CN109852672B - Method for screening acute myeloid leukemia DNA methylation prognosis marker - Google Patents

Abstract

The invention discloses a method for screening acute myeloid leukemia DNA methylation prognosis markers, which comprises the following steps: (1) collecting samples and collecting clinical data; (2) preparing a sample; (3) extracting genome DNA; (4) whole genome CpG site methylation capture sequencing: (5) and (5) analyzing methylation sequencing data, and screening out the marker. The invention uses the whole genome CpG locus methylation capture sequencing technology to detect the DNA methylation of acute myeloid leukemia and carry out prognosis analysis, and selects the gene marker which has the prognosis guiding significance and is regulated and controlled by the DNA methylation by integrating the genetic information and the epigenetic information to guide the accurate diagnosis and treatment of AML, thereby improving the curative effect and improving the prognosis of patients.

Description

Method for screening acute myeloid leukemia DNA methylation prognosis marker

Technical Field

The invention relates to a method for screening acute myeloid leukemia DNA methylation prognosis markers, belonging to the field of biological medicine.

Background

Leukemia is one of the most common malignant tumors in China, and the mortality rate of leukemia is the top in children and young cancer patients under 40 years old. 2016, Chinese scholars published an important article of Cancer Statistics in China,2015, which reveals that the morbidity of leukemia is ranked at 12 th, the mortality rate is increased to 9 th, about 5 ten thousand new cases of leukemia are developed each year, and the morbidity and the mortality rate show a trend of increasing year by year^[1]. Among leukemias, Acute Myelogenous Leukemia (AML) is the main cause, accounting for about 70% of the incidence rate of acute leukemia^[2]. AML has complex pathogenesis, large clinical heterogeneity and high relapse mortality^[3-7]The treatment cost is expensive, and heavy economic burden is brought to families and society. The accurate diagnosis and treatment of AML is a necessary choice for improving the curative effect, and becomes a hotspot and difficulty in current research.

AML is currently diagnostically and prognostically stratified primarily based on bone marrow Morphology (M), Immunology (I), Cytogenetics (C) and Molecular biology (M) (MICM) characteristics^[8,9]. Based on the current prognostic stratification, nearly 50% of patients remain normal karyotypes and lack prognostic-related gene mutations^[10-12]And accurate treatment under accurate prognosis guidance cannot be realized.

Research shows that epigenetic modifications such as DNA methylation, histone acetylation, chromatin remodeling and the like are closely related to the occurrence and development of AML^[13]. DNA methylation abnormality is a remarkable feature of AML^[13-15]Has important function in guiding the typing and the pre-and post-stratification of AML^[16-18]. However, due to the shortcomings of the current DNA methylation detection technology, DNA methylation detection cannot be clinically applied routinely, such as the Whole Genome Bisulfite Sequencing (WGBS) is expensive, the sequencing depth is insufficient, and the data analysis is difficult; methylated 450K chips (Illumina Human Methylation 450 ready array, Illumina 450K array) do not directly measure bases, have limited detection sites and are overly closed in data analysis systems; agilent SureSelect methylation assayThe sequence molecule has low diversity, large initial sample amount and single-strand capture; simplified representation bisulfite sequencing (RRBS) is limited by restriction of enzyme cleavage sites, region fixation, and data loss^[19,20]. Therefore, a brand-new DNA methylation detection method suitable for clinical application in AML is urgently needed, and by accurately determining the DNA methylation state of AML, prognosis stratification is perfected, and accurate diagnosis and treatment are promoted.

"Genome-wide CpG-locus methylation capture sequencing technology (MCC-Seq)" published in journal of Nature Communication in 2015 is a brand-new DNA methylation capture detection technology based on a second-generation sequencing platform^[20]. The technology not only makes up the defect of insufficient coverage of CpG sites of a 450K chip, but also solves the problems of overlarge coverage of WGBS, insufficient sequencing depth, insufficient site accuracy, time and labor consumption and high cost, has the advantages of unique probe design, capability of double-strand capture, low initial sample amount, accurate quantitative detection and the like, is an accurate, economic and efficient DNA methylation detection method, and can promote the popularization and application of DNA methylation research. So far, there is no report on the development of MCC-Seq sequencing technology in AML, and no report on AML prognosis research using the technology.

In conclusion, in view of the large heterogeneity of AML, the lack of the current prognosis stratification system, the close correlation between DNA methylation and AML, the defects of the previous methylation research methods, and the multiple advantages of MCC-Seq for methylation detection, we have a full basis that MCC-Seq can be used for detecting DNA methylation of AML, accurately describing methylation characteristics of AML, searching methylation regulation genes related to poor prognosis through DNA methylation difference analysis, perfecting the AML prognosis stratification system, and guiding accurate diagnosis and treatment of AML, thereby improving the curative effect, improving the prognosis of patients, reducing economic burden for the country and society, and making contribution to the development of accurate medicine.

The whole Genome CpG locus methylation capture sequencing technology (MCC-Seq) is a brand new DNA methylation system based on the enrichment system of the sequence of NimbleGen SeqCap EpiThe detection technology covers 550 ten thousand methylation sites, and CpG sites outside the CpG island and sites of disease related regions are added. Compared with the Illumina 450k chip, the SeqCap Epi comprises all sites on the DNA chip and more than 12 times of CpG sites, and the methylation level analysis result of the public site is high in consistency. In addition, novel sites for methylation modification alterations can be found, and allele-specific methylation patterns obtained^[21,22]. The MCC-Seq assay is effective in avoiding sample loss, with an initial sample size of only 1. mu.g, and 30% improvement in key sequencing indicators, such as homogeneity, which reflects the improvement in library complexity. In addition, multiple probes are designed aiming at the same position of a section to ensure that methylation under different conditions can be effectively captured, the probes are designed aiming at converted positive and negative double strands, the sequencing results of the positive and negative strands are used as references, and the sequencing result of the strand which is not methylated is used as a reference to judge whether the conversion of cytosine (C) into thymine (T) is mutation of a genome or the result of treatment of unmethylated cytosine by bisulfite, so that the accuracy of methylation analysis is improved.

