Application of Germline IGH Probes in Real-Time Quantitative PCR

Home , VH2

Leukemia (2000) 14, 1426–1435  2000 Macmillan Publishers Ltd All rights reserved 0887-6924/00 $15.00 www.nature.com/leu Application of germline IGH probes in real-time quantitative PCR for the detection of minimal residual disease in acute lymphoblastic leukemia OJHM Verhagen1, MJ Willemse2, WB Breunis1, AJM Wijkhuijs2, DCH Jacobs2, SA Joosten2, ER van Wering3, JJM van Dongen2 and CE van der Schoot1,4

1Department of Immunohematology, Central Laboratory of the Netherlands Red Cross Blood Transfusion Service and Laboratory of Experimental and Clinical Immunology, Academic Medical Center, University of Amsterdam, Amsterdam; 2Department of Immunology, Erasmus University Medical Center Rotterdam, Rotterdam; 3Dutch Childhood Leukemia Study Group, The Hague; and 4Department of Hematology, Academic Medical Center, Amsterdam, The Netherlands

Large-scale clinical studies on detection of minimal residual tative methods for the detection of clone-specific immunoglob- disease (MRD) in acute lymphoblastic leukemia (ALL) have ulin (Ig) and T cell receptor (TCR) gene rearrangements,5–8 shown that quantification of MRD levels is needed for reliable which are present in the vast majority of pediatric and adult MRD-based risk group classification. Recently, we have shown 5,9,10 that ‘real-time’ quantitative PCR (RQ-PCR) can be applied for ALL patients. The PCR technique has the ability to amplify this purpose using patient-specific immunoglobulin (Ig) and T target DNA up to a plateau, implying that after 35 to 40 cycles cell receptor (TCR) gene rearrangements as PCR targets with it is not possible to define precisely the initial amount of target TaqMan probes at the position of the junctional region and two DNA. Also more quantitative methods, like competitive PCR germline primers. Now, we tested an alternative approach on and limiting dilution are based on post-PCR ‘end-point’ analy- 35 immunoglobulin heavy chain (IGH) gene rearrangements, by 11–15 designing three germline J TaqMan probes to be used in com- sis. These techniques require serial dilutions and the H analysis of multiple replicates, which introduce variability and bination with one of six corresponding germline JH primers and one allele specific oligonucleotide (ASO) primer complemen- are too difficult and too time-consuming to be performed tary to the junctional region. In nine cases in which both routinely. approaches were compared, at least similar (n = 4) or slightly The novel ‘real-time’ quantitative PCR technology (RQ- = higher (n 5) maximal sensitivities were obtained using an ASO PCR) circumvents the above problems, because the PCR pro- primer. The ASO primer approach reached maximal sensitivities of at least 10−4 in 33 out of 35 IGH rearrangements. The duct accumulation is monitored throughout the complete PCR 16 Ј → Ј reproducible range for accurate quantification spanned four to process. The method is based on the 5 3 nuclease five orders of magnitude in 31 out of 35 cases. In 13 out of activity of Taq DNA polymerase and an internal dual-labeled 35 rearrangements the stringency of PCR conditions had to be fluorogenic probe with a 5Ј-reporter dye and a 3Ј-quencher increased to remove or diminish background signals; this only dye.17 During PCR, the 5Ј → 3Ј nuclease activity of Taq DNA concerned the frequently occurring JH4, JH5 and JH6 gene polymerase cleaves the hybridized probe and thereby separ- rearrangements. After optimization of the conditions (mainly by increasing the annealing temperature), only occasional aspec- ates the reporter dye from the quencher dye, resulting in emis- ific amplification signals were observed at high threshold cycle sion of a fluorescent signal that increases during each sub-

