CN114720615A - Pancreatic cancer early diagnosis marker and tissue screening method and application thereof - Google Patents

Abstract

The invention discloses a pancreatic cancer early diagnosis marker, a tissue screening method and application thereof, wherein the tissue screening method comprises the following steps: step 1, respectively taking clinical tissues PDAC of a pancreatic cancer patient in stage I-IV who has not received treatment and normal tissues PNT matched with the clinical tissues PDAC; step 2, analyzing and identifying primary differential metabolites between pancreatic cancer tissues and normal tissues by a chromatography-mass spectrometry combined metabonomics analysis method; step 3, visually showing the spatial distribution of the preliminary differential metabolites through mass spectrometry imaging, and finding metabolites which are abnormally distributed in pancreatic cancer tissue regions; and 4, performing auxiliary screening on the abnormally distributed metabolites: constructing a pancreatic cancer diagnosis Lasso model based on metabonomics based on tissue biopsy, screening early diagnosis markers in pancreatic cancer tissues from the abnormally distributed metabolites, wherein the early diagnosis markers obtained by the method can be used for early screening and diagnosis of pancreatic cancer.

Description

Pancreatic cancer early diagnosis marker and tissue screening method and application thereof

Technical Field

The invention relates to the field of biomedicine, in particular to a pancreatic cancer early diagnosis marker and a tissue screening method and application thereof.

Background

Pancreatic Ductal Adenocarcinoma (PDAC), which accounts for over 90% of Pancreatic malignancies, is the most aggressive and fatal cancer with a 5-year survival rate of less than 10%. While mortality from PDACs has risen over the past decade due to increased morbidity, delayed diagnosis and lack of significant progress in treatment, median survival time is less than 7 months. Currently, imaging techniques in imaging, such as MRI, ERCP, EUS, CT, etc., are routinely used in clinical practice for screening, diagnosis and staging of patients with PDAC. However, due to the lack of early symptoms and diagnostic markers, more than 80% of PDAC patients have advanced or metastatic disease at the time of diagnosis, missing the chance of radical surgery.

The use of the prior art presents difficulties in early diagnosis and discovery due to the lack of early symptoms and diagnostic markers in PDACs. However, despite considerable efforts to target potential PDAC-related metabolites in the blood of patients, such molecules have rarely been brought on the front-line of clinical practice, at least in part due to the lack of low cost and high throughput analytical tools.

Disclosure of Invention

The invention aims to overcome the defects in the prior art, provide a CPSI-MS and DESI-MSI-based metabonomics, and provide a powerful method for rapid molecular diagnosis and screening aiming at a characterization tool summarized in PDAC metabolism panorama so as to assist clinical practice.

In order to achieve the above object, the present invention provides a tissue screening method for an early diagnosis marker of pancreatic cancer, comprising the steps of: step 1, respectively taking a clinical tissue PDAC of a pancreatic cancer patient and a normal tissue PNT matched with the clinical tissue PDAC; step 2, analyzing and identifying primary differential metabolites between pancreatic cancer tissues and normal tissues by a chromatography-mass spectrometry combined metabonomics analysis method; step 3, visually showing the spatial distribution of the preliminary differential metabolites through mass spectrometry imaging, and finding metabolites which are abnormally distributed in pancreatic cancer tissue regions; and 4, performing auxiliary screening on the abnormally distributed metabolites: constructing a pancreatic cancer diagnosis Lasso model based on metabonomics based on tissue biopsy, and screening early diagnosis markers in pancreatic cancer tissues from the abnormally distributed metabolites.

Preferably, in step 2, the chromatography-mass spectrometry metabonomics is conductive polymer spray ionization mass spectrometry CPSI-MS metabonomics.

Preferably, in step 2, the analysis to identify preliminary differential metabolites between pancreatic cancer tissue and normal tissue comprises the steps of:

step 2.1, reserving sample ions by taking VIP larger than 1.0 as a standard;

Step 2.2, performing T test on the sample ions obtained in the step 2.1, and further reserving the sample ions by taking FDR <0.05 as a standard;

step 2.3, performing fold difference analysis on the sample ions obtained in the step 2.2, and further retaining the sample ions by taking the fold difference greater than 2.0 or less than 0.5 as a standard;

and 2.4, comparing the sample ions obtained in the step 2.3 with a metabolome database to obtain 76 primary differential metabolites in the tissues.

Preferably, in step 3, the mass spectrometry imaging technique is desorption electrospray ionization mass spectrometry imaging DESI-MSI, and the spatial distribution map of each of the preliminary differential metabolites is visualized for in situ verification.

