WO2025080800A1 - Protease activity sensing probes for detection and prognosis of cancer - Google Patents

Abstract

Provided herein is a method of detecting cancer in subject. The method includes providing a biological sample comprising one or more proteases from the subject and contacting the biological sample with probes capable of being cleaved by the proteases, wherein the probes comprise (i) at least 6 L-amino acids and (ii) a detectable label. The cleaved probes are the separated and detected. An increase in a signal from the detectable label of the cleaved probes as compared to a control indicates the subject has cancer. Compositions, kits and methods of using the probes to identify proteases are also provided.

Description

PROTEASE ACTIVITY SENSING PROBES FOR DETECTION AND PROGNOSIS OF CANCER

CROSS REFERENCE TO RELATED APPLICATION

This application claims the benefit of and priority to U.S. Provisional Application No. 63/589,201, filed October 10, 2023, the entire contents of which are hereby incorporated by reference.

FIELD

This application relates to compositions and methods for detecting cancer using protease activity sensing probes.

SEQUENCE LISTING

The Sequence Listing is submitted as an XML file in the form of the file named “Sequnce_Listing_102276-012410WO-1470149” (29,577 bytes), which was created on October 9, 2024, which is incorporated by reference herein.

BACKGROUND

Cancer is a leading cause of death in the United States and the world, thus early detection of lethal cancers is critical to saving lives and improving overall quality of life. Cancer is typically detected by either imaging or liquid biopsy. High-risk disease states are known to exist for many cancers due to genetic and environmental factors. The ability to continually screen these high-risk patients is limited by many factors including access, cost, and assay accuracy. There are numerous clinically approved screening tests for high-risk patients; however, these tests tend to be limited to specific tissues and many do not lower the chance of dying from cancer and lead to overdiagnosis. For example, pancreatic ductal adenocarcinoma (PDAC) is frequently not detected until it has progressed to a more lethal, late stage. Associated symptoms are typically non-specific, prolonging the time to diagnoses. There are no effective FDA approved, early detection assays for PDAC, but they are needed to transform survival from this deadly disease. The only FDA approved clinical biomarker for PDAC is carbohydrate antigen 19-9 (CA 19-9), which is used predominantly to measure disease burden across therapeutic treatment and is not used as an early detection biomarker due to its low positive predictive value (PPV) in the early detection setting. Thus, there is a need for the development of assays that are specific to lethal cancers, easily accessible, and inexpensive that can be applied to high-risk populations for continual monitoring of cancer risk or progression.

BRIEF SUMMARY

Provided herein is a method of detecting cancer in a subject. The method includes providing a biological sample comprising one or more proteases from the subject and contacting the biological sample with probes capable of being cleaved by the proteases, wherein the probes comprise (i) at least 6 L-amino acids and (ii) a detectable label. The cleaved probes are the separated and detected. An increase in a signal from the detectable label of the cleaved probes as compared to a control indicates the subject has cancer. Compositions, kits and methods of using the probes to identify proteases are also provided.

In some embodiments, the probes are attached to a plurality of beads and the biological sample is contacted with the plurality of beads comprising the probes to form a composition. In some embodiments, the probe further comprises at least one D-amino acid. In some aspects, the probe comprises at least 6 to 50 amino acids. In some embodiments, the beads further comprise a spacer located between the probes and the beads. In some cases, the spacer comprises polyethylene glycol (PEG), a peptide or a nucleic acid molecule. In some embodiments, the detectable label comprises a fluorescent label, a radioactive label, or an affinity label.

In embodiments, the composition comprises 0.5 to 150 pl of the biological sample. In some embodiments, the biological sample comprises whole blood. In some cases, the biological sample is a serum or plasma sample. In some aspects, the biological sample comprises extracellular vesicles. In embodiments, the one or more proteases are selected from the group of a matrix metalloproteinase (MMP), cathepsin and a caspase.

In some aspects, the method of detecting cancer further comprises performing an immunoassay. In embodiments, the immunoassay detects the presence of a polypeptide wherein an increase in the level of the polypeptide as compared to a control indicates the subject has cancer. In some cases, the polypeptide is CA-19-9. In some embodiments, the cancer is pancreatic cancer, liver cancer, head and neck, esophagus, bile, small intestine, stomach or colon cancer.

Also provided are kits comprising the probes and instructions for use. The kits may include a solid support. In embodiments, the solid support can be a bead, which, optionally, can be magnetic. The kit may also include one or more buffers, which can be lyophilized. The kit can also include water. In the provided kits, the kits can include, for example, one or more beads with the same or different types of probes.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 presents a bar graph plotting fold change signal analysis between PDAC and control for each set of probes based on a charge changing peptide (CCP) assay.

Figure 2A is a graph showing the Z-score of the protease signal for every probe with 95% confidence intervals. One-way ANOVA followed by Tukey’s multiple comparisons test for every probe was performed and significance annotated. (Control n=67, Pancreatitis (Pane) n=l 1, Neoplasia n=21, PDAC n=67).

Figure 2B is a graph showing PDAC patients subclassified by stage I-IV and their z- score protease signal plotted for every probe with 95% confidence intervals. One-way ANOVA followed by Tukey’s multiple comparisons test for every probe was performed and significance annotated.

Figure 2C is a graph showing accuracy, Area Under the Curve (AUC) of ROC (receiver operating characteristic curve), Sensitivity, and Specificity from cross-validation of probe-1.

Figure 2D is a graph showing cross-validated ROC of probe-1.

Figure 2E is a Kaplan-Meier plot showing log-rank statistic for cut-off population proportion of 0.5 for the combination of probes-2-3-4.

Figure 3A presents probe cleavage exploration and active protease identification, and depicts a graph indicating mass spectrometry results from isolated probe- 1 after incubation with serum to identify cleavage sites.

Figure 3B is a graph showing summary of cleavage sites acquired from mass spectrometry results for all probes.

Figure 4A is a graph showing comparison between z-score of protease signal for all probes for whole plasma, EVs and non-EV flow-through of PDAC patients (n = 5 for each group, Mean with 95% confidence interval, Kruskal-Wallis test with multiple comparison correction per probe).

Figure 4B is a graph showing comparison between z-score of protease signal for all probes for whole plasma, EVs and non-EV flow-through from healthy patients (n = 8, 10 and 4 for plasma, EVs, and flow-through respectively).

Figure 5A is a graph showing identification of number of amino acids needed for probe to cleave using d-Amino acid modified probes based on a nanosensor assay. Mean with SD, One-way ANOVA followed by Tukey comparisons. Figure 5B is a graph showing comparison of 27 PDAC patients, 27 controls and 38 neoplasia (IPMN/PanIN) using bead-based nanosensor assay (p < 0.0001, Kruskal-Wallis with Dunn’s correction).

Figure 5C is a graph showing correlation between CCP assay and nanosensor assay using the same 27 PDAC and 27 control patients.

Figure 5D is a graph showing correlation between two different batches of nanosensor assay.

Figure 6A shows NanoSensor blinded validation studies and presents a graph showing cross-validated sensitivity of the combination of the nanosensor and CAI 9-9 (computing the CI of a proportion using Wilson’s method). Indicated are blinded validation results for nanosensor or PAC*MANN-1, CA 19-9 and the combination using the most accurate cut-off in the training dataset (95% CI using Wilson/Brown’s method).

Figure 6B is a graph showing ROC of training dataset between PDAC and non- PDAC for proteases, CA 19-9 and the combination of both.

Figure 6C is a graph showing ROC of training dataset between PDAC and pancreatic neoplasias for proteases, CA 19-9 and the combination of both.

Figure 6D is a graph showing ROC of training dataset between PDAC and pancreatitis for proteases, CA 19-9 and the combination of both.

Figure 7 shows Pearson correlation plot between all CCP probes.

Figure 8A are graphs showing probe cleavage identification from mass spectrometry results for probe- 1 and probe-2 after incubation with sample with high signal, low signal and sample with high signal but no probe.

Figure 8B are graphs showing probe cleavage identification from mass spectrometry results for probe-3 and probe-4 after incubation with sample with high signal, low signal and sample with high signal but no probe.

Figure 8C are graphs showing probe cleavage identification from mass spectrometry results for probe-5 and probe-6 after incubation with sample with high signal, low signal and sample with high signal but no probe.

Figure 9 is a graph showing comparison between z-score of protease signal for all probes for neoplasia and early stage PDAC. The graph indicates analysis of early stage I and II patients compared to pancreatic neoplasia patients (Multiple t-test with False Discovery Rate). Figure 10A is a graph showing, in the training cohort for the nanosensor or PAC»MANN-1 results comparing PDAC, Healthy, Neoplasia, and pancreatitis.

Figure 10B is a graph showing, in the training cohort for CAI 9-9 results comparing PDAC, Healthy, Neoplasia, and pancreatitis.

DETAILED DESCRIPTION

For improved cancer early detection within high-risk disease states, patients will benefit from rapid, regular screening with minimally invasive blood microsampling. Initially, pancreatic ductal adenocarcinoma (PDAC) was explored because there are a number of high- risk pre-cancerous disease states that rapidly progress. As a liquid biopsy biomarker, active proteases were analyzed since they are important for cancer progression and their enzymatic activity allows for signal amplification. First, an agnostic screen for hundreds of proteases was created and a set of peptide probes were found that distinguished PDAC patients from those with pancreatitis, precursor lesions or controls as well as increased prognostic ability. Using this agnostic approach, it was discovered that the most likely, but not only, active proteases responsible in PDAC patient circulation were MMP2 and MMP14. Next, a rapid and high-throughput nanosensor that can distinguish PDAC from high-risk disease states was developed. Finally, samples were used for a training and blinded validation cohort (n=228) and the nanosensor data was combined with data from the current standard of care, CAI 9-9. Using logistic regression to find the most accurate cutoff point in the training set, the blinded validation cohort had 100% specificity (35/35) with 85% sensitivity (34/40). This nanosensor allows for the clinical application of a rapid and high-throughput assay to improve access to cancer detection tests for regular screening in high-risk individuals. It is to be noted that in the present disclosure, the nanosensor assay has also been referred to as a PAOMANN assay. The nanosensor assay or PAOMANN assay has been used interchangeably throughout the disclosure.

Provided herein are methods of detecting cancer in subject. The methods include providing a biological sample comprising one or more proteases from the subject, contacting the biological sample with a plurality of beads each bead comprising a plurality of probes to form a composition comprising the biological sample and beads, wherein each probe comprises (i) at least 6 L-amino acids and (ii) a detectable label, and wherein the probe is capable of being cleaved by the one or more proteases, separating the beads from the composition; and detecting the detectable label in the composition separated from the beads, wherein an increase in a signal from the detectable label as compared to a control indicates the subject has cancer. As used herein, the term nanosensor refers to a solid support to which the herein provided probes are attached.

Optionally, the methods include contacting the biological sample with the beads under conditions to generate cleavage products. The cleavage products are generated when a protease cleaves the probe from the bead. The cleavage or cleaved products retain the detectable label after being separated from the bead. Thus, the detecting can comprise detecting the detectable label of the cleavage or cleaved products.

Each bead or solid support to which the probes are attached can be the same or different. Likewise, a single bead or solid support can comprise one or more different types of probes. For example, a bead can comprise one or more types of probes, i.e. multiplexing or attaching the beads to more than one type of probe is also possible in embodiments described in the current disclosure. Thus, a bead can comprise 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15, different types of probes. When a bead has multiple types of probes, each probe can comprise a different type of detectable label. However, in some embodiments, two or more types of beads can contain the same type of detectable label.

A “control” or “standard control” refers to a sample, measurement, or value that serves as a reference, usually a known reference, for comparison to a test sample, measurement, or value. For example, a test sample can be taken from a patient suspected of having a given disease (e.g., cancer) and compared to a known normal (non-diseased) individual (e.g. a standard control subject). A standard control can also represent an average measurement or value gathered from a population of similar individuals (e.g. standard control subjects) that do not have a given disease (i.e. standard control population), e.g., healthy individuals with a similar medical background, same age, weight, etc. A standard control value can also be obtained from the same individual, e.g. from an earlier-obtained sample from the patient prior to disease onset or at different timepoints while monitoring a patient over the course of the disease. One of skill in the art will understand which standard controls are most appropriate in a given situation and be able to analyze data based on comparisons to standard control values.

Also provided are methods of identifying a protease in a biological sample. The methods include providing a biological sample; contacting the biological sample with a plurality of beads each bead comprising a plurality of probes to form a composition comprising the biological sample and beads, wherein each probe comprises (i) at least 6 L- amino acids and (ii) a detectable label and wherein the probe is cleaved by the protease thereby generating cleavage products comprising the detectable label; (c) separating the cleavage products from the beads; and (d) detecting the detectable label of the cleavage products thereby identifying the protease in the biological sample. For example, the cleavage products contain cleavage sites specific for certain proteases. Thus, the proteases will be identified by determining the cleavage products obtained by the method. Optionally, the protease is selected from the group consisting of a matrix metalloproteinase (MMP), cathepsin and a caspase. For example, the MMP can be MMP 1, 2, 3, 7, 8, 9, 12, 13, 14, 15, 16, 17, 19, 20, 21, 23, 24, 25, 26, 27, or 28. The cathepsin can be cathepsin, B, C, D, F, G, H, K, L or S. The caspase can be caspase 1, 2, 3, 4, 5, 6 7, 8, 9, 10, 12, or 14.

