Powered By Google

Supplementary Information

Treatment Specific Changes in Gene Expression Discriminate in vivo Drug Response in Human Leukemia Cells

F. TaqMan Real-Time RT-PCR (RT-PCR)

To further establish the validity of gene expression determined by microarray analyses we performed two additional experiments to validate the microarray analysis. (1) We determined gene expression by real-time RT-PCR (reverse transcription polymerase chain reaction) in four randomly selected post-treatment patient samples for three genes (AP1S2, BAX, and RBBP8) and (2) we determined the fold-change in gene expression by both RT-PCR for three genes in five paired patient samples (in both pre- and post treatment sample).

(1) We used TaqMan Universal PCR Master Mix Kit and the 7900-sequence-detection system (Applied Biosystems, Foster City, California). We designed primers and probes (Table 6) with Primer Express 2.0 software (Applied Biosystems, Foster City, California) and used the housekeeping gene RNase P (Applied Biosystems, Foster City, California) for normalization. One µg of total RNA was treated with DNase I and reverse transcribed using Superscript II Rnase H- reverse transcriptase and oligo dT primers (Invitrogen, Carlsbad, California). Additionally, we included controls that contained either no template or no reverse transcriptase, as negative controls in each run. We used aliquots (0.5 µl) of RT reaction mixture (20 µl) for quantification of AP1S1, BAX, RBBP8 and RNase P gene expression.

Table 6: Primers and probes used for real-time (TaqMan) RT-PCR

(1) sequence start length
AP1S2 (NM_003916)

Forward Primer
Reverse Primer
Probe
5'-AAGCAATTGAGCAGGCTGATC-3'
5'-TCAGTCCAATTTCTTCAAGAACACTAC-3'
5'-6FAM-ACTGCAGGAGGAAGCTGAAACCCCA-TAMRA-3'
395
469
417
21
27
25
BAX (NM_004f324)

Forward Primer
Reverse Primer
Probe
5'-TGGAGCTGCAGAGGATGATTG-3'
5'-GAAGTTGCCGTCAGAAAACATG-3'
5'-6FAM-AGAGGTCTTTTTCCGAGTGGCAGCTG-TAMRA-3'
221
315
267
21
22
26
RBBP8 (NM_002894)

Forward Primer
Reverse Primer
Probe
5'-GAACCCCCATGTCCGATACA-3'
5'-GGATGAGTTGAAGACTTGGAAACTTT-3'
5'-6FAM-ACATACTAAATTGGAGCACTCTGTGTGTGCAAATG-TAMRA-3'
838
936
868
20
26
35
(2) sequence assay-on-demand ID
ATM
Probe
5'-6FAM-CAACGCGCAGGACTTCTGCACGGAC-TAMRA-3' Hs00175892_m1
BAX
Probe
5'-6FAM-AAACTGGTGCTCAAGGCCCTGTGCA-TAMRA-3' Hs00180269_m1
FOS
Probe
5'-6FAM-TATCAACCAAAGGCCTTCTTGTATC-TAMRA-3' Hs00170630_m1

We checked primer quality by conventional PCR for amplification of the correct size and sequence of all transcripts. The total volume of the PCR reaction was 50 µl, containing 0.5 µl of RT-product, 400 nM each of the forward and reverse primers, 250 nM of probe, and 1X master mix. We used following thermal cycling parameters: two minutes at 50°C (activation of UNG enzyme to remove the carry-over PCR products), ten minutes at 95°C to activate AmpliTaq Gold DNA polymerase, 15 seconds at 95°C to denature and one minute at 60°C for annealing and extension, for a total of 45 cycles.

To estimate the amount of each of the three mRNAs in the four patient samples, we used linear regression analysis based on a standard curve representing six serial dilutions of cDNA made from the CEM human leukemia cells (American Type Culture Collection, Rockville, Maryland). In the standard curve, we plotted fluorescent signal intensities against the number of PCR cycles on a semi-logarithmic scale. Using CEM cDNA as standard, we achieved a high degree of linearity. We analyzed all unknown samples in triplicates in parallel with a standardization series using CEM cDNA. Based on the CT value and the corresponding standard curve the relative quantity of the specific mRNA for each sample was calculated. We plotted normalized and log transformed real-time RT-PCR gene expression level (AP1S2, BAX, and RBBP8) to the log-transformed signal from the Affymetrix MAS 5.0 output. The correlation between real-time RT-PCR and Affymetrix GeneChip was statistically significant (overall P=0.017, spearman rank test), as shown in Fig. 8 below and R2 of 0.973, 0.798, 0.971 for AP1S2, BAX, and RBBP8, respectively, thereby confirming expression levels determined by the gene expression array.