In 2015, the Nature Communication journal firstly reports the research of MCC-Seq for human adipose tissue Epigenome association analysis (EWAS), which proves that the MCC-Seq is an accurate, economical and efficient DNA methylation detection method, and is particularly suitable for the research of disease susceptibility/phenotype related functional differential methylation genes^[20]。

Disclosure of Invention

The technical problem to be solved by the invention is as follows: in view of the high incidence and mortality rate of AML, large heterogeneity, the current prognosis stratification system is insufficient, DNA methylation is closely related to AML, the existing methylation research method has defects, and the MCC-Seq detection methylation has multiple advantages, the invention applies the MCC-Seq technology to carry out DNA methylation sequencing on AML bone marrow specimens, quantitatively detects the methylation state of the CpG sites of the AML whole genome, searches methylation regulation genes related to poor prognosis through DNA methylation differential analysis, guides the accurate diagnosis and treatment of AML, improves the prognosis of patients, reduces economic burden for the country and society, and makes a contribution to the development of accurate medicine.

The technical scheme provided by the invention is as follows: a method for screening acute myeloid leukemia DNA methylation prognosis markers, comprising the following steps:

(1) collecting samples and collecting clinical data;

(2) preparing a sample;

(3) extracting genome DNA;

(4) sequencing by whole genome CpG locus methylation capture, comprising the following steps:

a) first, a library was constructed: fragmenting DNA, repairing tail end, adding 'A', connecting a connector, processing Bisulfit, performing PCR amplification and purifying a product to obtain a purified product;

b) target region capture sequencing: hybridizing the library with a chip in a warm bath before hybridization, eluting a target DNA fragment, carrying out PCR amplification on the target DNA fragment, and sequencing;

(5) methylation sequencing data analysis:

firstly, data filtering and comparison are carried out: filtering the sequencing data in the step (4), removing low-quality data to obtain available data, comparing the available data with a reference genome to obtain a comparison result after the data are detected to be qualified, obtaining C base methylation information by using the unique comparison data after the comparison quality is confirmed to be qualified, and performing information analysis processing to obtain a standard information analysis result and a personalized analysis result;

then, depth and coverage analysis of the capture area: analyzing the coverage of the capture region at different read sequencing depths and analyzing the cumulative distribution of the effective sequencing depths of the C bases, the distribution of the methylated C bases on the genome comprising three forms, CG, CHG and CHH, wherein H represents A or T or C bases;

carrying out methylation site identification and calculating methylation level, calculating the number of sites of each type of mC and the proportion of the sites of all mC, and calculating the methylation level;

analysis of differentially methylated regions: searching different methylation region DMRs meeting conditions among a plurality of different components and in different gene regions, annotating the found DMRs to a gene body, and comparing the distribution conditions of the DMRs in different regions of the gene body;

screening for differentially methylated genes: carrying out GO enrichment analysis and KEGG enrichment analysis on the differential methylated genes, determining main biochemical metabolic pathways and signal transduction pathways in which the differential methylated genes participate, and finally screening out a group of 12 functional differential methylated genes which are related to genetic prognosis and are regulated and controlled by DNA methylation, namely: BARD1, BCL9L, CLEC11A, DEFB1, FOXD2, IGF1, IL18, ITIH1, LSP1, P2RX6, RNASE3, and TUBGCP 2.

The method further comprises the step (6) of: clinically significant verification, calculating the DNA methylation level of each DMGs of each patient according to the 12 screened DMGs, and recording the average value of the methylation levels of the 12 DMGs of the same patient as the comprehensive methylation value M-value of the patient to represent the methylation level of the patient.

In the method, in the step (5), the methylation level of each methylated C base is calculated according to the following formula: methylation level at C-site ═ number of Reads supporting methylation/(number of Reads supporting methylation + number of Reads supporting non-methylation) × 100%;

the average methylation levels for the different gene regions were calculated as follows:

the average methylation level of a region is the number of Reads that the region supports methylation/the total number of Reads in the region x 100%.

In the step (5), DMR meets the following three conditions: at least 5 CpGs; wherein, at least 3 differential methylation sites exist, and the regional differential methylation level is more than or equal to 20 percent.

In step b), QPCR primer 1 and QPCR primer 2 for amplifying the target DNA fragment,

primer 1 sequence: upstream: GTTAGGTAGGGAAGAAGGGAGTAGT

Downstream: CCCAAAAATCAAATAATCAAAAAAA

The sequence of the primer 2 is as follows: upstream: GTGGTTAATTAATTTTTGAGTTTTGT

Downstream: TATTACCCTATAACCACCATCACC in step (7), the data filtering process includes the following steps:

(1) reads to decontaminate the joints, namely: the number of bases for removing the linker contamination in Reads is more than 5 bp;

(2) removing low quality Reads;

(3) the Reads with an N content of more than 5% are removed.

In the method, in the step (5), the available data is compared with the reference genome, in the information analysis process, the sequencing result and the reference genome are both subjected to C-to-T (forward) and G-to-A (reverse) conversion, the converted Reads and the converted genome sequence are compared, and the unique comparison Reads is used for the analysis of methylation information and is used for the comparison of genome database position University of California Santa Cruz HG 19.

The invention also provides an acute myeloid leukemia DNA methylation prognosis marker obtained by screening by the method, namely a functional differential methylation gene for DNA methylation regulation: BARD1, BCL9L, CLEC11A, DEFB1, FOXD2, IGF1, IL18, ITIH1, LSP1, P2RX6, RNASE3, and TUBGCP 2.

