KR20230018174A - An integrated magneto­electrochemical device for the rapid profiling of tumour extracellular vesicles from blood plasma - Google Patents

Abstract

The present invention relates to an integrated magneto-electro-chemical device which performs rapid profiling of CRC-derived detectable EVs (CRC-EVs) from plasma and a method thereof. To this end, provided are the magneto-electro-chemical device which comprises: a plurality of electrode arrays loaded with target analyte binding magnetic beads; an electrostatic potentiostat electrically coupled to each of the electrodes of the electrode arrays; a DAC electrically coupled to the input of the electrostatic potentiostat to convert a digital signal into an analog signal; an ADC electrically coupled to the output of the electrostatic potentiostat to convert a voltage signal of the electrostatic potentiostat into the digital signal; and a microcontroller electrically coupled to the DAC, transmitting a control command with the electrostatic potentiostat through the DAC or electrically coupled to the ADC to determine a target analyte through the digital signal transmitted from the ADC. According to the present invention, the total analysis time is less than 1 hour (96-well format), and the analysis signal is read in parallel from all detection probes (96) to exhibit very fast processing speed.

Description

An integrated magnetoelectrochemical device for the rapid profiling of tumor extracellular vesicles from blood plasma}

The present invention relates to a diagnostic kit or device, and more particularly, to a colorectal cancer diagnostic kit or device capable of diagnosing colorectal cancer.

Evaluation of circulating biomarkers, also known as liquid biopsy, is a novel approach to obtaining molecular information about a patient's tumor through repeated but minimally invasive sampling.

One attractive target for such evaluation is extracellular vesicles (EVs), membrane particles secreted by cells.

In addition to being generally abundant and stable, EVs have been reported to transport parental biomolecules (e.g. proteins, nucleic acids and lipids).

In particular, analysis of tumor-derived EVs has significant potential to improve current cancer diagnosis by elucidating the dynamic state of tumors.

However, setting up clinical EV testing is fraught with several technical challenges, such as (1) laborious manual sample preparation, (2) limited sensitivity and throughput of existing tools, and (3) high cost of test equipment or analysis (e.g., sequencing). are facing

Korean Patent Publication No. 10-2014-0050465 (published on April 29, 2014)

The present invention has been made to solve the above problems, and provides a colorectal cancer diagnosis kit capable of diagnosing colorectal cancer.

To this end, one aspect of the present invention includes a plurality of electrode arrays loaded with target analyte-binding magnetic beads; an electrostatic potential electrically coupled to each of the electrodes of the electrode array; a DAC electrically coupled to the input of the potentiometer to convert a digital signal into an analog signal; an ADC electrically coupled to the output of the potentiostat to convert the voltage signal of the potentiostat into a digital signal; And a microcontroller electrically coupled to the DAC to transmit a control command to a potentiometer through the DAC or electrically coupled to the ADC to determine a target analyte through a digital signal transmitted from the ADC; A magneto-electrochemical device is provided.

In addition, each electrode of the plurality of electrode arrays is characterized in that it is composed of three individual electrodes of a working electrode, a counter electrode and a reference electrode.

In addition, each electrode of the plurality of electrode arrays is connected to the static electricity and push-pin connectors, and is characterized in that it is disposed on a built-in magnet to concentrate the magnetic beads.

In addition, the electrostatic potential is electrically coupled to the working electrode and electrically coupled to the first operational amplifier maintaining a predefined potential difference between the working electrode and the reference electrode, the reference electrode and the counter electrode, and the working and a second operational amplifier consisting of a transimpedance amplifier that converts the induced current across the counter electrode from the electrode into a voltage signal.

In addition, a multiplexer formed in front of the ADC to select from a plurality of inputs when an output from the potentiometer is electrically coupled to an input of the ADC is further included, and the microcontroller is configured to select the multiplexer and the number of corresponding inputs. It is characterized in that the ADC is configured to sequentially access each electrode.

In addition, the target analyte is characterized in that the colorectal cancer extracellular vesicle (CRC-EV).

In addition, the extracellular vesicles are characterized by spiking from plasma.

In addition, the colorectal cancer diagnostic biomarker is any one of CD44, B7-H3, EGFR, ABCG2, GPA33, EpCAM, HER2, MUC1, MET, CD44v6, CD24, CD166, STEAP1 and ALDH1, MRP1, MDR1, TS and CD133 or It is characterized by being selected by a combination thereof.

Preferably, the colorectal cancer diagnostic biomarker is characterized in that it consists of a combination of EGFR, EpCAM, CD24 and GPA33.

In addition, the microcontroller determines a cutoff value that maximizes the sum of sensitivity and specificity in the ROC curve, defines the determined cutoff value as an EV _CRC score, and determines whether colorectal cancer is present based on the defined EV _CRC score It is characterized by doing.

Another aspect of the present invention is to (a) bind a target analyte to a plurality of magnetic beads, allow them to bind to a reactive enzyme after a plurality of washes, and combine with an electron mediator solution to form target analyte-binding magnetic beads on a plurality of electrode arrays. loading step;

(b) exposing the target analyte-binding magnetic beads to a magnetic field by electrically coupling the plurality of electrode arrays to a potentiostat;

(c) inducing an oxidation-reduction reaction between electron mediators in the target analyte-binding magnetic beads and the reactive enzyme; and

(d) a diagnosis method using a magneto-electrochemical device comprising the step of determining a target analyte based on the output result of the potentiostat is provided.

In addition, the reactive enzyme includes HRP, and the electron mediator includes 3,3', 5,5'-tetramethylbenzidine (TMB).

Also, in step (d), the target analyte is characterized in that the colorectal cancer extracellular vesicle (CRC-EV).

In addition, the extracellular vesicles are characterized by spiking from plasma.

In addition, the colorectal cancer diagnostic biomarker is any one of CD44, B7-H3, EGFR, ABCG2, GPA33, EpCAM, HER2, MUC1, MET, CD44v6, CD24, CD166, STEAP1 and ALDH1, MRP1, MDR1, TS and CD133 or It is characterized by being selected by a combination thereof.

Preferably, the colorectal cancer diagnostic biomarker is characterized in that it consists of a combination of EGFR, EpCAM, CD24 and GPA33.

In addition, a cutoff value that maximizes the sum of sensitivity and specificity in the ROC curve is determined, the determined cutoff value is defined as an EV _CRC score, and colorectal cancer is determined based on the defined EV _CRC score. .

The present invention provides a clinically adaptable high-throughput EV analysis device and method called high-throughout integrated magneto-electrochemical extracellular vesicle (HiMEX).

This simplifies EV analysis by combining EV enrichment and electrochemical detection in a single assay, enabling direct EV protein profiling from clinical samples.

The total analysis time is short (96-well format), less than 1 hour, and the analysis signal is read in parallel from all detection probes (96 probes), showing very fast processing speed.

In addition, HiMEX analysis revealed the versatile potential of EVs for CRC management.

First, EVs that reflect major protein features of the parental tumor are present in the circulation, and these CRC-derived detection EVs (CRC-EVs) have led to highly accurate cancer diagnosis (overall accuracy >96%).

Second, the continuous change in CRC-EV may be related to the patient's treatment response. In particular, the level of CRC-EV decreased in all patients after curative surgery, but rebounded from the baseline value in each patient with tumor recurrence. This makes it possible to predict the possibility of tumor recurrence.

Third, preoperative CRC-EV levels show a significant correlation with patients' 5-year disease-free survival (DFS) (n = 90), which effectively classifies patients into high-risk and low-risk groups, which can contribute to treatment prevention ..

Thus, these results can influence timely and better informed cancer treatment and expand the clinical utility of EVs in CRC diagnosis, recurrence monitoring and prognosis.