Preferably, in step 4, the auxiliary screening comprises the following steps:

step 4.1, taking the n-150 tissue sample points as a training set, and taking the n-90 tissue sample points as a test set, wherein the n-90 tissue sample points are used for evaluating the performance of a pre-trained Lasso model;

step 4.2, the 76 kinds of preliminary differential metabolites are included into initial input variables of a training set of the Lasso classifier, only variables contributing to classification are endowed with non-zero weight, and when only 22 variables in the 76 variables are reserved, the performance of the Lasso classifier achieves the highest accuracy on a test set;

And 4.3, referring to a receiver operating characteristic ROC curve to evaluate the optimal diagnostic performance of the Lasso model, and taking the highest true positive rate and the lowest false positive rate on the ROC curve as cut-off points to obtain the optimal diagnostic performance, which indicates that the CPSI-MS data acquisition combined with the Lasso model can be used as a potential in-vitro diagnostic strategy complementary with tissue biopsy.

The present invention also provides an early diagnosis marker for pancreatic cancer obtained by the method for tissue screening for an early diagnosis marker for pancreatic cancer according to any one of the above methods, comprising one or a combination of two or more of diglyceride, 1-N-butylamine, cytidine, histamine, phosphorylcholine, hydroxyllinolenic acid, linoleic acid, 3-phosphoglycerol, glutamic acid, acrylylcarnitine, glutamine, tyrosine, 3-hydroxypalmitic acid, lysophosphatidylcholine (P-16:0), butyrylcarnitine, pipecolic acid, lysophosphatidylcholine (18:2), acetylspermine, lactic acid, histidine, N-methyl histamine, and N-acetyl histamine.

Preferably, the pancreatic cancer early diagnosis marker can be applied to preparation of a pancreatic cancer early diagnosis reagent or kit.

The invention also provides a reagent for early diagnosis of pancreatic cancer, which can detect the expression level of the early diagnosis marker in a sample.

Preferably, the sample is a tissue sample.

The invention also provides a kit for early diagnosis of pancreatic cancer, which contains the early diagnosis marker for preparing a diagnostic reagent for early differentiation of pancreatic cancer.

The invention has the advantages that:

(1) the invention provides a pancreatic cancer early diagnosis marker, which can achieve the purposes of early screening and diagnosis of pancreatic cancer by detecting the expression level of the early diagnosis marker.

(2) The tissue screening method of the pancreatic cancer early diagnosis marker provided by the invention uses CPSI-MS combined metabonomics analysis, uses a tissue sample for detection, only needs to consume one drop of solvent (<10 mu L) for metabolite extraction and ionization, and is more suitable for high-throughput screening of the sample and rapid evaluation of the metabolic state of certain interested regions in the whole tissue; DESI-MSI incorporates electrospray probes with layer-by-layer scanning with flowing solvents and provides spatially resolved images of each metabolite distribution, more suitable for metabolite-based molecular pathology requiring complex region-dependent metabolic studies or more accurate surgical margin assessment; the tissue screening method of the pancreatic cancer early diagnosis marker provided by the invention combines CPSI-MS data acquisition and Machine Learning (ML) analysis to enable a workflow to be more economical and effective, and a Lasso model constructed by combining the CPSI-MS data acquisition can be used as a potential in-vitro diagnosis strategy complementary with tissue biopsy.

Drawings

Figure 1 is a non-targeted metabolomics result of PDAC tissue of the present invention.

In the figure, (A) CPSI-MSI in situ metabolism spectrogram;

(B) partial least squares discriminant analysis (PLS-DA) classification of PDAC and PNT;

(C) step-by-step statistical analysis of metabolites with significant changes compared to PNT specimens in PDACs is finally preserved;

(D) volcano figures highlight significant changes in metabolite ions in PDAC compared to PNT (FDR < 0.05);

(E) a heatmap of the relative expression levels of 40 representative metabolites in PNTs and PDACs.

FIG. 2 shows the average mass spectra of PNT (A) and PDAC (B) tissues under the full MS scanning mode.

FIG. 3 is a graph of the in situ validation and complementary diagnostic performance of metabolites in tissues of the present invention.

In the figure, (A) a DESI-MSI and machine learning process schematic diagram;

(B) a typical pair of optical images of PDACs with adjacent PNTs and normal control NC tissue;

(C) images of significant changes in representative metabolites of PDACs compared to PNTs;

(D) all 240 tissue score plots given by the Lasso classifier;

(E) a confusion matrix presenting classifications set by the developed Lasso model for testing of PDACs;

(F) ROC (receiver operating characterization) curve for evaluating Lasso model in training

Diagnostic performance on the set and test set;

(G) images of significant changes in the remaining metabolites in PDACs compared to PNTs.

FIG. 4 is a clinical application recommendation scenario of the CPSI-MS and DESI-MSI of the present invention in PDAC screening and diagnosis.