Meanwhile, the invention also provides a kit for screening the acute myeloid leukemia DNA methylation prognosis marker, which comprises the following components: DNA end repair system: the system comprises 10 XKAPA End Repair Buffer 7. mu.l, KAPA End Repair Enzyme 5. mu.l, dd H₂O8 mul, total volume 70 ul; adding an "A" system: the system comprises 5 mul of 10X KAPA A-labeling Buffer and 3 mul of APA A-labeling Enzyme, and the total volume is 50 mul; a connection system: the system comprises 10. mu.l of KAPA Ligation Buffer, 5. mu.l of KAPA T4DNA Ligation Buffer, 5. mu.l of Methylation Adapter (10. mu.M), and a total volume of 50 ul; PCR amplification System: the system comprises 25 mu l of KAPA HiFi HotStart Ready Mix, 2 mu l of KAPA Library amplification primers and 13 mu l of deionized water, wherein the total volume is 50ul, and the amplification primer sequence: upstream: AGTGGTTAATTAATTTTCGAGTTTC, downstream: TATTATTACCCTATAACCACCATCG, respectively; PCR amplification system after target fragment capture: the system comprises 1 ul of QPCR primer 1 and QPCR primer 2 respectively, 12.5 ul of HiFi DNA polymerase mixture and 25 ul of total amount, wherein the sequence of the QPCR primer 1 is as follows: upstream: GTTAGGTAGGGAAGAAGGGAGTAGT, downstream: CCCAAAAATCAAATAATCAAAAAAA, QPCR primer 2 sequence: upstream: GTGGTTAATTAATTTTTGAGTTTTGT, downstream: TATTACCCTATAACCACCATCACC are provided.

The invention has the following beneficial effects:

the invention mainly aims to solve the clinical problems from the practical clinical point of view and applies the whole genome CpG locus methylation capture sequencing technology to detect the acute myeloid leukemia DNA methylation and carry out prognostic analysis. The method has good conversion rate (> 99.5%), Q30 (95.03%), comparison rate (92.90%) and good stability verified by technical repeated samples. The invention screens the genetic marker which has the prognostic guidance significance and is regulated and controlled by DNA methylation by integrating the genetic information and the epigenetic information, and is beneficial to perfecting an AML prognostic layering system and guiding the precise diagnosis and treatment of AML, thereby improving the curative effect and improving the prognosis of patients. The method for detecting AML bone marrow sample data by applying the MCC-Seq technology is accurate and reliable, and has good feasibility and practicability. The DNA methylation level of AML patients is a stable and reliable index for evaluating the methylation state, and the AML patient promoter region has the most important functional differential methylation region and is suitable for screening DNA methylation regulation genes relevant to prognosis. Therefore, the average methylation level M-value of 12 DMGs screened by the method is a stable and reliable index for evaluating the DNA methylation level of a patient, the 12 DMGs can be used as a new biomarker for AML prognosis stratification, and the high M-value is a poor prognosis factor for inducing remission, disease-free survival and overall survival.

Drawings

FIG. 1 relationship between the methylation level of whole genome CpG sites and various clinical factors.

The promoter region of the gene of fig. 2 has the most important functional DNA methylation profile, among others: panel A, comparison of DNA methylation levels of different gene regions in naive AML and normal control bone marrow samples; panel B, 8 comparison of DNA methylation levels of different gene functional regions of naive-remission paired AML bone marrow samples; panel C, naive AML and normal control bone marrow samples comparing differentially methylated regions of different gene functional regions; panel D, differential methylation of different functional regions of genes compared between different genetic prognostic partitions in naive AML bone marrow samples.

Detailed Description

The invention is further illustrated by the following detailed description of specific embodiments, which are not intended to be limiting but are merely exemplary.

In the examples, unless otherwise specified, all experimental methods used were conventional methods; the experimental materials, reagents and the like used are commercially available.

First, experimental material

1. MCC-Seq sequencing cases

AML patients are selected according to the inclusion standard: (1) a definitive diagnosis of non-M3 type acute myeloid leukemia; (2) the age is greater than 18 years; (3) one who completes at least one course of chemotherapy and evaluates the curative effect; (4) MICM diagnosis data and clinical treatment data are complete; (5) (ii) mutational targeted sequencing results (to allow accurate molecular biological prognosis stratification); (6) at least the bone marrow specimen at the time of primary treatment is preserved. Finally, 21 primary AML patients (nos. C01, C02, C03 … …, C21) were selected and tested for 44 bone marrow samples, and the detailed information is shown in table 1.

TABLE 1 summary of DNA methylation sequenced patients and sample information

The "7 + 3" protocol, a standard induction protocol based on cytarabine and anthracycline, see in particular the NCCN guidelines; decitabine combination regimen, Decitabine 20mg/m²d1-5, Cytarabine 10mg/m²q12h d1-5, aclacinomycin 20mg d1,3,5, granulocyte stimulating factor 300 μ g/d used from one day before chemotherapyTo neutrophil recovery.

2. TCGA database for case verification

To better validate the clinical significance of differentially methylated genes screened from self-test cases, the inventors retrieved data from the TCGA database for AML patients with Illumina 450k chip methylation (II)https://tcga- data.nci.nih.gov/tcga/)^[23]. Screening was performed based on clinical information and the completeness of the DNA methylation assay data, resulting in a total of 169 AML patients enrolled in the study. The patient information is shown in table 2.

TABLE 2 summary of patient information in TCGA database

Second, Experimental methods

1. Sample collection and clinical data collection

In all cases (21 primary treatment AML, 2 relapse AML, 5 healthy controls) after informed consent was signed, 3-5ml bone marrow samples were collected by bone marrow aspiration. Clinical diagnosis and treatment information of the patient is collected, including general information, bone marrow MICM information, AML diagnosis, treatment and curative effect, survival condition and the like.

2. Technical replicate sample preparation

We randomly selected 2 relapsed AML patients (C22 and C23), collected fresh bone marrow specimens and extracted mononuclear cells, respectively, and then divided the mononuclear cells of the same patient into 2 aliquots to obtain 4 technical replicate samples (S22-repeat 1 and S22-repeat 2, S23-repeat 1 and S23-repeat 2) for DNA extraction and MCC-Seq sequencing to evaluate the stability of MCC-Seq assay bone marrow samples.