One advantage of the herein provided methods is that the biological sample does not need additional processing or manipulation before contacting with probes. Thus, the biological sample need not be manipulated or processed prior to the contacting.

The herein provided probes are peptides with cleavage sites of proteases. The probe can be of any suitable length of amino acids. For example, the probes can include at least 6 to 50 amino acids. Thus, the probes can be 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, or 50 amino acids in length. Optionally, the probes contain at least 6 to 15 amino acids.

The herein provided probes can include one or more D-amino acids. Optionally, the probe further comprises at least one D-amino acid. Optionally, the probes comprise at least 6 L-amino acids and at least two D-amino acids.

As discussed throughout, the herein provided probes contain cleavage sites by which they can be cleaved by a protease. For example, the probes can include PEPFAG (SEQ ID NO: 1) or GAFPEP (SEQ ID NO:7), GLAGGA (SEQ ID NO:2), AAGALG (SEQ ID NO:8), MARTLK (SEQ ID NO:3), KLTRAM (SEQ ID NO:9), PLGLVG (SEQ ID NO:4), GVLGLP (SEQ ID NO: 10), LRSVSG (SEQ ID NO: 5), GSVSRL (SEQ ID NO: 11), LRGGMP (SEQ ID NO: 6), PMGGRL (SEQ ID NO: 12) or any combination thereof. Thus, the probes in the provided methods can have an amino acid sequence comprising SEQ ID NO: 1 to 12 or any combination thereof.

In addition to cleavage sites, the probes can also contain a cysteine (C), lysine (K) and/or C and K amino acid residues. Cysteines and lysines can be used to chemically attach the probes to a solid support and/or for attaching a detectable label to the probes. The cysteine and/or lysine can be attached to one end of the probes or can be located internal to the ends of the probes. Optionally, the cysteine and/or lysine is located at one end of the probes, e.g., the N-terminus or the C-terminus. Optionally, the cysteine and/or lysine are located at the N-terminal end of the probes. Optionally, the cysteine and/or lysine are located at the C-terminal end of the probes. Such exemplary probes are set forth in SEQ ID NOs: 13 to 29 as shown in Table 6 below. Thus, the herein provided probes can have an amino acid sequence selected from the group consisting of SEQ ID NO: 13 to 29.

Also provided herein are also probes with at least 80%, 85%, 90%, 95% or 99% identity to any one of SEQ ID NOs: 1 to 29. The term, identity or substantial identity, as used in the context of a polynucleotide or polypeptide sequence described herein, refers to a sequence that has at least 60% sequence identity to a reference sequence. Alternatively, percent identity can be any integer from 60% to 100%. Exemplary embodiments include at least: 60%, 65%, 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99%, as compared to a reference sequence using the programs described herein; preferably BLAST using standard parameters. One of skill will recognize that these values can be appropriately adjusted to determine corresponding identity of proteins encoded by two nucleotide sequences by taking into account codon degeneracy, amino acid similarity, reading frame positioning and the like.

For sequence comparison, typically one sequence acts as a reference sequence to which test sequences are compared. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters can be used, or alternative parameters can be designated. The sequence comparison algorithm then calculates the percent sequence identities for the test sequences relative to the reference sequence, based on the program parameters. Algorithms that are suitable for determining percent sequence identity and sequence similarity are the BLAST and BLAST 2.0 algorithms. Software for performing BLAST analyses is publicly available through the National Center for Biotechnology Information (NCBI) web site.

As indicated throughout, the herein provided probes can have a detectable label. Optionally, the detectable label comprises a fluorescent label, a radioactive label, or an affinity label. A “label” or a “detectable label” is a composition detectable by spectroscopic, photochemical, biochemical, immunochemical, chemical, or other physical means. For example, useful labels include 32P, fluorescent dyes, electron-dense reagents, enzymes (e.g., as commonly used in an ELISA), biotin, digoxigenin, or haptens and proteins or other entities which can be made detectable, e.g., by incorporating a radiolabel into a peptide or antibody specifically reactive with a target peptide. A common method to introduce a detectable tag on a polypeptide involves chemical conjugation to amines or cysteines. Such conjugation methods are well known. As non-limiting examples, n-hydroxysuccinimide esters (NHS esters) are commonly employed to label amine groups that may be found on a polypeptide. Cysteines readily react with thiols or maleimide groups, while carboxyl groups may be reacted with amines by activating them with EDC (l-Ethyl-3-[3- dimethylaminopropyl]carbodiimide hydrochloride).

Exemplary fluorophores that can be used on the herein provided probes include, but are not limited to, fluorescent nanocrystals; quantum dots; d-Rhodamine acceptor dyes including dichlorofRl 10], dichloro[R6G], dichloro[TAMRA], dichlorofROX] or the like; fluorescein donor dye including fluorescein, 6-FAM, or the like; Cyanine dyes such as Cy3B; Alexa dyes, SETA dyes, Atto dyes such as atto 647N which forms a FRET pair with Cy3B and the like. Fluorophores include, but are not limited to, MDCC (7- diethylamino-3-[([(2- maleimidyl)ethyl]amino)carbonyl]coumarin), TET, HEX, Cy3, TMR, ROX, Texas Red, Cy5, LC red 705 and LC red 640. Fluorophores and methods for their use including attachment to polymerases and other molecules are described in The Molecular Probes® Handbook (Life Technologies, Carlsbad Calif.) and Fluorophores Guide (Promega, Madison, WI), which are incorporated herein by reference in their entireties.

While the above refers to a bead, the probes can be attached to any solid support. Optionally the polypeptide is bound to a solid support such as a slide, a culture dish, a multiwell plate, column, chip, array or stable beads. Optionally, the solid support is capable of being separated from the composition. Thus, the herein provided probes can be bound to a mobile solid support, e.g., beads, which can be sorted using sorting technology, e.g., magnetically. “Mobile solid support” refers to a set of distinguishably labeled microspheres or beads. Thus, in the provided methods, the beads can be separated from the composition magnetically.

In embodiments of the present disclosure, the beads may have a size ranging from about 40 nm to about 2000 nm in diameter. Exemplary embodiments may include a bead size ranging from about 40-100 nm, 40-300 nm, 40-500 nm, 100-300 nm, 100-500 nm, 200-500 nm, 300-600 nm, 400-800 nm, 500-1000 nm, 1000-1500 nm, 1500-2000 nm, 40-1000 nm, or 1000-2000 nm in diameter. Thus, the size of the beads may be 40 nm, 50 nm, 60 nm, 80 nm, 100 nm, 150 nm, 200 nm, 250 nm, 300 nm, 350 nm, 400 nm, 450 nm, 500 nm, 600 nm, 700 nm, 800 nm, 900 nm, 1000 nm, 1200 nm, 1400 nm, 1600 nm, 1800 nm, or 2000 nm. In a preferred embodiment of the disclosure, the bead size may comprise 50 nm.

“Biological sample” or “sample” refer to materials obtained from or derived from a subject or patient. A biological sample includes sections of tissues such as biopsy and autopsy samples, and frozen sections taken for histological purposes. Such samples include bodily fluids such as blood and blood fractions or products (e.g., serum, plasma, platelets, red blood cells, and the like), sputum, tissue, cultured cells (e.g., primary cultures, explants, and transformed cells) stool, urine, cancer cells and the like. Bodily fluids include without limitation blood, urine, serum, tears, breast milk, lymph, bile, cerebrospinal fluid, interstitial fluid, aqueous or vitreous humor, colostrum, sputum, amniotic fluid, saliva, anal and vaginal secretions, perspiration, semen, transudate, exudate, and synovial fluid. Optionally, the biological sample is a serum or plasma sample. Optionally, the biological sample is whole blood. In some embodiments, the biological sample may comprise extracellular vesicles isolated from bodily fluids. A biological sample is typically obtained from a eukaryotic organism, such as a mammal such as a primate e.g., chimpanzee or human; cow; dog; cat; a rodent, e.g., guinea pig, rat, mouse; rabbit; or a bird; reptile; or fish.

The herein provided methods is they can be carried out using a volume of a biological sample. Optionally, the composition comprises 0.5 to 150 pl of the biological sample. Thus, the composition can contain 100 to 150 pl of the biological sample. Optionally, the composition contains 0.5 to 50 pl of the biological sample. Optionally, the composition contains 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48 or 50 pl of the biological sample.

In the provided methods the beads comprising the probes can also comprise a spacer located between the probes and the beads. Suitable spacers include different lengths of polyethylene glycol (PEG), peptides or nucleic acid molecules.

The provided methods can include performing an additional assay for detecting cancer in the subject. The assay can detect the presence of a polypeptide wherein an increase in the level of the polypeptide as compared to a control indicates the subject has cancer. Optionally, the polypeptide is CA-19-9. Optionally, the assay is an immunoassay. Immunoassays are binding assays typically involving binding between antibodies and antigen. Examples of immunoassays include, but are not limited to, enzyme linked immunosorbent assays (ELISAs), radioimmunoassays (RIA), radioimmune precipitation assays (RIP A), immunobead capture assays, Western blotting, dot blotting, gel-shift assays, Flow cytometry, protein arrays, multiplexed bead arrays, magnetic capture, in vivo imaging, fluorescence resonance energy transfer (FRET), and fluorescence recovery /localization after photobleaching (FRAP/ FLAP).

As noted above, the herein provided probes can be used to determine whether a subject has cancer. As used herein, the term “cancer” refers to all types of cancer, neoplasm, or malignant tumors found in mammals, including leukemia, carcinomas and sarcomas. Exemplary cancers include cancer of the brain, breast, gastrointenstinal tract, cervix, colon, head & neck, liver, kidney, lung, esophagus, small intestine, stomach, bile, ducts, skin, pancreas, or ovary. Optionally, the cancer is pancreatic cancer, liver cancer or colon cancer. Optionally, the pancreatic cancer is pancreatic ductal adenocarcinoma (PDAC).

Also provided herein are compositions comprising one or more types of probes. For example, provided are compositions comprising a probe having an amino acid sequence comprising SEQ ID NOs: 1 to 12 or any combination thereof. Also provided are compositions comprising a probe comprising SEQ ID NO: 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 and/or 29. Optionally, the compositions comprise multiple types of probes. Thus, for example, a composition can comprise probes comprising SEQ ID NOs: 1-12, 13-24, or 25-28.

Also provided are kits comprising the probes and instructions for use. Thus, the kits can comprise compositions comprising one or more types of probes. The kits may also include a solid support. The solid support can be a bead, which, optionally, can be magnetic. The kit can also include one or more buffers, which can be lyophilized. The kit can also include water. In the provided kits, the kits can include, for example, one or more beads with the same or different types of probes. Optionally, the kits can include, for example, one or more beads wherein each bead comprises a different type of probe. Thus, the kit can include 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, or 12 pluralities of beads wherein each bead type comprises a single type of probe. Thus, the kits can include a first plurality of beads comprising a first probe type and a second plurality of beads with a second probe type and so on.

Disclosed are materials, compositions, and components that can be used for, can be used in conjunction with, can be used in preparation for, or are products of the disclosed methods and compositions. These and other materials are disclosed herein, and it is understood that when combinations, subsets, interactions, groups, etc. of these materials are disclosed that while specific reference of each various individual and collective combinations and permutations of these compounds may not be explicitly disclosed, each is specifically contemplated and described herein. For example, if a method is disclosed and discussed and a number of modifications that can be made to a number of molecules including the method are discussed, each and every combination and permutation of the method, and the modifications that are possible are specifically contemplated unless specifically indicated to the contrary. Likewise, any subset or combination of these is also specifically contemplated and disclosed. This concept applies to all aspects of this disclosure including, but not limited to, steps in methods using the disclosed compositions. Thus, if there are a variety of additional steps that can be performed, it is understood that each of these additional steps can be performed with any specific method steps or combination of method steps of the disclosed methods, and that each such combination or subset of combinations is specifically contemplated and should be considered disclosed.

Publications cited herein and the material for which they are cited are hereby specifically incorporated by reference in their entireties.

A number of embodiments have been described. Nevertheless, it will be understood that various modifications may be made. Accordingly, other embodiments are within the scope of the claims.

Examples

Example 1. A Nanosensor Measuring Protease Activity For Detection of Pancreatic Cancer.

Cancer is a leading cause of death in the United States and the world, thus early detection of lethal cancers is critical to saving lives and improving overall quality of life. Cancer is typically detected by either imaging or liquid biopsy [1], [2], High-risk disease states are known to exist for many cancers due to genetic and environmental factors. The ability to continually screen these high-risk patients is limited by many factors including access, cost, and assay accuracy. There are numerous clinically approved screening tests for high-risk patients; however, these tests tend to be limited to specific tissues and many do not lower the chance of dying from cancer and also lead to overdiagnosis [3]— [6] . Thus, there is a need for the development of biomarker assays that are specific to lethal cancers, easily accessible, and inexpensive that can be applied to high-risk populations for continual monitoring of cancer risk or progression.