In the figure, (a) metabolic profiling and rapid PDAC diagnosis can be performed by CPSI-MS which consumes only minute amounts of biopsy tissue (less than 1mg) from these highly suspicious patients;

(B) for patients surgically resected with PDAC tissue, more accurate diagnosis and spatial resolution metabolic profiling can be performed by DESI-MSI in conjunction with molecular pathology.

FIG. 5 is a demonstration of the deregulated enzymes, transporters and associated substrates of the present invention.

In the figure, (a) by searching for abundant metabolic pathways in metabolic analysis;

(B) enriched enzymes and transporters searched in KEGG and Reactome;

(C) the grade of affected biological function involved in the enzyme and transporter;

(D) subcellular location and functional schematic of PDZD, PISD, TYMP, and SLC1a 5;

(E) immunohistochemical detection of expression levels of PISD, PDZD11, TYMP, SLC1a5 on tissue chips;

(F) expression levels of PISD, PDZD11, TYMP and SLC1a5 in different tissues (PDAC and PNT);

(G) Correlation of high and low expression of SLC1a5 with progressive TNM staging of PDAC;

(H) survival analysis of SLC1A5 Low expression and high Table PDAC patients.

FIG. 6 shows that PISD-induced PS/PE transformation according to the present invention is a key metabolic process for PDAC progression and prognosis, inhibiting tumor metastasis.

In the figure, (A) a schematic diagram of PISD conversion PS/PE;

(B) taking PS (36:1) and PE (36:1) as typical examples, representing PS and PE images by using DESI-MSI;

(C) intensity ratio map of PE and PS on image pixel basis;

(D) performing a transwell experiment on tumor cells over-expressing PISD;

(E) migration experiments were performed on PISD overexpressing tumor cells. The scale bar is 10 μm. The test was performed in 3 replicates and subjected to a t-test. Values are expressed as mean ± SDs. P <0.05, p < 0.01.

FIG. 7 shows that SLC1A5 of the present invention up-regulates glutamine metabolism in PDACs.

In the figure, (a) SLC1a5 mediates transmembrane exchange of glutamine (Gln) with alanine (Ala) and GLS1 catalyzes the conversion of glutamine with glutamic acid (Glu) schematically, V9302 and CB839 are inhibitors against SLC1a5 and GLS1, respectively;

(B) DESI images of PDAC (depicted with white dashed lines) and PNT regions Ala, Gln, Glu distribution;

(C) relative levels of Gln and Ala in serum of PDAC patients at stages M0 and M1;

(D) Pixel-to-pixel intensity ratios of glutamine to Ala, glutamic acid to glutamine;

(E) differences in expression levels of SLC1a5 mRNA from normal pancreatic cell lines (H6C7) and pancreatic cancer cell lines (SW1990 and PANC 1);

(F) gln and Ala levels of the SW1990 pancreatic cancer cell line following SLC1a5 inhibitor V9302 treatment;

(G) treatment with different inhibitors, proliferation of SW1990 cells;

(H) gemcitabine (Gem) and CB-839 and V-9302 are combined for in vitro cell viability detection;

(I) dose-inhibitory relationship of gemcitabine to different treated SW1990 cells.

FIG. 8 shows that PDZD11 of the present invention upregulates purine and pyrimidine transport in PDACs and promotes tumor metastasis.

In the figure, (A) a schematic representation of the deoxyuridine transport and recovery of PDZD11 and TYMP;

(B) DESI images and corresponding box maps of deoxyuridine and uracil crossing PDAC (depicted with white lines) and PNT regions;

(C) a ratio image of deoxyuridine to uracil for indirect estimation of TYMP expression or activity.

Detailed Description

The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.

Experimental materials and methods

1.1 Experimental cell lines, plasmids and reagents

H6C7, SW1990, PANC1 cells (Manassas, VA, USA), glutaminase (GLS1) inhibitor CB-839, SCL1A5 inhibitor V-9302, CCK8(Cell Counting Kit) Kit (MedChemExpress, Monmouth Junction, NJ, USA), doxycycline (Sigma-Aldrich, St. Louis, MO, USA), 197011 glutamine assay Kit (Colorimetric), Amplite ^TMColorimetric L-alanine assay kit (Abcam, Waltham, MA, USA), 8um transwell chamber (BD Biosciences, New Jersey, USA), and the like.

1.2 Experimental instruments

LTQ Orbitrap Velos mass spectrometer (Thermo Scientific, San Jose, Calif., USA.), commercial 2D DESI system (Prosolia, Indianapolis, USA), microscope (Leica, Wetzlar, Germany), Infinite F500(Tecan) multifunctional microplate reader, incubator (Thermo Fisher, Waltham, MA, USA), and the like.