1 is a diagram showing a HiMAX approach for clinical EV analysis according to a preferred embodiment of the present invention.
1A is a conceptual diagram of a study design according to the present invention. Referring to FIG. 1A , blood samples were collected from CRC (n=91) patients during standard clinical treatment. EV and conventional markers (eg, CEA and CA19-9) were analyzed in blood samples and cross-compared with clinical outcomes including immunohistology, radiographic reports and survival.
Figure 1b is a diagram showing a two-step HiMEX analysis protocol according to the present invention. Referring to Figure 1b, EVs are enriched through immunomagnetic capture. Bead-bound EVs are marked with probing antibodies for signal generation by electrochemical reaction. The assay is simple and fast (<1 hour) and directly analyzes the plasma sample without a purification step.
Figure 1c shows the composition of the colorectal cancer diagnosis kit (HiMAX reader) according to the present invention it is a drawing Referring to Figure 1c, a small HiMEX reader with a touch screen interface was developed. The reader accommodated an array of 96 electrodes in a conventional 96-well plate. Beneath each electrode is a push-pin connector to make electrical contact and a magnet to focus the bead-bound EVs.
Figure 1d is a block diagram showing the configuration of the HiMAX reader of Figure 1c. Referring to Fig. 1d, the HiMEX reader is designed to perform 96 parallel measurements within 2 minutes. Each electrode was connected to its own potentiostat. The microcontroller initiates the electrochemical reaction by applying a potential through a digital-to-analog converter (DAC). The resulting current of the electrodes was read by an analog-to-digital converter (ADC) through rapid multiplexing (MUX). A microcontroller processes the data and displays it on the touchscreen. Readers can also communicate with external devices (e.g. smartphones) for data logging.
2 is a diagram showing optimization and characteristics of HiMAX analysis according to the present invention.
2a is a diagram showing a scanning electron micrograph image of a microbead after incubation with an EV sample. Referring to Figure 2a, beads (diameter, 3 μm) functionalized with an antibody to CD63 captured EVs isolated from cell culture medium (HCT116 cell line). Representative images are selected from technical replicate samples.
Figure 2b is a diagram of capturing EVs using three types of magnetic beads specific to different tetraspanins and their mixtures (Mix). Referring to Figure 2b, three types of magnetic beads specific to different tetraspanins and their mixture (Mix) were used to capture EVs, and then the expression of major tetraspanins in the captured EVs was measured. Bead cocktails most efficiently overcome EV heterogeneity. Representative images are selected from duplicate measurements. Full western blot images are shown in Supplementary Information.
Figure 2c is a graph showing the capture of EV (10 ⁷ ㎖ ^-1 ) from the HT29 cell line and additionally labeled with EGFR. Referring to Fig. 2c, the mixed bead type led to the highest HiMEX signal. Data are means±sd in technical triplicates.
Figure 2d is a graph comparing the HiMAX assay according to the present invention and the conventional ELISA assay. Referring to FIG. 2D , the HiMEX assay has better sensitivity and a wider dynamic range than ELISA (enzyme-linked immunosorbent assay). The LOD is about 10 ⁴ EVs ml ^-1 for HiMEX and about 10 ⁷ EVs ml ^-1 for ELISA in arbitrary units. EV numbers were estimated from nanoparticle tracking assays. Data points represent mean values of technical overlap.
Figure 2e is a graph profiling with a set of protein markers for EVs of different CRC cell lines. Referring to Figure 2e, the HiMEX results showed good agreement with the ELISA analysis results (Pearson's correlation coefficient, r = 0.82). To obtain the HiMEX expression (ξ) of the target protein marker (M), the marker-related current level was normalized to the current level of the loading control, CD63. For ELISA, the same amount of EV (10 ⁹ EVs ㎖ ^-1 ) was used for each marker. Data were presented on a logarithmic scale to better display small values. The blue dotted line represents the best linear fit on a log-log scale and the shaded gray area represents the 95% confidence band. HiMEX data represent mean values of technical replicates.
3 is a diagram showing EV profiling results for CRC detection according to the present invention.
3A is a graph showing images of tumor tissue (top) and plasma EV analysis (bottom). Referring to Figure 3a, three major CRC protein markers (EpCAM, EGFR and CD24) were evaluated in tumor tissue (immunohistochemistry) and blood samples (HiMEX) of each patient. Expression profiles showed qualitative agreement. Data from two representative patients are shown. Data are means ± sem in technical triplicate measurements. A full image of this sample is provided in Supplementary Information.
Figure 3b is a diagram showing the expression of EpCAM, EGFR and CD24 in tumor tissues (left) and plasma EV samples (right) of 12 CRC patients. Referring to FIG. 3B , tissue staining was graded by the percentage of marker-positive cells out of total cancer cells. Pathological scores and EV expression profiles showed a significant correlation (Spearman rank coefficient ρs=0.65, P<0.0001, two-tailed t-test).
3c is a HiMEX analysis diagram for CRC diagnosis. Referring to FIG. 3C , the upper drawing is a training set in which plasma samples from 58 preoperative CRC patients and 25 healthy controls were analyzed. The expression of 5 CRC markers (EpCAM, EGFR, CD24, CD133 and GPA33) was measured by HiMEX. In the lower figure, the diagnostic metric, EVCRC, was determined as the weighted sum of the four marker levels (EpCAM, EGFR, CD24 and GPA33) by logistic regression.
Figure 3d is a diagram showing the average levels of five CRC markers (EpCAM, EGFR, CD24, CD133 and GPA33) in CRC patients and controls. Referring to Fig. 3d, the average level for each of the indicated markers was higher in CRC patients than in healthy controls. However, marker distributions overlap between patients and controls, reducing the ability to classify single markers.
3e is a graph showing ROC curves for 5 CRC markers (EpCAM, EGFR, CD24, CD133 and GPA33). Referring to FIG. 3E, the area under the curve (AUC) of EV _CRC was 0.98, significantly larger than that of single markers (all P<0.001). The cutoff level that maximized the sum of sensitivity and specificity in the ROC curve was determined.
3f and 3g are diagrams showing EV _CRC for a control group and a CRC patient. Referring to Figures 3f and 3g, EV _CRC effectively classifies control and CRC patients (P=0.0009, unpaired two-tailed t-test) (f). The diagnostic accuracy in the training cohort was 98%. The cutoff value of the EV _CRC ROC curve was 4.96 (g).
4 is a diagram showing a prospective cohort analysis for CRC diagnosis according to the present invention.
The top of Fig. 4a shows 33 patients with CRC, 9 healthy donors, and 6 patients with non-CRC disease (gastrointestinal stromal tumor, n=2, small intestine prolapse, n=2) for expression of EpCAM, EGFR, CD24 and GPA33. 2, appendicitis, n=2) was analyzed for plasma EV. In the lower part of Fig. 4a, the EV _CRC was calculated by the same formula as for the training cohort.
Referring to Figure 4b, EV _CRC levels were significantly higher in CRC patients than in healthy controls validating the EV-based diagnostic algorithm. The same cutoff value (4.96) from the training set was applied.
Referring to Fig. 4c, EV _CRC was still superior in CRC detection with an AUC much greater than that of a single marker. Detailed statistics are provided in Table 2.
5 is a diagram showing HiMAX analysis of longitudinal samples from CRC patients.
Referring to FIG. 5A , blood samples of CRC patients were analyzed before and after surgery (n=13).
Molecular EV profiling showed that EV _CRC values were reduced in all patients (P<0.0001, paired two-tailed t-test) (middle left). whereas total EV concentration (P=0.59, paired two-tailed t-test) (left) and conventional serum markers CEA (P=0.19, paired two-tailed t-test) (middle right) and CA19-9 (P=0.19 , paired two-sided t-test) (right) showed no significant change. Each data point in the graph represents the mean value of technical overlap.
Referring to FIG. 5B , since some patients received chemotherapy, plasma EVs were additionally monitored after surgery. EV _CRC values rebounded in all patients (n=7) with recurrent tumors (left and middle). In contrast, in patients without recurrent tumors (n=4) (right), EV _CRC values continued to decrease or stabilize at postoperative levels. Data are means±s.e.m. from technical replicates.
Referring to FIG. 5C , EVs from patients with recurrent (n=3) and non-recurrent (n=3) CRC were analyzed for mRNA expression. Highly variable genes were plotted. Genes involved in DNA repair, protection against oxidative pressure and chemoresistance (5-FU) showed higher levels in EVs from patients with recurrent tumors.
Referring to FIG. 5D , CRC patients (n=90) were monitored for DFS for up to 5 years. Kaplan-Meier estimates for DFS were plotted stratified by preoperative EV _CRC level. High EV _CRC values (≥53.4) were associated with poor prognosis (P=0.02, two-tailed log-rank test).