3. Genomic DNA extraction

Bone marrow DNA extraction was performed according to the Genomi DNA extraction kit (promega) instructions.

Before the sample is checked, the absorbance value (A) is measured by using an ultraviolet spectrophotometer to determine the content and purity of the sample, and the integrity of the DNA is detected by agarose gel electrophoresis. The requirements of this experiment on DNA samples: the total amount of DNA is more than or equal to 3 mug, the concentration is more than or equal to 50 ng/mul, and the purity OD260/280 is more than or equal to 1.8. In addition, in terms of DNA integrity, no tailing and no bands of dispersion are required to be detected by agarose gel electrophoresis.

4. Whole genome CpG site methylation capture sequencing

4.1 library construction

First, fragmentation of DNA

5.8. mu.l of diluted thioconversion control reagent was added to 1. mu.g of DNA in a pre-cooled sonicator (4 ℃), mixed with sufficient EB to 50. mu.l and sonicated to fragment the 200bp DNA fragment. The specific operation method is described in Bioruptor Standard operation flow. Then, agarose electrophoresis is used for detecting whether the size of the DNA fragment is qualified or not.

Second, end repair

The repair system was formulated in an EP tube with the specific ingredients and amounts shown in Table 3. After mixing, the DNA was purified by magnetic bead recovery after a bath at 20 ℃ for about half an hour and dissolved by adding 42. mu.l of EB.

Three, plus "A"

An "A" addition system was prepared in an EP tube, and the specific ingredients and amounts used are shown in Table 4. After mixing, the mixture was incubated at 37 ℃ for about half an hour, and then the DNA was recovered and purified with magnetic beads and dissolved in 20. mu.l of EB.

Four, joint connection

The ligation system was further formulated with the DNA product after "A" addition, see Table 5. After being prepared, are mixed with

After bathing at 16 ℃ for about half an hour, the DNA was recovered and purified using magnetic beads, and dissolved by adding 30. mu.l of EB.

TABLE 3 repair System

TABLE 4 plus "A" system

TABLE 5 connection System

Fifthly, vulcanization treatment

Using EZ DNA Methylation-Gold from Zymo^TMKit sulfite-treated the ligation product to convert unmethylated C to T. The specific steps refer to EZ DNA Methylation-Gold^TMThe Kit step was performed and finally dissolved in 20ul ddH₂And (4) in O.

Sixthly, PCR amplification and product purification

The DNA product after the linker addition was amplified by PCR and further purified.

(1) The PCR amplification system is shown in Table 6

TABLE 6PCR amplification System

The primer sequence is as follows: upstream: AGTGGTTAATTAATTTTCGAGTTTC

Downstream: TATTATTACCCTATAACCACCATCG

(2) PCR reaction

Setting a PCR program:

(3) recovery and purification of PCR product

The DNA product after PCR amplification is recovered and purified by magnetic beads and dissolved in ddH₂O for the next step.

4.2 target region Capture sequencing

First, hybridization of pre-hybrid library with chip by warm bath

(1) Equal amounts of the library to be hybridized were mixed to prepare a sample library (1. mu.g).

(2) The corresponding Index for each sample library was mixed equimolar to generate an Index library (1000 pmol/L).

(3) Mu.l of the sulfurized capture enhancer with the mixture of step (1) was added to the EP tube, and 10. mu.l of the common primer (1000pmol/L) and 10. mu.l of the mixture of step (2) (1000pmol/L) were added thereto.

(4) The EP tube with the hybridized sample in step (3) is sealed by a closed membrane, and a plurality of holes are punched on the membrane, and the membrane is placed on a vacuum drying concentrator (the temperature is 50 ℃) to be dried by suction until the membrane is completely dried.

(5) The following two ingredients were added to the dry powder of the EP tube: 2 Xthe hybridization buffer and hybridization fraction A (Table 7), after shaking centrifugation for 10 seconds, were then denatured in a 95 ℃ incubator for 10 minutes.

TABLE 7 composition and amount of hybridization samples

(6) After time, the mixture in step (5) was added to a PCR tube containing a NimbleGen chip by rapid centrifugation. After shaking for 3 seconds, the mixture was centrifuged, placed on a 47 ℃ heating module for 68-72 hours, and a hot lid was set to 57 ℃.

Second, elution of the target DNA fragment

(1) The washing stock solutions in test tubes 1, 2, 3 and 4 were prepared as 1 Xbuffer (Table 7).

(2) Dynabeads M280 was placed at room temperature for half an hour in advance, shaken for 1 minute before the experiment, mixed well, dispensed 50. mu.l into a centrifuge tube, and placed on a magnetic rack to stand for 5 minutes to remove the supernatant, then 1 xBead washing buffer with twice the volume of the magnetic beads was added, shaken for 10 seconds and placed on the rack, and the supernatant was discarded. Resuspend with 50. mu.l of 1 XBead wash buffer and transfer to a new labeled EP tube and place again on the magnetic rack to remove supernatant.

TABLE 8 magnetic bead washing solution composition and content

(3) Adding the hybridized DNA library into the EP tube in the step (2), setting a heating module to be 47 ℃ for heating, and shaking and uniformly mixing for 45 minutes at intervals. Then 1 × Wash buffer I was added in a volume of 50 μ l at 47 ℃ and mixed well. The product was transferred to a new labeled EP tube and placed on a magnetic stand to remove the supernatant. Then, 100. mu.l of 47 ℃ 1 XStringent washing buffer was added, mixed well and then placed on a 47 ℃ heating module for a 5min warm bath.