1.3 Experimental methods

The experimental procedures not specifically mentioned in the present specification are all the procedures routine in the art.

1.3.1 CPSI-MS metabolomics analysis

Methanol-water is used as a solvent, and the volume ratio is 1: 1. For each tissue section, 5. mu.L of the solvent was dropped on the tissue surface. After sufficient extraction of the metabolite for 30 seconds, the droplet was aspirated back and transferred to the tip of the conducting polymer substrate at 13.0mm from the MS inlet. To reduce the effects of tissue heterogeneity, three aliquots of solvent droplets were randomly micro-transferred onto each tissue section and separately examined by CPSI-MS. And after the droplet extraction loading is finished, switching on the 4.5kV direct-current high voltage applied to the conductive polymer tip, and triggering the droplet spray ionization induced by the high electric field. This procedure brings metabolite ions into the LTQ Orbitrap Velos mass spectrometer and records non-targeted metabolic spectra in the range of m/z 50-1000 in positive mode. The MS capillary temperature was 275 ℃ and the S-lens voltage was 55V. The number of micro-scans was set to 1 scan with a maximum injection time of 400 microseconds. The data acquisition period for each case was 15 seconds to collect sufficient metabolomics data.

1.3.2 validation of DESI-MSI on target metabolites in tissue

For tissue imaging, a commercial 2D DESI system was used in a positive and negative ion scan mode. A commercial mass spectrometer was used to provide high voltage of 4.0kV, applied to a spray head to produce an electrospray that desorbs and ionizes components within frozen tissue sections. The spraying solvent is methanol-water (7:3, v/v), the gas pressure of the atomizer is 1.2MPa, and the flow rate is 3.0 muL/min. The impact angle of the showerhead with the substrate was set at 55 °. The height of the atomizer nozzle and the distance from the nozzle to the delivery tube were both set to 4.5 mm. In the above CPSI-MS experiments, the same parameters were used for MS acquisition. For tissue scanning, the raster speed was set at 0.2mm/sec and the width between scans was 0.2 mm. Target ion image reconstruction was performed using massager (Chemmind Technologies co., Ltd, Beijing, China) and self-written MATLAB (Mathworks, nature, MA, USA) scripts.

Example 1 discovery of PDAC-related Metabolic Profile and characterization of specific metabolites Using CPSI-MS

As shown in a of fig. 1, 40 previously untreated patients with stage I-IV PDAC were collected of fresh frozen paired PDAC tissue and adjacent normal tissue (PNT). Frozen sections 15 μm thick were taken for histopathological evaluation. Sections were labeled by two experimenters to determine the area of PDACs and PNTs, respectively, after hematoxylin and eosin (H & E) staining. Subsequently, the frozen sections were subjected to metabolic analysis using CPSI-MS. Three spots (spot diameter about 2.0mm) were randomly selected from each slice for the PDAC/PNT region for CPSI-MS data acquisition. As shown in fig. 2, according to the molecular weight distribution range of metabolite species, the metabolites recovered in both positive and negative modes within the range of m/z 50-1000 by full MS scan include amino acids, carboxylic acids, lyso-phosphoglycerides, nucleotides, nucleosides, acylcarnitines, diacylglycerolipids, fatty acids, glycerophospholipids and polyamines, and CPSI-MS obtains the relative abundance of 3829 ions in each sample.

As shown in B of FIG. 1, the sample points for PDACs and PNTs can be divided into two clusters as shown by the partial least squares discriminant analysis PLS-DA score map, and several criteria are used for stepwise variable selection in order to find PDAC specific metabolites, as shown in C of FIG. 1. First, a Variable Importance Projection (VIP) was used as a metric, retaining n-1151 highly contributing ions to the sample grouping (VIP >1.0), after which, between PDAC and PNT cryosections, significant differences were found for n-695 ions (FDR <0.05, T-test). To further select the most significant metabolites, the range was narrowed to n-626 ions with fold difference (FC) greater than 2.0 or less than 0.5(PDACs vs PNTs). By searching the human metabolome database and the Metlin metabolomics database, we speculated that n-222 ions were annotated, with the identified metabolite n-76 (as in table 1). As shown in D of fig. 1, the volcano plots highlight the first 10 small metabolites up and down regulated. As shown in fig. 1E, the heat map shows significant changes in representative small metabolites in PDACs. Down-regulated species include fatty acids (e.g., FA8:0, FA16:0, FA18:1, FA18:3), asparagine (Asn), sphingosine 1-phosphate (S1P), S-adenosylhomocysteine (SAH), cytidine, thymine, glutamine (Gln), while up-regulated species include glutamic acid (Glu), proline (Pro), leucine (Leu), citrulline, hypoxanthine (Hypo), phosphoserine (serp), and acyl carnitines (e.g., carnitine C2:0, C4:0, C16:0, C18:0, C18: 1).