Figure 8: Real-time (TaqMan) RT-PCR vs. Affymetrix GeneChip

Real-time (TaqMan) RT-PCR results are plotted versus expression levels determined by Affymetrix GeneChip® results for AP1S2, BAX and RBBP8, in four patients.

Figure 8: Real-time (TaqMan) RT-PCR vs. Affymetrix GeneChip

(2) We utilized the Assay-on-demandTM (part number 4331182, assay ID: ATM, Hs00175892_m1; BAX, Hs00180269_m1; FOS, Hs00170630_m1; Applied Biosystems, Foster City, California) according to the manufactures protocol (Table 6). To determine the relative quantity of gene expression of ATM, BAX, FOS in five pre- and five corresponding post-treatment patient samples we used the standard curve method and beta-actin, 18S-ribosomal RNA and RNAse P served as internal controls. We computed fold-change in expression for each gene based on post- to pre-treatment ratio as determined by RT-PCR, and we compared these fold-change values to the fold-change as determined by microarray analysis. As depicted in Fig. 9 below, the fold-change values determined by the two independent methods were highly correlated (R2=0.937).

Figure 9: Fold-change RT-PCR versus Affymetrix GeneChip®

Figure 9: Fold-change RT-PCR versus Affymetrix GeneChip®

RT-PCR results are plotted versus fold-change determined by Affymetrix GeneChip® results for ATM, BAX and FOS in five patients

Figure 10: Leave-one-out cross-validation of treatment classification.

Figure 10: Leave-one-out cross-validation of treatment classification.

Leave-one-out cross-validation using support-vector-machine (SVM) as the classifier, determined that prediction error rates were the smallest using linear discriminant analysis with variance (LDA) compared to ANOVA as the gene selection method. We constructed SVMs using the top-ranked discriminating genes selected by LDA or ANOVA. Leave-one-out cross-validation shows that classification error rate decreased with increasing number of genes, reaching zero only for fold-change in gene expression.

Table 7: Genes discriminating each treatment from all others.


Listed are the genes that were present in all three pair wise distinction calculations for each treatment group (all genes, P<0.05). There were 37 genes selected in common in each of the three pair wise comparisons of HDMTX vs. MP, HDMTX vs. HDMTX+MP and HDMTX vs. LDMTX+MP. There were 21 probe sets in common in the three pair wise comparisons for HDMTX+MP, 29 probe sets for MP alone and 9 probe sets for LDMTX+MP. The number assigned within each treatment group (e.g., HDMTX01) represents the rank of each gene by its distinction value in discriminating each treatment from the other treatments. The median fold-change in the individual treatment group (FC TX group) and the median fold-change in the other treatments combined (FC other TX groups) are shown for each gene, with minus (-) indicating genes that exhibited a decrease in expression, whereas a positive number indicates those genes that exhibited an increase in expression after treatment.