(4) Add 100. mu.l of 1 XWash buffer I to the EP tube from step (3), mix and place on a magnetic rack to remove supernatant. Then, 100. mu.l of 1 XWash buffer II was added, mixed and placed on a magnetic stand to remove the supernatant. Then, 100. mu.l of 1 XWash buffer III was added, and the mixture was placed on a magnetic stand to remove the supernatant. Finally, 50. mu.l ddH was added to the EP tube₂O, eluting the beads and capturing the DNA for lysis. The beads-DNA mixture was stored at-20 ℃.

Third, PCR amplification of target DNA fragment

(1) Preparation of post Capture PCR mix (Table 9)

TABLE 9 PCR mixture after Capture

Primer 1 sequence: upstream: GTTAGGTAGGGAAGAAGGGAGTAGT

Downstream: CCCAAAAATCAAATAATCAAAAAAA

The sequence of the primer 2 is as follows: upstream: GTGGTTAATTAATTTTTGAGTTTTGT

Downstream: TATTACCCTATAACCACCATCACC

(2) PCR reaction

Setting a PCR program:

(3) recovery and purification of PCR product

Recovering and purifying the amplified DNA product with magnetic beads,and dissolved in ddH₂O to be sequenced.

Four, Hiseq2500 sequencing

The hybridization capture samples were sequenced with Hiseq2500 PE 100.

5. Methylation sequencing data analysis

After the data is downloaded, data filtering is firstly carried out, low-quality data are removed, and usable data are obtained. And after the data are detected to be qualified, comparing the available data with the reference genome to obtain a comparison result. And after the comparison quality is determined to be qualified, obtaining C base methylation information by using the unique comparison data, and performing information analysis processing to obtain a standard information analysis result and a personalized analysis result.

5.1 data Filtering and alignment

First, original data

The Illumina high-throughput sequencing result exists in an original image data file at first, is converted into an original sequencing sequence after base recognition by CASAVA software, and is stored in a FASTQ (fq for short) file format. The FASTQ file contains the name of each sequenced sequence, the base sequence, and its corresponding sequencing quality information. In the FASTQ format file, each base corresponds to one base quality character, and 33(Sanger quality value system) is subtracted from the ASCII code value corresponding to each base quality character to obtain the sequencing quality score of the base. Different sequencing quality scores represent different base sequencing error rates, e.g., a sequencing quality score of 20 and 30 indicates a base sequencing error rate of 1% and 0.1%, respectively.

Second, data filtering

And filtering the original sequence obtained by sequencing to obtain high-quality Clean Reads, and then performing subsequent analysis, wherein the subsequent analysis is based on the Clean Reads.

The data processing steps are as follows:

(1) reads to remove linker contamination (number of linker-contaminated bases in Reads greater than 5 bp. for PE, if one end is linker contaminated, then remove both ends of Reads);

(2) removing low-quality Reads (bases with quality value Q less than or equal to 19 account for more than 15% of total bases in Reads, and for PE, if one end is low-quality Reads, both ends are removed);

(3) reads with an N content of more than 5% are removed (for PE, if one end has an N content of more than 5%, both ends are removed).

The data filtering statistical indexes are as follows:

reads Length (bp): the length of the sequence;

raw Reads: sequence number of original machine unloading;

raw Bases: the number of bases of the original starting sequence;

clean Reads: obtaining high-quality sequence numbers after filtering;

clean Reads Rate (%): and the high-quality sequence obtained after filtering accounts for the proportion of the original machine-descending sequence. The larger this value, the better the sequencing quality or library quality;

clean Bases: the number of bases of the filtered high quality sequence;

low-quality Reads: number of sequences removed due to excess low-mass bases;

low-quality Reads Rate (%): the removed sequence accounts for the proportion of the original off-machine sequence due to the excess of low-quality bases; ns Reads: the sequence number which is removed due to the over-high N content;

ns Reads Rate (%): because N is too high, the removed sequence accounts for the proportion of the original downloading sequence;

adapter poluted Reads: the number of sequences removed due to contamination of some sequences by linkers;

adapter poluted Reads Rate (%): because some sequences are contaminated by linkers, the removed sequences account for the proportion of the original machine-descending sequences;

origin Q30Bases Rate (%): the ratio of bases with sequencing quality value of more than 30 (error rate less than 0.1%) in Raw Reads to total bases;

clean Q30Bases Rate (%): the ratio of bases with sequencing quality value of more than 30 (error rate less than 0.1%) to total bases in Clean Reads.

Third, distribution of sequencing data quality values

The sequencing error rate is related to the base quality and is determined by the sequencer itself,Sequencing reagents, samples and the like. The error rate of sequencing per base was determined by sequencing the Phred number (Phred score, Q)_phred) The values are obtained by formula conversion, and the Phred values are calculated by a model for predicting the probability of error of Base discrimination in the Base recognition (Base Calling), and the corresponding relation is shown in the following table 10:

TABLE 10 concise correspondences between Illumina Casava base recognition and Phred scores

Fourth, base distribution of sequencing data

A base map was obtained by using the base positions of the filtered sequence as the abscissa and the ratio of ATCG bases (wherein N represents an unknown base) at each position as the ordinate.

Fifth, comparison analysis

In the information analysis process, the sequencing result and the reference genome are converted into C-to-T (Forward) and G-to-A (reverse), the converted Reads and the converted genome sequence are aligned, and the unique aligned Reads is used for the analysis of methylation information. Genomic database bit University of California Santa Cruz HG19 for alignment. Using software and parameters^[24](version: V0.9.0): bismark-p 5-N1-directional-o Directory-1fq1-2fq 2.

5.2 methylation site recognition and calculation of methylation levels

Methylation site recognition

The occurrence ratio of methylated C sites of different distribution types in genomes of different species is different^[25]Therefore, the number of sites of each type of mC (mCG, mCHG and mCHH) and the ratio of the sites of all mC (for example, the ratio of mCHG to the total number of mCHG/mC) reflect the characteristics of the whole genome methylation map of a specific species to a certain extent.