Proteases facilitate the degradation of the extracellular matrix in primary tumors to promote cancer progression [7], In addition to proteases role in primary cancer progression, secreted proteases also enable metastatic seeding making them a useful blood-based biomarker for cancer diagnoses as well as occult metastatic disease [8], [9], Current approved protein biomarkers for cancer focus on protein detection (e.g., ELISA) and do not leverage the functional enzymatic activity. Detection of functional enzymatic activity can increase assay sensitivity through targeted substrates and the ability of a single enzyme to cleave many substrates [10], For an initial screening of circulating protease activity, an innovative adaptation of a previously published protease activity detection assay was developed transforming it into a sensitive and agnostic screening tool for cancer-related protease activity from a small volume of blood for early detection of cancer [11], Since the human genome encodes for over 500 proteins with proteolytic activity, an agnostic approach allows for the examination of wide range of proteases that could be drivers of cancer [12], In addition, most proteases have numerous cleavage sites, making specific identification of the activity of a single protease in a complex solution cumbersome.

As described herein, pancreatic ductal adenocarcinoma (PDAC) and different disease states are known to range from low to high penetrance precursors of PDAC. Effective early detection assays for PDAC in high-risk patients have the potential to transform survival from this deadly disease. For example, non-resected Intraductal Papillary Mucinous Neoplasms (IPMNs), a known precursor of PDAC (P. Levy et al., “Natural history of intraductal papillary mucinous tumors of the pancreas: actuarial risk of malignancy,” Clin. Gastroenterol. Hepatol., vol. 4, no. 4, pp. 460-468, Apr. 2006, doi: 10.1016/J.CGH.2006.01.018), is typically monitored with expensive and difficult to access MRI at 1, 3, and 5 years [13], Liquid biopsy tests would reduce costs and improve access, but currently there is only one clinically approved biomarker for PDAC, carbohydrate antigen 19-9 (CAI 9-9), which has been shown to have a low positive predictive value (PPV) in early detection [14]— [18], Utilizing a previously published liquid biopsy-based protease assay for initial screening, it was found that all stages of PDAC showed upregulated circulating protease activity compared to normal, inflammatory, and neoplastic states (pancreatitis, IPMN ± Pancreatic Intraepithelial Neoplasia [PanlN]). To understand the specific proteases upregulated in PDAC different antibody based assays were used and it was found that MMP2 and MMP14 were drivers in PDAC patients. Finally, a nanosensor was developed that utilizes protease activity in serum for further clinical translation of this work by increasing throughput, increasing dynamic range, and improving the signal read-out. The nanosensor was tested alone and in combination with CAI 9-9 in a screening, training and blinded validation cohort. The screening nanosensor had a strong statistically significant difference between PDAC and precursor disease or healthy, as expected from our initial screening. The training set was used to establish cutoffs for the blinded validation and found 92% overall accuracy for PDAC and healthy patients when using both the nanosensor and CAI 9-9. Importantly, in the blinded validation cohort, 100% specificity with 85% sensitivity was obtained. In conclusion, these results show the potential of the nanosensor as a rapid and sensitive screening tool to differentiate PDAC from other high-risk disease states and improve early detection.

Results

Charge Changing Protease (CCP) assay for rapid and low volume detection of protease activity

The effectiveness of an enzymatic activity assay was analyzed to measure active proteases in peripheral blood specimens with improved sensitivity for cancer detection relative to traditional expression assays by allowing for a single enzyme to catalyze multiple targets. To detect protease enzymatic activity in serum or plasma, a CCP assay that previously showed thrombin activity detection for diabetes with high sensitivity from a small volume of blood was optimized [11], [19], [20], The CCP assay detects a negative-to-positive charge change of a peptide and cleavage products are visualized and quantified on a polyacrylamide gel. Uncleaved peptides and serum proteins are positively charged and do not enter the gel, thus enabling a physical separation with the cleaved, labeled peptide.

To understand the sensitivity of the CCP assay, a single CCP probe (Probe 11) with multi-protease enzyme specificity (e.g., Cathepsin G, Pepsin C, Chymase and Chymotrypsin, but not Trypsin) was tested. After incubating the probe with varying concentrations of the proteases, it was found that the absolute ICso for Cathepsin G was 20 pM; Chymase was 9 nM; Pepsin C was 7 pM; Chymotrypsin was 5 nM and Trypsin was unstable (no ICso). These results suggest that the CCP assay can detect protease activity in the low atto- to femtomoles depending on probe: protease combination, which is nearly 100 times more sensitive than a typical sensitive ELISA for protein expression.

Screening of PDAC and healthy samples for an optimal probe set for PDAC detection

Instead of assuming that a single probe is cleaved by a single protease, the goal was to develop a set of peptide probes with a wide range of cleavage sites for many different proteases to develop a cancer detection blood-based assay. To this end, twelve (12) peptide probes were created that collectively contained 38 unique cleavage sites and 18 shared cleavage sites (pl-pT) that supported the post-cleavage, change in charge paradigm (Table 1).

Table 1. Peptide Probes. Unique Unique Cancer

Shared Name Sequence cleavage cleavage sites Protease cleavage sites sites final assay Signal

GEPEPFAGAGK (SEQ 3 1.8

Probe-1 3 3

ID NO: 13)

DGLAGGAGGK (SEQ 1 5.4

Probe-2 1 6

ID NO: 14)

DGDGMARTLK (SEQ 5 1.0

Probe-3 5 2

ID NO: 15)

DPLGLVGPK (SEQ ID 6 4.9

Probe-4 1 7

NO:16)

DGDPSLRSVSGK 5 3.0

Probe-5 4 5

(SEQ ID NO: 17)

DGDLRGGMPGSGK 5 4.1

Probe-6 5 5

(SEQ ID NO: 18)

GEIGRLSAGK (SEQ ID 2.0

Probe-7 6 2

NO: 19)

DGPAGLVGPK (SEQ 5.8

Probe-8 0 9

ID NO:20)

DLEVLIVLGK (SEQ ID 0.2

Probe-9 4 2

NO:21)

DSSLYSSSGK (SEQ ID 0.3

Probe- 10 3 4

NO:22)

DGDGAGYSLPAAGGK 0.5

Probe- 11 3 7

(SEQ ID NO:23)

DGDPAELRAGK (SEQ 0.6

Probe-12 3 5

ID NO: 24)

To explore the utility of the 12-probe panel in differentiating cancer from healthy control subjects, the CCP assay was used to screen plasma from 6 PDAC (Stage 2-4) patients and 6 healthy controls. By using the average signal generated from cancer and healthy patients, a cancer: healthy ratio was calculated to identify probes better or worse at distinguishing cancer from healthy. Four probes had cancer: healthy ratios >2, and 4 probes that had cancerhealthy ratios <1 (Fig. 1). From the remaining 4 probes, we selected the 2 probes with the most unique sites and highest signal in PDAC samples. The 6-probe panel demonstrated strong positive signal in PDAC samples containing 26 unique cleavage sites and 11 shared cleavage sites. Based upon the MEROPS database [21], these 37 cleavage sites could be cleaved by 100s of different proteases. These results suggest that multiple CCP probes can be designed to distinguish cancer from healthy control patients.

Additional peptide nanosensors were developed that can detect different proteases present in cancer patients. The initial nanosensor testing was done with Probe 1, which can be cleaved by MMP2, 9 and 12 (Table 2) [30], A set of 4 nanosensors are shown in Table 2. Table 2. Nanosensor design.

a - lower case letters are D-amino acids, b - Mca 7-methocoumarin-4-acetic acid

Circulating Protease Activity distinguishes PDAC from other disease states and healthy The previously published CCP assay was limited by manual detection of cleavage products and a limited dynamic range. Increasing the dynamic range would allow for greater specificity. Therefore, an improved CCP (iCCP) assay was developed that utilizes a Gelbased Automated Detection software, which was developed as an automated R-based program for the quantification of proteolytic fluorescent signal. This approach provides an unbiased quantification of the cleavage product, resulting in a more accurate and precise evaluation of the total activity signal and increases the dynamic range by two-fold (p<0.0001). The software operates to detect individual wells and integrates the total signal for each specific probe. The key advantages of this software are the enhanced limit of detection and higher-throughput while limiting user-based bias.

Serum samples were analyzed from PDAC patients (n = 67) and healthy controls (n = 67) with the iCCP assay (Table 3). To individually evaluate each probe for efficacy in differentiating PDAC from healthy controls, the raw data was z-scored due to their different levels of total fluorescence signal between probes. For all probes, a statistically significant difference was identified between PDAC and healthy controls (p<0.0001, one-way ANOVA followed by Tukey’s multiple comparisons tests; Fig. 2A). PDAC patients were analyzed across stage (I-IV) (Table 3) to determine if circulating protease activity differed across cancer disease burden. Surprisingly, there was no statistically significant difference in protease activity across stage, however protease activity was statistically significant between stage II-IV tumors and healthy controls. Further, probe-1 showed a significant difference between controls and stage I tumors (Kruskal -Wallis test with Dunn’s multiple comparisons tests, Fig. 2B). Furthermore, probes- 1, 4 and 5 showed a significant difference between early stage I/II patients and pancreatic neoplasia patients (Multiple t-test with False Discovery Rate, Fig. 9). These results indicate that the protease probe panel can predict cancer presence, but their levels did not correlate with cancer progression.

To further analyze the ability of the iCCP panel to differentiate high-risk pre- cancerous diseases from PDAC, precursor lesions (IPMN and PanIN) was analyzed as well as pancreatitis patients. Along the PDAC disease axis, chronic pancreatitis is an inflammatory disease that harbors a very low probability of developing cancer, while within the precursor lesions, IPMN is a benign neoplasia with a low probability of progressing to cancer [22], whereas PanINs are small lesions that are thought to be direct precursors of PDAC [23], [24], Using scaled raw values from PDAC (n = 67), healthy controls (n = 67), precursor (n = 21, IPMN and PanIN) and pancreatitis (n = 11) sera, it was determined that all 6 probes were detected at levels significantly lower in sera from patients with pancreatitis compared to PDAC. Similarly, for precursor lesions, statistical significance in protease activity was detected in only probe-2 when compared to healthy control sera, and 5 out of 6 probes distinguished the precursor lesions from PDAC with statistical significance (Fig. 2A). These data indicate that protease activities are different across the disease axis and suggest that the probe panel can be used to distinguish PDAC from inflammation, precursor lesions and healthy controls, while also distinguishing precursor lesions from healthy controls.

Table 3. Demographic, clinical characteristics and assay data of patients. Continuous variables are presented as mean (SD); categorical variables are presented as frequency (%). IPMN - intraductal papillary mucinous neoplasia; PanIN - pancreatic intraepithelial neoplasia

Cross-validation reveals strong differentiation of PDAC from control samples for a single probe

Each probe was analyzed individually for AUC of the ROC, which revealed that probe 1 gives the strongest AUC compared to the other probes (Probe 1, AUC = 0.87; Probe 2, AUC = 0.71; Probe 3, AUC = 0.78; Probe 4, AUC = 0.82; Probe 5, AUC = 0.82; Probe 6, AUC = 0.83). To determine the clinical performance of the iCCP probe panel, a logistic regression (LR) model followed by 5-fold cross-validation (CV) (80%-20%, 200 random permutations) was tested. While the combination of all iCCP probes have an AUC greater than 0.8, the simplest model contains only Probe-1 due to multicollinearity (Fig. 7). After cross-validation with probe 1, the mean accuracy was 0.79 + 0.059, mean AUC 0.87 + 0.063, and sensitivity and specificity of 0.68 + 0.125 and 0.9 + 0.088 respectively (Figs. 2C and 2D). These results suggest that circulating proteases are a strong biomarker for PDAC early detection.

Prognostic ability of serum protease activity in PDAC patients

To determine if the iCCP panel had prognostic capacity, survival was evaluated using Kaplan-Meier (KM) with long-rank statistics. To improve the transparency of the KM test, a modified application of the KM test was developed to assess the Prognostic CApability of Biomarkers (ProCAB) for continuous variables. Instead of arbitrarily choosing a cut-off point, every possible cut-off in the dataset was tested using a sliding window approach, then tested a KM model for each step and annotated the significance of the log-rank test at every step. Using the sliding window approach, the prognostic capability of each iCCP probe was analyzed for overall survival (OS). Interestingly, only 3 (i.e., Probe-2, Probe-3, and Probe-4) of the 6 probes had multiple significant cutoffs in a row indicating reliable prognostic ability for OS. Next, using a weighted sum, three significant probes (Probe-2, Probe-3, and Probe-4) was combined into a single iCCP score and performed ProCAB. While individually, these probes have prognostic ability, when combined they show an increased spread in the number of cut-offs with p-value lower than 0.05 as well as a lower absolute minimum p-value. When using the cutoff of 0.5, this reveals a strong prognostic ability for this probe combination (probes-2, 3, and 4) (Fig. 2E). These results suggest that a subset of circulating protease activity prognostically identifies PDAC patients with more favorable OS.