Table 1: metabolites with significant changes found in PDAC tissue

Example 2 in situ validation and spatial visualization of PDAC feature metabolites using DESI-MSI

As shown in B of fig. 3, for study cases, PDAC cryosections paired with their PNT regions were prepared, and the spatial distribution of 76 metabolites in PDAC and CPSI-MS identified PNT neighboring and distant regions (PNT1 and 2) was successfully visualized (representative images are shown in C of fig. 3, the remainder are shown in table 2 and G of fig. 3). The PDAC region exhibits a high enrichment of lactic acid, leucine, Acylcarnitines (AC), Monoglycerides (MG), Phosphatidylcholine (PC), ceramides (Cer), Sphingomyelin (SM). Given that lipids can act as membrane components and signal molecules, the production of a large number of lipid molecules is consistent with the overgrowth and progression of tumors, the conversion between Diglyceride (DG) and PC is coupled with the conversion between SM and Cer. (quoted from Ogretmen B (2018) Sphingolipid metabolism in Cancer signalling and therapy. nat Rev Cancer 18:33-50.) the increased enrichment of SM, ceramide (Cer) and PC may infer the availability of the source of the Glycerolipid (GL) species. Glucose, Fatty Acids (FA), DG, Triglycerides (TG), PS, Phosphatidylinositol (PI), Phosphatidylglycerol (PG) and polyamines (except putrescine) are all down-regulated. Lipid headgroups, including glycerophosphocholine (GPCho), Glycerophosphoethanolamine (GPEA), glycerol and glycerol-3-phosphate (G3P) were observed to be very low in abundance in the PDAC region, while three phosphorylated headgroups were relatively high in abundance, including Phosphoethanolamine (PEA), phosphocholine (PCho) and phosphoserine (as in table 1), and these results also suggest abnormal lipid metabolism.

Table 2: the evidence of DESI-MSI shows that the metabolites in PDAC tissues are significantly altered

Example 3 metabolomics-based PDAC diagnostic model construction

To supplement histopathology-based diagnosis, a Lasso model was introduced to evaluate the probability of determining whether each CPSI-MS sample was from cancer or normal tissue. Metabolomics-based modeling was performed on a training set consisting of n-150 tissue sample points (n-75 PNT and n-75 PDAC tissue points from n-25 patients). An additional 90 sample points (n 45PNT and n 45PDAC tissue points from n 15 patients) were used as a test set to evaluate the performance of the pre-trained Lasso model on the untested cases. n-76 metabolites previously found by metabolic profiling and in situ validation were incorporated into the initial input variables of the training. The Lasso classifier is used because it embeds feature selection into the training process by adjusting the weighting coefficients of each input variable. Only variables that contribute to classification are given non-zero weight. Thus, the performance of the Lasso classifier achieved the highest accuracy on the test set when only 22 out of 76 variables were retained. The selected metabolic markers and weight coefficients are shown in table 3.

Table 3: metabolite markers of optimal Lasso model and their weight coefficients for complementary diagnosis

Most cross-validation samples of the PDAC group and PNT group were well distinguishable in the score plot with a threshold of 0.62 lasso prediction score, as shown in D of fig. 3. As shown in fig. 3E, the confusion matrix shows the classification results of the optimal Lasso model given, achieving an overall consistency (accuracy) of 93.3% across the test set. A Receiver Operating Characteristic (ROC) curve was introduced to evaluate the best diagnostic performance of the Lasso classifier, the area under the curve (AUC) on the test set was 0.96 (95% confidence interval: 0.93-1.00). As shown in fig. 3F, the optimal sensitivity and specificity of the diagnostic performance were 95.6% and 91.1%, respectively, with the highest true positive rate and the lowest false positive rate on the ROC curve as the cut-off point. These results indicate that PDAC-specific metabolomics shows excellent performance in differentiating tumor tissue from normal tissue. Therefore, CPSI-MS data collection in combination with the Lasso model can serve as a potential in vitro diagnostic strategy complementary to tissue biopsy and pathology.