Hereinafter, a colorectal cancer diagnosis kit according to a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

First, for EV diagnosis to be clinically useful, a new integrated method for the isolation and molecular analysis of EVs, ideally a method capable of high-throughput operation, is needed. Additionally, it is equally important to analyze large clinical samples to establish robust EV biomarker baselines for disease states.

Here we describe the development of a clinically adaptable high-throughput EV assay technology called high-throughout integrated magneto-electrochemical extracellular vesicle (HiMEX).

This technology simplifies EV analysis by combining EV enrichment and electrochemical detection in a single assay, enabling EV protein profiling directly from clinical samples.

The total assay time was less than 1 hour (96-well format) and assay signals were read from all detection probes 96 in parallel.

HiMEX is then applied to clinical sample analysis to demonstrate the practical benefits of HiMEX.

Specifically, we analyzed blood samples from 102 colorectal cancer (CRC) patients who underwent surgery and chemotherapy and 40 non-CRC controls (i.e., patients with non-CRC disease such as appendicitis or internal hernia in the small intestine). (n = 142 in total).

As follows, HiMEX analysis revealed the diverse potential of EVs for CRC management.

(1) EVs reflecting major protein features of the maternal tumor are present in the circulation, and these CRC-derived detection EVs (CRC-EVs) have led to highly accurate cancer diagnosis (overall accuracy >96%).

(2) The continuous change of CRC-EV may be related to the patient's treatment response. In particular, the level of CRC-EV decreased in all patients after radical surgery, but at the baseline value of each patient with tumor recurrence. rebounded

(3) Preoperative CRC-EV levels show a significant correlation with patients' 5-year disease-free survival (DFS) (n = 90), which effectively classifies patients into high-risk and low-risk groups.

These results could impact timely and better-informed cancer treatment and expand the clinical utility of EVs in CRC diagnosis, recurrence monitoring, and prognosis.

Accordingly, the HiMEX assay for clinical EV testing and the results thereof are described in detail below.

<result>

<HiMEX Approach for Clinical EV Testing>

As mentioned above, in this study, we aimed to evaluate HiMEX for application in CRC diagnosis, treatment monitoring and prognosis (Fig. 1a and Supplementary Fig. 1).

Overall, we analyzed plasma samples from 102 patients with CRC and 40 non-CRC controls (n=142 total).

For CRC diagnosis, we first defined CRC-EV signatures based on published results, in vitro studies and tissue immunohistochemistry, and then measured these markers in plasma EVs.

A total of 131 patient samples were used. That is, a training cohort consisting of 25 non-CRC controls and 58 CRC patients; and a test cohort of 15 non-CRC controls and 33 CRC patients.

An additional 11 CRC patients who received standard clinical care were followed for patient monitoring.

Serial blood samples were collected at defined time points (i.e., perioperatively and during chemotherapy) and analyzed for CRC-EV markers.

Then, the EV profiling results, for example, levels such as carcinoembryonic antigen (CEA) and carbohydrate antigen (CA19-9) as existing tumor markers, radiological evaluation and treatment results Comparison was made with clinical information included.

Finally, for the prognostic use of EV profiling, we analyzed the correlation between 5-year DFS and preoperative CRC-EV levels in 90 CRC patients.

To simplify EV analysis, the HiMEX technology was optimized to integrate sample preparation and measurement within a single assay (Fig. 1b).

We enriched target-specific EVs using immunomagnetic separation, allowing rapid (~15 min) isolation of EVs directly from native samples without further manipulation. Then, magnetic beads were used as a substrate for subsequent signal generation. In addition, bead-bound EVs were labeled with probing anti-bodies functionalized with catalytic enzymes for electrochemical reactions.

The HiMEX strategy has the following advantages.

(1) Simple sample handling (i.e., magnetic pulling) without the extensive purification of EVs required.

(2) It has fast binding kinetics and high-efficiency assay results in EV capture and labeling (about 30 minutes in total).

(3) the assay signal was amplified by an enzymatic reaction;

(4) The detection modality (i.e., electrical measurements) was scalable for high-throughput measurements.

<HiMEX detection system for parallel measurement>

To exploit the unique technical advantages of HiMEX, a compact system capable of parallel measurement was developed (Fig. 1c).

That is, the system is compatible with 96-well plates and enables batch processing.

For EV capture and labeling, a conventional 96 well plate, i.e., a custom-designed magnet array (Supplementary Fig. 2a), was used for bead processing (e.g., washing and buffer exchange).

EV-bound beads were then spotted on a 12x8 probe array (Supplementary Fig. 2b) loaded into the detection device.

Each probe in the array containing three electrodes (i.e., a working electrode, a counter electrode, and a reference electrode) was connected to a potentiostat with quick push-pin connectors and a small built-in to focus the magnetic bead. was placed over a magnet (Fig. 1c and Supplementary Fig. 2c).

The detection system also featured a touchscreen interface for device control and data display, and communicated wirelessly (via Bluetooth) with an external server for data storage.

A constant voltage was applied between the working electrode and the reference electrode while monitoring the current between the working electrode and the counter electrode as an analysis signal.

The system electronics were designed to quickly (100 Hz) poll each probe during measurement to enable fast readings (Fig. 1d).

That is, the embedded microcontroller sequentially accessed individual probes through a configuration of 6 analog-to-digital converters (ADCs) and a 6-to-1 multiplexer (MUX).

This scheme effectively executed real-time, parallel measurements (Supplementary Fig. 2d).

The total read time for all 96 probes was less than 2 minutes.

The performance of the system was benchmarked using K ₄ Fe(CN) ₆ samples at various concentrations, with particular emphasis on evaluating the reproducibility between the 96 probes.

For a given K ₄ Fe(CN) ₆ concentration, the different probes in the array reported consistent signals (Supplementary Fig. 3a).

Four standard curves generated from parallel measurements of the K ₄ Fe(CN) ₆ dilution series showed an excellent linear relationship between current levels and K ₄ Fe(CN) ₆ concentrations (Supplementary Fig. 3b).

Notably, these curves are statistically identical, demonstrating high probe-to-probe fidelity (i.e., low intra-probe variability) for parallel detection.

<HiMEX analysis features>

Next, the key assay steps, i.e. EV capture and electrochemical detection, were optimized to maximize the assay signal.

For EV capture, three types of magnetic beads specific to one of the EV-abundant tetraspanins (i.e., CD63, CD9, and CD81) were prepared.

When mixed with EV, these beads were effectively covered with vesicles (Fig. 2a). That is, control beads conjugated with IgG antibodies showed minimal non-specific binding (Supplementary Fig. 4).

In the present invention, a single type or capture bead cocktail was additionally used for EV pull-down and the level of EV tetraspanin was compared (Fig. 2b).

The mixture of capture beads (CD63, CD9 and CD81) produced higher signals at all tetraspanin levels than the single marker beads alone (Fig. 2b and Supplementary Fig. 5a), presumably reflecting EV heterogeneity.

Single EV imaging (Supplementary Figure 6) supported the observation. That is, the number of EVs visible under the microscope was highest when a cocktail of CD63, CD9 and CD81 antibodies was used for fluorescent staining.

Immunomagnetic capture was also efficient in plasma (Supplementary Fig. 5b).

Based on CD63 levels before and after immune capture, the pull-down yield was estimated to be 82%.

To generate a detection signal, we use a probing antibody to label the captured EVs with horseradish peroxidase (HRP) and conjugate the conjugate with a chromogenic electron mediator (3,3′, 5,5′- tetramethylbenzidine (TMB)).