Identification ofMMP2 and MM Pl 4 as the predominant circulating proteases in PDAC Table 4. Probes and Proteases that Cleave them.

While the protease probe panel was created to target a wide number of proteases, it was desired to find the specific protease(s) that are cleaving the iCCP probes. The specific cleavage sites within each probe were identified. LC-MS was used to analyze the cleavage sites of the iCCP probes after incubation with cancer (high activity) or healthy (low activity) serum. For example, Probe-1 shows a higher intensity LC-MS peak at site P/F compared to the other possible cleavage sites and the high activity sample is orders of magnitude greater than the low activity sample (Fig. 3A). Using this analysis, the most likely cleavage site for each probe was calculated (Fig. 3B and Figs. 8A-8C). These results revealed that all 6 peptide probes were cleaved at unique amino acid sites. To further understand the protease required for cleavage of these probes, the online database ProCleave was used and focused on three protease families previously identified to be upregulated in cancer, MMPs, Cathepsins and Caspases (Table 4). The cleavage sites identified via LC-MS were used and the number of probes that a single protease could theoretically cleave was analyzed. It was observed that MMPs were the most likely family to cleave the highest number of probes and MMPs were the only proteases that could theoretically cleave probe- 1 at the P/F site, the best performing probe. These results suggested that MMPs could be an important driver in PDAC but it also suggested that a combination of multiple proteases from different families may be present in PDAC.

To further identify the proteases that are enzymatically active in circulation of PDAC patients, it was tested if inhibition of MMPs altered circulating protease activity. It was found that two pan-MMP inhibitors (Actinonin and NNGH) blocked the majority of enzymatic activity in serum for 5 of 6 iCCP probes with minimal reduction in activity for probe-2, which correlates with the Procleave data to theoretically identify the possible proteases that could cleave each probe. A more specific inhibitor against MMP-2/9/14 (ND-336) for the 5 probes was tested and a similar inhibition of three probes (probe- 1, 4 and 6), but only a partial inhibition of the other two probes (probe-3 and 5) was observed. The inhibitor data combined with the cleavage site data strongly suggest that the MMP family of proteases is important in PDAC, but also that multiple protease families are present in circulation of PDAC patients (e.g. Probe 2).

To further identify the specific MMP in PDAC, the Sensolyte Plus Assay was used with enhanced selectivity which uses antibodies to capture target proteases, then measures the protease activity. MMP2 (with minimal cross-reactivity with MMP 14), MMP9 and MMP 13 were tested and significant differences were observed in activity between PDAC and control patients for MMP2, but not for MMP 13 or MMP9 (multiple t-test with FDR correction). To determine the relation between MMP2 specific activity and the iCCP assay, a correlation matrix was performed for the average of the three probes that showed highest inhibition level using specific inhibitor against MMP-2/9/14 (Probe- 1,4 and 6) and the lowest inhibition (Probe-2, 3 and 5). It was observed that iCCP probes with higher inhibition correlated better to the MMP2 specific assay (R=0.532) compared to iCCP probes with lower inhibition (R=0.374) (linear regression comparison, p=0.022). Next, the expression of MMP2, MMP9, MMP 14, MMP 13 and TIMP2 (an MMP inhibitor protein) was analyzed and slight changes were found in the levels of proteins, but no statistically significant difference in PDAC compared to healthy serum (Mann-Whitney test with multiple comparison correction). Finally, it was identified that normalizing activity by expression revealed that not only MMP2, but also MMP13, can separate PDAC from control patients since MMP expression and activity are not directly related (Mann-Whitney test with multiple comparison correction). These results strongly suggest that MMP2 is a key protease for the detection of PDAC, but that again this is not the only protease upregulated in PDAC patient blood.

Since MMP2 is biologically associated to MMP14 and both MMP2 and MMP14 have been known to be on extracellular vesicles (EVs), their expression and activity on EVs was examined [25]— [30] . EVs are nanometer sized, contain cargo from the cell of origin, are actively and passively secreted from cells, and are highly stable in circulation [7], [31], EVs have a diverse cargo from mRNA, microRNA and protein including proteases and have been identified with cancer development and metastasis. To test if EVs have active MMPs, the ExoTIC platform was used which allows for gentle isolation of specifically sized EVs and also isolation of free floating proteins [32], After isolating EVs between 50-200nm, this fraction contained tetraspanins (CD63 and CD9) that are typical markers of exosomes. Both PDAC and control EVs had the expected size range of exosomes with no significant differences in EV concentration between PDAC and control samples. Super-resolution microscopy was used to measure the percentage of both MMP14 and MMP2 on EVs. It was observed that both MMP14 and MMP2 proteins were present on EVs as shown by colocalization with a membrane stain. In addition, transmission electron microscopy (TEM) showed typical EV sized objects. The fraction of MMP14⁺ EVs were elevated in PDAC compared to control samples (p<0.05), but there was no difference in the fraction of MMP2⁺ EVs between control and PDAC samples (p>0.05). Finally, by analyzing total protein levels, a > 10-fold increase in MMP14 expression on PDAC EVs compared to control EVs, but similar levels of MMP2 expression in control EVs compared to PDAC EVs was found. In addition to overall expression, protein size was also analyzed because pro-MMP proteins are larger in size and inactive ([33]— [35]), but upon cleavage of the pro-peptide, the resulting active MMP protein is smaller in size. By examining the Western blots for protein size, a >20-fold increase in MMP14 cleaved product and 1.7-fold increase in MMP2 cleaved product was found. These results strongly suggest that MMP2 and MMP 14 are present on EVs and are correlated with PDAC status.

Finally, the ExoTIC isolated EV samples were analyzed for active proteases using our iCCP panel by comparing plasma, EV-isolated fraction and EV depleted flow-through in PDAC and control samples. In PDAC samples, plasma and EV fractions had similar amounts of protease signal while the flow through signal was significantly lower for probes 1, 2, 4 and 6. Interestingly, in control samples, plasma and flow-through had similar levels, but the EV fraction was significantly lower for probes 1, 2, and 4 (Figs. 4A and 4B, Kruskal -Wallis test with multiple comparison correction). In addition, when correlating the plasma signal with each one of the fractions, the increased plasma signal was due to the increased signal from the EV fraction, and independent of the flow-through. These results suggest three things: 1) PDAC patients are continually secreting EVs that are associated with MMP2 and MMP 14, 2) these EVs contain proteases that remain functional in circulation, and 3) serum or plasma can be used as a surrogate for isolated EVs for building a point-of-care device.

Protease activity nanosensor development for high-throughput PDAC detection

To optimize the clinical translation of the iCCP assay, the cleavage product detection approach was modified to be run with unaltered serum, without the need for charge changing peptide sequences or cumbersome electrophoresis. To do this, a nanosensor assay (also known as PAOMANN assay or protease activity -based assay using a magnetic nanosensor) was developed where a fluorescent modification of probe- 1 was covalently attached to the surface of a 50nm iron oxide nanoparticle (termed PAOMANN- I ) and after incubation with serum the non-cleaved peptide remaining on the beads can be magnetically removed from the solution. First, to increase the specificity of the assay and reduce non-targeted cleavage, Probe- 1 was designed to have D-amino acids at different positions away from the cleavage site since D-amino acids cannot be cleaved by most human proteases. These experiments showed that at least 6 L-amino acids are required for cleavage, whereas 5 and 2 L-amino acids are not sufficient for cleavage (Fig. 5A, p<0.0001, one-way ANOVA).

To demonstrate the benefit of using the nanosensor specifically for high-risk patient monitoring in unaltered serum, a subset of PDAC (n = 27) and healthy controls (n = 27) was analyzed and the number of precursor/pancreatic neoplasia samples (n = 38) with similar average age and sex was analyzed. For each sample, the nanosensor was incubated with 8 pL of serum and a magnet was used to remove the non-cleaved probe from the solution containing the cleaved product. Similarly to the iCCP assay for probe- 1, a non-significant difference between healthy and precursor/pancreatic neoplasia samples but a significant difference between both of those and the PDAC samples was observed (Fig. 5B). The results from nanosensor and iCCP assays were strongly correlated (R = 0.69) suggesting a robust use of both assays for detecting circulating proteases specific to cancer (Fig. 5C). Finally, to assess the reproducibility and variability of the assay, the batch-to-batch effects with all the PDAC and controls samples showing strong correlation between two separate runs (R = 0.94) was analyzed (Fig. 5D). These results suggest that this new assay can distinguish high-risk disease states from cancer. In addition, the dynamic range (LowestSignakHighestSignal) increased by 7-fold (0.6: 11.3 for CCP and 384:50000 for PAC*MANN-1) and the background resolution (Background:LowestSignal) increased by 3-fold (0.5:0.6 for CCP and 91 :384 for PAC*MANN-1) with the PAC»MANN-1 assay compared to the CCP assay. Finally, to assess the reproducibility and variability of cleavage of PAOMANN- I , all PDAC and healthy control sera were reanalyzed and a strong correlation between the technical replicates from the same batch of magnetic nanoparticles (R = 0.94, linear regression) were observed. These initial results demonstrate the clinical potential of PAOMANN- I in distinguishing high-risk disease states from PDAC.

To further evaluate PAOMANN-1, a larger cohort of patient samples were obtained that included an entirely new set of healthy control samples (n=170), 95 new PDAC patient samples (n=l 10), pancreatitis patient samples (n=45), and pancreatic neoplasia (IPMN+PanIN) patient samples (n=31). When added to the previously tested specimens, a total of 356 samples were evaluated, of which 110 were PDAC samples and 246 were non- PDAC samples. To test if PAOMANN- I would specifically detect PDAC and not other comorbidities, we grouped the pancreatitis and neoplasia with healthy controls into the non- PDAC group. The 356 samples were run in 6 groupings by 4 experimenters with PDAC and non-PDAC samples interspersed among each grouping. The samples were not statistically different for age (p=0.74, Mann-Whitney test), sex (p=0.72, Fisher’s exact test) and diabetes mellitus (p=0.16, Fisher’s exact test) when comparing PDAC to non-PDAC patients. Healthy patients were not significantly different in age compared to PDAC or high-risk patients (p=0.053, Kruskal-Wallis test with Dunn’s multiple comparison). Never smokers were more prevalent in the non-PDAC patients compared to PDAC patients (p=0.01, Fisher’s exact test). In addition to measuring protease activity, all samples were also analyzed for CA 19-9 protein levels. To study batch effects of the PAC»MANN-1 assay, we generated two different sets of PAOMANN-1 magnetic particles, one with NHS group on the surface and the other with a carboxyl group on the surface. We found a strong correlation in both the PDAC (p<0.0001, r=0.92, Spearman, n=109) and the non-PDAC (p<0.0001, r=0.69, Spearman, n=53) samples with an overall Spearman correlation of r=0.89. We also ran independent serum aliquots of 39 pairs of PDAC and healthy patients and found a coefficient of variation of <20% similar to other reported values. We compared fresh versus frozen samples and found no significant difference between these groups. These results show the high reproducibity and robust performance of the PAC*MANN-1 assay.

