Title: Novel real‐time PCR assay using a universal molecular marker for diagnosis of hematologic cancers
Abstract: Hematologic cancers, the third most common cause of cancer deaths, will kill 46,600 Americans this year.1 While most lymphomas and leukemias (LL), which consist of several types and subtypes, can be diagnosed by flow cytometric analysis of surface antigens, a large panel of markers (˜50) is required.2 In 5–10% of cases, an initial diagnosis of malignancy is in doubt. In addition, detection of minimal residual disease (MRD) after chemotherapy is of major clinical importance as it may predict future recurrence.2 Routine MRD assays such as histology and flow cytometry are of low sensitivity, while PCR assays require identification of the specific form of LL and may need individualized assay conditions.2, 3 Thus, a single assay for the diagnosis of all LLs and for the detection of MRD would be of major clinical benefit. Cytokines are secreted proteins that play a central role in the regulation of a wide array of cellular functions in the lymphohematopoietic system, including proliferation, differentiation and cell survival.4 Cytokines bind to and activate their cognate cell surface receptors. For certain cytokines, activated receptors recruit members of the Janus kinase (JAK) family of cytoplasmic tyrosine kinases, which in turn phosphorylate and activate transcription factors known as signal transducers and activators of transcription (STATs).4 Activated and dimerized STATs enter the nucleus and induce transcription of multiple genes having STAT recognition sites in their promoters. The pathway is negatively regulated by protein tyrosine phosphatases (SHP1) and by other proteins including SOCS1 and SYK,5 whose inactivation may result in up regulation. Upregulation of the JAK/STAT signaling pathway is frequent in LL.6 Epigenetic inactivation of tumor suppressor genes through aberrant methylation of their promoters is a very frequent event.7 While downregulation of expression of a candidate suppressor gene may be demonstrated by RNA or protein assays, these substances are labile and may not be of sufficient quality or quantity for detection in routine clinical samples. However, assays utilizing DNA (such as methylation-specific PCR [MSP]) are more robust and can be applied to a wide variety of samples, even those in alcohol fixatives. We initially determined the promoter methylation status of 3 negative regulators (SHP1, SOCS-1 and SYK) of JAK-STAT pathway in 90 LLs. using a standard PCR-based MSP assay. These initial analyses showed that SHP1 may be a near universal candidate marker for the diagnosis of malignancy. Thus, we developed a highly sensitive and specific semiquantitative real-time PCR assay for methylated SHP1 based on TaqMan probe technology. To develop real time PCR assay, LL samples (n = 90) were obtained from the Flow Cytometry Facility of the Department of Pathology at the UT Southwestern Medical Center, Dallas, TX and its affiliated hospitals, after receiving Institutional Review Board permission. In brief, composition of the LL samples was as follows: 36 B-cell lymphoma (Burkitt, diffuse large B-cell, follicular, marginal zone and mantle cell lymphomas); 6 T-cell lymphomas (large cell anaplastic, peripheral T-cell); 36 lymphoid leukemias (acute and chronic); and 12 myeloid leukemias (acute and chronic). We also obtained 26 control samples (peripheral blood mononuclear cells [PBMC], lymph node or bone marrow) from healthy volunteers and hematologic patients without cancers. In addition, we used 12 Epstein-Barr virus transformed B-cell lymphoblastoid (BL) lines from the PBMCs of healthy individuals that had been in continuous culture for about 5 months when tested. To further validate and check the reproducibility of the real-time PCR assay, an additional 44 LL samples (27 B-cell lymphoma, 2 T-cell lymphoma, 9 acute lymphocytic leukemia [ALL], 6 acute myelogenous leukemia [AML]) were obtained from the Department of Pathology at the UT Southwestern Medical Center, Dallas, TX and its affiliated hospitals. We also obtained 60 control samples: 40 PBMC obtained from healthy volunteers and 20 bone marrows obtained from hematologic patients without cancers. Genomic DNA was extracted from cell lines, primary tumors and nonmalignant cells by using a DNA extraction kit (Serological Corporation, Norcross, GA). Aberrant promoter methylation of SHP1, SOCS1 and SYK was determined on bisulfite treated DNA by MSP, as reported by Herman and Baylin,7 using primers for methylated sequences and methodologies previously described for SHP1,8 SOCS19 and SYK.10 Primer sequences specific for the unmethylated form of the p16 gene were used to check the integrity of bisulfite-modified DNA from tumors. DNAs from peripheral blood lymphocytes (n = 14) from healthy volunteers were used as negative controls for methylation specific assays. Genomic DNA, treated with Sss1 methylase (New England Biolabs, Beverly, MA) and after bisulfite modification was used a positive control and distilled water as a negative control. Sodium bisulfite-treated genomic DNA was amplified by fluorescence-based real-time MSP by using TaqMan technology (Perkin Elmer, Foster City, CA) as described previously.11 We performed real-time MSP assays with the Gene Amp 5700 Sequence Detection System (Perkin Elmer, Foster City, CA). In