II, calculating the methylation level

The methylation level of each methylated C base was calculated as follows: methylation level at C-site ═ number of Reads supporting methylation/(number of Reads supporting methylation + number of Reads supporting non-methylation) × 100%. The capture region average methylation levels reflect the overall characteristics of the genomic methylation profile. Only methylation sites covering a depth of not less than 5 are considered here.

The average methylation levels for the different gene regions were calculated as follows: the average methylation level of a region is the number of Reads that the region supports methylation/the total number of Reads in the region x 100%.

5.3 differential methylation region analysis

The present study searched for Differentially Methylated Regions (DMRs) among multiple different components, in different gene regions, and the DMR required the following three criteria: at least 5 CpGs; wherein, at least 3 differential methylation sites (differential methylation sites are sites with differential methylation level more than or equal to 20 percent); the regiodifferential methylation level is more than or equal to 20 percent. Using software and parameters (V0.9.2): r-package methyKit^[26]And eDMR^[27]。

The found DMRs are annotated to the gene body, and the distribution of the DMRs in different regions of the gene body is compared, so that the influence of the change of the methylation modification on the gene regulation can be better understood.

5.4 differential methylation Gene screening and clinical significance verification

The invention aims to search genes with poor hypermethylation prognosis in AML patients, so that differential methylation comparison is carried out between different prognosis groups on the basis of cytogenetics and molecular genetics prognosis groups, hypermethylation differential areas in groups with high risk are searched, the differential genes are annotated according to the differential areas, gene function analysis is further carried out, target genes are screened and determined, and then clinical significance verification including relation with clinical factors, remission rate and survival significance analysis and the like is carried out in research cases and TCGA database cases.

6. Statistical method

Continuous variables were tested by student's t test, analysis of variance, Mann-Whitney U rank sum test; the categorical variables were tested using Fisher's accuracy test or Chi-squared test. Survival analysis used the Kaplan-Meier curve. Overall Survival (OS) definition: from first visit to patient death or termination of follow-up; disease Free Survival (DFS) is defined as from complete remission to relapse, death, or termination of follow-up. Two-sided test P <0.05 is a statistical difference in significance. The above statistical analysis was performed on Software such as SPSS 19.0(IBM Corp., Armonk, NY, USA), GraphPad Prism 6(GraphPad Software Inc., San Diego, California, USA).

Third, experimental results

1. Test data coverage results

All 44 samples sequenced on the machine were used for sequencing data quality control analysis. Our captured targets covered 240,513 regions of the whole genome, containing more than 5Mb CpG sites, sharing 80M base pairs. The sequencing read was 125bp in length. The average percentages of CpG sites with coverage of not less than 1X, 5X, 10X, 20X were 92.73%, 80.32%, 67.20%, and 43.81%, respectively. All 44 cases of secondary DNA samples covered more than 30 Xwith data size of 430 Gb. The conversion of all samples exceeded 99.5%.

2. Test data filter results

After filtering the poor quality data, the average total clean reads number of the 44 samples was 58,147,036 (range: 40,663,694-79,302,058), and the average sequencing quality value clean Q30base rate was 95.03% (range: 92.65% -98.05%).

3. Test data comparison result

Compared with the reference genome, Clean reads aligned to the target capture CpGs of the genome on average 72.32% (range: 38.54% -85.31%), and the overall alignment rate was 92.90% (range: 80.52% -96.40%).

4. Technical duplicate sample results

To verify the stability of MCC-Seq assays for DNA methylation in bone marrow samples, we performed independent, parallel assays on sample replicates of two AML patients. The results show that two duplicate DNA methylation sequencing results of the same patient at different sequencing depths have high consistency, and the correlation coefficient is gradually increased and stronger with the increase of the sequencing depth (S22-repeat 1 and S22-repeat 2 comparison: R0.959, 0.971, 0.978, 0.985 at the cut-off values of the sequencing depth of 1, 5, 10, 20, respectively; S23-repeat 1 and S23-repeat 2 comparison: R0.954, 0.967, 0.974, 0.982 at the cut-off values of the sequencing depth of 1, 5, 10, 20, respectively).

5. Factors affecting the DNA methylation level of AML patients

We calculated the methylation levels (DMI) of captured CpGs across the genome of 21 naive AML samples and analyzed the relationship of each clinical signature to DMI to determine if specific factors affected the DMI of the patients. The results show that DMI is not affected by bone marrow sample primary cell proportion, patient age, sex, FAB typing, cytogenetic risk grouping, molecular genetic risk grouping, and number of somatic mutations. It was also found that patients aged > 50 years had significantly higher DMI compared to patients aged <50 years (49.39% ± 3.43% vs. 46.75% ± 2.31%, p ═ 0.048) (fig. 1).

6. DNA methylation levels in different gene regions

We analyzed the DNA methylation levels of different gene regions, such as all CpG site capture regions (all CpGs), CpG islands (CpG islands), promoter regions (promoters), exon regions (exons), first exons (exon 1), introns (introns), enhancers (enhancers), 5 ' non-coding regions (5 ' untranslated regions, 5 ' UTR), etc. By comparing the methylation levels of these regions of bone marrow in treatment naive AML patients and normal controls, it was found that the DNA methylation levels (DMI) of the promoter and enhancer regions alone were significantly higher than those of the normal controls (p ═ 0.025 and p ═ 0.021) (fig. 2A); further, by analysis of 8 pairs of naive-remission paired samples, only promoter region DNA methylation levels were significantly lower than naive (p 0.018) while enhancer region DNA methylation levels did not change significantly (p 0.145) during induction remission (fig. 2B); the results of 3 additional treatment-relapse paired samples suggested that there was no significant difference in the level of DNA methylation in the promoter region at relapse from that at treatment (p ═ 0.305). The results show that the DNA methylation level of the promoter region of AML patients is higher than that of the normal control, the remission stage is remarkably reduced, and the DNA methylation level of the promoter region is increased again when the patients relapse, which indicates that the DNA methylation level of the promoter region is a remarkable characteristic of AML pathogenesis and is closely related to clinical treatment response.

