Cancer Research SABCS Telomeres
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sun, B.
Right arrow Articles by Ehrlich, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sun, B.
Right arrow Articles by Ehrlich, M.
[Cancer Research 61, 6931-6937, September 15, 2001]
© 2001 American Association for Cancer Research

Molecular Biology and Genetics

ABL1 Promoter Methylation Can Exist Independently of BCR-ABL Transcription in Chronic Myeloid Leukemia Hematopoietic Progenitors1

Baodong Sun, Guanchao Jiang, Muhammad-Ali A. Zaydan, Vincent F. La Russa, Hana Safah and Melanie Ehrlich2

Tulane Cancer Center [B. S., M-A. A. Z., V. F. L. R., H. S.], Human Genetics Program [B. S., G. J., M. E.], and Departments of Medicine [V. F. L. R., H. S.], and Biochemistry [M. E.], Tulane Medical School, New Orleans, Louisiana 70112


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Formation of the hybrid BCR-ABL gene is responsible for >95% of chronic myeloid leukemia (CML). The alternative, downstream ABL promoter (Pa), which is usually retained in this chimeric oncogene, was reported to be methylated in many CML patients, but there has been controversy as to whether this methylation is a frequent change in bone marrow (BM) in early chronic phase (CP) or only past this stage. Also, the relevance of Pa promoter methylation to BCR-ABL expression in CML is unclear. We examined methylation of the ABL Pa promoter in uncultured BM samples and in colonies derived from their hematopoietic precursor cells by bisulfite and PCR-based assays (combined bisulfite restriction analysis and methylation-specific PCR). BM from seven CP CML patients at diagnosis had about 20–60% of the copies of the ABL Pa promoter methylated. No Pa methylation was detected in normal BMs or colonies derived from them. In contrast, most colonies from CP CML patients had Pa methylation. Surprisingly, 18–49% of the CML-derived colonies with this methylation reproducibly had no detectable BCR-ABL RNA on nested reverse transcription-PCR. Furthermore, the percentage of BCR-ABL RNA-positive colonies was almost same among the colonies not displaying Pa methylation as among the colonies in which this methylation was found. We conclude that ABL Pa methylation is often an early marker of CML in hematopoietic precursors and in total mononuclear BM cells but that it is not associated with an increased frequency of BCR-ABL RNA-positive cells. This methylation might be emblematic of cancer-associated hypermethylation elsewhere in the genome with the consequent silencing of tumor suppressor genes seen in many malignancies.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The t(9;22; q34;q11) translocation that produces the activated oncogene-containing Ph chromosome3 is responsible for >95% of the cases of CML (1, 2, 3) . This oncogene contains the 5' end of the BCR gene fused to the second exon, exon 1a, of the ABL gene. It encodes a fusion protein, BCR-ABL, which, like ABL, is a tyrosine kinase. However, the fusion protein has altered kinase activity, which is the source of its oncogenicity by affecting signal transduction pathways and gene expression (4 , 5) . The normal ABL gene is constitutively expressed with transcription initiating either from the Pa promoter or the upstream Pb promoter (6) . These transcripts specify functional proteins, albeit with different NH2-termini. The ABL Pa promoter is usually retained within the chimeric gene on the Ph chromosome (7) . Ben-Yehuda et al. (8) have proposed that transcription from Pa could interfere with expression of the BCR-ABL product from the upstream BCR promoter and that Pa methylation might abrogate this down-regulation and thereby facilitate blastic transformation.

Methylation of CpGs, especially in promoter regions, can repress gene expression in vivo (9) , including in cancers. Abnormal promoter hypermethylation is frequently one of the steps of oncogenesis (10, 11, 12) . Methylation of the ABL Pa promoter has been found in CML BM and PB samples but not in analogous samples from normal individuals nor in patients with acute lymphocytic leukemia, acute myelogenous leukemia, lymphoma, or myelodysplastic syndrome (8 , 13 , 14) . There is controversy about the frequency of this methylation in CML. Ben-Yehuda et al. (8) reported that only 26% of 148 CP CML patients displayed ABL Pa methylation as assessed by the ability of this DNA region to be amplified by PCR after digestion with CpG methylation-sensitive restriction endonucleases. In contrast, Issa et al. and Nguyen et al. (13 , 14) found that 78% (73/93) and 77% (10/13) of CP patients were positive for such methylation by Southern blot analysis or a bisulfite and PCR-based assay (methylation-sensitive single nucleotide primer extension). Ben-Yehuda et al. (8 , 15) observed large increases in Pa methylation correlated with the stage of disease in analyses of uncultured BM and PB samples and in derivative hematopoietic colonies. Recently, Issa et al. reported only modest increases (13) and Nguyen et al. reported no increase (14) in Pa methylation with the stage of the disease in studies of blood and BM. For evaluation of the hypotheses that this methylation contributes to blastic transformation and may be a predictor of the response to IFN-{alpha} therapy (8 , 15) , it is important to assess the frequency of ABL Pa methylation and its relationship to BCR-ABL RNA levels in hematopoietic colonies as well as in uncultured BM and PB samples. These colonies arise from hematopoietic progenitor cells, which are the repository for maintenance and progression of CML.

In the present study, we examined hematopoietic colonies both for their ABL Pa methylation status and the presence of BCR-ABL transcripts. During the course of this study we established conditions for RT-PCR and methylation analysis from individual colonies that give reproducible results circumventing problems described previously with stochastic effects on PCR of hematopoietic colonies (16) . We also quantitated ABL Pa methylation in BM samples from CP CML patients and correlated our findings with the relative concentration of BCR-ABL transcripts in these samples and with clinical cytogenetic observations. We confirmed the high incidence of Pa methylation in CP BM samples and the lack of such methylation in normal BM (13 , 14) . However, unexpectedly, many BM-derived hematopoietic colonies from CML patients displayed ABL Pa promoter methylation but not BCR-ABL transcripts.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Clinical Samples, Colony Formation, and DNA Isolation.
BM samples were from 10 early CP CML patients (7 female and 3 male; ages 42–68). Patient CML10 had been treated with IFN-{alpha} and low-dose AraC for 6 months. The only PB sample was a stem cell pheresis product from a patient (CML9) after 8 months of IFN-{alpha}/AraC therapy and then treatment with idarubicin/AraC and granulocyte colony-stimulating growth factor (protocol approved by the Investigational Review Board of Tulane Medical Center). The other patient samples were obtained at diagnosis. The CML patients and four healthy BM donors (two female and two male, ages 9, 31, 42, and 43) gave informed consent for study of their tissues. The MNC fraction was immediately plated for colony formation with a 14-day incubation in methylcellulose medium containing GM-CSF, stem cell factor, erythropoietin, and interleukin 3 (MethoCult GF H4434, Stem Cell Technologies, Vancouver, British Columbia, Canada) or frozen for viability and later plated and incubated for 14 or 21 days. Only well-separated colonies with the typical features of either CFU-GM or BFU-E colonies were harvested. As a control for PCR, methylcellulose without cells was aspirated and processed analogously. Genomic DNA was isolated from the MNC by standard methods (DNeasy kit; Qiagen, Valencia, CA).

In Vitro Methylation of Normal DNA.
Normal lymphocyte DNA (12 µg) was incubated with 120 units of M.SssI CpG methylase (New England Biolabs, Beverly, MA) and 320 µM AdoMet at 37°C in a 1.2-ml reaction volume. After 3 h, an equal amount of AdoMet was added, and incubation continued for 2 h. The extent of in vitro methylation was determined by methylation in parallel of a plasmid containing a single SmaI site, digestion with SmaI, electrophoresis, and analysis of the ratio of signal in the intact plasmid band to the total signal. This ratio was 0.95, and so we estimated the extent of in vitro methylation of the ABL Pa promoter in normal lymphocyte DNA, which is unmethylated at tested sites in vivo (14) , to be ~95%.

Bisulfite Modification of DNA.
BM or PB DNA (at least 2 µg in 25 µl of TE buffer) was treated with bisulfite (17) at 55°C for 4 h and 2 µg of salmon sperm DNA was used as the carrier for ethanol-precipitation. The DNA was dissolved in 50 µl of 10 mM Tris (pH 8)-1 mM EDTA and stored at -20°C for up to 1 week before PCR. For hematopoietic colonies, cells from half of a colony were incubated for 60 min at 37°C in 25 µl of water containing 2 µg of salmon sperm DNA, 1 mM SDS, and 280 µg/ml proteinase K (18) ; heated in a boiling water bath for 15 min under mineral oil; and treated with bisulfite as described above.

COBRA Analysis of ABL Pa Methylation.
PCR was performed on 10 µl of bisulfite-treated BM or PB DNA in 50 µl of reaction mixture containing 1.5 mM MgCl2, 0.2 mM of each deoxynucleotide triphosphate, 2.5 units of Taq polymerase (HotstarTaq, Qiagen), and 150 ng each of COBRA1 and COBRA2 primers4 for the noncoding strand of the Pa promoter (Fig. 1Citation ; Table 1Citation ). DNA was denatured at 94°C for 15 min; amplified for 35 cycles of 94°C for 45 s; 48°C for 45 s; 72°C for 45 s; and incubated at 72°C for 10 min. An aliquot was electrophoresed to check that only the expected 281-bp band was seen. The DNA was purified from the PCR mixture (Qiagen) and digested with 4 units of TaqI (New England BioLabs) at 65°C for 3 h, electrophoresed on a 2% agarose gel, and the ethidium bromide fluorescence quantitated by digital analysis. Water or methylcellulose extract served as a negative control for each PCR set for COBRA and MSP. A 1:1 mixture of M.SssI-methylated lymphocyte DNA and the analogous untreated DNA served as the positive control. The percentage of methylation is expressed as the ratio of signal in the 162-plus 113-bp bands divided by that in the 162- and 113-plus 281-bp bands (Fig. 2)Citation and by the correction factor 0.89 (Ref. 19 ; Fig. 3Citation ). A plasmid (0.5 µg) was incubated with 10 µl of each PCR product-containing TaqI digestion mixture and shown to be completely digested.



View larger version (67K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. The ABL Pa promoter region and sequences within it that are complementary to the primers for MSP and COBRA after bisulfite modification. The two major transcription start sites (a and b) for expression from the Pa promoter (6) are indicated by broken arrows, and primers are denoted by horizontal arrows. Sequences surrounding Pa transcription start sites are shown with position numbers given from GenBank accession no. U07563. In COBRA and MSP, unmethylated C residues are converted to U and then amplified to T residues so the primers for MSP and COBRA (Table 1)Citation have C-to-T or G-to-A substitutions relative to the corresponding sequence in this figure. The two underlined and italicized four-base sequences denoted Pre-TaqI are the sites at which methylation was assayed in COBRA (Fig. 2)Citation .

 

View this table:
[in this window]
[in a new window]

  		Table 1 Sequences of primers for COBRA, MSP, and RT-PCR

 


View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. Scheme for analysis of ABL Pa methylation by COBRA. The indicated sequence is 133–142 bp upstream of the proximal major transcription initiation site of the ABL Pa promoter (Fig. 1)Citation . This sequence is amenable to COBRA, because bisulfite treatment, followed by incubation with alkali, will convert the two CCGA sites to UCGA sites that become TCGA (TaqI) sites on PCR but only if they were methylated in vivo. If there were methylation of both sites, 162- and 113-bp fragments would be obtained after incubation with TaqI. If only one or the other of the two CpGs in this sequence were methylated, then 113- and 168- or 119- and 162-bp fragments would be generated. On agarose gel electrophoresis, 113- and 119-bp fragments could not be distinguished nor could 162- and 168-bp fragments.