(CT) values above 42 cycles and at least six cycles above the sequent PCR cycle. The real-time detection of fluorescence CT value of the detection limit. Hence, these rare aspecific sig- intensity generates quantitative data based on the early cycles nals could be easily discriminated from specific signals. We when the fidelity of PCR amplification is the highest. This conclude that the here presented set of three germline JH Taq- quantification can be performed over a large dynamic range Man probes and six corresponding germline JH primers can be used to develop patient-specific RQ-PCR assays, which allow of four to five orders of magnitude. accurate and sensitive MRD analysis in almost all IGH gene Previously, we have shown that RQ-PCR can be used for rearrangements. These results will facilitate standardized RQ- the quantification of MRD levels using patient-specific Ig/TCR PCR analysis for MRD detection in large clinical studies. Leuke- gene rearrangements as PCR targets with allele-specific oli- mia (2000) 14, 1426–1435. gonucleotide (ASO) probes.18 All 12 patient-specific ASO Keywords: acute lymphoblastic leukemia (ALL); IGH gene probes developed for Ig/TCR gene rearrangements in four rearrangements; real-time quantitative PCR (RQ-PCR); minimal residual disease (MRD); germline J TaqMan probes precursor-B-ALL were able to detect and to quantify leukemia- H derived DNA. Sensitivities of 10−3 to 10−5 could be obtained, which were comparable to the dot-blot method but less sensitive than liquid hybridization. Other groups have shown that Introduction RQ-PCR can be used for quantification of MRD levels using leukemia-specific chromosome aberrations as PCR targets, ie Several prospective clinical ‘minimal residual disease’ (MRD) t(9;22), t(14;18), t(8;21) and t(12;21).19–23 studies in childhood acute lymphoblastic leukemia (ALL) have Now, we tested an alternative RQ-PCR approach for IGH shown that it is important to determine precisely the level of gene rearrangements, by positioning the TaqMan probe and MRD at early remission time points for discrimination one of the primers at germline J segments in combination between low- and high-risk patients.1–4 Information about the H with an ASO primer complementary to the junctional region. kinetics of tumor load reduction has allowed the recognition This ASO primer approach theoretically results in more sensi- of three MRD-based risk groups with significant differences in tive MRD detection compared to usage of germline primers, relapse rate.3 because no competition can occur with the amplification of Until now, most PCR-based MRD studies used semi-quanti- similar rearrangements in normal cells. Furthermore, the design of germline TaqMan probes is more cost-effective, since it would not be necessary to design TaqMan probes per Prof JJM van Dongen, MD, PhD, Department of Immunology, Erasmus University Rotterdam, PO Box 1738, 3000 DR Rotterdam, The patient but per type of Ig/TCR gene rearrangement. Firstly, we Netherlands; Fax: 31-10-4089456 compared the alternative ASO primer approach with our Received 5 January 2000; accepted 18 February 2000 initial ASO probe approach in nine IGH gene rearrangements. RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1427 Secondly, we performed RQ-PCR analysis of 35 IGH gene DH–JH junctional regions (Figure 1, Table 1). In contrast to rearrangements using the ASO primer approach. germline oligonucleotides, primer–dimer formation with ⌬G higher than −3.5 kcal/mol, stretches of four or more Gs, and GC-rich 3Ј-ends, could not always be avoided in ASO pri- Materials and methods mers. Amplicon size ranged from 74 to 178 bp. In two cases (5978 and 6264; Table 1) shorter ASO primers with a Tm of Patients and cell samples 53°C were developed to increase specific annealing. For the ASO probe approach we designed a probe comp- Bone marrow (BM) samples were derived from 34 childhood lementary to the junctional region (Table 1) and matching 18 precursor-B-ALL patients at diagnosis. The mononuclear cell germline VH- and JH-primers (Table 2). For the ASO primer (MNC) fractions were isolated by Ficoll–Hypaque (1.077 approach we designed three probes to the germline JH gene 3 g/cm ; Pharmacia, Uppsala, Sweden) density gradient centri- segments. Probe T-JH-1.2.4.5 was designed to the JH1, JH2, fugation prior to cryopreservation. The MNC usually con- JH4 and JH5 gene segments, with one mismatch for JH2 in the tained у90% leukemic blasts as assessed by routine immuno- middle of the probe (C instead of T). Probe T-JH3 and probe

phenotyping. Genomic DNA was extracted from MNC with T-JH6 were designed to the JH3 and JH6 gene segments, the QIAamp Blood kit (Qiagen, Chatsworth, CA, USA), as respectively. Six JH gene segment-speciﬁc reverse primers described before.24 Clonality of the B cells was proven by were designed complementary to the intron downstream of

Southern blot analysis of IGH genes in BglII digests with the each JH gene segment (Table 2). The ASO forward primers 25,26 H24-JH or IGHJ6 probe. were designed at the same strand as the germline JH probes. Patients included in this study were randomly chosen At least a part of each ASO primer was complementary to the except for positive selection of at least two different rearrange- junctional region.