Example 5 functional characteristics of several metabolic enzymes and transporters

As shown in Table 1, to further determine the biological consequences and metabolic network of 76 metabolites identified by CPSI-MS and DESI-MSI in PDAC tissue, Pathway Enrichment Analysis (PEA) was performed on Metabioanalysis (cited in Pang Z, Chong J, Zhou G, de Lima Morais DA, Chang L, et al (2021) Metabioanalysis 5.0: evolving the gap beta raw spectra and functional insights nucleic Acids Res.) to predict the relevant metabolic pathways during PDAC development. As shown in a of fig. 5, these differentially enriched metabolites were found to be involved in different metabolic pathways in PDAC tissues. Specifically, as shown in table 7, taurine and hypotaurine metabolism (e.g., cysteine, taurine, hypotaurine), glycerophospholipid metabolism (e.g., PE, PC, lysophosphatidylcholine, DG, PCho, PS, GPCho, GPEA, G3P), histidine metabolism (e.g., glutamic acid, urinary sugar, histidine, methyl histamine, aspartic acid), glutamic acid and glutamine metabolism (e.g., glutamic acid, glutamine, ketoglutaric acid), linoleic acid metabolism (linoleic acid, phosphatidylcholine), cysteine and methionine metabolism (e.g., serine, methionine, cysteine, phosphoserine), aromatic amino acid biosynthesis (e.g., phenylalanine, tyrosine), and arginine-related metabolites (e.g., arginine, citrulline, aspartic acid, ornithine) were enriched in PDAC tissues.

Table 7: overview of alterations in pancreatic ductal adenocarcinoma tissue metabolic pathways

PEA was further performed by searching for all PDAC related metabolic markers in KEGG and Reactome. As shown in fig. 5B, fig. 5C, table 8, initially, 56 enzymes or proteins with varying degrees of inhibition or activation were predicted, with 4 metabolic enzymes and transporters ranked highest among the candidate genes, with the differences statistically significant (P < 0.05). Phosphatidylserine decarboxylase (PISD) is down-regulated, PDZ domain protein 11(PDZD11), thymidine phosphorylase (TYMP) and solute carrier family 1 member 5(SLC1a5) are up-regulated. As shown in D of FIG. 5, PDZD11 and SLC1A5 are primarily involved in The transport of water soluble Amino acids, vitamins, free purines and pyrimidines, nucleotides and nucleosides (From Scale M, Pochini L, Console L, Losso MA, INDIVri C (2018) The Human SLC1A5(ASCT2) Amino Acid Transporter: From Function to Structure and Role in Cell biology. front Cell Dev Biol 6: 96.). PISD Is mainly responsible for the conversion of Phosphatidylserine (PS) to Phosphatidylethanolamine (PE) (from Thomas HE, ZhangY, Stefely JA, Veiga SR, Thomas G, et al (2018) Mitochondrial Complex I Activities Is Required for Maximula Autophagym. Cell Rep 24: 2404. sup. 2418.), while TYMP catalyzes the conversion of thymine/deoxyuracil to thymine/uracil, respectively (from Tetsuhiro GoKS, Kazuki Yokomei, Kato Sakuraba, Youhei Kitamura, Atsushi Shirahamata, Miveluo Sakurto, Gaku Kigawa, Hiroshi Nemoto, Yutuya adada, Kenji Hibaron (Expression) Expression of polysaccharide, pH 1757. promoter). As shown in fig. 5E and 5F, Immunohistochemistry (IHC) analysis was performed on 92 PDAC primary cancer specimens using a tissue chip, confirming this result, showing that expression of PISD was reduced in PDAC tissues, while expression of SLC1a5, TYMP and PDZD11 was significantly increased compared to PNT.

Table 8: prediction of related transporters and enzymes from enriched metabolic pathways

Example 6 biological relevance of PDAC-specific metabolizing enzymes

As shown in A of FIG. 6, PISD catalyzes The conversion of PS to PE (from Yawen Ma LW, Renbing Jia (2020) The role of mitochondral dynamics in human cameras. am J Cancer Res 10: 1278-1293.). To further explore its biological significance, PISD substrates and products were detected on paired PDAC and PNT tissue cryosections by DESI-MSI in situ imaging. As shown in B of FIG. 6, DESI-MSI analysis showed that the relative abundance of PE (a product of PISD) in PDAC was lower than in the PNT region. As shown in C of fig. 6, although there was not much difference in abundance of PDAC substrate PS in the PDAC and PNT regions in the two regions, the PE/PS ratio was significantly reduced (P < 0.05). According to the research, PISD has the tumor inhibition effect (introduced from Thomas HE, Zhang Y, Stefely JA, Veiga SR, Thomas G, et al (2018) Mitochondrial Complex I Activity Is Required for Maximal Autophary. cell Rep 24: 2404-. As shown in D of fig. 6, the Transwell experiment showed that the cell density of the PISD overexpression group was significantly lower than that of the control group. As shown in E of fig. 6, cell migration experiments showed that the distance of cell migration was short for SW1990 over PISD expression. These results are consistent with the downregulation of PISD in PDAC tissues and tumor suppressor function of PDAC.