To find functional differentially methylated genes, we first performed a Differential Methylation Region (DMR) comparison of the functional regions of the relevant genes (e.g., promoter, first exon, enhancer and 5 'noncoding region, i.e., promoter, exon 1, enhancer, 5' UTR) between the naive AML sample and the normal control sample. As a result, the promoter region was found to have the most differentially methylated regions (60.9%, 669/1099, p <0.001)) and to be predominantly hypermethylated differentially regions (5.0%, 502/669, p <0.001) (fig. 2C); from the perspective of integrating genetic and epigenetic prognoses, we further compared the Differential Methylation Regions (DMR) between different cytogenetic and molecular genetic prognosis groups for AML naive and annotated the corresponding differential methylation genes according to the DMR of the different gene regions, which showed that the most DMGs were annotated to the promoter region (p <0.001) (fig. 2D). The above results suggest that the promoter region has the most important functional differential methylation region and has a link to known genetic prognostic groupings. Screening for methylation prognosis-related genes based on differential methylation of promoter regions is of utility.

7. Differential methylation Gene screening results

In order to screen genes whose promoter regions are hypermethylated to cause poor patient prognosis, we determined a set of prognostic gene markers integrating genetics and epigenetics in order to develop a kit for determining the prognosis of AML, and we performed screening and validation. Firstly, dividing 21 cases of primarily treated AML subjected to MCC-Seq sequencing into different cytogenetic and molecular genetic prognosis groups according to NCCN guideline prognosis risk stratification standard, and comparing methylation differences of promoter regions among different prognosis groups, wherein the comparison direction is as follows: high-risk vs. medium-risk, medium-risk vs. low-risk, and high-risk vs. low-risk. The comparison among cytogenetics groups finds 100 high differential methylation genes (hyper-DMGs), and the comparison among molecular genetics groups finds 44 high differential methylation genes. Then, a method of 'taking intersection' is further adopted, namely, the hyper-DMGs in 100 hyper-DMGs which are commonly compared between cytogenetics groups and 44 hyper-DMGs which are compared between molecular genetics groups are selected, so that 18 hyper-DMGs with poor prognosis in cytogenetics group comparison and molecular genetics group comparison are obtained, and further, the characteristics of 18 selected genes are definitely screened by retrieving a Gene database (NCBI Gene database), retrieving literature reports of related genes, carrying out GO function analysis and KEGG channel analysis on the genes, and finally determining 12 genes for subsequent clinical significance verification. Of the 18 genes, 3 pseudogenes (GUCY1B2, HNRNPA1P33 and TUBA3FP) and 2 genes without molecular functions (MIR3150B and MIR4638) were deleted, and 1 gene without DNA methylation detection results in the TCGA database (PLEC). Finally, 6 genes are knocked out, a group of 12 functional differential methylated genes (BARD1, BCL9L, CLEC11A, DEFB1, FOXD2, IGF1, IL18, ITIH1, LSP1, P2RX6, RNASE3 and TUBGCP2) which are related to genetic prognosis and are regulated by DNA methylation are determined, and clinical significance verification is carried out in research cases and TCGA cases, so that the test strip has significance for judging AML prognosis and can be used for the development of follow-up prognosis stratification kits.

8. Clinical significance verification result of differential methylation gene

From the 12 DMGs screened, we calculated the DNA methylation level of each DMGs of each patient. The mean of the methylation levels of 12 DMGs in the same patient is taken as the integrated methylation value (M-value) of that patient, and is used to represent the methylation level of the patient. We then verified the clinical significance of M-value in 21 cases of MCC-Seq and 169 cases of TCGA.

It was found that M-value was not affected by patient age, sex and sample primary cell proportion, similar to the results where methylation levels of CpGs within the capture range were not affected by these factors (fig. 1).

We evaluated the relationship of M-value to clinical short-term efficacy, i.e., induced remission rate. The results show that the mean M-value for patients with complete remission is lower than that of those without remission (37.42% ± 15.79% vs. 49.69% ± 12.51%, p ═ 0.09). To further evaluate the relationship of M-value to CR, we demarcated 21 patients as the median of M-value, and divided them into low M-value groups, i.e., M-value ≦ median (n ═ 11); and a high M-value group, i.e., M-value > median (n ═ 10). In the low M-value group, up to 90.9% (10/11) of patients achieved CR with significantly higher remission rates than in the high M-value group (40.0%, 4/10; p ═ 0.024). In addition, 5 of 6 patients with intermediate-risk (IR-AML) and low M-value in prognosis reached CR with a remission rate of 83.3%, while only 3 of 8 patients with IR-AML and high M-value reached CR with a remission rate of only 37.5%. The above results suggest that patients with low M-value are more likely to acquire CR for all non-M3 AML and IR-AML.

Evaluation of M-value in relation to Long term efficacy, i.e., survival: we divided both study cases and TCGA database cases into low M-value and high M-value groups by their median M-value. First, the overall median Overall Survival (OS), median disease-free survival (DFS), 1-year cumulative OS, 1-year cumulative DFS were 23.8 months, unachieved, 78.9%, and 69.1% of 21 study cases, respectively. In the low M-value group and the high M-value group, the median OS was not reached and 14.93 months (p ═ 0.062), respectively; median DFS was not reached and 10.97 months (p ═ 0.039), respectively; the 1 year cumulative OS was 88.9% and 68.6%, respectively (p ═ 0.145); cumulative DFS was 90.9% and 30.0% for 1 year (p <0.001), respectively. Survival analysis showed that high M-value is a risk factor for DFS (HR: 6.83, 95% CI: 1.07-40.28; p ═ 0.039); for OS, a high M-value tends to be worse in prognosis. In the TCGA cases, the OS of patients in the high M-value group was significantly worse than in the low M-value group (median OS: 15.1months vs.16.4 months; HR: 1.491, 95% CI: 1.043-2.151, p ═ 0.038). For DFS, high M-value tends to have poor prognosis. Also, the 2-year cumulative OS and DFS for high M-value patients were significantly lower than those of the low M-value group (OS: 35.9% vs. 45.9%, p ═ 0.001; DFS: 30.2% vs. 44.2%, p < 0.001). The above results indicate that high M-value is a poor prognostic factor.