ments for each of the JH gene segments (11 JH6, eight JH5, nine JH4, two JH3, two JH2 and three JH1 gene rearrangements). Five patients (5538, 5627, 5807, 6198 and 6437) showed an oligo- RQ-PCR analysis clonal IGH gene rearrangement pattern as assessed by South- ern blot analysis (Table 1). All studied IGH gene rearrange- For RQ-PCR analysis the TaqMan 1000 Reactions Gold with ments represented major clones and were representative for Buffer A kit (PE Biosystems) was used. Reaction mixtures of rearrangements found in precursor-B-ALL. In one patient 50 ␮l contained the TaqMan buffer A with the ROX dye as ␮ (5978) two IGH gene rearrangements were used. passive reference, 5mM MgCl2, 200 M each dNTPs, 300 nM primers (ASO primer approach) or 900 nM primers (ASO probe approach), 100 nM probe, 1.25 U AmpliTaq Gold (PE Identification of PCR targets Biosystems), 10% glycerol and 500 ng DNA. The reaction conditions were 10 min 95°C followed by 50 cycles of 15 s ° ° Complete IGH gene rearrangements were amplified with VH- 95 C and 1 min 60–69 C. Fluorescence data were collected family-specific framework region 1 (FR1) primers VH1/7, during the annealing/extension phase of every cycle, using the VH2, VH3, VH4/6, VH5 and a consensus FR3 in combination ABI PRISM 7700 Sequence Detection System containing a 96- with a consensus joining primer JH21.27–29 Usually, major well thermal cycler (PE Biosystems). All RQ-PCR experiments PCR bands on ethidium bromide-stained polyacrylamide gels were performed in duplicate. were excised and eluted in 10mM Tris pH 8.0. After a second The fluorescence intensity was normalized using the passive round PCR with the same primer combination, the PCR pro- reference ROX. A real-time amplification plot was generated duct was directly used for the sequence reaction using the Big using the normalized reporter signal (Rn). The PCR product Dye Terminator Cycle Sequencing Ready Reaction kit and an yield or ⌬Rn was defined as the Rn minus the baseline signal ABI PRISM 377 Automated Sequencer (PE Biosystems, Foster established in the first few cycles of the PCR. The threshold

City, CA, USA). VH,DH and JH gene segments were identiﬁed cycle (CT) was that cycle where PCR product was ﬁrst using DNAPLOT software (W Mu¨ller, H-H Althaus, University detected. The entire RQ-PCR analysis was performed in a of Cologne, Germany) by searching for homology with all closed tube.

known human germline VH,DH and JH sequences obtained To check for the quantity and quality (ampliﬁability) of from the VBASE directory of human Ig genes (http://www.mrc- DNA, we used RQ-PCR analysis of the albumin gene, as cpe.cam.ac.uk/imt-doc/).30–32 described before.18 The RQ-PCR results of the albumin gene in the various DNA samples differed less than a factor 2, implying that the amount of DNA per tube and the ampliﬁ- Design of ASO primers and ASO probes ability of the DNA samples were very comparable.

All primers and probes were designed with the Primer Express version 1.0 software (PE Biosystems) and the OLIGO 6.1 PCR target sensitivity software (Wojciech Rychlik, Molecular Biology Insights, Cas- cade, CO, USA). In general, TaqMan probes did not have a To determine the efﬁciency of ampliﬁcation and sensitivity of G at the 5Ј-end, contained more Cs than Gs, and the melting the PCR target, diagnosis DNA was diluted according to temperature (Tm, nearest neighbour method) was 67 to 70°C, square root 10-fold dilutions into MNC DNA, from 10−1 down and 8 to 10°C above the Tm of the primers, according to the to 3.16 × 10−6. To avoid skewed gene rearrangement patterns manufacturer’s guidelines. FAM was chosen as reporter dye and to obtain a bulk of the polyclonal control, the normal at the 5Ј-end of the probe and TAMRA as the quencher dye MNC DNA was isolated from equivalent mixtures of PB-MNC at the 3Ј-end. Oligonucleotides that were likely to form from 10 different healthy donors. The dilution series of diag- secondary structures were avoided. nosis DNA was subjected to RQ-PCR analysis together with

ASO’s were developed complementary to the VH–DH or negative controls (H2O and MNC).