brief, oligonucleotide primers were designed to specifically amplify bisulfite-converted DNA within the promoter of the SHP1 gene and a probe was designed to anneal specifically within the amplicon during extension. The primers and probes for real time MSP of methylated SHP1 are as follows: 5′-AACGTTATTATAGTATAGCGTTCG-3′ (forward primer, nt 397–420); 5′-TCACGCATACGAACCCAAACG-3′ (reverse primer, nt 534–554); and 5′ 6-FAM AGCGTGGGTTAGGGAGGGTTGCG BHQ-1 3′ (probe, nts 397–420). For the internal reference gene, MYOD1, the primers and probe were designed to avoid CpG nucleotides. Amplification of MYOD1 occurs independent of its methylation status, whereas the amplification of SHP1 is proportional to the degree of cytosine methylation within the amplicon. The methylation ratio was defined as the ratio of the fluorescence emission intensity values for the SHP1 PCR products to those of MYOD1 PCR products multiplied by 100. This ratio was used as a measure for the relative level of methylated SHP1 alleles in the particular sample. Semiquantitative real-time PCR assays were performed in a reaction volume of 25 μl using components supplied in a TaqMan PCR Core Reagent kit. Separate amplification assays were performed for SHP1 and MYOD1; each assay was performed in duplicate. The final reaction mixture contained the forward and reverse primers at 600 nmol/L each; the probe at 200 nmol/L; 200 μmol/L each of the 4 nucleotide phosphates; 5.5 mmol/L MgCl2; 1× TaqMan Buffer A; 1U of Hot Star Taq DNA Polymerase (Qiagen, Valencia, CA) and 3 μl bisulfite-converted genomic DNA. PCR was performed under the following conditions: 95°C for 12 min, followed by 50 cycles at 95°C for 15 sec and 60°C for 1 min. We used serial dilutions of the SHP1 positive lymphoma cells with normal lymphocytes (0.001–10%) to create a standard curve. The frequencies of methylation between groups were compared using the χ2 test. Correlation value was analyzed by simple regression analysis. For all tests probability values of p < 0.05 were considered statistically significant. Initially, 90 LL samples were analyzed by standard conventional MSP for methylation of the 3 genes. SHP1 was methylated in 97% (87/90), SOCS1 in 24% (22/90) and SYK in 7% (6/90) of the cases (Fig. 1a). We also analyzed 26 control samples and 12 BL lines (as described above). All of the control samples and BL lines were negative for methylation of all 4 genes. All test and control samples showed distinct bands for unmethylated P16 indicating the integrity of the bisulfite treated DNA. To further confirm the frequency of methylated bands resulting from the PCR amplification of 25 LLs including 1 high-grade myelodysplastic syndromes (MDS) were gel purified directly sequenced (ABI sequencer; ABI, Foster City, CA) in forward and reverse directions using methylation specific primers described by Oka et al.8 Additionally, methylated bands from 5 methylation positive samples in 96-well real-time PCR plates were gel purified and directly sequenced (ABI sequencer) in forward and reverse direction using real-time PCR primers. Based on forward and reverse sequencing analysis all CpG sites (shown to be critical by Oka et al.8) were found to be consistently methylated in amplicons generated by primers used in standard PCR and real-time PCR analysis. To further confirm the specificity of methylation bisulfite, DNAs from 5 tumors (found to be positive based on the analysis described above) and 5 lymphocyte specimens from normal volunteers were subjected to PCR amplification using primers designed to exclude binding to any CpG dinucleotides to ensure amplification of both methylated and unmethylated forms. The resultant amplicons were cloned into plasmid vector using Topo TA cloning kit (Invitrogen, Carlsbad, CA). The analysis showed uniform methylation (rate of methylation, 54–73%) of the 11 CpG dinucleotides in the methylated sequences and no methylation (rate of methylation, 0%) of any of the CpG nucleotides in normal lymphocytes (Fig. 1b), further confirming the specificity of our assay. (a) Frequencies of methylation of JAK-STAT inhibitor genes (SHP1, SOCS-1 and SYK) in LLs. A total of 90 LLs were analyzed for methylation using standard MSP assay. (b) MSP analysis of CpG island in the SHP1 promoter (U15536), nucleotides 396 to 554. Methylation of this region has been shown to be critical in silencing of the gene expression.8 Closed circles represent methylated CpG, while open circles represent unmethylated CpG. The positions of the primers used in real-time PCR, standard PCR and sequencing are shown. The CpG dinucleotides that are part of the primers and probes are shown as boxed. LL: hematologic cancers; NL: normal lymphocytes. A total of 5 independent clones were analyzed for tumors as well as lymphocytes. The figure shows uniform methylation of the region in the LL (73% rate of methylation in the case shown here), while no methylation was detectable in normal lymphocytes. (c) Real time MSP assay standard curve CT against tumor cell percentage. CT represents the threshold cycle in the exponential phase of amplification. Closed circles represent CT values at different tumor cell percentages. Serial dilutions of SHP1 positive lymphoma cells with normal lymphocytes (0.001–10%). (d) Correlation between real-time MSP quantitative ratio and tumor cell percentage values. Open circles represent quantitative