rank1 by
TX group
probe set ID accession
number
gene name median FC
TX group
median
FC other
TX groups
HDMTX01 37781_at AB023138 neurexin 2 -1.7 1.1
HDMTX02 35926_s_at AF004230 leukocyte immunoglobulin-like receptor 2.4 -1.7
HDMTX03 34751_at AI970189 KIAA0997 protein -1.4 1.5
HDMTX04 2067_f_at L22475 BCL2-associated X protein 2.5 1.1
HDMTX05 41471_at W72424 S100 calcium-binding protein A9 (calgranulin B) 4.6 -1.9
HDMTX06 38994_at AF037989 STAT induced STAT inhibitor-2 -1.6 1.0
HDMTX07 31793_at AL036554 defensin, alpha 3, neutrophil-specific 5.4 -1.4
HDMTX08 41096_at AI126134 S100 calcium-binding protein A8 (calgranulin A) 4.2 -1.3
HDMTX09 38363_at W60864 TYRO protein tyrosine kinase binding protein 1.9 -1.1
HDMTX10 41598_at AA890010 SEC22, vesicle trafficking protein -1.7 -1.2
HDMTX11 32749_s_at AL050396 filamin A, alpha (actin-binding protein-280) 2.1 1.4
HDMTX12 31506_s_at L12691 defensin, alpha 3, neutrophil-specific 9.4 -1.4
HDMTX13 39286_at D64109 transducer of ERBB2, 2 -1.4 2.0
HDMTX14 35621_at L77213 phosphomevalonate kinase 1.6 1.0
HDMTX15 41126_at AA978353 phosphoserine aminotransferase 1.7 1.0
HDMTX16 402_s_at X69819 intercellular adhesion molecule 3 1.8 1.1
HDMTX17 40362_at X61498 nuclear factor of kappa light polypeptide gene enhancer 2.4 1.1
HDMTX18 36789_f_at AF025534 leukocyte immunoglobulin-like receptor 1.8 -1.7
HDMTX19 37984_s_at M57763 ADP-ribosylation factor 6 -1.9 -1.1
HDMTX20 38973_at AB028943 HIC1-related gene on chromosome 22 -1.2 1.6
HDMTX21 679_at J04990 cathepsin G 1.7 -1.5
HDMTX22 32227_at X17042 proteoglycan 1, secretory granule 1.8 -1.4
HDMTX23 38999_s_at M86707 N-myristoyltransferase 1 -1.1 1.5
HDMTX24 34965_at AF031824 cystatin F (leukocystatin) 1.5 -1.2
HDMTX25 40877_s_at AF041080 D15F37 (pseudogene) 1.1 1.5
HDMTX26 39119_s_at AA631972 natural killer cell transcript 4 2.1 -1.3
HDMTX27 1403_s_at M21121 small inducible cytokine A5 (RANTES) 1.7 -1.2
HDMTX28 34702_f_at M27826 endogenous retroviral protease 1.1 -2.7
HDMTX29 33371_s_at U59877 RAB31, member RAS oncogene family 1.4 -1.6
HDMTX30 37105_at M16117 cathepsin G 2.4 -1.1
HDMTX31 31510_s_at Z48950 H3 histone, family 3B (H3.3B) -1.5 -1.2
HDMTX32 33661_at U66589 ribosomal protein L5 -1.8 1.0
HDMTX33 40450_at L09561 polymerase (DNA directed), epsilon -1.6 1.7
HDMTX34 41360_at AA044787 CCR4-NOT transcription complex, subunit 8 -1.5 1.0
HDMTX35 1402_at M16038 v-yes-1 Yamaguchi sarcoma viral related oncogene 1.7 -1.1
HDMTX36 2000_at U26455 ataxia telangiectasia mutated 2.6 1.2
HDMTX37 40520_g_at Y00638 protein tyrosine phosphatase, receptor type, C 1.7 1.1
HDMTX+MP01 782_at U93867 polymerase (RNA) III (DNA directed) 1.7 -1.2
HDMTX+MP02 40478_at AL021396 hypothetical protein 2.3 1.1
HDMTX+MP03 39489_g_at W27720 protocadherin 9 1.4 -1.5
HDMTX+MP04 37662_at AI701164 ubiquitin-conjugating enzyme E2G 1 -1.8 1.1
HDMTX+MP05 491_at U46116 HSPTPRG28 Human receptor tyrosine phosphatase 1.3 -2.1
HDMTX+MP06 33870_at AB029005 chromosome 5 open reading frame 7 -1.7 -1.1
HDMTX+MP07 32257_f_at AF003001 telomeric repeat binding factor (NIMA-interacting) 1 2.9 -1.1
HDMTX+MP08 322_at D88532 phosphoinositide-3-kinase, regulatory subunit, 2.0 -2.1
HDMTX+MP09 40383_at AB023200 gene from NF2/meningioma region of 22q12 2.2 -2.2
HDMTX+MP10 41667_s_at AJ006068 dTDP-D-glucose 4,6-dehydratase -1.6 1.3