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1428 Table 1 PCR targets and ASO’s used for RQ-PCR analysis

Patient Southern Gene rearrangementb ASO codec ASO sequenced code blot dataa 5Ј → 3Ј

3954 ND VH1.18(0)-7-(0)JH5b F-VH3-3954a TGTGCGAGAGCCACCCA 5181 R/R VH3.43(2)-8-(5)D6.19(2)-(5)JH4b F-VH3-5181a ATTACTGTGCAAAAGTAAGTAGGTAG 5221 R/G VH4(4)-9-(6)D6.6(3)-(0)JH6b F-VH4-5221a GCTTACTGGAAAGCAGCTCGATTAC 5255 R/R VH3(1)-19-(0)D2.2(4)-14-(5)D2.15(4)-4-(3)JH4 F-VH3-5255a GGGGGAAAAGCTTACTAGGATATTGT 5282 R/R VH1(2)-19-(2)JH5 F-VH1-5282a TGCGAGAAGAACAGCTACAGTACA 5292 R/R VH3.13(0)-1-(3)D1.7(0)-3-(0)JH3b F-VH3-5292a CAAGAGATAACTGGAACTACGTGAT e 5400 R/R/R VH3(1)-9-(4)D2.15(1)-11-(0)JH1 F-VH3-5400a CTGTGCGAGGTCCCCTGTT 5414 R/R VH(0)-9-(0)D.2.2(12)-6-(3)JH5b F-VH-5414a TTACTGTGCGAGAGGCCTTAGAA f w 5538 R/R/R VH3(0)-10-(8)D2.8(9)-2-(0)JH4a F-VH3-5538a GAGAGGGTCCCGGGGTACTA 5566 R/R VH2(0)-20-(4)D(2)-4-(7)JH6 F-VH2-5566a ATCCCCGCGTTTCGACTAT 5579 R/G VH(0)-6-(0)D2.15(0)-5-(12)JH4b F-VH-5579a TGTATTACTGTGCGAGAGATTACAGGA 5594 R/R/G VH1(0)-28-(2)D3.22(6)-16-(2)JH5b F-VH1-5594a GAGATACACAACCCCCCACAAT f 5627 R/Rw VH3(0)-20-(6)D2.2(5)-9-(9)JH6 F-VH3-5627b AGGGTAGGCTAGGGGGTTGTA e 5648 R/R/R VH3.33(8)-11-(1)D1.7(4)-2-(5)JH5b F-VH3-5648a CGGCTGTGTATTACTATACCTAAGGGGTA 5771 R/R VH3(12)-9-(8)JH2 F-VH3-5771a ACGGCTGTGTATTATTCCTGGAG 5772 R/R VH(1)-17-(0)D2.2(8)-5-(12)JH6c F-VH-5772a GGGCGAGAGATCGTTGGA f w w 5807 R/R/R /R VH6.1(1)-2-(6)D2.2(9)-10-(1)JH5 F-VH6-5807a TACCAGCCGCAAATGTTCA 5902 R/G VH(0)-11-(4)D3.3(10)-5-(2)JH6b F-VH-5902a CGATTTTTGGAGTGGCTCCTTA g 5978 R/R VH3(0)-4-(2)D6.6(1)-3-(5)JH4 F-VH3-5978a GCAGGTATAGCAGCTCGTCGTATT g 5978 R/R VH1(0)-4-(2)JH5 F-VH1-5978a GTGTATTACTGTGCGAGAGGTTTTAAC VH1(0)-4-(2)JH5 F-VH1-5978b ATTACTGTGCGAGAGGTTTTAAC 6110 R/G VH(3)-19-(8)D2.8(2)-9-(2)JH6c F-VH-6110a ATTACTGTGCGGTACGAGAGTCCTA 6135 R/R VH1.8(9)-6-(2)D3.22(2)-3-(5)JH1 F-VH1-6135a GCCGTGTATTACGGGCTCATT 6148 R/R VH2.10p(0)-9-(6)JH6c Tr-VH2-6148a TGTAGTAGTAGTAGTCCCCAGTTGTCTCCTTGC VH2.10p(0)-9-(6)JH6c F-VH2-6148a GCAAGGAGACAACTGGGGACTA 6154 NP VH3.15(0)-21-(2)D7-27(0)-10-(4)JH4b T-VH3-6154a CCACAGATCCCATAGGTATTCCCTCCCG VH3.15(0)-21-(2)D7-27(0)-10-(4)JH4b F-VH3-6154a CTGTACCACAGATCCCATAGGTATTC f w w 6198 R/R/R /R VH6.1(5)-8-(7)D2.2(3)-(8)JH6 F-VH6-6198a TTACTGTGCAGGGTCCGATGT 6216 R/G VH3.15(3)-9-(6)D3.22(5)-(6)JH4b Tr-VH3-6216a CAGTAGTCAATAACCACTACTATCATAG GGACATGGTACA