The redox reaction quickly reached equilibrium with the current stabilizing within 1 min (Supplementary Fig. 7).

The current was recorded between 50 and 55 seconds after the initiation of the reaction, and the average value (I) was used as an analysis index.

A three bead cocktail was used for EV capture in the rest of the study, as it consistently produced maximal HiMEX signals (Fig. 2c).

The next step was to characterize the analytical capabilities of HiMEX.

First, samples were prepared with various concentrations of EV.

EVs were captured and labeled with CD63 using a mixture of CD63, CD9 and CD81 beads. That is, a control sample was generated by labeling the captured EV with an IgG antibody.

These samples were then measured using the HiMEX detector (Supplementary Fig. 7).

To minimize the effects of common-mode errors (e.g. temperature fluctuations and non-specific binding), the net current, i.e., ΔI _CD63 =I _CD63 - I _IgG , was used as an analytic metric. .

From the titration results (Fig. 2d), the limit of detection (LOD) of the HiMEX technology was estimated to be about 10 ⁴ EVs, more than 1,000 times lower than that of conventional ELISA assays.

HiMEX also had a much wider dynamic range (about 10 ⁵ EVs) than the ELISA assay (less than 10 ³ EVs).

Next, assay reproducibility was evaluated.

Over 3 days, we prepared EV samples and probed them for CD63, CD9 and CD81 using HiMEX arrays.

Signals observed for a given marker were statistically identical on different days (Supplementary Figure 8).

In addition, signal levels were further compared with EVs spiked in either plasma or serum (Supplementary Fig. 9) and similar signal changes were observed in both sample types.

Plasma was also chosen because it is considered a physiological vehicle for EVs and has fewer platelet-derived vesicles.

A competition assay was performed to test whether EVs captured by the bead mixture contained CRC-related markers (Supplementary Fig. 10a).

Samples were prepared by spiking human plasma with EVs from a CRC cell line (SW480 cells, positive for EpCAM and CD24).

EV immunomagnetic capture was performed in the presence of excess free-floating capture antibody (mixture of CD63, CD9 and CD81).

Bead-bound EVs were then examined for expression of CD63, EpCAM and CD24.

The measured HiMEX signal decreased with increasing concentration of free antibody (Supplementary Fig. 10b), indicating that the bead-captured EVs have CRC markers.

The following protocol was set up for quantitative HiMEX analysis.

The HiMEX signal (ΔI _M = I _M - I _IgG ) was measured after EV capture and labeling for a given marker of interest (M).

ΔI _CD63 was also measured using an aliquot of the same sample as an EV loading control.

The expression level of the target marker (ξ _M ) was estimated by scaling the I _M relative to the loading control (ξM = ΔI _M /ΔI _CD63 ).

Methods for profiling EVs for other protein markers were applied.

Results showed a good linear correlation (Pearson r=0.82) with ELISA (Fig. 2e).

<High-throughput screening of cell-line derived EVs>

A custom-designed bioinformatics pipeline was applied to a public database to select early EV markers relevant to CRC detection (Methods and Supplementary Fig. 11).

Protein expression profiles for CRC and healthy tissue were retrieved from the Human Protein Atlas and markers overexpressed in CRC were selected. This collection is then cross-referenced with UniProt annotations to select proteins containing extracellular domains.

The list was further narrowed by applying two criteria: (1) presence of reported markers in EVs (Vesiclepedia database), and (2) availability of ELISA-compatible antibodies (Supplementary Table 1).

This algorithm selected 9 markers (CD44, B7-H3, EGFR, ABCG2, GPA33, EpCAM, HER2, MUC1 and MET). In addition, the list was added to include markers found to be overexpressed in CRC (CD44v6, CD24, CD166, STEAP1 and ALDH1) and markers used for monitoring drug resistance (MRP1, MDR1, TS and CD133). Candidate markers were then measured in EVs derived from CRC cells using the high throughput of HiMEX.

This preclinical study focused on two investigations: That is, (1) marker expression correlation between EVs and their cells of origin; and (2) confirming the presence of candidate markers in EVs.

The cell-line panel included various CRC stages (SW480 cells, Dukes sorted B; DLD-1 and SW620 cells, C: Colo201 cells, D) and drug resistance (irinotecan) phenotypes (HT29 and HCT116). Indicative cells were included.

As a negative control, we profiled EVs from a healthy colon cell line (CCD-18Co) and plasma from healthy donors.

The inclusion of healthy plasma samples was justified because the goal of the analysis was to identify markers overexpressed in CRC-EVs.

EVs were collected and then profiled by HiMEX (Supplementary Fig. 12a), and cell expression was measured by flow cytometry.

Overall, the expression profiles of selected markers for EVs and their parental cells were in close agreement (r = 0.89) (Supplementary Fig. 12b), supporting the use of EVs as cellular surrogates.

EpCAM, EGFR, CD24 and CD133 were selected for further testing on clinical EV samples based on their relatively higher expression in CRC cell lines (Supplementary Fig. 12c).

In addition, GPA33 was included for clinical relevance in CRC diagnosis.

<Definition of EV protein signature for CRC diagnosis in clinical samples>

Next, HiMEX was applied to detect CRC EVs in clinical samples.

First, we investigated whether circulating EVs from CRC patients contained key protein markers found in CRC tissues.

Both preoperative blood samples and intraoperative tissue specimens were obtained from a pilot cohort of 12 patients with CRC.

In addition, the expression of the top three markers (EpCAM, EGFR and CD24) identified in cell line studies was evaluated.

Tissue specimens were stained for these markers (immunohistochemistry), and expression was graded by a pathologist as the fraction of marker-positive cells out of total cancer cells (0 to 1 in increments of 0.1).

EVs were screened for the same markers via HiMEX (Fig. 3a).

Overall, marker expression between tissue and EV samples showed good concordance (Fig. 3b and Supplementary Fig. 13) and a positive correlation (Spearman's rank coefficient ρs=0.65, P<0.0001).

These results indicate the presence of CRC-derived EVs in the circulation, supporting the rationale for evaluating tumor molecular status with EV profiling.

To establish a CRC-EV diagnosis, the cohort was expanded and the entire marker set was used in EV profiling.

Blood was drawn from CRC patients before surgery.

Control samples were obtained from healthy donors as well as patients with non-CRC disease.

All blood samples were collected according to established protocols (Methods) and processed consistently.

Circulating EVs were evaluated by HiMEX for EpCAM, EGFR, CD133, GPA33, CD24 and CD63 levels.

Aliquots of the samples were also analyzed for the clinically relevant serum markers CEA and CA19-9.

3C summarizes the EV marker expression profiles of the initial training cohort (58 CRC patients and 25 non-CRC controls).

Mean levels of all five markers were higher in CRC patients than controls (Figure 3d).

However, because the two distributions overlapped, it was difficult to classify based on a single marker, so a combination of multiple markers was considered. In particular, we used the weighted sum of EV markers whose optimal weight was determined by logistic regression (Methods).

ROC curves were constructed for both individual markers and their combinations (Fig. 3e and Supplementary Fig. 14).

A cutoff value maximizing the sum of sensitivity and specificity was determined for each ROC curve, and a combination of EGFR, EpCAM, CD24 and GPA33 was found to be optimal. Therefore, this metric was defined as the EV _CRC score.

The AUC of EV _CRC was significantly higher than individual EV markers (Figure 3e) and combinations of two or three markers (all P<0.001, DeLong test).

Compared to the five marker combinations, EV _CRC showed statistically equal (P=0.17, DeLong test) classification performance.

Overall, the EV _CRC with a cutoff value of 4.96 showed high diagnostic power in the training cohort (Fig. 3f, 3g), achieving 97% detection sensitivity, 100% specificity and 98% accuracy (Methods and Tables 1 and 2). .

We further tested the statistical model by analyzing samples from a prospective test cohort (33 patients with preoperative CRC and 15 non-CRC controls) (FIG. 4A).

When the same cutoff values of the training set were applied (FIGS. 4b and 4c), the EV _CRC maintained good diagnostic statistics (sensitivity 94%, specificity 100%, accuracy 96%).