Training and blinded validation of the nanosensor for PDAC detection

Next, a larger cohort was obtained with an untested set of control patients that were screen negative for breast or prostate cancer and the addition of 65 untested PDAC patient samples. This resulted in a total of 228 samples (124 PDAC and 104 control). With the high- throughput and rapid nature of the nanosensor, all samples were run in less than 6 hours in a single day. Since CAI 9-9 is the current standard of care biomarker for PDAC diagnosis, all samples were analyzed for CAI 9-9 protein levels from the same serum sample. The samples were randomly divided into 50% non-PDAC (healthy, pancreatitis, and neoplasia) and 50% PDAC for a training set (n=178) and the other half went into a blinded validation set (n=178). In the training cohort for PAC*MANN-1, strong statistically significant differences between healthy controls, pancreatitis or neoplasia compared to PDAC samples (p<0.001 Kruskal- Wallis test with Dunn’s multiple comparison) were detected (Fig. 10A). However, no statistically significant differences were found between healthy controls compared to pancreatitis or neoplasia samples (Fig. 10A). CAI 9-9 showed strong statistically significant differences between healthy controls or pancreatitis compared to PDAC samples (p<0.001 Kruskal -Wallis test with Dunn’s multiple comparison) (Fig. 10B). However, no statistically significant differences were found between neoplasias and PDAC samples (Fig. 10B). Within PDAC patients, there was no difference in PAOMANN- I levels between PDAC patients with diabetes mellitus or without (p=0.22, Mann-Whitney test), with pancreatitis or without (p=0.14, Mann-Whitney t-test), and never, former or current smokers (p=0.9, Kruskal-Wallis test with Dunn’s multiple comparison). In addition, we found no differences in the healthy population with or without diabetes (p=0.32, unpaired t-test) and never, former or current smokers (p=0.47, Kruskal-Wallis test with Dunn’s multiple comparison). The data from PAOMANN- I were analyzed with and without CA 19-9 values in our training and blinded validation cohorts. Logistic regression was used to combine PAC*MANN-1 with CA 19-9 into a single model. The AUC of the training set comparing PDAC to non-PDAC for PAC*MANN-1 alone was 0.91+0.03, for CA 19-9 alone it was 0.77+0.04, and for the combination it was 0.94+0.02 (Fig. 6B). The AUC of the training set comparing neoplasias to PDAC for PAC‘MANN-1 was 0.93+0.04, for CA 19-9 alone it was 0.56+0.08, and for the combination it was 0.89+0.04 (Fig. 6C). Finally, the AUC of the training set comparing pancreatitis to PDAC for PAC*MANN-1 was 0.9+0.04, for CA 19-9 was 0.82+0.05 and 0.94+0.02 for the combination (Fig. 6D). These results show that PAC*MANN-1 performs better than CA 19-9 alone and that the combination of both slightly improves distinguishing PDAC status from healthy and pancreatitis but doesn’t improve performance for pancreatic neoplasia patients. For the blinded validation, logistic regression was performed on the training set data and then cutoffs were established for the blinded validation set based on the model with the highest accuracy in the training set, set specificity of 99%, and set sensitivity of 90% (Table 5). Utilizing the most accurate training model, we found a 90% accuracy for PDAC and non- PDAC patients in our blinded validation cohort using only PAOMANN-1, while there was only 79% accuracy for CA 19-9 alone. Combining PAOMANN- I and CA 19-9 together resulted in an overall accuracy of 93%. The PAC*MANN-1 assay alone identified 98% of non-PDAC samples and 73% of all PDAC samples (Table 5). When divided by cancer stage, the PAC*MANN-1 assay alone resulted in sensitivities of 62% for Stage I, 56% for Stage II, 92% for Stage III, and 85% for Stage IV cancers (Fig. 6A). By comparison, CA 19-9 alone resulted in sensitivities of 31% for Stage I, 31% for Stage II, 54% for Stage III, and 69% for Stage IV (Fig. 6A). Importantly, the combination of PAOMANN- I and CA 19-9 resulted in improved sensitivity for early stage disease with 85% for Stage I, 81% for Stage II, 92% for Stage III, and 92% for Stage IV (Fig. 6A). In addition, the PAC*MANN-1 assay can distinguish 100% of neoplasia (16/16) and 100% of pancreatitis (22/22) patients with 73% sensitivity for PDAC patients, while CA 19-9 only distinguishes 81% of neoplasia (13/16) patients with only 45% sensitivity for PDAC. These results suggest that the PAC*MANN-1 assay can be used in high-risk populations for PDAC detection.

In order to determine the effectiveness of the PAC*MANN-1 assay in an early detection setting, we analyzed the blinded dataset for extremely high specificity to eliminate false positives and high sensitivity to eliminate false negatives (Table 5). At 99% specificity, the PAC*MANN-1 assay correctly called all non-PDAC cases (123/123), while still correctly calling 58% of PDAC cases (32/55). By comparison, CA 19-9 correctly identified 98% of non-PDAC cases (121/123), but only correctly identified 5% of PDAC cases (3/55). Importantly, the combination of PAOMANN- I and CA 19-9 correctly identified all non- PDAC cases (123/123) and improved identification of PDAC cases to 71% (39/55). At 90% sensitivity, the PAOMANN-1 assay correctly identified 91% of PDAC cases (50/55) and still correctly called 68% of non-PDAC cases (81/123). By comparison, CA 19-9 correctly identified 87% of PDAC cases (48/55), but only correctly called 35% of non-PDAC cases (43/123). Again, the combination of PAOMANN-1 and CA 19-9 correctly identified 93% of PDAC cases (51/55), but improved the identification of non-PDAC cases to 78% (96/123). Finally, one of the major problems with CA 19-9 as an early detection biomarker is approximately 20% of people with PDAC either have low CA 19-9 secretion or do not express the Lewis antigen for CA 19-9 detection. To determine if the PAC*MANN-1 assay can predict cancer in patients that are low for CA 19-9, we found that in patients with low CA 19-9 (<37 U/mL), the PAC*MANN-1 assay correctly identified 79% (11/14) with 100% specificity. These results demonstrated that the PAC*MANN-1 assay is far superior and not correlated to CA 19-9 and when combined with CA 19-9 shows strong specificity and sensitivity of early stage PDAC detection from a small volume of blood in a high-throughput and rapid manner.

Table 5. Blinded validation. Percent correctly called in the blinded validation study for non-PDAC and PDAC samples, (a - Most accurate model was calculated from the test set with no set specificity or sensitivity, b - High-Risk patients are pancreatitis plus neoplasia, c - Non-PDAC are Healthy plus pancreatitis plus neoplasia, d - PDAC stage is pathological stage.)

Discussion

Development of novel biomarkers to diagnose lethal cancer early and predict its progression is an unmet need for impacting survival from cancer. Most cancer detection biomarkers are assayed using hallmarks of apoptosis or rely upon substantial disease burden, and some blood-based biomarkers are too rare for reproducible reliability. Current clinical assays using CAI 9-9 can distinguish cancer from healthy, but there are still a significant number of false positives and false negatives [36], [37], An ideal blood-based biomarker would harbor the ability to amplify a small signal or event. Thus, measurement of a secreted active enzyme that is specific to neoplastic cells and not precursor cells have exciting promise. Proteases secreted into peripheral blood offers an intriguing target, since protease activity can specifically amplify the assay target signal. Moreover, protease activity is essential for tumor cell invasion, extravasation and intravasation making its presence ubiquitous in cancer progression. A set of protease activated nanosensors were developed that allow for the detection of pancreatic cancer at early stages relative to healthy subjects and precursor lesions. The nanosensor adds another biomarker for early detection of PDAC and reduces false positives and negatives. The nanosensor allows for the detection of early-stage pancreatic cancer relative to healthy subjects and patients with pancreatitis or pancreatic neoplastic lesions. The present disclosure demonstrates that a single peptide probe (PAC*MANN-1) has high specificity and sensitivity, thus reducing the number of false positives and negatives and significantly better at detecting early stage PDAC compared to CA 19-9. Further, the PAOMANN- I assay is independent of CA 19-9 and can be utilized in the 20% of patients that do not express or release CA 19-9 into circulation. Another advantage of the PAC’MANN assay is that the assay is not limited to a specific protease, such as ELISA based methods, but can measure a broad range of proteases to improve sensitivity. Finally, when compared to methylated cfDNA based tests, the PAC’MANN assay uses significantly less blood volume (10pL versus lOmL), using only a single informative probe compared to >100,000 methylation sites, and having similar detection sensitivities as a previously reported method for pancreatic cancer stage I (63% versus 62% [PAOMANN alone] or 85% [PAC«MANN+CA19-9]), stage II (83% versus 56% or 82%), stage III (75% versus 92% or 92%) and stage IV (100% versus 85% or 92%). In addition, by combining the nanosensor with CA-19-9, diagnostic power was improved. This nanosensor lays the foundation for a sensitive, low cost, low volume and scalable cancer early detection test that can be used regularly in high-risk patient populations or populations with less access to medical tests. With the high-throughput, low volume and low cost of the nanosensor assay, the results suggest this assay will be very efficient at regularly screening different high-risk populations.

Materials and Methods

Protease conjugation and storage

Peptides were purchased from GenScript and designed with N-terminal acetylation and C-terminal amidation with >98% purity and Trifluoro acetate (TFA) removal. Peptides were dissolved in 100 mM of NaHCCh pH 8.2 to a final concentration of 10 mg/mL. In parallel, we dissolved BODIPY FL NHS Ester (Lumiprobe) in DMSO to a final concentration of 10 mg/mL. Next, we combined equal volumes of both solutions and incubated them at room temperature for 1 hour protected from light. After incubation, the solution was diluted 1 : 10 in ultrapure water and aliquoted into smaller stock volumes to a final concentration of 500 pg/mL of peptide. Each aliquot was only thawed once to avoid peptide degradation.

Human Research Participants and Biological Material

All ethical guidelines for human research participants were applied. The Institutional Review Board at Oregon Health and Science University provided guidelines for study procedures and protocol was approved (IRB00005169). Informed consent was obtained from all study participants. PDAC, Neoplasia and Pancreatitis patients were fasted the night before blood draw. Breast and prostate screening patients (Healthy) were not specifically told to fast before the blood draw. PDAC, Neoplasia and Pancreatitis patient blood draw was done before the initial endoscoptic ultrasound (EUS) or surgery, not under anesthesia. For non-EV samples, serum was used. Blood was collected in serum tubes and allowed to clot for 30 minutes, then the clot was removed by centrifugation at 2,000 x g for 10 minutes. For the EV samples, plasma was used. Blood was collected in heparin tubes, centrifuged at 1,500 x g for 10 minutes, then 2,500 x g for 10 minutes. All samples were stored at -80°C. Retrospective samples were used. Sample size of PDAC was based upon obtaining samples that were untreated with similar numbers of each stage. Healthy control samples were obtained to match PDAC sample numbers. High-risk comorbidity samples of IPMN/PanIN and pancreatitis samples were obtained to give similar numbers to each stage of PDAC. Sex was as close to evenly distributed between male and female. Diabetes mellitus and smoking status were also acquired to understand low-risk co-morbidities.

CCP protease assay

The general protease assay, unless indicated otherwise, was performed as follows. First 2 pL of 50 mM of calcium chloride is added into a tube or a well of a 96 well plate followed by 4 pL of working stock of peptide and 4 pL of serum or plasma. Finally, the solution was mixed and incubated with 150 rpm agitation for 45 minutes at room temperature in the dark. After incubation 8 pL are loaded per well to a 20% acrylamide TBE gel (ThermoFisher) and ran at 250V for 60 minutes with inverted polarity. After the gel has been processed, the gel was imaged using the iBright FL1000 (ThermoFisher) using 488nm channel with 100 ms exposure time unless otherwise specified. Assay was performed nonblinded.

Quantification for screening

For peptide screening and selection of best targets, we used the previously published methodology for the quantification [11], Briefly, ImageJ was used to create a box around the signal of 10x10 pixels and a signal integration was performed and was further used for analysis. Another 10x10 pixel box was created in a region without signal for background subtraction.

Quantification Software Development

An R script was developed to be able to automatically quantify the fluorescence signal from electrophoretic gels and was specially adapted for each peptide developed in this study. Briefly, the script will take an image where the wells are marked with white lines and will quantify the fluorescence pixel intensity of the signal using an automated integration methodology. The first step is to load the images in tiff file and the script will convert them automatically into a png file. Next, it will compute an image gradient in both the x and y direction and merge them into a single image to be able to detect any borders that are present in the image. After, the script will recognize the y pixel where the wells begin (where the gel start) and the pixels of x where every well starts and ends. After finding the wells, the script will take the average 20 pixels of the center of every well and average them. After averaging them, we will plot a density histogram with the average 20-pixel intensities for all pixels on the y-axis. Next, using experimental data we quantify the length that cleaved probes run into the gel and using that distance in a probe-specific manner and the Simpson rule the script performs a signal integration over 150 pixels. Finally, the integration results are plotted into a histogram where values are indicated and the original gel image will be returned with a rectangle around the areas where the signal was integrated for every well. For every gel processed a pdf is created with a summary of all the plots as well as the final intensity used for further processing.

Logistic regression model and cross-validation

For selecting the best probes, all-subsets logistic regression was performed where all possible models of individual and combinations of probes were tested using cross-validation. Briefly, the caret package in R was used to perform a 5-fold cross-validation with 200 random permutations for each possible combination of variables. For each model, the mean cross-validated AUC of the ROC was annotated. Then, we compared the mean AUC of all the models and selected the variable(s) with the highest AUC. To evaluate the performance of the selected probe more in-depth, we repeated the 5-fold cross-validation with 200 permutations with just the selected variable (in this case Probe-1 only). For each permutation, we annotated the accuracy, AUC of the ROC, and the ROC sensitivities and specificities. The data was summarized and plotted using Graphpad prism 10.0.

For the blinded validation, we evaluated three different cut-offs of a logistic regression model, highest accuracy cut-off, 99% specificity cut-off and 90% sensitivity cutoff. Using R and the caret package, we first created a logistic regression model utilizing only the training dataset. Then, to determine the most accurate cut-off, we tested the model on the same dataset and binarized the data utilizing different cut-offs ranging from 0.2 to 0.9 with 0.01 increments. We iterated through this process and for each cut-off, the predicted probabilities were binarized and a confusion matrix was computed. For each iteration, we used this formula to calculate accuracy:

TP + TN

^Accuracy = _{TP + TN + FP + FN}

Where TP is true positive, TN is true negative, FP is false positive, and FN is false negative. The selected cut-off with highest accuracy was used for the blinded dataset. In the scenario where there are multiple cut-offs with the same highest accuracy value, the average between the cut-offs was used. Similarly, utilizing the same logistic regression model, we selected the cut-off that led to 99% specific (99% of controls being correctly called) and 90% sensitive (90% of non-control individuals being correctly called) in the training dataset.