SLC1A5 protein expression was detected by tissue chip immunohistochemistry, followed by construction of nomograms based on independent predictors determined by multiplex analysis. As shown in H of fig. 5, high expression of SLC1a5 was significantly associated with late TNM and higher stratification in PDAC patients. The Kaplan-Meier method compares survival rate curves, and shows that the expression level of SLC1A5 is in negative correlation with the overall survival time of PDAC patients, and the 1-year and 3-year survival rates of the SLC1A5 high expression group are also lower than those of the SLC1A5 low expression group.

As shown in A of FIG. 7, SLC1A5 primarily mediates transmembrane exchange of extracellular glutamine with cytoplasmic alanine (From Scale M, Pochini L, Console L, Losso MA, INDIVri C (2018) The Human SLC1A5(ASCT2) Amino Acid Transporter: From Function to Structure and roll in Cell biology. front Cell Dev Biol 6: 96). By DESI-MSI and CPSI-MS analysis, as shown in B of fig. 7, significantly lower glutamine and alanine in PDAC than in the PNT region was observed. As shown in C of fig. 7, serum metabolite marker validation further showed that glutamine and alanine were significantly lower in patients with metastatic (M1) than in patients with non-metastatic PDAC (M0). The above results indicate that glutaminase 1(GLS1) -mediated glutamine hydrolysis increases glutamine consumption. As shown in FIG. 7B, FIG. 7D, the higher glutamate levels in the PDAC region, higher glutamate to alanine (SLC1A5 substrate) and glutamate to glutamine (GLS1 product and substrate) ratios support glutamine consumption. Since the product to substrate ratio indicates the relative levels or activities of the corresponding enzymes (cited in Brett R. Hamilton DLM, Nichols R. Casewell, RobertA. Harrison, Stephen J. Blanksby, Eivind A. B. Undheim (2020) Mapping Enzyme Activity on Tissue by Functional Mass Spectrometry imaging. Angew Chem Int Ed 59: 3855-3858), these results are consistent with the IHC results, with the protein water average for SLC1A5 and GLS1 being significantly higher than for PNT in PDAC Tissue. Furthermore, as shown in E of fig. 7, SLC1a5 mRNA levels were higher in PDAC cells SW1990 than in non-tumorigenic cells H6C 7. To determine whether this was caused by SLC1A5, SW1990 PDAC cell lines were treated with the SLC1A5 inhibitor V-9302 of the present invention, as shown in FIG. 7F, with reduced glutamine and increased alanine levels. As shown in FIG. 7G, FIG. 7H, FIG. 7I, the V-9302 and GLS1 inhibitors CB-839 inhibited cell proliferation and increased sensitivity of the cells to the chemotherapeutic drug gemcitabine.

Also, consistent with the increased expression of PDZD11 and TYMP (e.g., D in fig. 5, F in fig. 5), many of the differentially enriched small metabolites were associated with over-transport and metabolism of purines and pyrimidines as shown in a in fig. 8. As shown in FIG. 8B, FIG. 8C, FIG. 8D, DESI-MSI results show that most of the purines and pyrimidines detected, including uric acid, xanthine, hypoxanthine, adenine, inosine, cytosine, deoxyuracil and thymine, were higher in the PDAC region, with high levels of deoxyuridine (product), thymine (product) and low levels of uracil (substrate) consistent with high expression of TYMP. This is consistent with a previous report that TP-dependent Thymidine kinase catabolism contributes to cancer cell survival under low nutrient conditions (from Toi M, Rahman MA, Bando H, Chow LWC (2005) Thymidine phosphoridase (planar-driven end-cell growth factor) in cancer biology and evaluation. the Lancet Oncology 6: 158-166.).

In the examples, the early diagnosis marker obtained by screening according to the tissue screening method for early diagnosis markers of pancreatic cancer comprises: any one or a combination of any two or more of diglyceride, 1-N-butylamine, cytidine, histamine, phosphorylcholine, hydroxy linolenic acid, linoleic acid, 3-phosphoglycerol, glutamic acid, acrylylcarnitine, glutamine, tyrosine, 3-hydroxypalmitic acid, lysophosphatidylcholine (P-16:0), butyrylcarnitine, piperidinecarboxylic acid, lysophosphatidylcholine (18:2), acetylspermine, lactic acid, histidine, N-methyl histamine, and N-acetyl histamine.

It is understood that the early diagnosis marker of pancreatic cancer can be used for preparing a detection product for detecting the expression level of the early diagnosis marker.

In some embodiments, the test product can be an in vitro test reagent.

In some embodiments, the test product can also be a test kit, and in particular, the test kit can contain a test reagent.