The diagnostic accuracy of EV _CRC was superior to conventional serum markers (Supplementary Fig. 15a) due to the CRC-specific nature of the EV assay.

EV _CRC did not show a significant correlation (Supplementary Fig. 15b) with either CEA (ρ _S = -0.12, P = 0.24) or CA19-9 (ρ _S = -0.11, P = 0.29). Combining either of these serum markers with EV _CRC (by logistic regression) did not significantly improve AUC (P=0.20 for CEA addition, P=0.18 for CA19-9 addition, DeLong test).

<Sequential CRC-EV monitoring during clinical treatment>

Next, longitudinal EV profile changes were followed when patients received standard clinical treatment.

First, EV levels before and after surgery were compared (Fig. 5a).

Blood samples (n=13) were collected 24 hours before surgery and within 1 week after surgery.

Total EV load showed no significant change (P=0.634, paired t-test) or directionality.

In contrast, EV _CRC decreased in all patients (P<0.0001, paired t-test), probably due to fewer CRC-derived EVs after curative tumor resection.

The inference was further supported when we analyzed blood samples from patients (n = 4) who underwent abdominal surgery for non-CRC diseases (eg, appendicitis and internal hernia in the small intestine).

For this patient, EV _CRC values remained statistically the same before and after surgery (P=0.77, paired t-test) (Supplementary Fig. 16).

Although the values of CEA and CA19-9 in the patient-wide mean decreased after surgery, the downward trend was not statistically significant in pairwise comparisons (before and after surgery) (P=0.19 for CEA, P=0.19 for CA19-9, P for paired t-test). =0.190).

Additional patients with CRC (n=11) were monitored for 6 months postoperatively to better confirm temporality.

Eight patients in stage II (inadequate lymph node sampling, perforation, or lymphovascular invasion) or stage III patients with risk factors were treated with 5-fluorouracil (5-FU) plus leucovorin and oxaliplatin. Adjuvant treatment (FOLFOX chemotherapy) was used.

The remaining three patients were all stage II and were excluded from treatment (see Supplementary Table 2 for patient details).

Blood samples were collected before and after treatment in patients who received treatment, and at about 6 months after surgery in patients who did not receive treatment.

Tumor recurrence status was later confirmed by surgical excision or radiographic detection of lesions that increased in size over time.

Of the treated patients (n=8), 4 were classified as non-recurring, the rest were classified as recurring, and tumors recurred in all untreated patients (n=3).

HiMEX EV profiling was performed blindly to these classifications, and postoperative EV _CRC values decreased in all patients (Supplementary Fig. 17).

At postoperative EV _CRC baseline, EV _CRC increased in all relapsed patients regardless of treatment status.

In contrast, EV _CRC decreased or stabilized in patients who did not relapse (FIG. 5B).

Thus, continuous changes in EV _CRC (ΔEV _CRC ) showed promise as an indicator of short-term tumor recurrence (P=0.0004, unpaired two-sided t-test) (Supplementary Fig. 18).

Also, changes in CEA (P=0.13, t-test) and CA19-9 (P=0.76, t-test) levels were similar regardless of recurrence.

<EV capture and RNA analysis>

The immunomagnetic capture used in HiMEX allowed us to enrich EVs for CRC protein detection (HiMEX) as well as other molecular analyses.

Here, we applied this technique to assist with EV-RNA analysis in blood samples collected after adjuvant chemotherapy.

EVs were captured using a tetraspanin magnetic bead cocktail as in the HiMEX assay.

Plasma samples from relapsed (n=3) and non-relapsed (n=3) patients were processed. Captured EVs were then lysed and RNA content was sequenced for quantitative gene expression profiling (Methods).

Principal component analysis based on the mRNA readout separated 2 cohorts (i.e., relapsed versus non-relapsed), with more genes upregulated in the relapsed patient group (Supplementary Fig. 19a).

In addition, differential analysis was performed to identify highly variable genes in the samples (Supplementary Fig. 19b). In particular, all of the selected genes showed higher expression in relapsed patients (Fig. 5c). These genes are responsible for DNA damage repair (GADD45A and CDKN1A), protection against oxidative stress (GPX1, GPX4, PRDX6 and UCP2), fatty acid oxidation (ACOT7 and CD36) and resistance to 5-FU (TGFB1, DNAJB6, DNAJC6 and ATF4). is involved in resistance to

<Assessment of CRC-EV signature for prognosis of 5-year DFS>

Finally, the ability of the EV assay to predict DFS was evaluated.

A high CRC-EV burden may indicate a large primary lesion or malignant phenotype.

Therefore, it can be judged that a high level of EV _CRC before surgery may be correlated with the risk of disease recurrence, possibly caused by residual tumor presence or metastasis after surgery.

In addition, 90 patients with CRC (Supplementary Table 3), all of whom had preoperative blood tests and who underwent tumor resection, were prospectively followed for up to 5 years.

EV analysis found that the initial EV _CRC score was higher in the group of patients whose tumors eventually progressed to metastasis (P = 0.041, unpaired two-sided t-test) (Supplementary Fig. 20).

Survival analysis was able to more robustly confirm the association between high EV _CRC values and poor DFS.

Patients were stratified into two groups according to EV _CRC score and a score threshold of 53.4 was chosen to maximize the absolute log-rank statistic between the two strata.

We also found that 13 of 18 patients with a high EV _CRC score reported tumor metastasis during the 5-year follow-up, whereas 21 of 72 patients with an EV _CRC score below the threshold reported tumor metastasis. That is, a comparison of the two survival curves suggested that there was a significant prognostic difference between the two groups (P = 0.02, two-sided log-rank test) (Fig. 5d and methods).

In contrast, preoperative CEA concentrations (cutoff value of 7 ngml−1), often used for CRC prognosis, displayed borderline performance with respect to DFS in this cohort (P = 0.07, 2-sided log-rank test). (Supplementary Fig. 21).

<Discussion>

Analyzing tumor-derived EVs can provide a real-time snapshot of the molecular composition of tumors, potentially providing opportunities for timely and informed clinical intervention.

In the present invention, HiMEX was developed to place these EV-based tests closer to clinical reality.

The HiMEX approach integrates EV isolation and protein detection into a continuous workflow, enabling batch processing of clinical samples.

In particular, HiMEX provides the following technical advantages.

(1) direct use of plasma samples for target-specific EV protein analysis;

(2) excellent detection sensitivity by magnetic concentration and enzymatic signal amplification (about 1,000 times higher than ELISA);

(3) It offers the advantages of a compact device that performs parallel measurements without the use of a mechanical scanner.

In the present invention, a compact HiMEX device using a 96 electrode array was implemented to prove this concept, and it was also shown that the entire HiMEX analysis from EV isolation to analysis was completed within 1 hour.

HiMEX makes EV profiling simple in standard laboratory settings. That is, this technique is cost-effective, fast, and enables parallel EV protein detection.

Immunomagnetic pulling, an initial analytical step in HiMEX, may facilitate EV collection for other molecular analyses. For example, when investigating drug resistance status, we immunomagnetically capture EVs directly from plasma. Then, a portion of the bead-captured EVs was used for CRC protein detection by HiMEX and the rest for target mRNA sequencing.

To further improve the system, detection electrodes can be designed to improve assay sensitivity and throughput.

Miniaturization of the electrodes to the micrometer scale improves the mass detection limit (by reducing the sample volume) and facilitates the construction of dense arrays.

The surface of the sensing electrode may be textured with a nanostructure to obtain exquisite sensitivity.

These improvements enable HiMEX to detect rare EV targets (e.g., mutant proteins, protein modifications, and nucleic acids inside vesicles), enabling comprehensive molecular profiling in a single assay system.

To use HiMEX for CRC detection, bioinformatic investigations were initially performed to determine overexpressed protein biomarkers for CRC diagnosis.

Although many proteins have been associated with CRC, we narrowed it down to a panel consisting of EGFR, EpCAM, CD24 and GPA33.

