|
|
Blood, Vol. 113, Issue 19, 4512-4520, May 7, 2009
Acquired variation outweighs inherited variation in whole genome analysis of methotrexate polyglutamate accumulation in leukemia
Blood French et al.
113: 4512
Supplemental materials for: French et al
Measurement of MTXPG accumulation
MTXglu2 – MTXglu7 standards were obtained from Dr. B. Schircks (Jona, Switzerland). After HPLC separation of MTXPGs, one of two possible detection methods was used to quantify MTXPGs: a radioenzymatic assay (REA) or fluorescence following post-column photo-oxidation. For the patient samples analyzed by the HPLC-REA method, the MTXPGs were first separated by HPLC using a previously published method.1 The fractions corresponding to each glutamylated metabolite were collected separately, dried to completion, and quantified by a previously published radioenzymatic assay method2 modified from that reported by Kamen and Winick.3 For the cell lines and patient samples analyzed by HPLC-fluorescence, the supernatants were thawed, transferred to a clean tube, evaporated in a SpeedVac® (Thermo Fisher Scientific Inc, Waltham, MA) until dry, reconstituted in 80 µl of ddH2O, vortexed, put on ice for 10 minutes, and centrifuged. 50 µl of the supernatant was injected into the HPLC (Shimadzu, Columbia, MD); a post-column online photochemical reactor was used to convert MTXPGs to fluorescent derivatives for enhanced detection (Aura Industries Inc, New York, NY). The mobile phase buffer contained 5mM sodium phosphate (dibasic anhydrous), 2.5mM tetrabutylammonium hydroxide (1M in water) and was equilibrated to pH7.3 with phosphoric acid. Two mobile phases were utilized in a gradient; A: 10% acetonitrile, 90% buffer with 0.2% hydrogen peroxide and B: 40% acetonitrile, 60% buffer with 0.2% hydrogen peroxide; the gradient was: 0–36 minutes: 18% mobile phase B, 36–44 minutes: 100% mobile phase B, 46–60 minutes: 18% mobile phase B, 61 minutes: end of run or inject new sample. The flow rate was 0.7ml/min. All results were expressed as picomoles of MTXPGs per 109 cells (pmol/109 cells); the lower limit of detection for this assay was 0.2 pmol/109 cells and the upper limit was 16 pmol/109 cells. The HPLC-radioenzymatic assay (REA) method was used for samples from patients on Total XIIIA, XIIIB, and for the first one-third of Total XV.2;4–8 The remaining two-thirds of Total XV samples and all of the HapMap cell line samples were assayed using the HPLC-fluorescence method.9 The main reason for the change in detection methods was that reagents used for the REA assay, previously purchased from the Enzyme Center (Malden, MA 02148) became completely commercially unavailable during the course of the Total XV trial, and it was not possible to re-assay all of the prior Total XIIIA, Total XIIIB, and early Total XV samples by the newer HPLC-fluorescence method because many of these precious bone marrow samples were depleted by the first (HPLC-REA) assay. Based on controlled experimental samples, the HPLC-fluorescence generated results were adjusted to make them comparable to the HPLC-REA results, so that the results reported herein are comparable to our prior reports in patient samples. To determine differences between these two methods of detection, we spiked Nalm6 cell-line extracts with 30 known concentrations (0.4 to 20 pmol) of MTXPG1 through MTXPG7, and then processed the extracts as above for each assay. The median quartiles difference between the measured vs spiked predicted concentrations for the sum of MTXPG1–7 was −0.2% −7%, 6% for the HPLC-fluorescence method and −29% −37%, −22% for the HPLC-REA method. To account for these observed differences in the spiked samples due to assay method we determined the relation between the two methods by comparing the amount of MTXPG for each of the 7 polyglutamates at each spiked concentration. There was not a significant difference between MTXPG length (1–7) and the linear relation between the HPLC-REA–determined concentrations and the HPLC-fluorescence determined concentration (supplemental Figure 2); thus, we used a single equation to adjust the sum of the HPLC-fluorescence measured MTXPGs to render them comparable to the sum of the HPLC-REA–measured MTXPGs: Estimated HPLC-REA method MTXPG1–7 concentration=0.6918+0.6179*HPLC-fluorescence method measured MTXPG1–7 concentration (r2=0.92). We used this equation to normalize the HPLC-fluorescence measured concentrations to be comparable to the HPLC-REA–measured concentrations; thus, all MTXPG concentrations are either measured by HPLC-REA or measured by HPLC-fluorescence but normalized to be comparable to HPLC-REA–measured concentrations. Assessment of inherited copy number variation (CNV) in patients