VH3.15(3)-9-(6)D3.22(5)-(6)JH4b F-VH3-6216a GGCCATGTATTACTGTACCATGTCC 6264 D/R VH6.1(0)-13-(13)D4(2)-2-(7)JH6c Tr-VH6-6264a TGGTACTCCCCTAACCCCACTCTTGCA VH6.1(0)-13-(13)D4(2)-2-(7)JH6c F-VH6-6264a GTATTACTGTGCAAGAGTGGGGTTAG VH6.1(0)-13-(13)D4(2)-2-(7)JH6c F-VH6-6264b TACTGTGCAAGAGTGGGGTTAG 6379 ND VH3.30(0)-24-(5)D6.13(0)-3-(5)JH6b Tr-VH3-6379a TACCAGCTGCTGCTATCTGAAGGGC VH3.30(0)-24-(5)D6.13(0)-3-(5)JH6b F-VH3-6379a GGGAGGGGCCCTTGAGAT 6396 R/R VH3.53(0)-28-(2)JH5b Tr-VH3-6396a CAATTGTAGCCATCTCTGGTTGATCTCTCG VH3.53(0)-28-(2)JH5b F-VH3-6396a CAGAGATGGCTACAATTGAGGGTAA 6401 R/R VH4.39(0)-11-(8)D3.10(3)-2-(3)JH3b F-VH4-6401a TGTGCGAGACATTAACCGCTTAT f w 6437 R/R/R VH3.13(4)-4-(5)D2.21(5)-10-(17)JH6b Tr-VH3-6437a CCATACCACACGCGGAGAGCAGTCA VH3.13(4)-4-(5)D2.21(5)-10-(17)JH6b F-VH3-6437b GTGACTGCTCTCCGCGTGT 6480 R/R VH1.2(0)-33-(8)JH4b Tr-VH1-6480a CCACTTATTTTAAATGATGGGGTACCTCTCGC VH1.2(0)-33-(8)JH4b F-VH1-6480a GCGAGAGGTACCCCATCATTT 6501 R/R VH6.1(16)-18-(2)D7.27(1)-5-(4)JH4b T-VH6-6501a ACGGCTGTACCCTTTAAAAACCCGTAAACTG VH6.1(16)-18-(2)D7.27(1)-5-(4)JH4b F-VH6-6501a GCTGTACCCTTTAAAAACCCGTAA 6504 R/R VH3.g(19)-18-(5)D2.21(8)-(16)JH2 F-VH3-6504a GTCTTCCGGGATCGACCAT 6513 D/R VH3-15(0)-39-(6)JH1 F-VH3-6513a CGTGGGAGCACCTTTCAGA

a 25,26 Southern blots with BglII digested DNA were hybridized with the H24-JH or IGHJ6 probes. G, allele in germline conﬁguration; R, rearranged allele; Rw, weak rearranged allele; D, deleted allele; ND, not done. bThe most recent nomenclature was used to designate IGH gene segments.30–32 The number in parentheses indicates the loss of nucleotides of the gene segment. The number between dashes indicates the amount of N-nucleotides inserted. The position of the ASO in either

the VH–DH or DH–JH junctional region is underlined. cF, forward primer; T, TaqMan probe; Tr, TaqMan probe designed at the reverse strand. dUnderlined sequences represent randomly inserted nucleotides of the junctional region. ePatients 5400 and 5648 have a trisomy for chromosome 14. fFive patients (5538, 5627, 5807, 6198 and 6437) were oligoclonal, as assessed by Southern blot analysis of the IGH genes. gBoth IGH gene rearrangements of patient 5978 were used for the RQ-PCR study.

⌬ To be able to compare the CT and Rn values of the differ- was 1.5 cycles, with a maximal CT value of 40 cycles. In this ent primer/probe combinations, the RQ-PCR data have also reproducible range the standard curve should have a corre- been analyzed using a fixed threshold of 0.01. For analyzing lation coefficient of at least 0.95 for precise quantification. the slope and correlation coefficient of each standard curve, The maximal sensitivity (sensitivity threshold) of a automatically set (optimal) thresholds were used. primer/probe combination was defined as the last dilution of The reproducible range of a primer/probe combination was diagnosis DNA in which at least one of the duplicate dilution defined as that part of the standard curve in which the maxi- samples resulted in a positive fluorescent signal with a maxi-

mal difference in CT value of the duplicate dilution samples mal CT value of 40 cycles. Furthermore, these CT values had

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1429

Figure 1 Schematic diagram of an IGH gene rearrangment and ASO probe/primer design. For the ASO probe approach, TaqMan probes were designed complementary to the junctional region (VH–DH or the DH–JH), in combination with two germline VH- and JH-primers. For the Ј ASO primer approach, TaqMan probes were designed to the 3 -end of the JH gene segments in combination with one ASO primer complementary to the junctional region and one primer complementary to the downstream sequence of the JH gene segments. Primers are indicated by arrows and probes by bars (see Table 2 for sequences).

