benign pancreatic neoplasms, such as intraductal papillary mucinous neoplasms (IPMNs) and neuroendocrine tumors, that arguably are pre-malignant and not truly “false-positives”. Yang et al. found a similar elevation of IPMN over control samples using a nanoplasmonic assay with GPC1 and four other markers. 28 It is of interest to note that in another recent study, the KRASG12D DNA mutations characteristic of pancreatic cancer were detected also in exosomes from certain patients with pancreatitis or IPMN,29 suggesting that such pathologies may be on a pre-cancerous pathway, and thus detected by our assay.

Figure 5. ACE integrated biomarker assay of glypican-1 and CD63 rapidly discriminates between PDAC and normal patient plasma. Plasma samples from PDAC patients (N=20), benign pancreatic disease (N=7), or normal, healthy persons (N=11) were assessed for the presence of glypican-1 (Panel A) and CD63 (Panel B). Glypican-1 levels in both PDAC and BPD groups were elevated as compared to the normal group (unpaired student’s t-tests, glypican-1: PDAC, p

Integrated Analysis of Exosomal Protein Biomarkers on Alternating

Recommend Documents