Once these cut-offs were established, we used the logistic regression model developed for the training dataset on the blinded data for testing. Then, the predicted probabilities for each patient in the blinded dataset were binarized utilizing the different cut-offs and performance metrics were calculated utilizing the clinical ground truth (Table 5).

Kaplan-Meier prognosis analysis

When the KM test is used for continuous variables, an arbitrary cut-off must be chosen to separate high and low levels for the test variable. Different ways to choose this cutoff have been investigated, such as the intermediate point between two normal distributions in the dataset or arbitrarily determining a 1/2, 1/3, or 2/3 cut-off. This arbitrary cut-off presents a clear bias since is up to the researcher to pick the cut-off and in many cases, there is no rational design as to why such a cutoff was chosen. Survival was the only clinical endpoint considered. Patients were followed from time to diagnosis to death for up to 10 years. In lieu of choosing a specific cut-off arbitrarily, we tested every possible cut-off in the dataset using a sliding window approach, then developed a KM model for every step and annotated the significance of the log-rank test for every model created. Next, we plot the population proportion used to create the model and the log of the p-value. In these plots, there are two main features that are noticeable. The first feature is the lowest p-value of the dataset which determines the best possible cut-off that led to the high prognostic capability of the biomarker. The second feature is the spread of cut-offs that lead to a log-rank statistic p-value lower than 0.05. In essence, this feature is a representation of the robustness of the biomarker since the bigger the spread, the more likely the biomarker is to withstand variability with increased sample sizes. The methodology developed here, termed ProCAB, for KM prognosis analyses represents a transparent way of understanding the overall performance of biomarkers when Cox Regression cannot be used.

To do this, a survival R package and a sliding window approach for every 0.01 z- score in the middle 5-95% of the dataset was used to perform the Kaplan Meier plot and the log-rank test was annotated as a function of the proportion of patients binarized as high. This methodology works under the premise that a good prognostic biomarker will have a wide distribution of possible cut-offs that still separate the two populations in the dataset significantly. In order to assess the performance of each biomarker we looked at two metrics, the number of possible significant cut-offs in the dataset as well as the minimum and median p-value in the middle 50% of the dataset. Finally, in order to combine multiple biomarkers into a single index we used the following weighted sum equation where B is the number of biomarkers, PSC is the possible significant cut-offs in the middle 50% of the dataset, and min.pval is the minimum p-value also in the middle 50% of the dataset.

Liquid chromatography-mass spectrometry (LC-MS)

After incubation of plasma samples with each respective peptide, the 8 pL incubation mixtures were frozen before mass spectrometric analysis. Upon thawing, proteins were precipitated by the addition of 24 pL of acetonitrile, samples vortexed, incubated for 5 min, centrifuged at 7,800 x g for 10 min at 20°C, and supernatants removed. A peptide assay was then performed on a portion of the supernatants using a Pierce quantitative colorimetric assay (Thermo Scientific) and indicated that approximately 9 pg of peptide was recovered from each sample. Peptides were then dried by vacuum concentration, dissolved in 5% formic acid and an equal volume of each sample containing approximately 100 ng of peptide was analyzed by liquid chromatography-mass spectrometry (LC-MS). An NCS-3500RS UltiMate RSLCnano UPLC system was used for peptide separation and an Orbitrap Fusion Tribrid instrument with an EasySpray nano source for mass analysis (Thermo Scientific). Peptides were injected onto an Acclaim PepMap 100 pm x 2 cm Nano Viper Cl 8, 5 pm trap column on a switching valve. After 5 min of loading and washing, the trap column was switched on-line to a PepMap RSLC C18, 2 pm, 75 pm x 25 cm EasySpray column (Thermo Scientific). Peptides were then separated using a 7.5-30% acetonitrile gradient over 60 min in a mobile phase containing 0.1% formic acid at a 300 nl/min flow rate. Survey mass spectra were acquired over a m/z 175-1600 range in the Orbitrap at a resolution of 120,000 and targeted MS2 scans were performed in the Orbitrap mass analyzer at 7,500 resolution using the calculated m/z values of each peptide, as well as all possible N- and C-terminal fragments calculated using Skyline software (version 21.2.0.425) (MacCoss lab, https://skyline.ms/project/home/software/Skyline/begin.view). Either +1 or +2 charge states for each peptide was used, depending on whether the fragment contained either 1 or 2 basic residues. For each peptide substrate, between 17-35 precursor ions were specified with a loop count set so a survey scan would occur after each group of 10 MS2 scans. Survey scans used a maximum ion time of 50 ms, AGC target of 4 x 105, and were mass corrected using the internal ETD calibrant. Targeted MS2 scans using a quadrupole isolation set at 1.6 m/z, HCD activation at 30% collision energy, maximum injection time of 22 ms, and AGC setting of 5 xl04. Results were analyzed using Skyline software, and when possible, the identities of proteolyzed peptide fragments were confirmed by coelution of expected fragment ions in MS2 scans and precursor peptides in survey MSI scans. Due to their lack of retention on the reverse phase column, single amino acids from the N-terminus were not targeted. However, the BODIPY labeled C-terminal lysine was included (+1 charge state m/z = 420.2377).

Extracellular vesicles (EVs) isolation

EV isolation and processing was performed using the Exosome Total Isolation Chip (ExoTIC) according to previously published by the authors [32], Briefly, the ExoTIC device was assembled housing a stack of membranes: a 50-nm polycarbonate nanoporous low protein binding filter membrane, a 200-nm PES support membrane, and a thick paper pad. 300 pL of plasma was diluted in PBS and filtered through a 0.22-pm PES syringe-type filter before being introduced to the ExoTIC system. After pre-filtration, the diluted samples were processed via ExoTIC with a flow rate of 5 mL/hr using a syringe pump in a cold room. EVs, putatively sized 50-220 nm were retained in the isolation chamber in front of the filter membrane, whereas other small molecules smaller than 50 nm, such as free nucleic acids and proteins, passed through the outlet and was also collected for controls. PBS was then used to wash the retained EVs and remove remaining contaminants. After the washing step, the EV isolate was collected from the inlet and then analyzed by Nanoparticle Tracking Analysis (NTA).

Nanoparticle tracking analysis (NTA)

NTA was performed using the Nanosight NS300 (Malvern, UK). The sample was diluted with PBS according to the manual’s recommendation. The following settings were used: camera level = 14, capture script = 3 x 30 s, and detection threshold = 5. Transmission electron microscopy (TEM)

A Tecnai EM microscope was used to capture TEM pictures (FEI). 10 L of the material was dropped on the TEM grid while it was being held by forceps, and it was incubated for 5 minutes. Using a Whatman paper, extra sample on the grid was blotted. Before each usage, 2% uranyl acetate was prepared in distilled water and filtered using a 0.1- micron syringe filter. The TEM grid was floated on a droplet of 20 pL of 2% uranyl acetate for seven minutes. The sample was blotted using Whatman paper to remove extra uranyl acetate, and it was then allowed to air dry at room temperature.

Super-resolution microscopy ONI

Oni EV staining and loading was performed according to manufacture (ONI EV Profiler kit). Pre-conjugated CD63 from the manufacturer was used and MMP14 was conjugated in house (ThermoFisher Scientific # PA5-13183, R&D Systems # MAB918 clone 5H2)[38], [39], For membrane staining, ExoBrite EV membrane staining kit was used (Biotium 30112-T). Merged images were analyzed with the CODI software. First, drift correction was performed at minimum entropy (DME). Next background signal (from channel 5) was removed by setting a background filter to remove any photons >5-10,000 photons/ pm^A2 depending on the level of background signal within the image. For localization-based super-resolution microscopy, the photo-switching kinetic for temporal grouping was set to a maximum distance of 20nm and a frame gap of 2. Temporal grouping reduces the noise by combining duplicated localizations from a single blinking event. Next, HDBSCAN was used to perform clustering to find biologically relevant structures and quantify statistical and morphological properties. The HDBSCAN was set to 5 as the minimum density of the clusters and 5 and the minimum number of localizations per cluster with channel 1 and channel 2 merged for cluster analysis. Finally, the cluster characteristics were output via the counting tool in CODI. For a cluster to be scored as positive, it had to have at least 5 binned signals per channel with a maximum cluster radius of 210 nm. All parameters not specifically mentioned were set as the default setting in the CODI software. Western Blot EV markers

Pooled PDAC EV samples and pooled healthy EV samples were exposed to reducing conditions and loaded into a NuPAGE 4-12% Bis-Tris Protein Gel (Invitrogen) for 22 minutes at 200V. The proteins were then transferred from the gel to PVDF membrane by Mini Blot Module (Invitrogen) for 1 hour at 20V. Membranes were blocked in 5% fat-free milk in TBST and then incubated with primary antibodies for EV markers, rabbit monoclonal to CD9 (Abeam, ab236630, clone EPR23105-121 knockout validated) and rabbit monoclonal to CD-63 (R&D Systems, MAB50482 clone 2585J) [40], or MMP primary antibodies, mouse monoclonal to MMP2 (Abeam, ab86607, clone 6E3F8) [41] and rabbit polyclonal to MMP- 14 (Invitrogen, PA5-13183) in 5% fat-free milk overnight at 4°C. Membranes were washed in TBST and incubated with goat anti-rabbit IgG HRP (Cell Signaling, 7074) or anti-mouse IgG HRP (Invitrogen, Al 6072) for 1 hour at room temperature. PVDF membranes were then exposed to chemiluminescent substrates (ThermoFisher Scientific) and exposed for chemiluminescence by iBright imaging system (Invitrogen).

ELISA

ELISA kits were purchased from MyBioSource for MMP-2, MMP-9, MMP-13, MMP-14, TIMP-2 (MBS260339, MBS175780, MBS160467, MBS2516058, MBS355424). The protocol was performed as described by the manufacturer. Assays were performed nonblinded. and from RayBiotech for CAI 9-9 (ELH-CA19-9). The protocol was performed as described by the manufacturer. Assays were performed non-blinded.

Antibody activity assay

Antibody activity assays were purchased from Anaspec and the protocol was performed as described by the manufacturer and the signal was measured after a 1-hour incubation (MMP 13 AS-72019, MMP2 AS-72224, MMP9 AS-72017).

Nanosensor assay or PAC»MANN-1 assay

Absolute Mag™ NHS-Activated or Carboxyl Magnetic Particles Conjugation Kit, 50 nm (CD Bioparticles, WHM-X019 or WHM-K024), was used for conjugation, and the protocol was performed as described by the manufacturer with small modifications. Briefly, 50 mg of powder was weighed and added into a 1.5 mL low-bind tube and resuspended with 0.4 mL of activation buffer (25mM MES, 0.01% Tween 20, pH 6.0) and vortexing for 15 minutes. Next, a NAP-5 desalting column was used as described by the manufacturer using the activation buffer and after all the beads are inside the column it was moved to a 2 mL low-binding tube and 1 mL of activation buffer added to the column and the beads eluted. For the Carboxyl particles, 0.5 mg of particles were placed in 250 pL of activation buffer. 0.125mg of l-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) (Pierce, A35391) and 0.125mg of N-hydroxy sulfosuccinimide (sulfo-NHS) (Pierce, A39269) were added and incubated for 15 min. 20mM of 2-Mercaptoethanol (Gibco, 21985-023) was used to inactivate the EDC. The following is for both magnetic particles. Next, 125 nmol of peptide (Nterm-H {PEG4}GePEPFAGaGC(Cy5)K(Biotin)-OH-Cterm) [lower case letters are D-amino acids] (Pepscan) in phosphate buffered saline pH 7.4 were added to a volume of 250 pL (final) of activation buffer and added to the eluted beads. The solution was vortexed, incubated, and protected from light for 2 hours. After incubation 100 pL of quenching buffer (lOOmM Tris-HCl, PH 7.4) was added to the solution and incubated for 30 minutes at room temperature protected from light. After the incubation, 6 washes of the beads were performed using 400 pL storage buffer (lOmM sodium phosphate, 15mM NaCl, 0.01% Tween 20, 0.05% NaNs, pH 7.2) and waiting progressively shorter time for each wash for the beads to go into the magnet (30 minutes, 30 minutes, 15 minutes, 15 minutes, 5 minutes, and 5 minutes). After washing, beads were resuspended in 0.5 mL of storage buffer and stored at 4°C until use.