 


View larger version (41K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Standardization of COBRA assays for ABL Pa promoter methylation. Calibration of COBRA was done on mixtures of various ratios of in vitro methylated and untreated normal lymphocyte DNA containing the indicated percentage of ABL Pa molecules methylated. M, 100-bp ladder of marker DNAs. A, gel electrophoresis analysis of the mixtures after TaqI digestion of the bisulfite-treated and amplified DNA. B, the expected and deduced COBRA curves. The small difference between these curves led to a correction factor of 0.89 that was used for the results obtained from CML BM samples.

 
MSP Analysis of ABL Pa Methylation.
For BM and PB samples, one round of PCR was performed on 5 µl of bisulfite-treated DNA using MSP1-unmeth and MSP2 primers or MSP1-meth and MSP2 primers (Table 1Citation ; Ref. 15 ). PCR was done as for COBRA except that the Tanneal was 58°C and 34 cycles were used for the former, unmethylation-specific primer pair. For the latter, methylation-specific primer pair, reagents that improved the efficiency of PCR, 25 mM of tetramethylammonium chloride (Sigma Chemical Co.) and 1 x Q solution (Qiagen), were added; 54°C was the Tanneal, and PCR was for 36 cycles. After electrophoresis on a 1.5% agarose gel, ethidium bromide fluorescence was quantitated, and the ratio of the intensity of the band from a given sample amplified with the methylation-specific primer pair and that from the same sample analogously amplified with unmethylation-specific primer pair was calculated. MSP analysis of the ABL Pa promoter in colonies was done by seminested PCR on half of a colony and the other half was subjected to RT-PCR. The first-round product (2 µl) from MSP2 and MSP1meth primers was subjected to 32 more cycles of PCR with MSP3 and MSP1meth primers (54°C, Tanneal). Aliquots from colonies that gave no band with the methylation-specific primer pair were always assayed with the unmethylation-specific primer pairs (MSP1unmeth and MSP2 followed by MSP1unmeth and MSP3; 58°C, Tanneal); colonies that gave no band with either set of primers were discarded from the analysis. For one of the untreated patients, PCR was done with both unmethylation-specific and methylation-specific primer pairs on all of the colonies.

RNA Isolation and RT-PCR for BCR-ABL RNA.
RNA was extracted (20) from MNC (2–5 x 106) and reverse transcribed with M-MLV reverse transcriptase (300 units; Life Technologies, Inc., Gaithersburg, MD) and random hexamer primers in a 40-µl reaction. PCR was done as described above but with primers NB+ and ABL3- (Table 1Citation ; Ref. 21 ); 4 µl of the reverse-transcriptase reaction product; and 32 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 45 s. Samples that were negative for the BCR-ABL transcript were amplified with primers A2N (for exon 2 sequences) and A4- (for exon 4 sequences; Table 1Citation , 58°C, Tanneal)5 for the ubiquitously expressed ABL transcript. For quantitation of BCR-ABL transcripts, competitive RT-PCR to determine equivalence points was done with ABL as an internal standard, as described previously (22) . For the MNC sample that gave no BCR-ABL product after the first round of PCR (CML9), nested PCR was done with a 200-fold dilution of the first-round product and primers B2A and CA3- (Table 1Citation ; 30 cycles; 60°C, Tanneal); competitors were added only during the first round. For BCR-ABL RT-PCR on colony samples, RNA was isolated after lysing the cells in 0.5 ml of denaturing buffer (20; 4 M guanidium thiocyanate, 5 mM EDTA, 25 mM sodium citrate, 0.5% N-lauroyl sarosine) containing 4 µg of Escherichia coli tRNA carrier. cDNA was prepared and nested PCR was done as above in 25-µl reactions using 5 µl of the reverse-transcriptase reaction product and, in the second round of PCR, 2 µl of the first-round product. For BCR-ABL RNA-negative colonies, two rounds of PCR (30 cycles each) were performed with the ABL primers. The colonies that did not give the expected product with ABL primers (~10% of colonies) were excluded from the analysis.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
ABL Pa Promoter Methylation Analyses.
To standardize conditions for quantitation of methylation of the ABL Pa promoter (Fig. 1)Citation in BM and PB samples, we needed analogous methylated and unmethylated human DNA samples with known amounts of methylation in this promoter region. First, we prepared normal lymphocyte DNA exhaustively methylated at CpG sites with M.SssI, which was deduced to be ~95% methylated at its CpGs, as described in "Materials and Methods." The unmethylated standard was the untreated normal lymphocyte DNA, which we showed to be unmethylated at the site monitored by COBRA (Figs. 2Citation and 3Citation ). COBRA and MSP assays of DNA methylation involve bisulfite modification and PCR and rely on the resistance of 5-methylcytosine residues and the sensitivity of unmethylated C residues to bisulfite modification (23 , 24) . On treatment with bisulfite and then alkali, an unmethylated C residue is converted to U, on which PCR becomes T.

COBRA is a quantitative assay of C methylation in which restriction sites can be created because of C->T conversions only at unmethylated C residues (25) . The bisulfite and alkali treatments and PCR are followed by digestion with the appropriate restriction endonuclease. In the ABL Pa promoter there is a 5'-CCGAGGCCGA-3' sequence amenable to methylation analysis by COBRA with TaqI (Fig. 2)Citation . With various mixtures of the M.SssI-methylated and untreated lymphocyte DNA preparations, we showed that this COBRA assay was quantitative (Fig. 3)Citation . There was no appreciable bias for amplification of the unmethylated or the methylated ABL Pa sequence, a complication that is sometimes found when other DNA sequences are subjected to bisulfite treatment and PCR even using primers from CpG-deficient regions of the DNA (19) .