HDMTX+MP11 39307_s_at X81637 clathrin light chain b -2.1 1.0
HDMTX+MP12 36729_g_at M76446 adrenergic, alpha-1D-, receptor 2.2 -1.1
HDMTX+MP13 1814_at D50683 transforming growth factor, beta receptor II 1.1 1.6
HDMTX+MP14 37563_at AB007871 KIAA0411 gene product 2.3 -1.1
HDMTX+MP15 1368_at M27492 interleukin 1 receptor, type I 2.5 -1.4
HDMTX+MP16 579_at M95724 centromere protein C 1 -1.7 1.1
HDMTX+MP17 36514_at U66469 cell growth regulatory with ring finger domain -3.5 -1.4
HDMTX+MP18 41826_at W28287 KIAA1467 protein <-2.3 <1.1
HDMTX+MP19 33353_at W26466 cDNA /gb=W26466 1.6 -2.3
HDMTX+MP20 33427_s_at AF106861 Attractin 2.6 1.0
HDMTX+MP21 33958_at T06733  cDNA /clone=HFBDX74 1.8 -1.7
HDMTX+MP22 32443_at U28687 zinc finger protein 157 (HZF22) 1.4 -1.7
LDMTX+MP01 37036_at AB002299 KIAA0301 protein 1.9 1.1
LDMTX+MP02 41218_at AB018272 KIAA0729 protein <-1.5 <1.1
LDMTX+MP03 37391_at X12451 cathepsin L -2.2 1.0
LDMTX+MP04 2081_s_at L07032 protein kinase C, theta -2.7 -1.1
LDMTX+MP05 38859_at AL080141 secretory pathway component Sec31B-1 2.7 1.5
LDMTX+MP06 795_s_at X66358 cyclin-dependent kinase-like 1 (CDC2-related kinase) -2.6 -1.1
LDMTX+MP07 1186_at D49493 growth differentiation factor 10 -2.7 -1.1
LDMTX+MP08 40607_at U97105 dihydropyrimidinase-like 2 1.4 1.0
LDMTX+MP09 40377_at AB014582 KIAA0682 gene product 2.8 1.3
MP01 35432_at AF074723 RNA polymerase II transcriptional regulation mediator 2.4 -1.0
MP02 37244_at AA746355 ubiquitin carboxyl-terminal esterase L3 1.8 -1.1
MP03 40942_g_at W27026 vesicle-associated membrane protein-associated protein -2.4 1.0
MP04 38390_at Z34975 low density lipoprotein receptor defect C 1.7 1.0
MP05 38414_at U05340 CDC20 (cell division cycle 20) -1.1 -1.9
MP06 33134_at AB011083 adenylate cyclase 3 1.5 1.0
MP07 35116_at X80821 KIAA0874 protein -1.8 1.1
MP08 39927_at U17032 Rho GTPase activating protein 5 1.5 -1.2
MP09 1944_f_at AF001359 DNA mismatch repair protein (hMLH1) alternatively spliced 2.3 -1.2
MP10 39199_at W28661 cDNA /gb=W28661 <-2.4 <-1.4
MP11 34886_at L02320 Radixin -1.6 1.1
MP12 36694_at AF043472 potassium voltage-gated channel, delayed-rectifier 1.8 -1.2
MP13 36297_at X55544 activating transcription factor 1 1.4 -1.2
MP14 35227_at U72066 retinoblastoma-binding protein 8 -1.5 1.0
MP15 36225_s_at W27611 splicing factor proline/glutamine rich 1.7 -1.4
MP16 37077_at D13243 pyruvate kinase L -2.7 -1.3
MP17 39637_at U14528 solute carrier family 26 (sulfate transporter) 2.5 -1.3
MP18 37624_at M29458 Human carbonic anhydrase III -2.9 1.1
MP19 37967_at AF000424 lymphocyte antigen 117 -1.0 1.5
MP20 35005_at AF051941 nucleoside diphosphate kinase type 6 1.6 -1.3
MP21 1106_s_at M12959 T cell receptor alpha locus -1.3 1.5
MP22 39775_at X54486 serine (or cysteine) proteinase inhibitor, -3.2 1.0
MP23 37553_at D50863 testis-specific kinase 1 2.0 1.3
MP24 41188_at W28186 putative integral membrane transporter -2.4 1.0
MP25 41244_f_at X80910 protein phosphatase 1, catalytic subunit, beta isoform -2.1 -1.2
MP26 39674_r_at AB011792 extracellular matrix protein 2 -2.4 -1.2
MP27 34927_at M28826 CD1B antigen, b polypeptide 1.5 -1.6
MP28 32904_at M28393 perforin 1 (pore forming protein) -1.9 1.2
MP29 33490_at L27071 TXK tyrosine kinase 1.8 -1.2

1rank ordered by distinction values