In particular, the CRC-EV panel achieved high accuracy in the discovery (98%) and test (96%) cohorts, which may be due to the avid release of tumor EVs into the circulation.

Moreover, serial EV profiling revealed a correlation between the molecular signature of EVs and temporal changes in tumor burden.

CRC-EV levels dropped in all patients immediately after surgery.

In patients without short-term relapse, CRC-EV levels continued to decline over time, whereas in patients with relapse, CRC-EV levels rebounded with tumor progression.

We further observed that initial EV burden may be a predictive risk factor for short-term (5 years) tumor recurrence.

Combined, these results will broaden the potential of EVs as CRC biomarkers for diagnosis as well as treatment monitoring and prognosis.

However, several limitations of the current work should be addressed in future studies.

Comparisons between tissue and circulating EVs were performed for three markers (EpCAM, EGFR and CD24) and were limited by insufficient amounts of tissue for immunohistochemistry optimization.

Expansion of both markers (e.g., CD133 and GPA33) and the number of samples will firmly establish the existence of CRC-derived EVs representative of tumor cells.

Serial EV tracking, especially for treatment monitoring, also requires more frequent sampling and larger prospective cohorts than the current study to obtain robust statistics.

For the patient class, resectable patients with established disease were enrolled because we were interested in monitoring EV changes during treatment.

Invariably, many patients were in stage II and III and few were in stage I.

In future studies, it may be interesting to apply the panel to CRC screening tools.

It is contemplated that additional molecular markers (e.g. KRAS(G12D), KRAS(G13D), KRAS(G12V), BRAF(V600E), MLH1, MSH3, MSH6 and IGF2) can be tested for this purpose.

In the future, we plan to extend the current study to:

expand its influence

First, the study cohort can be expanded to both CRC patients and controls to further validate EV-based CRC diagnosis.

Collecting samples of CRC patients from multiple institutions can account for racial and geographic diversity.

Including non-CRC patients with potentially confounding intestinal conditions (e.g. irritable bowel syndrome or Crohn's disease) will help improve EV biomarkers and their cutoffs.

With high throughput and low equipment cost, the HiMEX rapidly handles these sample volumes even in standard laboratory settings.

Second, EV tests can be designed to distinguish different CRC stages to complement existing imaging.

Improved non-invasive CRC staging and molecular profiling will (1) be a powerful tool for selecting patients for neoadjuvant rather than adjuvant chemotherapy, (2) allow for the identification of groups at high risk of recurrence; 3) select patients for targeted clinical trials;

To achieve this classification ability, a larger number of patients at each CRC stage than in the current study is required.

Third, EV screening can be expanded to obtain various clinically relevant information.

For example, tumor-derived EVs have been shown to display distinct patterns of integrins.

These EVs can be taken up by resident cells in an organ-specific manner preparing the pre-metastatic niche.

Others have hypothesized that chemotherapy may further induce primary tumors to release EVs with metastatic potential.

HiMEX's ability to enrich tumor-specific EVs will facilitate the detection of these EV subpopulations to help understand, monitor and predict metastasis.

<Method>

<Collecting clinical samples>

The present invention was approved by the Institutional Review Board (IRB) of Kyungpook National University Medical Center (Director: J.S.P.), which collected all clinical samples.

Subsequent procedures followed institutional guidelines.

Informed consent was obtained from all subjects.

Peripheral blood (~15 ml) was drawn from the patient and centrifuged at 400 g for 15 minutes to separate plasma from red blood cells and pale yellow.

20 μl of plasma was used to analyze each marker.

<HiMEX system construction>

The HiMEX device consists of a microcontroller (ATSAMD21G18, Atmel Corporation), a digital-to-analog converter (DAC8552, Texas Instruments), an analog-to-digital converter (AD7490, Analog Devices), a multiplexer (ADG708, Analog Devices), and 96 potentiometers. (ptentiostats).

Each potentiostat has two operational amplifiers (AD8606, analog device).

That is, one amplifier maintained a potential difference between the working electrode and the reference electrode, and the other functioned as a transimpedance amplifier that converted current into a voltage signal.

The current measurement range of the transimpedance amplifier was ±7.5 ㎂, and a 12×8 electrode array (96×220, DropSens) was used.

<Manufacture of analytical reagents>

The first step was to produce immunomagnetic beads.

For 10 minutes at room temperature, 5 mg of magnetic beads (eg 14302D, Dynabeads M-270 Epoxy, Invitrogen) with an epoxy group were suspended in 1 ml of 0.1 M sodium phosphate solution.

Magnetic beads were separated from the solution with a permanent magnet and resuspended in 100 μl of the same solution.

Then 100 μg of antibodies against CD63, CD9 or CD81 (see Supplementary Table 1 for details) were added and mixed thoroughly.

Next, 100 μl ammonium sulfate solution (3M) was added, and the whole mixture was incubated at room temperature for 2 hours and then incubated overnight at 4° C. with slow tilt rotation.

The beads were washed twice with PBS and finally resuspended in 2 ml of PBS containing 1% bovine serum albumin (BSA).

The second step was to label the antibody.

A 10 mM solution of sulfo-NHS-biotin (A39257, Pierce) in PBS was incubated (incubated) with the antibody for 2 hours at room temperature.

Unreacted Sulfo-NHS-Biotin was removed using a Zeba spin desalting column, 7K MWCO (89882, Thermo Scientific).

Biotinylated antibodies were stored at 4°C until use.

<HiMEX Analysis>

We mixed EV samples (10 μl of EV spiked PBS or 20 μl of plasma) with 50 μl of immunomagnetic bead solution for 15 minutes.

Then, they were collected with a permanent magnet, the beads were washed, the supernatant was discarded, and fresh buffer (50 μl of PBS containing 1% BSA) was added.

Then, 10 μl of the biotinylated antibody of interest (20 μg ml ⁻¹ in PBS) was added and the mixture was incubated for 15 minutes.

We dissociated and washed the beads as described above and added 5 μl of streptavidin-conjugated HRP enzyme (21130, Pierce, 1:100 dilution in PBS, 15 min at 21°C).

The beads were separated, washed and finally suspended in 7 μl of PBS.

Prepared bead solutions and 20 μl of UltraTMB solution (34028, ThermoFisher Scientific) were loaded onto the screen printed electrodes to generate signals.

After 3 minutes, chronoamperometry measurements were started.

Current levels between 50 and 55 seconds were averaged.

Total assay time was approximately 1 hour.

All analytical steps were performed at 21 °C.

<EV capture>

EV capture antibodies (anti-CD63, anti-CD9 or anti-CD81) bind to magnetic beads at a ratio of 10 μg of total antibody to 100 μl of beads by incubation overnight at 4° C. with rotation.

Beads were washed 3 times with 500 μl of PBST buffer (PBS + 0.001% Tween 20) and resuspended in 100 μl of the same buffer.

EVs were isolated by OptiPrep density gradient ultracentrifugation, as previously described.

The final pellet was resuspended in 100 μl of PBS.

The collected EVs (5 μl, 5 × 10 ⁹ vesicles ml ⁻¹ ) were mixed with 25 μl of antibody-coated beads, and then 170 μl of PBST was added and incubated at 4° C. while rotating.

Beads with pull-down EVs were collected and washed 3 times with 500 μl PBST.

The flow-through was concentrated to 50 μl using an Amicon Ultra 10-kDa filter (UFC501096, Millipore Sigma).

For Western blotting, samples were dissolved in non-reducing LDS sample buffer, boiled at 95° C. for 5 minutes, and then loaded onto a gel.

Proteins were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and immunostained with the following antibodies. Namely, the following antibodies were anti-CD63 (clone H5C6, BD Biosciences, 1:200 dilution); anti-CD9 (clone D8O1A, Cell Signaling, 1:1,000 dilution); and anti-CD81 (clone 1.3.3.22, Thermo Fisher, 1:500 dilution).

After adding the HRP-conjugated secondary antibody, a chemiluminescent substrate (WesternBright Sirius, Advansta) was added and blots were developed using autoradiographic film (Fig. 2b and Supplementary Fig. 22).

Films were digitized and quantification of signal intensities was performed using ImageJ.

<ELISA>

The CD63 antibody (Ancell) and the IgG1 antibody (Ancell) were diluted in PBS (5 μg ml ⁻¹ ) and transferred to a Maxisorp 96-well plate (Nunc) and incubated overnight at 4° C.

After washing with PBS, 2% BSA in PBS blocking solution was added to the plate (1 hour incubation at room temperature).

EV samples (in 100 μl PBS) were then added to each well for 1 hour incubation at room temperature.

The blocking solution was discarded and antibodies against various markers (1 μgml ⁻¹ ) were inserted into each well and incubated for an additional 1 hour at room temperature.

Unbound antibody was washed three times with PBS.

Streptavidin-HRP molecules were then added to each well and the mixture was incubated for 1 hour at room temperature.

After washing with PBS, the chemiluminescence signal was measured with a plate reader (Tecan).

<Flow cytometry>

Approximately 10 ⁶ cells per marker were used for flow cytometry experiments.

Cells were fixed with 4% paraformaldehyde for 10 minutes at room temperature and then washed with PBS (0.5% BSA).

Next, cells were blocked with BSA (0.5% in PBS) and incubated with primary antibody (4 μgml ⁻¹ ).

The labeled cells were washed, incubated with a fluorophore-conjugated secondary antibody (2 μg ml ⁻¹ ; Abcam), and washed again.

Control samples were similarly labeled using isotype-matched IgG and secondary antibodies.

An aliquot of cells was incubated with the secondary antibody as a background estimator.

Blocking and incubation with antibodies (primary and secondary) were performed for 30 minutes each at room temperature.

All washing steps consisted of washing 3 times for 5 minutes at 300 g with PBS (0.5% BSA).

The fluorescence signal of the labeled cells was measured using a BD LSRII Flow Cytometer (BD Biosciences).

Mean fluorescence intensity (MFI) was obtained from three types of samples (ie target, IgG and secondary-only).

The expression level of the target marker was calculated as (MFI _target - MFI _IgG )/MFI _secondary .

Gating information is provided in Supplementary Fig. 23.

<Cell culture>

A panel of CRC cell lines was purchased (ATCC) and grown in the following vendor-recommended medium: DLD-1 (RPMI-1640, Cellgro); HT29 (MacCoy's 5a modified with 2% NaHCO3, Cellgro); HCT116 (MacCoy's 5a, Cellgro); Colo201 (RPMI-1640 modified with 1% sodium pyruvate, Cellgro); SW480 and SW620 (DMEM, Cellgro); CCD-18Co (Eagle's MEM, Cellgro); CCD-112Con (Eagle's MEM, Cellgro); CCD-33Co (MEM from Eagle, Cellgro) was grown in medium.

All media were supplemented with 10% FBS and penicillin-streptomycin (Cellgro).

All cell lines were tested and determined to be free of mycoplasma contamination (MycoAlert Mycoplasma Detection Kit, Lonza, LT07-418).

<Isolating EVs from cell culture>

Pure EVs were prepared in cell culture for spike-in experiments.

Cells from passages 1-15 were cultured in vesicle-depleted medium (with 5% FBS depleted) for 48 hours.

Conditioned media from approximately 10 ⁷ cells was collected and centrifuged at 300 g (5 min).

The supernatant was filtered through a 0.2 μm membrane filter (Millipore) and concentrated by centrifugation at 100,000 g (1 h).

After removing the supernatant, the EV pellet was washed with PBS and centrifuged at 100,000 g (EV).

Collected EVs were resuspended in PBS and stored at 4°C.

For stock samples, EV concentrations were estimated through nanoparticle tracking analysis (NTA).

<Immunohistochemistry>

Immunohistochemical staining was performed on tumor and non-tumor tissue sections from CRC patients (Fig. 3a and Supplementary Fig. 24).

Primary monoclonal antibodies against EGFR (EGFR1, Abcam, 1:50 dilution), EpCAM (VU-1D9, Abcam, 1:100 dilution) and CD24 (eBioSN3, ebioscience, 1:100 dilution) were used, and the manufacturer's specifications were used. Staining was performed via a Ventana BenchMark XT automated slide stainer (Ventana Medical Systems) according to recommendations.

Stained images were scored based on the percentage of positive cells out of total cancer cells by a pathologist (G.Y.) blinded to clinicopathological parameters and EV profiling results.

Positive cells were defined as a membranous pattern within cancer cells in the presence of concurrent negative labeling in positively stained cytoplasm or non-neoplastic tissues.

Marker expression was ranked in increments of 0.1 from 0 (negative) to 1 (positive).

This level was used as an ordinal variable in Spearman's rank test with EV profiling results.

<Determination of tumor recurrence status>

Recurrence was diagnosed pathologically by surgical excision or radiographic detection of lesions that increased in size over time.

Radiologists and pathologists independently assessed radiographic images and pathological specimens.

Local recurrence was defined as any recurrence within the pelvic cavity or perineum.

Systemic recurrence was defined as any recurrence outside the pelvic cavity.

<EV mRNA analysis (EV mRNA analyses)>

Approximately 3 mL of patient plasma sample was used as recommended by the supplier.

EVs from patient plasma were captured on magnetic beads and RNA was isolated using the exoRNeasy Serum/Plasma Starter kit (77023, Qiagen).

Libraries were prepared with the QIAseq Targeted RNA panel according to the manufacturer's instructions (Human Molecular Toxicology Transcriptome, Qiagen).

Libraries were determined using the Qubit dsDNA HS Assay kit (Q32850, ThermoFisher Scientific) and the QIAseq Library Quant Assay kit, and library size was determined using the Agilent 2100 Bioanalyzer (High Sensitivity DNA Analysis Kit, Agilent Technologies).

Prepared libraries were pooled (4 nM for each library) and the pooled mixture (6.5 pM) was run on an Illumina MiSeq (MiSeq Reagent Kit v2, Illumina).

Data were analyzed using the DESeq2 package in R version 3.6.1.

<Single EV imaging>

EVs of CRC cell lines (SW480, HT29; 10 ⁸ vesicles ㎖ ^-1 ) were incubated for 30 minutes at room temperature on glass slides (63429-04, Electron Microscopy Science).

The solution was drained onto a paper towel and then incubated for 15 minutes in a 4% paraformaldehyde solution (AAJ19943K2, Thermo Scientific) at room temperature for EV fixation.

Slides were washed twice with PBS.

Perm/Wash buffer (554723, BD Science) was added for EV permeability and the slides were incubated for 5 minutes at room temperature.

After washing 5 times with PBS, the slides were blocked with 2% BSA in PBS for 30 minutes.

Then 10 μl of AF647 (1434, Click Chemistry Tools)-labeled anti-CD63 (215-020, Ancell), CD9 (555675, BD science) or CD81 (555370, BD Science) antibodies (containing 1% BSA) One 10 μgml ⁻¹ PBS ) was added for EV labeling overnight at 4°C.

After washing three times in PBS, slides were covered with cover slips and EVs were imaged with a BX63 fluorescence microscope (Olympus).

<Bioinformatic algorithm for selecting initial markers>

Immunohistochemistry data for CRC and normal tissues were downloaded from the Human Protein Atlas portal.

A numerical value was assigned to the degree of staining (high, 3, medium, 2, low, 1, not detected, 0) and the average staining value for each protein was obtained.

For each tissue type (i.e., CRC and normal) these values were normalized z-scores.

Markers were then ranked according to differential z-score (CRC-normal).

This list was narrowed down by cross-referencing UniProt to select proteins in specific cellular domains (eg transmembrane).

We further filtered the list using the EV database to collect markers reported to be found in EVs.

Analysis was performed in R using a custom script.

<Statistical analyzes>

Statistical analysis was performed using Stata version 12.1 (StataCorp LP), GraphPad Prism version 8.0 (GraphPad Software Inc.) or R version 3.6.1.