In conclusion, PDAC metabonomics based on CPSI-MS and DESI-MSI are a convenient characterization tool, can perform overall overview on cancer metabolism at an in-situ level, is a robust method for rapid molecular diagnosis and screening, and differential metabolites in pancreatic cancer tissues and normal tissues obtained through experiments can be used as early diagnosis markers of pancreatic cancer, so that the early diagnosis markers of pancreatic cancer have wide clinical application prospects.

While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims.

Claims (10)

1. A tissue screening method for an early diagnosis marker of pancreatic cancer, comprising the steps of:

step 1, respectively taking a clinical tissue PDAC of a pancreatic cancer patient and a normal tissue PNT matched with the clinical tissue PDAC;

step 2, analyzing and identifying primary differential metabolites between pancreatic cancer tissues and normal tissues by a chromatography-mass spectrometry combined metabonomics analysis method;

step 3, visually showing the spatial distribution of the preliminary differential metabolites through mass spectrometry imaging, and finding metabolites which are abnormally distributed in pancreatic cancer tissue regions;

and 4, performing auxiliary screening on the abnormally distributed metabolites: constructing a pancreatic cancer diagnosis Lasso model based on metabonomics based on tissue biopsy, and screening early diagnosis markers in pancreatic cancer tissues from the abnormally distributed metabolites.

2. The method for tissue screening of early diagnosis markers of pancreatic cancer according to claim 1, wherein in step 2, the chromatogrAN _ SNhy/mass spectrometry-metabolomics is conductive polymer spray ionization mass spectrometry-CPSI-MS-metabolomics.

3. The method of claim 1, wherein the step 2 of analyzing and identifying preliminary differential metabolites between pancreatic cancer tissue and normal tissue comprises the steps of:

Step 2.1, reserving sample ions by taking VIP larger than 1.0 as a standard;

step 2.2, performing T test on the sample ions obtained in the step 2.1, and further reserving the sample ions by taking FDR <0.05 as a standard;

step 2.3, performing fold difference analysis on the sample ions obtained in the step 2.2, and further retaining the sample ions by taking the fold difference greater than 2.0 or less than 0.5 as a standard;

and 2.4, comparing the sample ions obtained in the step 2.3 with a metabolome database to obtain 76 primary differential metabolites in the tissues.

4. The method for tissue screening of markers for early diagnosis of pancreatic cancer according to claim 1, wherein in step 3, the mass spectrometric imaging technique is desorption electrospray ionization mass spectrometric imaging DESI-MSI, visualized showing the spatial distribution map of each of said preliminary differential metabolites for in situ validation.

5. The method for tissue screening of early diagnosis markers of pancreatic cancer according to claim 1, wherein the auxiliary screening in step 4 comprises the following steps:

step 4.1, taking the n-150 tissue sample points as a training set, and taking the n-90 tissue sample points as a test set, wherein the n-90 tissue sample points are used for evaluating the performance of a pre-trained Lasso model;

Step 4.2, the 76 kinds of preliminary differential metabolites are brought into initial input variables of a training set of the Lasso classifier, only variables contributing to classification are endowed with non-zero weight, and when only 22 variables out of the 76 variables are reserved, the performance of the Lasso classifier achieves the highest accuracy on a test set;

and 4.3, referring to a receiver operating characteristic ROC curve to evaluate the optimal diagnostic performance of the Lasso model, and taking the highest true positive rate and the lowest false positive rate on the ROC curve as cut-off points to obtain the optimal diagnostic performance, which indicates that the CPSI-MS data acquisition combined with the Lasso model can be used as a potential in-vitro diagnostic strategy complementary with tissue biopsy.

6. The early diagnosis marker for pancreatic cancer according to any one of claims 1 to 5, which comprises: any one or a combination of any two or more of diglyceride, 1-N-butylamine, cytidine, histamine, phosphorylcholine, hydroxyl linolenic acid, linoleic acid, 3-phosphoglycerol, glutamic acid, acrylylcarnitine, glutamine, tyrosine, 3-hydroxypalmitic acid, lysophosphatidylcholine, butyrylcarnitine, piperidinecarboxylic acid, lysophosphatidylcholine, acetylspermine, lactic acid, histidine, N-methyl histamine, and N-acetyl histamine.

7. The use of the early diagnosis marker of pancreatic cancer according to claim 6 in the preparation of an early diagnosis reagent or kit for pancreatic cancer.

8. An agent for early diagnosis of pancreatic cancer, which is capable of detecting the expression level of the early diagnosis marker of claim 6 in a sample.

9. The reagent for the early diagnosis of pancreatic cancer according to claim 8, wherein said sample is a tissue sample.

10. A kit for early diagnosis of pancreatic cancer is characterized in that the kit contains an early diagnosis marker for preparing a diagnostic reagent for early differentiation of pancreatic cancer.

Images