DNA was extracted from normal leukocytes of 137 patients. 500 ng of each DNA sample was digested with HindIII and XbaI or NspI and StyI restriction enzymes, amplified, labeled and hybridized to the Affymetrix GeneChip® Human Mapping 100K set and 500K set respectively. DNA copy number analysis was performed using dChipSNP 10 as described previously.11 The CNV results data set comprised normalized single SNP intensities including 615,922 SNPs.11,12 Statistical analysis of inherited CNV vs leukemia cell MTXPG accumulation
To determine the association between MTXPG accumulation and single SNP inherited copy number variation, we used the normalized signal intensity for each SNP as the independent variable and performed a linear regression analysis between MTXPGs and the signal intensities for each of the ~600,000 SNPs on the SNP Chip. Results Inherited CNV associated with leukemia cell MTXPG accumulation in patients
We assessed the association of inherited single SNP CNV with MTXPG accumulation in 137 patients. Of those SNPs with copy number variation (n=14,889 SNPs), 49 were associated with MTXPG accumulation (p<0.01), implicating 9 genes (Table S4). None of the 9 genes overlap with the 7 genes implicated in the overlap of the 3 whole-genome variation methods (p15 of the main manuscript; Figure 2 and Table S5), the 3 genes found on chromosome 18 (p17 of main manuscript; Table 2) or the 10 genes found on chromosome 10 (pp17–18 of main manuscript; Table S9). Additional chromosomes with somatic leukemia cell copy number variation associated with leukemia cell MTXPG accumulation
In addition to chromosomes 10 and 18 that had copy number variation associated with MTXPG accumulation, SNP chip-defined CNV (n=82 patients) of chromosomes 4, 5, 6, 14, and 21 was significantly associated with MTXPG accumulation (p=7 × 10−4, p=0.0060, p=0.0016, 1.25 × 10−6 and p=4.65 × 10−6 respectively, ANOVA) in univariate analyses, but not when the analyses were adjusted for ALL subtype (p=0.38, p=0.97, p=0.50, p=0.22 and p=0.94 respectively, ANOVA) (Table S6 and Fig. S7). Using cytogenetics, more samples were evaluable (n=213) than using SNP chip-defined CNV, and cytogenetically-defined gains of chromosomes 4 and 21 were significantly associated with MTXPG accumulation (p=5.06 × 10−8 and p=6.42 × 10−10 respectively, ANOVA; Fig. S8) but not when we adjusted for ALL subtype (p=0.62 and p=0.31 respectively, ANOVA).
Files in this Data Supplement:
- Table S1 (PDF, 36.9 KB) -
MTXPG (normalized to the HPLC-REA method of detection) accumulation in patient leukemia cells is significantly associated with ALL subtype (in this table we compared all subtypes independently to B-hyperdiploid patients), window treatment arm (4 hr vs 24 hr), methotrexate clearance (estimated from plasma MTX concentrations),6 and method of detection (REA vs fluorescence).
- Table S2. Genes on chromosome 18 whose expression is significantly associated with MTXPG accumulation in patient leukemia cells (n=145; p<0.05) (PDF, 26.8 KB)
- Table S3. Top 10 regions* of SNPs with copy number variation associated with MTXPG accumulation in patient leukemia cells (n=82; p<0.01) (PDF, 21.6 KB)
- Table S4. Top 10 regions* of inherited SNPs with CNV (determined by 600K SNP chip data) associated with MTXPG accumulation in patient leukemia cells (n=137; p<0.01) (PDF, 54.6 KB)
- Table S5. Seven genes were implicated in all 3 types of genomic variation (PDF, 40.2 KB) -
Leukemia cell methotrexate polyglutamates (MTXPGs) vs leukemia cell gene expression, somatic leukemia cell copy number variation and inherited SNP genotype analyses.
- Table S6. Somatic whole chromosome CNV of chromosomes 10 and 18 (determined by SNP chip) in the leukemia cells of 82 patients is significantly associated with leukemia cell MTXPG accumulation even after adjusting for ALL subtype (PDF, 59.2 KB)
|
|