Just before running the assay, 10 pL from the stock were transferred to a 0.2 mL PCR tube and diluted to 100 pL with storage buffer. Next, 5 washes of the diluted stock were performed using the magnet and 100 pL of storage buffer per wash. After washing the particles were resuspended into 100 pL of storage buffer and the reaction was set up. The reaction was set up in a 0.2 mL PCR tube or 96 well plate low bind. For each reaction, we added 8 pL of beads from the final washed stock, 4 pL of CaCh, and 8 pL of serum. We pipetted up and down until mixed but with no vortex and placed in a rotator for 45 minutes at room temperature. After incubation, we quickly centrifuged the tubes and added them to the magnet until there was a visible pellet at the back of the tube for around 3 minutes. Finally, we took 15 pL of the supernatant and transferred it to a 384-well plate (Coming, 3544) for fluorescence read-out. We used the TECAN Spark plate reader to analyze Cy5 fluorescence intensity in each condition. Statistical Analyses

Measurements were performed on distinct samples. Statistical analyses were performed using GraphPad 9.0 if not indicated otherwise. Normality was assessed using D’Agostino & Pearson test using alpha = 0.05. All statistical analyses were two-sided. Significance nomenclature was used as follows: * = pval < 0.05, ** = pval < 0.01, *** = pval < 0.001, **** = pval < 0.0001. All boxplots above are Min to Max if not indicated otherwise. Study Design

Sample size was based upon obtaining >100 samples divided almost evenly between stages. Power analysis was not used for calculating this number. This was a retrospective study, thus there was no stopping of collection, endpoints or inclusion or exclusion criteria. All data were excluded in the analysis. Each experiment was performed a single time from a blood draw at initial visit and diagnosis of cancer. The research subjects were cancer patients, healthy patients, pancreatitis patients, and PDAC precursor lesion patients. This was a controlled laboratory experiment where we measured protease activity, CAI 9-9 and protein levels. For the screening set, all samples obtained were used. For the training and blinded validation, 2/3 of samples were assigned to the training set and 1/3 were assigned to the validation. Randomization was performed in excel by assigning a random number to each data point, then sorted and divided 2: 1. For the training and validation cohorts, all the data was collected at the same time, then the samples were randomized. For the training and blinded validation set, all samples were mixed between cancer and healthy and given a different ID number for blinding during the experiment. For the blinded validation, the samples were assigned a random number and the person analyzing the data and performing experiments was blinded to the sample ID until all analyses were finished.

References

[1] J. J. Underwood et al., “Liquid biopsy for cancer: Review and implications for the radiologist,” Radiology, vol. 294, no. 1, pp. 5-17, Nov. 2020, doi:

10.1148/RADIOL.2019182584/ASSET/IMAGES/LARGE/RADIOL.2019182584. FIG3.JPE G.

[2] J. Loud and J. Murphy, “Cancer screening and early detection in the 21 st century HHS Public Access,” Semin Oncol Nurs, vol. 33, no. 2, pp. 121-128, 2017, doi: 10.1016/j.soncn.2017.02.002.

[3] U. Menon et al., “Ovarian cancer population screening and mortality after long-term follow-up in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): a randomised controlled trial,” Lancet (London, England), vol. 397, no. 10290, pp. 2182-2193, Jun. 2021, doi: 10.1016/S0140-6736(21)00731-5.

[4] D. Ilic, M. M. Neuberger, M. Djulbegovic, and P. Dahm, “Screening for prostate cancer,” Cochrane database Syst. Rev., vol. 2013, no. 1, Jan. 2013, doi:

10.1002/14651858. CD004720.PUB3.

[5] E. F. Patz et al., “Overdiagnosis in low-dose computed tomography screening for lung cancer,” JAMA Intern. Med., vol. 174, no. 2, pp. 269-274, Feb. 2014, doi:

10.1001 /JAMAINTERNMED.2013.12738.

[6] D. Crosby et al., “Early detection of cancer,” Science (80- )., vol. 375, no. 6586, Mar.

2022, doi: 10.1126/SCIENCE. AAY9040/ASSET/0A3087E0-148F-45C4-B9D9- 5558E774A080/ASSETS/IMAGES/LARGE/SCIENCE. AAY9040-F6.JPG.

[7] K. C. Park, M. Dharmasivam, and D. R. Richardson, “The Role of Extracellular Proteases in Tumor Progression and the Development of Innovative Metal Ion Chelators That Inhibit Their Activity,” Int. J. Mol. Sci., vol. 21, no. 18, pp. 1-22, Sep. 2020, doi: 10.3390/IJMS21186805.

[8] M. Vizovisek, D. Ristanovic, S. Menghini, M. G. Christiansen, and S. Schuerle, “The Tumor Proteolytic Landscape: A Challenging Frontier in Cancer Diagnosis and Therapy,” Int. J. Mol. Sci., vol. 22, no. 5, pp. 1-30, Mar. 2021, doi: 10.3390/IJMS22052514.

[9] H. W. Verspaget, “Proteases as prognostic markers in cancer: Proteolytic enzymes and their inhibitors influence the spread of cancer,” BMJ Br. Med. J., vol. 316, no. 7134, p. 790, Mar. 1998, doi: 10.1136/BMJ.316.7134.790.

[10] L. Wu and X. Qu, “Cancer biomarker detection: recent achievements and challenges.,” Chem. Soc. Rev., vol. 44, no. 10, pp. 2963-2997, Mar. 2015, doi: 10.1039/C4CS00370E.

[11] A. E. Modestino et al., “Elevated Resting and Postprandial Digestive Proteolytic Activity in Peripheral Blood of Individuals With Type-2 Diabetes Mellitus, With Uncontrolled Cleavage of Insulin Receptors,” J. Am. Coll. Nutr., vol. 38, no. 6, pp. 485-492, Aug. 2019, doi: 10.1080/07315724.2018.1545611.

[12] M. Vizovisek, D. Ristanovic, S. Menghini, M. G. Christiansen, and S. Schuerle, “The Tumor Proteolytic Landscape: A Challenging Frontier in Cancer Diagnosis and Therapy,” Int. J. Mol. Sci., vol. 22, no. 5, pp. 1-30, Mar. 2021, doi: 10.3390/IJMS22052514.

[13] M. Nakamura, Y. Miyasaka, Y. Sadakari, K. Date, and T. Ohtsuka, “Comparison of guidelines for intraductal papillary mucinous neoplasm: What is the next step beyond the current guidelines?,” Ann. Gastroenterol. Surg., vol. 1, no. 2, pp. 90-98, Jan. 2017, doi: 10.1002/AGS3.12012.

[14] S. Hasan, R. Jacob, U. Manne, and R. Paluri, “Advances in pancreatic cancer biomarkers,” Oncol. Rev., vol. 13, no. 1, pp. 69-76, Jan. 2019, doi: 10.4081/ONCOL.2019.410.

[15] S. L. Ivry el al., “Global protease activity profiling provides differential diagnosis of pancreatic cysts,” Clin. Cancer Res., vol. 23, no. 16, pp. 4865-4874, Aug. 2017, doi:

10.1158/1078-0432. CCR-16-2987/274481/AM/GLOBAL-PROTEASE-ACTIVITY- PROFILING-PRO VIDES .

[16] L. L. Isaacs, “Pancreatic Proteolytic Enzymes and Cancer: New Support for an Old Theory,” Integr. Cancer Ther., vol. 21, 2022, doi: 10.1177/15347354221096077.

[17] K. Soreide, M. Roalso, and J. R. Aunan, “Is There a Trojan Horse to Aggressive Pancreatic Cancer Biology? A Review of the Trypsin-PAR2 Axis to Proliferation, Early Invasion, and Metastasis,” https://home.liebertpub.com/pancan, vol. 6, no. 1, pp. 12-20, Feb. 2020, doi: 10.1089/PANCAN.2019.0014.

[18] K. Felix and M. M. Gaida, “Neutrophil-Derived Proteases in the Microenvironment of Pancreatic Cancer -Active Players in Tumor Progression,” Int. J. Biol. Sci., vol. 12, no. 3, p. 302, Jan. 2016, doi: 10.7150/IJBS.14996.

[19] R. B. Lefkowitz, J. Y. Marciniak, C. M. Hu, G. W. Schmid-Schonbein, and M. J. Heller, “An electrophoretic method for the detection of chymotrypsin and trypsin activity directly in whole blood,” Electrophoresis, vol. 31, no. 2, pp. 403-410, Jan. 2010, doi: 10.1002/ELPS.200900424.

[20] R. B. Lefkowitz, G. W. Schmid-Schonbein, and M. J. Heller, “Whole blood assay for elastase, chymotrypsin, matrix metalloproteinase-2, and matrix metalloproteinase-9 activity,” Anal. Chem., vol. 82, no. 19, pp. 8251-8258, Oct. 2010, doi:

10.1021 /AC 101462C/AS SET/IMAGES/L ARGE/AC-2010-01462C_0005.JPEG.

[21] N. D. Rawlings, A. J. Barrett, P. D. Thomas, X. Huang, A. Bateman, and R. D. Finn, “The MEROPS database of proteolytic enzymes, their substrates and inhibitors in 2017 and a comparison with peptidases in the PANTHER database,” Nucleic Acids Res., vol. 46, no. DI, pp. D624-D632, Jan. 2018, doi: 10.1093/NAR/GKX1134.

[22] S. H. Choi, S. H. Park, K. W. Kim, J. Y. Lee, and S. S. Lee, “Progression of Unresected Intraductal Papillary Mucinous Neoplasms of the Pancreas to Cancer: A Systematic Review and Meta-analysis,” Clin. Gastroenterol. Hepatol., vol. 15, no. 10, pp. 1509-1520. e4, Oct. 2017, doi: 10.1016/J.CGH.2017.03.020. [23] R. H. Hruban, A. Maitra, and M. Goggins, “Update on Pancreatic Intraepithelial Neoplasia,” Int. J. Clin. Exp. Pathol., vol. 1, no. 4, p. 306, Jan. 2008, Accessed: Mar. 28, 2023. [Online], Available: /pmc/articles/PMC2480542/

[24] S. J. Murphy et al, “Genetic alterations associated with progression from pancreatic intraepithelial neoplasia to invasive pancreatic tumor,” Gastroenterology, vol. 145, no. 5, 2013, doi: 10.1053/J.GASTR0.2013.07.049.

[25] H. Tang et al., “Exosomal MMP2 derived from mature osteoblasts promotes angiogenesis of endothelial cells via VEGF/Erkl/2 signaling pathway,” Exp. Cell Res., vol. 383, no. 2, Oct. 2019, doi: 10.1016/J.YEXCR.2019.111541.

[26] A. Das et al., “Inhibition of extracellular vesicle-associated MMP2 abrogates intercellular hepatic miR-122 transfer to liver macrophages and curtails inflammation,” iScience, vol. 24, no. 12, Dec. 2021, doi: 10.1016/J.ISCI.2021.103428.

[27] K. Y. Han, J. Dugas-Ford, M. Seiki, J. H. Chang, and D. T. Azar, “Evidence for the involvement of MMP14 in MMP2 processing and recruitment in exosomes of corneal fibroblasts,” Investig. Ophthalmol. Vis. Sci., vol. 56, no. 9, pp. 5323-5329, 2015, doi:

10.1167/IOVS.14-14417.

[28] S. Runz et al., “Malignant ascites-derived exosomes of ovarian carcinoma patients contain CD24 and EpCAM,” Gynecol. Oncol., vol. 107, no. 3, pp. 563-571, Dec. 2007, doi: 10.1016/J.YGYN0.2007.08.064.

[29] H. Sato, T. Kinoshita, T. Takino, K. Nakayama, and M. Seiki, “Activation of a recombinant membrane type 1 -matrix metalloproteinase (MT1-MMP) by furin and its interaction with tissue inhibitor of metalloproteinases (TIMP)-2,” FEBS Lett., vol. 393, no. 1, pp. 101-104, Sep. 1996, doi: 10.1016/0014-5793(96)00861-7.

[30] K. Lehti, J. Lohi, H. Valtanen, and J. Keski-Oja, “Proteolytic processing of membrane-type- 1 matrix metalloproteinase is associated with gelatinase A activation at the cell surface.,” Biochem. J., vol. 334, no. Pt 2, p. 345, Sep. 1998, doi: 10.1042/BJ3340345.

[31] M. Nawaz et al, “Extracellular Vesicles and Matrix Remodeling Enzymes: The Emerging Roles in Extracellular Matrix Remodeling, Progression of Diseases and Tissue Repair,” Cells, vol. 7, no. 10, Oct. 2018, doi: 10.3390/CELLS7100167.

[32] F. Liu et al, “The Exosome Total Isolation Chip,” ACS Nano, vol. 11, no. 11, pp. 10712-10723, Nov. 2017, doi:

10.1021/ACSNANO.7B04878/ASSET/IMAGES/LARGE/NN-2017-04878E_0006. JPEG. [33] K. I. Shimokawa et aL, “Matrix metalloproteinase (MMP)-2 and MMP-9 activities in human seminal plasma,” Mol. Hum. Reprod., vol. 8, no. 1, pp. 32-36, Jan. 2002, doi: 10.1093/MOLEHR/8.1.32.

[34] J. Christensen and V. P. Shastri, “Matrix-metalloproteinase-9 is cleaved and activated by Cathepsin K,” BMC Res. Notes, vol. 8, no. 1, p. 322, Jul. 2015, doi: 10.1186/S 13104-015- 1284-8.