MSP assays are usually done in a nonquantitative manner. In MSP, at least one of the primer regions on the genomic DNA contains multiple CpGs (e.g., the region for primer MSP1-meth; Fig. 1Citation ) so that a methylation-specific primer and an unmethylation-specific primer (e.g., MSP1-unmeth and MSP1-meth, Table 1Citation ) can be designed that will anneal to this sequence after bisulfite treatment depending on its state of methylation. The second primer for MSP of the Pa promoter (Fig. 1Citation , primer MSP2 or MSP3) had no CpGs so that it could be used in both methylation-specific and unmethylation-specific amplifications. Unlike COBRA, MSP involves the comparison of products from two exponential amplifications. To make MSP analysis of ABL Pa promoter methylation semiquantitative, we varied the cycle number for amplification of ~0.2 µg of bisulfite-treated DNA that was a 1:1 mixture of M.SssI-methylated and untreated normal lymphocyte DNA using either the methylation-specific or unmethylation-specific primer pairs. Comparison of the amount of PCR product from the methylation-specific reaction for 32, 34, or 36 cycles and the nonmethylation-specific reaction that had been amplified for two cycles less gave the same results, namely, that 51 ± 2% of the assayed Pa promoter sequence was methylated. That these conditions permit semiquantitative MSP was confirmed with different ratios of M.SssI-methylated and untreated normal lymphocyte DNA (Fig. 4)Citation .



View larger version (33K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 4. Standardization of MSP assays for ABL Pa promoter methylation. Mixtures of various ratios of in vitro-methylated and untreated normal lymphocyte DNA were prepared to calibrate MSP with primers MSP1meth plus MSP2 for 36 cycles or MSP1unmeth plus MSP2 for 34 cycles. A, gel electrophoresis analysis of the mixtures after amplification of the DNA with the methylation-specific primer pair or with the primer-pair specific for the same sequences when they are unmethylated at their CpGs. B, the empirical MSP curve. Because of the close correspondence between this calibration curve and the expected results, no correction factor was used for the semiquantitative MSP results from CML BM samples.

 
Quantitation of Methylation of the ABL Pa Promoter Region versus BCR-ABL Gene Expression in CP CML BM Samples.
Using both COBRA and MSP, methylation of the ABL Pa promoter in BM samples obtained from four normal BM donors, seven untreated early CP CML patients, and one IFN-{alpha}/AraC-treated patient was quantitated. A stem cell pheresis PB sample of a IFN-{alpha}/AraC- and idarubicin/AraC-treated patient was also analyzed (Table 2)Citation . The normal BM samples showed no evidence of methylation of this promoter by either assay (Fig. 5)Citation . In contrast, all BM samples from the CML patients had appreciable methylation of this promoter at the assayed sites (Fig. 5)Citation . The results from semiquantitative MSP and COBRA were generally in agreement about the approximate percentage of copies of the Pa promoter that were methylated (Table 2)Citation .


View this table:
[in this window]
[in a new window]

  		Table 2 Comparison of cytogenetics, BCR-ABL RT-PCR, and methylation analyses (COBRA and MSP) on CML samples

 


View larger version (60K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 5. Analysis of ABL Pa methylation in normal BM samples and representative CML BM samples by COBRA and MSP. A, COBRA analysis of Pa methylation in CML BM samples (2, 10, 7, and 3) and normal BM samples (1 2 3 4) . The 162- and 113-bp bands for CML2 and CML10 were clearly seen in the original gel. B, MSP analysis of Pa methylation in normal BM samples using methylation-specific or methylation-unspecific primer pairs and 36 or 34 cycles, respectively. No bands were visible for normal BM samples amplified with the methylation-specific primer pair. C, MSP analysis of Pa methylation in the same CML samples as in A. Std. is the standard mixture of in vitro-methylated and untreated lymphocyte DNA with ~50% of the copies of the ABL Pa promoter methylated (Figs. 3Citation and 4Citation ) and M is the marker DNA.

 
The similar percentages of methylation deduced from MSP and COBRA suggest that methylation is often widespread along the ABL Pa promoter region as concluded from genomic sequencing of CML patient samples in a previous study of accelerated and blast phase patients (15) . The COBRA assay monitored methylation of sequences in the Pa promoter at position -133 to -142 (all positions given relative to the downstream Pa transcription start site; Ref. 6 ). At least one of the two CpGs at the assayed sequence should be methylated for the sequence to score as methylated in the COBRA assay (Fig. 2)Citation . For MSP, most or all of the seven CpGs at the genomic sequence from position -16 to +14 (MSP1 primer region, Fig. 1Citation ) should be methylated if the bisulfite-treated sample is to hybridize specifically to the MSP1-meth primer and not to the MSP1-unmeth primer and thereby score as methylated. Despite the requirement for more m5C residues at the assayed sequence in the MSP than in the COBRA assay, similar results were obtained from these two analyses (Table 2)Citation suggesting that when the promoter region is methylated, it is usually extensively methylated.

The frequency of methylation of the Pa promoter in these CML samples was compared with the steady-state level of BCR-ABL transcripts determined by competitive RT-PCR, as illustrated in Fig. 6Citation . The constitutively expressed transcripts initiated from either the Pa or Pb promoter from the unrecombined ABL gene were the internal standard. Both of these will be amplified by the ABL RT-PCR primer pair. All of the samples from patients having a ratio of leukemic BCR-ABL transcripts to normal ABL transcripts of >=0.15 had about 20–60% of the copies of their Pa promoter methylated at the assayed sequences (Table 2)Citation . By cytogenetic analysis of unstimulated BM, those patients had >=45% Ph+ cells, which is consistent with their high level of BCR-ABL transcripts. As expected, only the PB sample from CML9 obtained after treatment with idarubicin/AraC and the BM sample from patient CML10 obtained after treatment with IFN-{alpha}/AraC showed low levels of BCR-ABL transcripts as well as the absence of detected Ph-positive cells on routine cytogenetic analysis (Table 2)Citation . The former had no detectable methylation, and the latter had only ~5% of the copies of the Pa promoter methylated.