For all statistical tests, a P value <0.05 was considered significant.

Marker selection for CRC detection

Candidate predictive markers were selected from the EV profiling data (Fig. Sa) by applying minimum absolute contraction and selection operator (Lasso) regression.

Non-CRC controls were augmented with EV data from healthy plasma samples.

The cross-validation error was calculated and the tuning parameter (λ) that minimized the cross-validation error was determined.

We used the glmnet package in R.

CRC diagnosis

ROC analysis was performed using a training cohort (n=83) to evaluate the ability of selected protein markers individually or in combination to predict CRC cases.

Specific individual biomarkers under consideration were EGFR, EpCAM, CD24, GPA33 and CD133.

Also, 2 markers (EGFR + CD24, EGFR + EpCAM and EpCAM + CD24), 3 markers (EGFR + EpCAM + CD24), 4 marker combinations (EGFR + EpCAM + CD24 + GPA33 and EGFR + EpCAM + CD24 + CD133 ), and a linear combination of five marker combinations (EGFR + EpCAM + CD24 + GPA33 + CD133) were considered.

For each biomarker combination, the optimal weight was determined through logistic regression.

In the present invention, an ROC curve was constructed and a level cutoff value maximizing the sum of sensitivity and specificity was determined.

Standard formulas were used to define sensitivity (true positive rate), specificity (true negative rate) and accuracy [(true positive + true negative)/(positive + negative)].

AUCs were compared according to Delong's method.

Next, we analyzed a prospective cohort (n=48) using the selected cutoff values and determined the classification performance of each biomarker and biomarker combination.

We estimated exact 95% confidence intervals for sensitivity, specificity and accuracy.

Analysis was performed using the pROC package in R.

<Survival analyzes>

In the present study, 90 CRC patients who underwent curative surgery for up to 61 months were monitored.

DFS was defined as the interval between surgery and first radiographic recurrence or death due to CRC.

Survival functions stratified by EV _CRC or CEA level were calculated using the Kaplan-Meier method.

The EV _CRC cut-off value selected for patient classification was the value that maximized the absolute log-rank score statistic between the two classification groups. That is, the P value for comparing the corresponding survival curves for the two groups above and below the selected EV _CRC cutoff value was maximally calculated using conditional Monte Carlo simulations with 10,000 replicates to approximate the null distribution of the selected rank statistic. got it

The log-rank test was used to compare survival curves at the original CEA cutpoint.

Analysis was performed using R's maxstat and survival packages.

<Reporting Summary>

Additional information on the study design can be found in the Nature Research report summary linked to this article.

<Data availability>

The main data supporting the findings of this study are available in the paper and supplementary information.

Raw patient data sets generated and analyzed during the study period will be provided to the corresponding author upon reasonable request and approved by the Institutional Review Board of Kyungpook National University Hospital.

Public databases Human Protein Atlas (https://www.proteinatlas.org) and UniProt (https://www.uniprot.org) were used for initial marker selection. Source data are provided with this paper.

<Code availability>

The source code for marker selection is available at https://csb.mgh.harvard.edu/bme_software.

As described above, the technical ideas described in the embodiments of the present invention may be implemented independently, or may be implemented in combination with each other. In addition, the present invention has been described through the embodiments described in the drawings and detailed description of the invention, but these are only examples, and various modifications and other equivalent embodiments may be made by those skilled in the art in the art to which the present invention belongs. possible. Therefore, the technical protection scope of the present invention should be defined by the appended claims.

Claims (17)

a plurality of electrode arrays loaded with target analyte-binding magnetic beads;
an electrostatic potential electrically coupled to each of the electrodes of the electrode array;
a DAC electrically coupled to the input of the potentiometer to convert a digital signal into an analog signal;
an ADC electrically coupled to the output of the potentiostat to convert the voltage signal of the potentiostat into a digital signal; and
A microcontroller that is electrically coupled to the DAC and transmits a control command to a potentiometer through the DAC or is electrically coupled to the ADC and determines a target analyte through a digital signal transmitted from the ADC; A characterized magneto-electrochemical device. According to claim 1,
Each electrode of the plurality of electrode arrays is a magneto-electrochemical device, characterized in that composed of three individual electrodes of a working electrode, a counter electrode and a reference electrode. In paragraph 1,
The magneto-electrochemical device according to claim 1 , wherein each electrode of the plurality of electrode arrays is connected to the electrostatic and push-pin connectors and disposed on a built-in magnet to focus the magnetic bead. According to claim 1,
The potentiostat comprises: a first operational amplifier electrically coupled to a working electrode and maintaining a predefined potential difference between the working electrode and a reference electrode;
and a second operational amplifier comprising a transimpedance amplifier electrically coupled to the reference electrode and the counter electrode and converting an induced current across the counter electrode from the working electrode into a voltage signal. . According to claim 1,
Further comprising a multiplexer formed in front of the ADC to select from a plurality of inputs when an output from the potentiometer is electrically coupled to an input of the ADC,
The magneto-electrochemical device, characterized in that the microcontroller configures the multiplexer and the ADC of the selected number of inputs to sequentially access each electrode. According to claim 1,
The magneto-electrochemical device, characterized in that the target analyte is colorectal cancer extracellular vesicles (CRC-EV). According to claim 6,
The magneto-electrochemical device, characterized in that the extracellular vesicles spiked from plasma. According to claim 6,
The colorectal cancer diagnosis biomarker is any one of CD44, B7-H3, EGFR, ABCG2, GPA33, EpCAM, HER2, MUC1, MET, CD44v6, CD24, CD166, STEAP1 and ALDH1, MRP1, MDR1, TS and CD133, or any of these A magneto-electrochemical device, characterized in that selected by a combination. According to claim 8,
The colon cancer diagnostic biomarker is a magneto-electrochemical device, characterized in that composed of a combination of EGFR, EpCAM, CD24 and GPA33. According to claim 9,
The microcontroller,
determine a cutoff value that maximizes the sum of sensitivity and specificity in the ROC curve;
By defining the determined cutoff value as an EV _CRC score,
A magneto-electrochemical device characterized in that it determines whether or not there is colorectal cancer based on the defined EV _CRC score. (a) binding a target analyte to a plurality of magnetic beads, allowing them to bind to a reactive enzyme after washing a plurality of times, and combining with an electron mediator solution to load the target analyte-binding magnetic beads onto a plurality of electrode arrays;
(b) exposing the target analyte-binding magnetic beads to a magnetic field by electrically coupling the plurality of electrode arrays to a potentiostat;
(c) inducing an oxidation-reduction reaction between electron mediators in the target analyte-binding magnetic beads and the reactive enzyme; and
(d) a diagnosis method using a magneto-electrochemical device comprising the step of determining a target analyte through the output result of the potentiostat. According to claim 11,
The reactive enzyme includes HRP, and the electron mediator includes 3,3', 5,5'-tetramethylbenzidine (TMB). According to claim 11,
In step (d), the target analyte is a diagnosis method using a magneto-electrochemical device, characterized in that the colorectal cancer extracellular vesicles (CRC-EV). According to claim 13,
The extracellular vesicle is a diagnostic method using a magneto-electrochemical device, characterized in that spiking from plasma. According to claim 13,
The colorectal cancer diagnosis biomarker is any one of CD44, B7-H3, EGFR, ABCG2, GPA33, EpCAM, HER2, MUC1, MET, CD44v6, CD24, CD166, STEAP1 and ALDH1, MRP1, MDR1, TS and CD133, or any of these A diagnostic method using a magneto-electrochemical device, characterized in that selected by a combination. According to claim 15,
The colon cancer diagnosis biomarker is a diagnosis method using a magneto-electrochemical device, characterized in that composed of a combination of EGFR, EpCAM, CD24 and GPA33. According to claim 16,
determine a cutoff value that maximizes the sum of sensitivity and specificity in the ROC curve;
By defining the determined cutoff value as an EV _CRC score,
A diagnostic method using a magneto-electrochemical device, characterized in that it determines whether or not there is colorectal cancer based on the defined EV _CRC score.

Images