ratios at different tumor cell percentages. Tumor cell percentages were determined by flow cytometric analysis. (e) Comparison between standard MSP assay (represented by presence or absence of amplicon bands in the gel analysis) and real-time MSP assay (represented by quantitative ratios above the gel). MSHP1 and umSHP1 represent methylated and unmethylated forms of the gene, respectively. CTS represents CT value for SHP1, while CTS represents CT value for MYOD1. These data show that the CT values for MYOD1 among treated and untreated samples (A,B,C) were close, while the CT values for SHP1 in these samples were significantly different. A–F: 6 lymphoma samples; 1: before chemotherapy; 2: after chemotherapy. Based on the standard MSP assay, SHP1 showed potential as a universal marker for LLs. Thus, we developed a real-time PCR assay to quantitate methylated SHP1 from LLs. When real-time data for different percentages of LL tumor cells mixed with normal lymphocytes were compared we observed a linear relationship from 10% to 0.001% (R = 0.99) (Fig. 1c). The assay was repeated and the results were found to be reproducible. Thus, using our real-time PCR technique, it may be possible to reliably detect and quantitate methylated SHP1 up to 1 methylated cell in 100,000 nonmethylated cells. To further validate our assay, 45 LLs (for which tumor cell percent estimate was available from flow-cytometric analysis) were analyzed for methylated SHP1 by real-time PCR assay. These 45 LLs include 3 LLs that were methylation-negative by standard PCR assay. The tumor cell percentage in these tumors ranged from 2% to 91%. The tumor cell percent in the 3 methylation-negative tumors ranged from 70% to 90%. The real-time PCR analysis clearly showed excellent correlation (R = 0.87; p < 0.0001) between the quantitative ratio (fluorescent unit) for methylated SHP1 and tumor cell percent estimate. Analysis of 26 control samples and 12 BL lines were below the lowest level of detection. All the 42 LL tumor samples (Fig. 1d) were positive by both standard and real-time MSP assays. All the 26 control samples and 12 BL lines and methylation-negative tumors were negative by both MSP assays. Thus, in these 83 samples (tumors and normal controls) there was 100% concordance between the 2 assays. The quantitative ratio values for real-time MSP assay of the 42 LL tumor samples ranged from 2.5 at a tumor cell percentage of 2% to 62.4 at a tumor cell percentage of 91% (mean value, 37.3). The real-time PCR analysis clearly showed excellent correlation (R = 0.87; p < 0.0001) between the quantitative ratio (fluorescent unit) and tumor cell percent estimate. To further check the specificity and reproducibility of the real-time PCR assay, additional 44 LLs and 60 control samples were obtained and analyzed for methylated SHP1. Our results showed 41 of 44 LLs (93%) were methylated, with the quantitative ratio values for real-time MSP assay at a mean value of 28.3. All the control samples were below lowest level of detection. These results further validated the specificity and reproducibility of our assay. Based on the standard curve (Fig. 1c) using our real-time PCR technique, it may be possible to reliably detect and quantitate methylated SHP1 up to 1 tumor cell in 100,000 cells. Our standard MSP assay can detect methylated SHP1 up to 1 tumor cell in 10,000 cells. Thus, our real-time MSP assay appears to be at least 10 times more sensitive than the conventional MSP assay This increased sensitivity was further evident when we could not detect methylated SHP1 by standard MSP or flow cytometric analysis in bone marrow from 3 of 8 lymphoma cases that had undergone chemotherapy, while we could detect it by real-time MSP assay. (Fig. 1e). Additionally, real-time analysis of 14 MDS showed 6 of 14 (43%) samples positive for methylated SHP1 with quantitative ratios ranging from 35.7 to 0.21. A total of 3 of the 6 MDS were high grade lesions and had higher ratios than others. The MDS with quantitative ratio of 0.21 could be detected by real-time PCR but not by standard PCR. This observation further demonstrates the increased sensitivity of the real-time PCR assay. Based on our standard MSP assay, SHP1 was found to be a candidate universal marker for the diagnosis of malignancy and identification of minimal residual disease. We developed a real-time PCR assay for methylated SHP1 that can quantitate 1 copy of the methylated gene in 100,000 copies of unmethylated genes, while standard PCR assay could detect 1 copy of aberrant gene in 10,000 copies of normal genes. This assay was found to be highly specific based on 100% concordance between standard PCR data and real-time PCR data from the analysis of LLs and control tissues. Sensitivity of the assay was further demonstrated by detection of methylated SHP1 in cases in apparent remission, only by real-time PCR assay and not by conventional PCR assay. Sensitivity of the assay was also apparent in detection of methylated SHP1 in potentially premalignant stages (MDS) of leukemia by real-time PCR assay. Our study shows that real-time PCR quantitation of the methylated SHP1 may have significant potential for diagnosis of hematologic cancers and identification of minimal residual disease. Yours sincerely, Narayan Shivapurkar, Victor Stastny, Takao Takahashi, Makoto Suzuki, Chinyere Echebiri, Jyotsna Reddy, Adi F. Gazdar.