View larger version (36K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 6. Quantitation of BCR-ABL transcripts by competitive RT-PCR: a representative analysis. A, RT-PCR with BCR-ABL primers on CML1 BM (Table 2)Citation . B, RT-PCR on the same sample with ABL primers as an internal control. The number of molecules of competitor plasmid added before PCR is indicated above the Lanes for A. For this sample, the equivalence point, the concentration of competitor DNA at which the competitor and sample-derived PCR product bands are of equal intensity (22) , was estimated to be ~3 x 103 molecules for BCR-ABL transcripts and 1 x 104 for the internal control, ABL transcripts. Therefore, the BCR-ABL/ABL ratio for this sample was 0.3 (Table 3)Citation . The negative control that was included in this set of assays was electrophoresed on another gel.

 
Comparison of Methylation of the ABL Pa Promoter and Expression of BCR-ABL Transcripts in CFU-GM and BFU-E Colonies from CML Samples.
Because concerns have been raised about the reliability of PCR assays on hematopoietic precursor-derived colonies (16) , we first checked the reproducibility of such assays. For each analyzed BM-derived CFU-GM colony (~800–1,500 cells/colony) or BFU-E colony (~8,000–10,000 cells/colony), we assayed half of the cells by MSP and half by BCR-ABL RT-PCR by amplifying 1/10 of the cDNA or bisulfite-treated genomic DNA. Triplicate MSP amplifications of DNA from three Pa methylation-positive CML colonies and three Pa methylation-negative colonies were uniformly positive or negative, and the positive replicate samples gave PCR product bands of similar intensities. Also, the same reproducibility among replicate aliquots from a given colony was observed in triplicate or quintuplicate amplifications of cDNA for BCR-ABL using 12 CML colonies. Furthermore, we tested the reproducibility of the RT-PCR not only by doing replicate amplifications on cDNA from a colony, but also by isolating RNA in quintuplicate from half of the cells of a colony and preparing replicate cDNA samples for RT-PCR. For this analysis we used five CFU-GM and five BFU-E colonies from BM of an untreated CML patient. All of these colonies were found to be positive for the BCR-ABL transcript when amplifying one-fourth of the cDNA product from each of these quintuplicate RNA preparations. The reproducibility of these findings was established by showing that whenever an RNA preparation was verified to be intact by amplification of the control ABL transcript, which was the case for 3–5 of the replicate aliquots from each of the colonies, the RT-PCR for BCR-ABL was also positive. In the other replicate RNA preparations where the BCR-ABL primers gave no PCR product, the ABL primers also gave no product, probably indicating a problem with RNA preparation on such a small scale, smaller than we routinely used. Such ABL RNA-negative samples were never incorrectly scored as BCR-ABL-negative, because in Tables 3Citation and 4Citation we included colonies that were negative for BCR-ABL RNA only if they were positive for ABL RNA. Only 1 exceptional aliquot of the 50 tested was negative for BCR-ABL RNA and positive for ABL RNA, whereas the replicate aliquots from the same colony were positive for both RNA species. Therefore, although we were examining BCR-ABL expression and ABL Pa methylation in rather small numbers of cells, we did not have previously reported problems with stochastic effects on RT-PCR from colony samples (16) nor on MSP.


View this table:
[in this window]
[in a new window]

  		Table 3 Analysis of hematopoietic progenitor-derived colonies from CML BM for ABL Pa methylation and BCR-ABL (Ph chromosome) RNA

 

View this table:
[in this window]
[in a new window]

  		Table 4 CML-derived colonies with methylation of ABL Pa are no more likely to express BCR-ABL RNA than are colonies without this methylation

 
MSP and RT-PCR were done on about 20–40 colonies each from three BM samples from untreated CML patients who were strongly cytogenetically Ph-positive, a BM sample of a patient in complete cytogenetic remission after treatment with IFN-{alpha}/AraC (CML9 BM), and a stem cell pheresis PB sample obtained from this patient after additional treatment with idarubicin/AraC (CML9 PB). For the three untreated patients, 65–100% of the CFU-GM colonies were positive for BCR-ABL transcripts (Table 3)Citation . About 40–60% of their BFU-E colonies were positive for BCR-ABL RNA. For the cytogenetically negative PB and BM samples from the treated patients, about 20–40% of the CFU-GM or BFU-E colonies contained BCR-ABL transcripts (Table 3)Citation .

Most of the colonies from the untreated patients gave the expected size band with the methylation-specific primer pair (Table 3)Citation indicating Pa promoter methylation. None of the 60 assayed colonies from normal BM samples displayed any Pa methylation. When duplicate aliquots of DNA from CML colonies with methylated copies of the Pa promoter were amplified with the unmethylation-specific primer pair, they always gave the same size product indicating that both unmethylated and methylated copies of the Pa promoter were present. Therefore, unmethylated copies of the Pa promoter were present along with the methylated copies; the colony assays did not allow quantitation of their relative levels within a colony.

The percentages of colonies from CML6 and CML7 that were positive for Pa methylation were even higher than the percentage of Pa molecules that were methylated in the corresponding uncultured BM samples (Tables 2Citation and 3Citation ). However, in comparing these results it should be noted that only a very small percentage of the cells in the MNC fraction of BM are clonogenic hematopoietic progenitors. In contrast with the results obtained from untreated patient samples, methylation of the Pa promoter was seen in only a small number of colonies after IFN-{alpha}/AraC treatment and not in any colonies after idarubicin/AraC treatment (Table 3)Citation .