Reference to the literature

1.Chen,W.,et al.,Cancer statistics in China,2015.CA Cancer J Clin,2016.

2.Hasserjian,R.P.,Acute myeloid leukemia:advances in diagnosis and classification.Int J Lab Hematol,2013.35(3):p.358-66.

3.Walter,R.B.,et al.,Significance of FAB subclassification of"acute myeloid leukemia,NOS"in the 2008WHO classification:analysis of 5848newly diagnosed patients.Blood,2013.121(13):p.2424-31.

4.Prada-Arismendy,J.,J.C.Arroyave,and S.Rothlisberger,Molecular biomarkers in acute myeloid leukemia.Blood Rev,2017.31(1):p.63-76.

5.Papaemmanuil,E.,et al.,Genomic Classification and Prognosis in Acute Myeloid Leukemia.N Engl J Med,2016.374(23):p.2209-21.

6.Taylor,J.,W.Xiao,and O.Abdel-Wahab,Diagnosis and classification of hematologic malignancies on the basis of genetics.Blood,2017.130(4):p.410-423.

7.Dohner,H.,et al.,Diagnosis and management of acute myeloid leukemia in adults:recommendations from an international expert panel,on behalf of the European LeukemiaNet.Blood,2010.115(3):p.453-74.

8.Bennett,J.M.,et al.,Proposals for the classification of the acute leukaemias.French-American-British(FAB)co-operative group.Br J Haematol,1976.33(4):p.451-8.

9.Daniel A.Arber,et al.,The 2016revision to the World Health Organization classification of myeloid neoplasms and acute leukemia.Blood,2016.127(20):p.2391-405.

10.Bullinger,L.,et al.,Identification of acquired copy number alterations and uniparental disomies in cytogenetically normal acute myeloid leukemia using high-resolution single-nucleotide polymorphism analysis.Leukemia,2010.24(2):p.438-49.

11.Mardis,E.R.,et al.,Recurring mutations found by sequencing an acute myeloid leukemia genome.N Engl J Med,2009.361(11):p.1058-66.

12.Shen,Y.,et al.,Gene mutation patterns and their prognostic impact in a cohort of 1185patients with acute myeloid leukemia.Blood,2011.118(20):p.5593-603.

13.Gutierrez,S.E.and F.A.Romero-Oliva,Epigenetic changes:a common theme in acute myelogenous leukemogenesis.J Hematol Oncol,2013.6:p.57.

14.Figueroa,M.E.,et al.,Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype,disrupt TET2 function,and impair hematopoietic differentiation.Cancer Cell,2010.18(6):p.553-67.

15.Schoofs,T.,W.E.Berdel,and C.Muller-Tidow,Origins of aberrant DNA methylation in acute myeloid leukemia.Leukemia,2014.28(1):p.1-14.

16.Figueroa,M.E.,et al.,DNA methylation signatures identify biologically distinct subtypes in acute myeloid leukemia.Cancer Cell,2010.17(1):p.13-27.

17.Deneberg,S.,et al.,Gene-specific and global methylation patterns predict outcome in patients with acute myeloid leukemia.Leukemia,2010.24(5):p.932-41.

18.Deneberg,S.,et al.,Prognostic DNA methylation patterns in cytogenetically normal acute myeloid leukemia are predefined by stem cell chromatin marks.Blood,2011.118(20):p.5573-82.

19.Sun,Z.,et al.,Base resolution methylome profiling:considerations in platform selection,data preprocessing and analysis.Epigenomics,2015.7(5):p.813-28.

20.Allum,F.,et al.,Characterization of functional methylomes by next-generation capture sequencing identifies novel disease-associated variants.Nat Commun,2015.6:p.7211.

21.NimbleGen.www.nimblegen.com.

22.Walker,D.L.,et al.,DNA methylation profiling:comparison of genome-wide sequencing methods and the Infinium Human Methylation 450 Bead Chip.Epigenomics,2015.7(8):p.1287-302.

23.Cancer Genome Atlas Research,N.,Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia.N Engl J Med,2013.368(22):p.2059-74.

24.Krueger,F.and S.R.Andrews,Bismark:a flexible aligner and methylation caller for Bisulfite-Seq applications.Bioinformatics,2011.27(11):p.1571-2.

25.Lister,R.,et al.,Human DNA methylomes at base resolution show widespread epigenomic differences.Nature,2009.462(7271):p.315-22.

26.Akalin,A.,et al.,methylKit:a comprehensive R package for the analysis of genome-wide DNA methylation profiles.Genome Biol,2012.13(10):p.R87.

27.Li,S.,et al.,An optimized algorithm for detecting and annotating regional differential methylation.BMC Bioinformatics,2013.14 Suppl 5:p.S10.

Claims (2)

1. The application of a methylation reagent for detecting a combination of 12 acute myeloid leukemia DNA methylation prognosis markers in preparing an acute myeloid leukemia DNA methylation prognosis detection reagent, wherein the 12 acute myeloid leukemia DNA methylation prognosis markers are DNA methylation regulated functional differential methylation genes: BARD1，BCL9L，CLEC11A，DEFB1，FOXD2，IGF1，IL18，ITIH1，LSP1，P2RX6，RNASE3andTUBGCP2。

2. the use of claim 1, wherein the methylation level of 12 different methylated genes of the same patient is detected by a detection reagent, and the average value obtained is recorded as the methylation value M-value of the patient, which is used to represent the methylation level of the patient, wherein a larger value indicates a poor prognosis factor.

Images