[35] M. A. Lafleur, M. D. Hollenberg, S. J. Atkinson, V. Knauper, G. Murphy, and D. R. Edwards, “Activation of pro-(matrix metalloproteinase-2) (pro-MMP-2) by thrombin is membrane-type-MMP-dependent in human umbilical vein endothelial cells and generates a distinct 63 kDa active species,” Biochem. J., vol. 357, no. Pt 1, pp. 107-115, Jul. 2001, doi: 10.1042/0264-6021 :3570107.

[36] J. E. Kim, K. T. Lee, J. K. Lee, S. W. Paik, J. C. Rhee, and K. W. Choi, “Clinical usefulness of carbohydrate antigen 19-9 as a screening test for pancreatic cancer in an asymptomatic population,” J. Gastroenterol. Hepatol., vol. 19, no. 2, pp. 182-186, 2004, doi: 10.1111/J.1440-1746.2004.03219.X.

[37] D. Qu et al., “Doublecortin-Like Kinase 1 Is Elevated Serologically in Pancreatic Ductal Adenocarcinoma and Widely Expressed on Circulating Tumor Cells,” PLoS One, vol.

10, no. 2, p. eOl 18933, Feb. 2015, doi: 10.1371/JOURNAL.PONE.Ol 18933.

[38] A. Hryniewicz-Jankowska, J. Wierzbicki, R. Tabola, K. Stach, K. Sossey-Alaoui, and K. Augoff, “The Effect of Neddylation Inhibition on Inflammation-Induced MMP9 Gene Expression in Esophageal Squamous Cell Carcinoma,” Int. J. Mol. Set., vol. 22, no. 4, pp. 1- 17, Feb. 2021, doi: 10.3390/IJMS22041716.

[39] R. Kaimal et al., “Selective blockade of matrix metalloprotease- 14 with a monoclonal antibody abrogates invasion, angiogenesis, and tumor growth in ovarian cancer,” Cancer Res., vol. 73, no. 8, pp. 2457-2467, Apr. 2013, doi: 10.1158/0008-5472.CAN-12-1426.

[40] S. Shan et al., “Extracellular vesicle-derived long non-coding RNA as circulating biomarkers for endometriosis,” Reprod. Biomed. Online, vol. 44, no. 5, pp. 923-933, May 2022, doi: 10.1016/J.RBM0.2021.11.019.

[41] T. Kanamoto et al., “Integrin ct.2|31 plays an important role in the interaction between human articular cartilage-derived chondrocytes and atelocollagen gel,” Sci. Rep., vol. 11, no. 1, Dec. 2021, doi: 10.1038/S41598-021-81378-2. Table 6. Informal Sequence Listing

Each embodiment disclosed herein can comprise, consist essentially of or consist of its particular stated element, step, ingredient or component. Thus, the terms “include” or “including” should be interpreted to recite: “comprise, consist of, or consist essentially of.” The transition term “comprise” or “comprises” means has, but is not limited to, and allows for the inclusion of unspecified elements, steps, ingredients, or components, even in major amounts. The transitional phrase “consisting of’ excludes any element, step, ingredient or component not specified. The transition phrase “consisting essentially of’ limits the scope of the embodiment to the specified elements, steps, ingredients or components and to those that do not materially affect the embodiment.

Unless otherwise indicated, all numbers expressing quantities of ingredients, agent concentrations, experimental conditions, and so forth used in the specification and claims are to be understood as being modified in all instances by the term “about.” Accordingly, unless indicated to the contrary, the numerical parameters set forth in the specification and attached claims are approximations that may vary depending upon the desired properties sought to be obtained by the present invention. At the very least, and not as an attempt to limit the application of the doctrine of equivalents to the scope of the claims, each numerical parameter should at least be construed in light of the number of reported significant digits and by applying ordinary rounding techniques. When further clarity is required, the term “about” has the meaning reasonably ascribed to it by a person skilled in the art when used in conjunction with a stated numerical value or range, i.e. denoting somewhat more or somewhat less than the stated value or range, to within a range of ±20% of the stated value; ±19% of the stated value; ±18% of the stated value; ±17% of the stated value; ±16% of the stated value; ±15% of the stated value; ±14% of the stated value; ±13% of the stated value; ±12% of the stated value; ±11% of the stated value; ±10% of the stated value; ±9% of the stated value; ±8% of the stated value; ±7% of the stated value; ±6% of the stated value; ±5% of the stated value; ±4% of the stated value; ±3% of the stated value; ±2% of the stated value; or ±1% of the stated value.

Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the invention are approximations, the numerical values set forth in the specific examples are reported as precisely as possible. Any numerical value, however, inherently contains certain errors necessarily resulting from the standard deviation found in their respective testing measurements.

The terms “a,” “an,” “the” and similar referents used in the context of describing the invention (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated herein or clearly contradicted by context. Recitation of ranges of values herein is merely intended to serve as a shorthand method of referring individually to each separate value falling within the range. Unless otherwise indicated herein, each individual value is incorporated into the specification as if it were individually recited herein. All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., “such as”) provided herein is intended merely to better illuminate the invention and does not pose a limitation on the scope of the invention otherwise claimed. No language in the specification should be construed as indicating any non-claimed element essential to the practice of the invention. Groupings of alternative elements or embodiments of the invention disclosed herein are not to be construed as limitations. Each group member may be referred to and claimed individually or in any combination with other members of the group or other elements found herein. It is anticipated that one or more members of a group may be included in, or deleted from, a group for reasons of convenience and/or patentability. When any such inclusion or deletion occurs, the specification is deemed to contain the group as modified thus fulfilling the written description of all Markush groups used in the appended claims.

Certain embodiments of this invention are described herein, including the best mode known to the inventors for carrying out the invention. Of course, variations on these described embodiments will become apparent to those of ordinary skill in the art upon reading the foregoing description. The inventor expects skilled artisans to employ such variations as appropriate, and the inventors intend for the invention to be practiced otherwise than specifically described herein. Accordingly, this invention includes all modifications and equivalents of the subject matter recited in the claims appended hereto as permitted by applicable law. Moreover, any combination of the above-described elements in all possible variations thereof is encompassed by the invention unless otherwise indicated herein or otherwise clearly contradicted by context.

Furthermore, numerous references have been made to patents, printed publications, journal articles and other written text throughout this specification (referenced materials herein). Each of the referenced materials are individually incorporated herein by reference in their entirety for their referenced teaching.

It is to be understood that the embodiments of the invention disclosed herein are illustrative of the principles of the present invention. Other modifications that may be employed are within the scope of the invention. Thus, by way of example, but not of limitation, alternative configurations of the present invention may be utilized in accordance with the teachings herein. Accordingly, the present invention is not limited to that precisely as shown and described.

The particulars shown herein are by way of example and for purposes of illustrative discussion of the preferred embodiments of the present invention only and are presented in the cause of providing what is believed to be the most useful and readily understood description of the principles and conceptual aspects of various embodiments of the invention. In this regard, no attempt is made to show structural details of the invention in more detail than is necessary for the fundamental understanding of the invention, the description taken with the drawings and/or examples making apparent to those skilled in the art how the several forms of the invention may be embodied in practice.

Definitions and explanations used in the present disclosure are meant and intended to be controlling in any future construction unless clearly and unambiguously modified in the example(s) or when application of the meaning renders any construction meaningless or essentially meaningless.

Claims

WHAT IS CLAIMED IS:

1. A method of detecting cancer in a subject, the method comprising:

(a) providing a biological sample comprising one or more proteases from the subject;

(b) contacting the biological sample with a plurality of beads each bead comprising a plurality of probes to form a composition comprising the biological sample and beads, wherein each probe comprises (i) at least 6 L-amino acids and (ii) a detectable label, and wherein the probe is capable of being cleaved by the one or more proteases;

(c) magnetically separating the beads from the composition; and

(d) detecting the detectable label in the composition separated from the beads, wherein an increase in a signal from the detectable label as compared to a control indicates the subject has cancer.

2. The method of claim 1, wherein the probe further comprises (iii) at least one D-amino acid.

3. The method of claim 1, wherein the probe comprises at least 6 to 50 amino acids.

4. The method of claim 1, wherein the probe comprises at least 6 to 15 amino acids.

5. The method of claim 1, wherein the probe comprises at least 6 L-amino acids and at least two D-amino acids.

6. The method of claim 1, wherein the probe has an amino acid sequence comprising SEQ ID NO: 1 to 12 or any combination thereof.

7. The method of claim 1, wherein the probe has an amino acid sequence selected from the group consisting of SEQ ID NO: 13 to 29.

8. The method of claim 1, wherein the biological sample is a serum or plasma sample.

9. The method of claim 1, wherein the biological sample is whole blood.

10. The method of claim 1, wherein the biological sample is not manipulated or processed prior to the contacting.

11. The method of claim 1, wherein the detectable label comprises a fluorescent label, a radioactive label, or an affinity label.

12. The method of claim 1, wherein the composition comprises 0.5 to 150 pl of the biological sample.

13. The method of claim 12, wherein the composition comprises 100 to 150 pl of the biological sample.

14. The method of claim 12, wherein the composition comprises 0.5 to 50 pl of the biological sample.

15. The method of claim 1, wherein the beads further comprise a spacer located between the probes and the beads.

16. The method of claim 15, wherein the spacer comprises polyethylene glycol (PEG), a peptide or a nucleic acid molecule.

17. The method of claim 1, further comprising performing an additional assay for detecting cancer in the subject.

18. The method of claim 17, wherein the assay is an immunoassay.

19. The method of claim 17, wherein the assay detects the presence of a polypeptide wherein an increase in the level of the polypeptide as compared to a control indicates the subject has cancer.

20. The method of claim 17, wherein the polypeptide is CA-19-9.

21. The method of claim 1, wherein the cancer is pancreatic cancer, liver cancer, head and neck, esophagus, bile, small intestine, stomach or colon cancer.

22. A composition comprising a probe having an amino acid sequence comprising SEQ ID NOs: 1 to 12 or any combination thereof.

23. A kit comprising the composition of claim 22 and instructions for use.

24. The kit of claim 23, further comprising a solid support.

25. The kit of claim 24, wherein the solid support is a bead.

26. The kit of claim 25, wherein the bead is magnetic.

27. The kit of claim 23, further comprising one or more buffers.

28. The kit of claim 23, further comprising water.

29. The kit of claim 27, wherein the one or more buffers are lyophilized.

30. A method of identifying a protease in a biological sample, the method comprising: providing the biological sample; contacting the biological sample with a plurality of beads each bead comprising a plurality of probes to form a composition comprising the biological sample and beads, wherein each probe comprises (i) at least 6 L-amino acids and (ii) a detectable label and wherein the probe is cleaved by the protease thereby generating cleavage products comprising the detectable label;

(c) separating the cleavage products from the beads; and

(d) detecting the detectable label of the cleavage products thereby identifying the protease in the biological sample.

31. The method of claim 30, wherein the probe further comprises (iii) at least one D- amino acid.

32. The method of claim 30, wherein the probe comprises 6 to 50 amino acids.

33. The method of claim 30, wherein the probe comprises 6 to 15 amino acids.

34. The method of claim 30, wherein the probe comprises at least 6 L-amino acids and at least two D-amino acids.

35. The method of claim 30, wherein the probe has an amino acid sequence comprising SEQ ID NO: 1 to 12 or any combination thereof.

36. The method of claim 30, wherein the probe has an amino acid sequence selected from the group consisting of SEQ ID NO: 13 to 29.

37. The method of claim 30, wherein the biological sample is a serum or plasma sample.

38. The method of claim 30, wherein the biological sample is whole blood.

39. The method of claim 30, wherein the biological sample is not manipulated or processed prior to the contacting.

40. The method of claim 30, wherein the detectable label comprises a fluorescent label, a radioactive label, or an affinity label.

41. The method of claim 30, wherein the composition comprises 0.5 to 150 pl of the biological sample.

42. The method of claim 41, wherein the composition comprises 100 to 150 pl of the biological sample.

43. The method of claim 41, wherein the composition comprises 0.5 to 50 pl of the biological sample.

44. The method of claim 30, wherein the beads further comprise a spacer located between the probes and the beads.

45. The method of claim 44, wherein the spacer comprises polyethylene glycol (PEG), a peptide or a nucleic acid molecule.

46. The method of claim 30, wherein the protease is selected from the group consisting of a matrix metalloproteinase (MMP), cathepsin and a caspase.

47. The method of claim 46, wherein the MMP is MMP 1, 2, 3, 7, 8, 9, 12, 13, 14, 15, 16, 17, 19, 20, 21, 23, 24, 25, 26, 27, or 28.

48. The method of claim 46, wherein the cathepsin is cathepsin B, C, D, F, G, H, K, L or S.

49. The method of claim 46, wherein the caspase is caspase 1, 2, 3, 4, 5, 6 7, 8, 9, 10, 12, or 14.

50. The method of claim 30, wherein the biological sample comprises extracellular vesicles (EVs).