A surprising result from the comparison of the ABL Pa methylation and the BCR-ABL transcript status of CML BM-derived colonies was that about 15–40% of the total examined colonies from three BM samples obtained at diagnosis contained Pa methylation but no detectable BCR-ABL RNA even after two rounds of PCR on the cDNA (Table 3Citation , column D). Moreover, the percentage of BCR-ABL RNA-positive colonies was no higher among the colonies in which the Pa promoter was methylated than in the colonies in which it was not (Table 4)Citation . The RT-PCR and MSP results from three colonies that were BCR-ABL RNA-negative and Pa methylation-positive and two colonies that were BCR-ABL RNA-positive and Pa methylation-negative were among those verified by quintuplicate RT-PCR and triplicate MSP assays, as described above. Furthermore, the percentage of BCR-ABL RNA-negative colonies was consistently higher for BFU-E than for CFU-GM colonies for untreated patient samples despite the larger size of the BFU-E colonies and the use of about 7–10-fold more cells for RT-PCR on these colonies than on the CFU-GM. This additionally validates the significance of our finding of BCR-ABL RNA-negative and Pa methylation- positive colonies from CML samples.


   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
ABL Pa is the second downstream promoter of the ABL gene, and it is the promoter that is generally retained within the chimeric BCR-ABL gene of the Ph chromosome in CML hematopoietic cells. Methylation of this promoter has been proposed to be important in disease progression by repressing expression of the normal ABL transcript from the translocated gene thereby allowing more synthesis of the leukemic BCR-ABL transcript (26) . If so, there should be only a low percentage of methylation of this promoter in early CP CML patients. With two assays for DNA methylation, COBRA and MSP, we have shown that a high percentage of early CP CML patients have Pa methylation in BM samples. Similar results were obtained by Issa et al. and Nguyen et al. (13 , 14) with different methylation assays. Furthermore, in this study we demonstrated that most of the hematopoietic progenitor-derived colonies from early CP CML patients had ABL Pa methylation. This contrasts with the results of Asimakopoulos et al. (15) , who found that only one CP patient of six displayed such methylation using the same MSP assay for analysis.

In this study, we quantitated both the level of ABL Pa methylation and of BCR-ABL transcripts in uncultured BM samples from CML patients. BM samples from untreated CP CML patients with high concentrations of this leukemic transcript had about 15–60% of the copies of their ABL Pa promoter methylated as assayed by COBRA or MSP (Table 2)Citation , although we cannot state whether this methylation is on the translocated or the normal ABL allele. One patient (CML10) who had been treated with IFN-{alpha}/AraC and was in cytogenetic remission at the time of analysis had a low level of BCR-ABL transcripts in the uncultured BM and only ~5% of the copies of the Pa promoter methylated at the examined sites. Another uncultured CML sample (CML9) that was the stem cell pheresis product from the only patient in this study who had been treated with idarubicin/AraC chemotherapy had no detectable Pa methylation and only trace amounts of BCR-ABL RNA, apparently as a result of therapy. As reported previously for some other treated patients (27 , 28) , ~25% of hematopoietic colonies cultured from this PB sample were positive for BCR-ABL transcripts despite the negative cytogenetic findings on BM and the trace levels of BCR-ABL RNA in the sample before plating (Table 3)Citation . The patient relapsed 18 months after treatment, which indicates that the analysis of progenitor-derived colonies for these transcripts was a better indicator of residual disease than were the cytogenetic findings on BM. However, the colonies of this patient displayed no Pa methylation.

ABL Pa promoter methylation independent of expression of BCR-ABL RNA was seen in progenitor-derived hematopoietic colonies from the CML samples. Not only did we find colonies with no detectable ABL Pa methylation despite the presence of the BCR-ABL transcript, but also, all of the CML BM samples yielded some colonies with strong signals in MSP assays for ABL Pa methylation but no detectable BCR-ABL RNA in nested RT-PCR (Table 3)Citation . With careful controls, we demonstrated that we should have negligible numbers of false negative results. Whether cells in these Pa-methylated, BCR-ABL RNA-negative colonies had a transcriptionally silent Ph chromosome or have CML-associated Pa methylation in the absence of the Ph chromosome remains to be determined by three-way cytogenetic analysis, Pa methylation assays, and BCR-ABL RT-PCR on hematopoietic progenitor-derived colonies having Pa methylation but no BCR-ABL transcripts. It has been reported that there are no BCR-ABL RNA-negative hematopoietic progenitor-derived colonies from CML samples that contain the Ph chromosome (29 , 30) , although evidence to the contrary has also been presented (31) .

Importantly, we found that among the CML samples plated for colony formation, hematopoietic precursor-derived colonies that displayed ABL Pa methylation were no more likely to express BCR-ABL RNA than those without such methylation (Table 4)Citation . Nonetheless, ABL Pa methylation was never seen in this or other studies in uncultured non-CML samples (Fig. 5Citation ; Refs. 8 , 13 , 14 ) nor in cultured non-CML BM samples (Table 3)Citation . Therefore, methylation of this promoter is clearly associated with CML, although it does not appear to be linked to BCR-ABL expression.

Increases (and decreases) in DNA methylation in cancer cells (and in embryos) occur in many genomic sequences, only some of which affect the cellular phenotype (9 , 11) . Cancer-linked and transcription-repressing de novo methylation of promoters of tumor suppressor genes is very often found in diverse types of malignancies (10, 11, 12) . Although some tumor suppressor gene promoters were examined for hypermethylation in CP CML and found to be normal (14 , 15 , 32 , 33) , hypermethylation in the 5' region of the calcitonin1 gene has been observed to track with CML progression (34 , 35) , and hypermethylation of BCR sequences has been reported in ~40% of CP CML patients (36) . Because our results indicate that ABL Pa methylation does not favor BCR-ABL expression and our results and others (13 , 14) demonstrate that it shows little or no correlation with leukemic progression, this methylation might just be a marker for coordinate methylation-regulated silencing of tumor suppressor genes in CP CML. Such DNA methylation might play a role in the generation of CML or in determining the variability in the clinical course of the disease.