Table 2 Germline primers and probes used for RQ-PCR analysis Results

a Ј b Ј Ј Oligonucleotide 5 position Sequence 5 –3 Comparison of ASO probe vs ASO primer

ASO probe approach F-D2.21-cons1 −49 GGGGCTCGTGTCACTGTGAG To compare the ASO probe approach using germline VH and − F-VH1-cons1 51 GCTGAGCAGGCTGAGATCTGA JH primers with the ASO primer approach using germline JH F-VH2-cons1 −56 AACAGTGATCAACATGGACATTGTG probes, we tested nine IGH gene targets (four JH4, one JH5 F-VH3-cons2 −32 GAGGACACAGCCGTGTATTACTGT and four J 6 rearrangements) obtained from precursor-B-ALL − H F-VH3-cons3 60 CTATCTGCAAATGGACAGTCTGAAA samples (Table 3). For all nine IGH rearrangements, it was F-VH3-cons4 −99 CCGATTCACCATCTCCAGACA F-VH3-cons5 −26 ACGGCTGTGTATTACTGTGCGA possible to develop primer/probe combinations for both F-VH6-cons1 −35 CCCGAGGACACGGCTGT approaches, which resulted in successful amplification and F-VH6-cons2 −84 CCCAGACACATCCAAGAACCA real-time detection. R-JH4b-cons1 +45 AGGAGACGGTGACCAGGGTT The reproducible ranges of the ASO primer approach were R-JH5b-cons1 +53 ACCTGAGGAGACGGTGACCA comparable to those of the ASO probe approach; in three R-JH6b-cons1 +52 GTGACCGTGGTCCCTTGG cases it was equal for both approaches, in five cases the ASO R-JH6b-cons2 +37 TGGCCCCAGACGTCCATA primer approach resulted in a longer range, and in one case R-JH6c-cons1 +32 GCCCCAGACGTCCATGTAGTA the ASO probe approach resulted in a longer range. The C + T R-JH6c-cons2 41 GGTCCCTTTGCCCCAGAC values at a fixed threshold of 0.01 had in all cases a range ASO primer approach between 20 and 36, except for case 6148 (CT range of 29– −5 T-JH1.2.4.5 +31/32/27/30c CCCTGGTCACCGTCTCCTCAGGTGd 37). Only twice, in the 10 dilutions of cases 6264 and 6437, T-JH3 +20 CAAGGGACAATGGTCACCGTCTCTTCA was a positive signal measured above 40 cycles (Table 3). T-JH6 +43 CACGGTCACCGTCTCCTCAGGTAAGAA Both positive signals were found using the ASO probe R-JH1-intron +25 CGCTATCCCCAGACAGCAGA approach in the absence of background signals in four control R-JH2-intron +47 GGTGCCTGGACAGAGAAGACT MNC DNA samples. Hence, we assume that these positive R-JH3-intron +23 AGGCAGAAGGAAAGCCATCTTAC signals represented specific detection. R-JH4-intron +36 CAGAGTTAAAGCAGGAGAGAGGTTGT In all cases except for cases 6148 and 6264, the C values + T R-JH5-intron 24 AGAGAGGGGGTGGTGAGGACT of the 10−1 dilution were one to three cycles lower in the ASO R-JH6-intron +35 GCAGAAAACAAAGGCCCTAGAGT primer approach as compared to the ASO probe approach. The maximal sensitivity of the ASO primer approach was in aF, forward primer; R, reverse primer; T, TaqMan probe. bPosition of the primer/probe upstream (−) or downstream (+) rela- five of the nine cases comparable to those of the ASO probe tive to the RSS of the gene segment, including the oligonucleotide. approach, while in four cases the ASO primer approach was c Position of the germline T-JH1.2.4.5 probe per JH gene segment. more sensitive, either 3.16-fold (two cases) or 10-fold (two dThe single mismatch in the T-JH1.2.4.5 probe (C instead of a T) cases). Only in case 6148 the maximal sensitivity (10−3) was with the JH2 gene segment is underlined. too low with respect to the theoretical maximum. In all other eight cases a maximal sensitivity of 10−4 to 3.16 × 10−6 was reached (Table 3). ