Dynamics of circulating t(14;18)-positive cells during first-line and subsequent lines of treatment in follicular lymphoma.

In follicular lymphoma the t(14;18) might be useful as a tumor marker in predicting the quality of the response to treatment. We investigated whether analyzing numbers of t(14;18)-positive cells in peripheral blood correlated with remission status in individual patients receiving a variety of treatments. Numbers of circulating t(14;18)-positive cells were determined by real-time polymerase chain reaction (PCR) technique. Disease parameters and response to treatment were related to the pre- and post-treatment numbers of circulating t(14;18)-positive cells for 53 follicular lymphoma patients. In these 53 patients, 70 treatment episodes were investigated. A content of more than 328 t(14;18)-positive cells per 75,000 cells prior to therapy correlated with the more advanced stage IV disease ( P=0.01), bone marrow involvement ( P<0.01), and overt leukemic lymphoma ( P=0.04). Therapy episodes that cleared circulation from t(14;18)-positive cells with more than one log resulted in a significantly longer progression-free survival than treatment episodes with less than one log decline (26 versus 12 months, respectively) ( P<0.01). After first-line treatment episodes, numbers of circulating t(14;18)-positive cells declined in fairly all cases, irrespective of the clinical response. However, for second or later lines of treatment, declining numbers of lymphoma cells correlated with a clinical remission, whereas increasing numbers of lymphoma cells were associated with clinically stable or progressive disease. From this, we conclude that quantitation of circulating t(14;18)-positive cells in peripheral blood is of only limited clinical significance in predicting treatment efficacy for the individual follicular lymphoma patient.

Acquired skewing of Lyonization remains stable for a prolonged period in healthy blood donors.

The pattern of X-chromosome inactivation (XCIP), or Lyonization, can be used to distinguish monoclonal from polyclonal cell populations in females. However, a skewed XCIP exists in hematopoietic cells in approximately 40% of healthy elderly females, interfering with interpretation of clonality assays. In hematopoiesis, an active stem cell pool is assumed to be present within a larger population of inactive stem cells, with a continuous exchange of cells between the two compartments. The assumption that the active stem cell pool size decreases with age may explain the phenomenon of acquired skewing occurring by chance and predicts the XCIP of this population to fluctuate. This fluctuation should be reflected in the XCIP of peripheral granulocytes. We examined the XCIP for fluctuations in time in peripheral granulocytes, monocytes and T cells of young, middle-aged and elderly healthy females. We used an optimized HUMARA PCR assay that eliminates unbalanced DNA amplification. We found no fluctuations in XCIP in any age group in up to 18 months follow-up. We conclude that acquired skewing arises gradually in life without fluctuations in XCIP and that analysis at multiple time points cannot distinguish monoclonal hematopoiesis from normal, skewed hematopoiesis.

Analysis of expression of chorionic gonadotrophin transcripts in prostate cancer by quantitative Taqman and a modified molecular beacon RT-PCR.

Expression of human chorionic gonadotrophin (hCG) is associated with trophoblastic, testicular and other malignancies such as bladder, pancreatic, cervical, breast and prostate cancer. In the prostate, however, hCG expression, associated with neuroendocrine cells, is also found in normal tissue. Of the six highly homologous genes that all encode the beta-subunit of hCG, the beta 7 gene is reportedly the only gene expressed in several non-transformed tissues. The beta 3, 5 and 8 genes would be variably expressed in malignant tissue and placenta, but not in normal tissue. To assess to what extent this expression difference can also be found in the prostate, we compared the levels of the different hCG beta transcripts in concurrent normal and cancerous prostate tissues obtained from 17 patients. To this end, we developed a Taqman real-time fluorescent RT-PCR assay for hCG beta, and a quantitative assay specific for the beta 3, 5 and 8 genes, modified from the molecular beacon principle. This latter assay proved highly specific and capable of reliably distinguishing between these hCG beta transcripts that differ in only one nucleotide. Surprisingly, median expression levels of hCG beta were lower in prostate cancer when compared with normal tissue from the same patient. In contrast, hCG beta 3, 5 and 8 transcripts were found in normal tissue and did not differ in prostate cancer, arguing against a specific role of these transcripts in the development of prostate cancer.

Degradable dU-based DNA template as a standard in real-time PCR quantitation.

Development of real-time quantitative PCR assays requires suitable positive controls. For assays with clinical applications, these controls may be difficult to obtain because some molecular aberrations are rare and patient material may be available in limited amounts. Because of the risk of introducing contaminations in the laboratory, cloned DNA is not a desirable alternative. We describe the use of dU-containing DNA as a positive control template in real-time quantitative PCR. dU-DNA constructs can be decontaminated by adding uracil N-glycosylase (UNG) to the reaction mixture. In addition, dU-DNA can be used for accurate quantification, because it allows quantification to be expressed in numbers of molecules. Since synthetic dU-DNA constructs can easily be quantitated spectroscopically, they provide a more accurate control than arbitrary cell line units. We applied this method for the detection of the E2A-Pbx1 gene fusion and show that UNG-containing reactions can be employed for diagnostics without loss of sensitivity, and that for positive and quantitative controls UNG negative reactions can be used. The use of dU-DNA provides a novel type of control template that can easily be integrated into existing PCR protocols.

Lack of correlation between numbers of circulating t(14;18)-positive cells and response to first-line treatment in follicular lymphoma.

In follicular lymphoma, the t(14;18) status of the peripheral blood and bone marrow analyzed by polymerase chain reaction (PCR) is assumed to correlate with disease activity in patients with relapsed disease. The clinical significance of quantitating circulating lymphoma cells by real-time PCR is reported in patients on first-line treatment. Thirty-four consecutive patients with previously untreated follicular lymphoma and detectable t(14;18)-positive cells in pretreatment peripheral blood samples were monitored. All patients were treated with standard chemotherapy in combination with interferon alfa-2b. Before and after induction therapy, blood samples were taken for quantitative analysis of t(14;18). At presentation, a median of 262 t(14;18)-positive cells per 75,000 normal cells was found (range, 1-75 000). Patients with lower numbers of circulating tumor cells more frequently had bulky disease (P =.02). Seventy-nine percent of the patients responded clinically to treatment. In 22 of 28 patients, including 4 patients in whom treatment had failed clinically, the number of circulating t(14;18)-positive cells decreased to undetectable or low levels after therapy. In the remaining responding patients, circulating tumor cells persisted after therapy. These quantitative data on circulating t(14;18)-positive cells call into question the usefulness of molecular monitoring of the blood in a group of patients with follicular lymphoma uniformly treated with a noncurative first-line regimen. T(14;18)-positive cells decreased in peripheral blood after treatment, irrespective of the clinical response. Therefore, the significance of so-called molecular remission should be reconsidered in follicular lymphoma. (Blood. 2001;98:940-944)

Transcription of the MAGE-1 gene and the methylation status of its Ets binding promoter elements: a quantitative analysis in melanoma cell lines using a real-time polymerase chain reaction technique.

The human MAGE gene family comprises at least 12 highly homologous genes. This makes it very difficult to assess expression of a single member quantitatively by means of Northern blotting. In order to investigate expression of the MAGE-1 gene quantitatively we therefore used the recently developed real-time polymerase chain reaction (PCR), a novel fluorescence-based quantitative PCR technique. This powerful technique enables detection of expression levels which differ by as much as a factor of 10(5) in magnitude. MAGE-1 expression is known to correlate with demethylated status of the Ets binding sites of its promoter. In a panel of 19 melanoma and nine non-melanoma cell lines we were able to confirm the relationship between MAGE-1 expression and demethylation of the Ets binding promoter region. Five cell lines, however, showed only very slight expression, while the two essential Ets promoter elements were largely demethylated. Earlier studies have shown that treatment of MAGE-1-negative cell lines with the demethylating agent 5-aza-2'-deoxycytidine (DAC) is sufficient to induce MAGE-1 expression. We were able to induce clear MAGE-1 expression in two of the non-expressing cell lines by incubation with DAC, although this expression did not reach very high levels. Consistent with this low level of induction is the observation that the Ets binding sites of the MAGE-1 promoter were not completely demethylated in the DAC-treated cell populations. In conclusion, we show in this study that the real-time PCR technique is a very useful tool for the quantification of expression of highly homologous genes.

Reproducibility of detection of tyrosinase and MART-1 transcripts in the peripheral blood of melanoma patients: a quality control study using real-time quantitative RT-PCR.

In recent years, large discrepancies were described in the success rate of the tyrosinase reverse transcription polymerase chain reaction (RT-PCR) for detecting melanoma cells in the peripheral blood of melanoma patients. We present a quality control study in which we analysed the reproducibility of detection of tyrosinase and MART-1 transcripts in 106 blood samples from 68 melanoma patients (mainly stages III and IV). With this study, we aimed to improve insight in the reproducibility of a RT-PCR for the detection of (minimal) amounts of circulating melanoma cells. We performed two reverse transcriptions on each mRNA sample and performed tyrosinase and MART-1 nested PCRs in duplicate per cDNA sample. Thus, four tyrosinase and four MART-1 measurements were performed per blood sample. In our study, the majority of blood samples was negative for tyrosinase (80%) or MART-1 (66%). Only four samples were positive in all four determinations for tyrosinase and seven for MART-1. Variable results (1-3 times positive results) were obtained for tyrosinase and MART-1 in 16% and 27% respectively. MART-1 PCR had a better performance than tyrosinase PCR. Sensitivity increased when both markers were used. We reasoned that the low number of melanoma marker PCR-positive blood samples can be explained by differences in mRNA quality. By using real-time quantitative PCR, we found that this was not the case: amplification of porphobilinogen deaminase (PBGD), a low copy household gene, was not different in blood samples in which a melanoma marker was not detected from groups in which this marker was detected more or less consistently (1-4 times). When applying real-time quantitative PCR for tyrosinase and MART-1, we found that a low amount of SK-MEL-28 cell equivalents was present in the blood of melanoma patients, with a higher number of equivalents in the group with a consistently positive result. We conclude that low reproducibility of a repeated assay for the detection of circulating melanoma cells is not caused by differences in mRNA quality between the samples, but due to low numbers of amplifiable target mRNA molecules in the mRNA sample. Use of more than one marker and repetition of the assay will increase the probability of finding positive PCR results.

Specific detection of myeloma plasma cells using anti-idiotypic single chain antibody fragments selected from a phage display library.

Bone marrow plasma cells constitute the bulk of malignant cells in multiple myeloma patients. B-lymphocytes having immunoglobulin heavy chain gene rearrangements identical to those of the malignant clone (clonally related B-lymphocytes) may function as malignant plasma cell precursors. We and others proposed the use of anti-idiotypic antibodies to isolate and study clonally related B-lymphocytes. This strategy failed until now because anti-idiotypic antibodies raised by conventional hybridoma techniques proved to react frequently with epitopes shared by different idiotypes. Recently, we succeeded in selecting specific single chain Fv antibodies from phage libraries. To select single chain Fv bearing phages specifically directed against the immunoglobulin idiotype expressed by myeloma tumor cells we panned a semisynthetic phage library against purified myeloma paraprotein Fab fragments. The selection was performed in the presence of soluble polyclonal immunoglobulin as a competitor. Three independent selections for three myeloma patients yielded 10-26 clones. Between two and seven of the selected clones were reactive with patient Fab and not with polyclonal immunoglobulin in enzyme-linked immunosorbent assays. Five out of six anti-idiotypic single chain Fvs were able to specifically stain fixed monoclonal plasma cells in myeloma bone marrow. Idiotype specificity of these single chain Fvs was confirmed by flow cytometry since they did not react with monoclonal plasma cells of other patients, a panel of nine myeloma cell lines, isolated polyclonal bone marrow plasma cells and cultured B-lymphocytes. Using these anti-idiotypic reagents we were able to detect 25 myeloma plasma cells in a background of 50000 immunoglobulin isotype-matched cells of the myeloma cell line UM-1 or 50000 donor bone marrow cells (sensitivity 0.05%). This paper shows that highly specific anti-idiotypic single chain Fv antibody fragments selected from a phage display library can be used to detect rare idiotypic cells in patient samples.

Contamination-free and automated composition of a reaction mixture for nucleic acid amplification using a capillary electrophoresis apparatus.

The acceptance of the polymerase chain reaction (PCR) as an amplification method in molecular diagnostics and the rapid development of capillary electrophoresis (CE) as an analysis method of those PCR products was a reason for us to investigate further integration of those two techniques. Using a fused-silica capillary as a pipette we were able to compose a PCR mixture in the CE apparatus. Because a capillary can be thoroughly rinsed and the CE apparatus is a closed system, the risk of contamination and therefore the occurrence of false positive results is minimized. The fact that a CE system can be fully automated contributes to a more reproducible and standardized PCR composition protocol.

Hematopoietic malignancies demonstrate loss-of-function mutations of BAX.

The BCL-2 gene family regulates the susceptibility to apoptotic cell death in many cell types during embryonic development and normal tissue homeostasis. Deregulated expression of anti-apoptotic BCL-2 can be a primary aberration that promotes malignancy and also confers resistance to chemotherapeutic agents. Recently, studies of Bax-deficient mice have indicated that the pro-apoptotic BAX molecule can function as a tumor suppressor. Consequently, we examined human hematopoietic malignancies and found that approximately 21% of lines possessed mutations in BAX, perhaps most commonly in the acute lymphoblastic leukemia subset. Approximately half were nucleotide insertions or deletions within a deoxyguanosine (G8) tract, resulting in a proximal frame shift and loss of immunodetectable BAX protein. Other BAX mutants bore single amino acid substitutions within BH1 or BH3 domains, demonstrated altered patterns of protein dimerization, and had lost death-promoting activity. Thus, mutations in the pro-apoptotic molecule BAX that confer resistance to apoptosis are also found in malignancies.

Molecular beacons: a new approach for semiautomated mutation analysis.

Molecular beacons are oligonucleotide probes that become fluorescent upon hybridization. We designed molecular beacons to detect a point mutation in the methylenetetrahydrofolate reductase (MTHFR) gene, a mutation that has been related to an increased risk for cardiovascular disease and neural tube defects. The application of molecular beacons enables fast, semiautomated, accurate mutation detection. Moreover, the procedure is performed in a closed tube system, thereby avoiding carryover contamination. We believe these probes will find their way into nucleic acid research and diagnostics.

An antisense Bcr-Abl phosphodiester-tailed methylphosphonate oligonucleotide reduces the growth of chronic myeloid leukaemia patient cells by a non-antisense mechanism.

The specificity of antisense oligonucleotides targeted to the mRNA breakpoint region of the Bcr-Abl oncogene, found in leukaemic cells from patients with chronic myeloid leukaemia, remains controversial due to non-specific effects. To prevent protein binding of oligonucleotides we designed and tested a methylphosphonate oligonucleotide with an attached 3' soluble phosphodiester tail. Growth of chronic myeloid leukaemia (CML) cell lines BV173, KCL-22 and cells of CML patients tested was inhibited by the b2a2 type antisense Bcr-Abl oligonucleotide and not with controls. Also the growth of control CD34+ cells of two healthy donors, control cell lines and cells from AML patients was only moderately affected or not affected. Bcr-Abl protein studies in combination with growth-determination experiments indicated that the antisense methylphosphonate Bcr-Abl oligonucleotide tested is a potent inhibitor of the growth of CML cells but works in a non-antisense manner.

Analysis at the DNA level of human Fc gamma RII isoforms in relation to the polymorphic binding of murine IgG2b to human B cells.

Three classes of human leucocyte Fc gamma receptors (hFc gamma R) have been identified so far: hFc gamma RI, hFc gamma RII, and hFc gamma RIII. Previous studies have demonstrated that genetically determined differences between individuals exist both with respect to the binding of murine IgG1 (mIgG1) to hFc gamma receptors, and with respect to the binding of murine IgG2b (mIgG2b). The polymorphism in binding of mIgG1 could be ascribed to hFc gamma RIIA, an isoform of hFc gamma RII. The authors have now investigated whether one of the isoforms of hFc gamma RII is also responsible for the polymorphism in binding of mIgG2b. In these studies the authors used EBV-transformed human B cells that demonstrated either binding or no binding of mIgG2b in EA-rosetting assays. mRNA obtained from these cells was amplified by reverse transcriptase and polymerase chain reaction (RT-PCR). Hybridization experiments with the RT-PCR products revealed that the hFc gamma RIIB but not the hFc gamma RIIA isoform was present in these cells. DNA sequencing further demonstrated that the nucleotide sequence of both the extracellular part and the cytoplasmic moiety of hFc gamma RIIB was identical for all individuals tested, regardless of their ability to bind mIgG2b. These findings indicate that the polymorphic binding of mIgG2b cannot be ascribed to one of the isoforms of hFc gamma RII. Since hFc gamma RI and hFc gamma RIII are not present on the cell surface of these cells, the authors conclude that an Fc receptor different from the known hFc gamma receptors must be responsible for the polymorphic binding of mIgG2b. These data further expand the complexity of hFc gamma R.

A long-lasting, complete hematologic and cytogenetic remission of chronic myelogenous leukemia after treatment with busulfan alone.

A 44-year-old man suffering from cytogenetically and molecularly proven Philadelphia translocation-positive chronic myelogenous leukemia in chronic phase was treated with busulfan for 18 months and studied during a follow-up period of 13 years. Hematologically and cytogenetically, he attained a continuing complete remission, although at one point (9.5 years) at least, after attaining complete remission molecular analysis indicated the presence of minimal residual disease.

Bias in nucleotide composition of antisense oligonucleotides.

In this study we investigated if specific sequence motifs occur with a higher frequency in antisense oligonucleotides than can be expected on the basis of the mRNA composition to get an impression of the importance of these motifs for antisense effects. Computer analysis of 206 antisense oligonucleotides extracted from the literature and from sequence databases, all targeted against human mRNA, was performed. We compared the sequence composition of these oligonucleotides with the average of 100 equally large and randomly selected sequences from sequence databases and of their target mRNA. We found that the frequency of sequence motifs containing GG, CCC, CC, GAC, and CG is significantly higher and TT and that TCC is significantly lower in antisense oligonucleotide sequences than in the randomly selected mRNA sequences. We conclude that there is a bias in the nucleotide composition of antisense oligonucleotides. Some of these biased sequence motifs have been reported to induce nonantisense effects mediated by protein binding. Further analysis of the biologic function of these motifs is necessary to investigate if they should be avoided or incorporated into future designs of therapeutic effective oligonucleotides.