   ACKNOWLEDGMENTS
 
We thank Sherry Price for help with BM and PB sample preparation, Jean-Pierre Issa for sharing his unpublished sequences for COBRA primers with us, and Nicholas Cross and John Goldman for detailed information about competitive RT-PCR conditions and for plasmids.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 Supported by NIH Grant CA78639. Back

2 To whom requests for reprints should addressed, at Human Genetics Program SL31, Tulane Medical School, 1430 Tulane Avenue, New Orleans, LA 70112. Phone: (504) 584-2449; Fax: (504) 584-1763; E-mail: ehrlich{at}tulane.edu Back

3 The abbreviations used are: Ph chromosome, Philadelphia chromosome; ABL, ABL1 gene; CML, chronic myeloid leukemia; BM, bone marrow; PB, peripheral blood; CP, chronic phase; Pa, alternative downstream promoter of ABL; IFN-{alpha}, {alpha}-interferon; AraC, 1-ß-D-arabinofuranosylcytosine; MNC, mononuclear cell fraction; CFU-GM, granulocyte-macrophage colony-forming cells; BFU-E, erythroid burst-forming cells; COBRA, combined bisulfite restriction analysis; MSP, methylation-specific PCR; Tanneal, annealing temperature; RT-PCR, reverse transcription-PCR; I-FISH, interphase fluorescence in situ hybridization. Back

5 N. Cross, personal communication. Back

4 J. P. Issa, unpublished observations. Back

Received 5/ 3/01. Accepted 7/17/01.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Rowley J. D. The Philadelphia chromosome translocation. A paradigm for understanding leukemia. Cancer (Phila.), 65: 2178-2184, 1990.[Medline]
  2. Westbrook C. A. The ABL oncogene in human leukemias. Blood Rev., 2: 1-8, 1988.[Medline]
  3. Era T., Witte O. N. Regulated expression of P210 Bcr-Abl during embryonic stem cell differentiation stimulates multipotential progenitor expansion and myeloid cell fate. Proc. Natl. Acad. Sci. USA, 97: 1737-1742, 2000.[Abstract/Free Full Text]
  4. Zou X., Calame K. Signaling pathways activated by oncogenic forms of Abl tyrosine kinase. J. Biol. Chem., 274: 18141-18144, 1999.[Free Full Text]
  5. Deininger M. W., Vieira S., Mendiola R., Schultheis B., Goldman J. M., Melo J. V. BCR-ABL tyrosine kinase activity regulates the expression of multiple genes implicated in the pathogenesis of chronic myeloid leukemia. Cancer Res., 60: 2049-2055, 2000.[Abstract/Free Full Text]
  6. Shtivelman E., Lifshitz B., Gale R. P., Roe B. A., Canaani E. Alternative splicing of RNAs transcribed from the human abl gene and from the bcr-abl fused gene. Cell, 47: 277-284, 1986.[Medline]
  7. Jiang X. Y., Trujillo J. M., Dao D., Liang J. C. Studies of BCR and ABL gene rearrangements in chronic myelogenous leukemia patients by conventional and pulsed-field gel electrophoresis using gel inserts. Cancer Genet. Cytogenet., 42: 287-294, 1989.[Medline]
  8. Ben-Yehuda D., Krichevsky S., Rachmilewitz E. A., Avraham A., Palumbo G. A., Frassoni F., Sahar D., Rosenbaum H., Paltiel O., Zion M., Ben-Neriah Y. Molecular follow-up of disease progression and interferon therapy in chronic myelocytic leukemia. Blood, 90: 4918-4923, 1997.[Abstract/Free Full Text]
  9. Ehrlich M. DNA hypomethylation and cancer Ehrlich M. eds. . DNA Alterations in Cancer: Genetic and Epigenetic Changes, : 273-291, Eaton Publishing Natick, Massachusetts 2000.
  10. Issa J. P. Hypermethylator phenotypes in aging and cancer Ehrlich M. eds. . DNA Alterations in Cancer, : 311-322, Eaton Publishing Natick, Massachusetts 2000.
  11. Baylin S. B., Herman J. G. Epigenetics, and Loss of Gene Function in Cancer Ehrlich M. eds. . DNA Alterations in Cancer, : 293-309, Eaton Publishing Natick, Massachusetts 2000.
  12. Robertson K. D., Jones P. A. DNA methylation: past, present and future directions. Carcinogenesis (Lond.), 21: 461-467, 2000.[Abstract/Free Full Text]
  13. Issa J. P., Kantarjian H., Mohan A., O’Brien S., Cortes J., Pierce S., Talpaz M. Methylation of the ABL1 promoter in chronic myelogenous leukemia: lack of prognostic significance. Blood, 93: 2075-2080, 1999.[Abstract/Free Full Text]
  14. Nguyen T. T., Mohrbacher A. F., Tsai Y. C., Groffen J., Heisterkamp N., Nichols P. W., Yu M. C., Lubbert M., Jones P. A. Quantitative measure of c-abl and p15 methylation in chronic myelogenous leukemia: biological implications. Blood, 95: 2990-2992, 2000.[Abstract/Free Full Text]
  15. Asimakopoulos F. A., Shteper P. J., Krichevsky S., Fibach E., Polliack A., Rachmilewitz E., Ben-Neriah Y., Ben-Yehuda D. ABL1 methylation is a distinct molecular event associated with clonal evolution of chronic myeloid leukemia. Blood, 94: 2452-2460, 1999.[Abstract/Free Full Text]
  16. Melo J. V., Yan X. H., Diamond J., Lin F., Cross N. C., Goldman J. M. Reverse transcription/polymerase chain reaction (RT/PCR) amplification of very small numbers of transcripts: the risk in misinterpreting negative results. Leukemia (Baltimore), 10: 1217-1221, 1996.[Medline]
  17. Raizis A. M., Schmitt F., Jost J. P. A bisulfite method of 5-methylcytosine mapping that minimizes template degradation. Anal. Biochem., 226: 161-166, 1995.[Medline]
  18. Warnecke P. M, Mann J. R, Frommer M, Clark S. J. Bisulfite sequencing in preimplantation embryos: DNA methylation profile of the upstream region of the mouse imprinted H19 gene. Genomics, 51: 182-190, 1998.[Medline]
  19. Warnecke P. M., Stirzaker C., Melki J. R., Millar D. S., Paul C. L., Clark S. J. Detection and measurement of PCR bias in quantitative methylation analysis of bisulphite-treated DNA. Nucleic Acids Res., 25: 4422-4426, 1997.[Abstract/Free Full Text]
  20. Chomczynski P., Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem., 162: 156-159, 1987.[Medline]
  21. Cross N. C., Hughes T. P., Feng L., O’Shea P., Bungey J., Marks D. I., Ferrant A., Martiat P., Goldman J. M. Minimal residual disease after allogeneic bone marrow transplantation for chronic myeloid leukaemia in first chronic phase: correlations with acute graft-versus-host disease and relapse. Br. J. Haematol., 84: 67-74, 1993.[Medline]
  22. Cross N. C. P., Feng L., Chase A., Bungey J., Hughes T. P., Goldman J. M. Competitive polymerase chain reaction to estimate the number of BCR-ABL transcripts in chronic myeloid leukemia patients after bone marrow transplantation. Blood, 82: 1929-1936, 1993.[Abstract/Free Full Text]
  23. Wang R. Y. H., Gehrke C. W., Ehrlich M. Comparison of bisulfite modification of 5-methyl-deoxycytidine and deoxycytidine residues. Nucleic Acids Res., 8: 4777-4790, 1980.[Abstract/Free Full Text]
  24. Clark S. J., Harrison J., Paul C. L., Frommer M. High sensitivity mapping of methylated cytosines. Nucleic Acids Res., 22: 2990-2997, 1994.[Abstract/Free Full Text]
  25. Xiong Z., Laird P. W. COBRA: a sensitive and quantitative DNA methylation assay. Nucleic Acids Res., 25: 2532-2534, 1997.[Abstract/Free Full Text]
  26. Zion M., Ben-Yehuda D., Avraham A., Cohen O., Wetzler M., Melloul D., Ben-Neriah Y. Progressive de novo DNA methylation at the bcr-abl locus in the course of chronic myelogenous leukemia. Proc. Natl. Acad. Sci. USA, 91: 10722-10726, 1994.[Abstract/Free Full Text]
  27. Schulze E., Krahl R., Thalmeier K., Helbig W. Detection of bcr-abl mRNA in single progenitor colonies from patients with chronic myeloid leukemia by PCR: comparison with cytogenetics and PCR from uncultured cells. Exp. Hematol., 23: 1649-1654, 1995.[Medline]
  28. Reiter A., Marley S. B., Hochhaus A., Sohal J., Raanani P., Hehlmann R., Gordon M. Y., Goldman J. M., Cross N. C. BCR-ABL-positive progenitors in chronic myeloid leukaemia patients in complete cytogenetic remission after treatment with interferon-{alpha}. Br. J. Haematol., 102: 1271-1278, 1998.[Medline]
  29. Diamond J., Goldman J. M., Melo J. V. BCR-ABL, ABL-BCR, and ABL genes are all expressed in individual granulocyte-macrophage colony-forming unit colonies derived from blood of patients with chronic myeloid leukemia. Blood, 85: 2171-2175, 1995.[Abstract/Free Full Text]
  30. Savoldo B., Sammarelli G., Dotti G., Garau D., Regazzi E., Cilloni D., Tabilio A., Rizzoli V., Carlo-Stella C. Reverse transcription polymerase chain reaction is a reliable assay for detecting leukemic colonies generated by chronic myelogenous leukemia cells. Leukemia (Baltimore), 12: 434-440, 1998.[Medline]
  31. Keating A., Wang X. H., Laraya P. Variable transcription of BCR-ABL by Ph+ cells arising from hematopoietic progenitors in chronic myeloid leukemia. Blood, 83: 1744-1749, 1994.[Abstract/Free Full Text]
  32. Herman J. G., Civin C. I., Issa J. P., Collector M. I., Sharkis S. J., Baylin S. B. Distinct patterns of inactivation of p15INK4B and p16INK4A characterize the major types of hematological malignancies. Cancer Res., 57: 837-841, 1997.[Abstract/Free Full Text]
  33. Katzenellenbogen R. A., Baylin S. B., Herman J. G. Hypermethylation of the DAP-kinase CpG island is a common alteration in B-cell malignancies. Blood, 93: 4347-4353, 1999.[Abstract/Free Full Text]
  34. Malinen T., Palotie A., Pakkala S., Peltonen L., Ruutu T., Jansson S. E. Acceleration of chronic myeloid leukemia correlates with calcitonin gene hypermethylation. Blood, 77: 2435-2440, 1991.[Abstract/Free Full Text]
  35. Nelkin B. D., Przepiorka D., Burke P. J., Thomas E. D., Baylin S. B. Abnormal methylation of the calcitonin gene marks progression of chronic myelogenous leukemia. Blood, 77: 2431-2434, 1991.[Abstract/Free Full Text]
  36. Litz C. E., Vos J. A., Copenhaver C. M. Aberrant methylation of the major breakpoint cluster region in chronic myeloid leukemia. Blood, 88: 2241-2249, 1996.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Hum Mol GenetHome page
M. Ehrlich, K. L. Buchanan, F. Tsien, G. Jiang, B. Sun, W. Uicker, C. M.R. Weemaes, D. Smeets, K. Sperling, B. H. Belohradsky, et al.
DNA methyltransferase 3B mutations linked to the ICF syndrome cause dysregulation of lymphogenesis genes
Hum. Mol. Genet., December 1, 2001; 10(25): 2917 - 2931.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sun, B.
Right arrow Articles by Ehrlich, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sun, B.
Right arrow Articles by Ehrlich, M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online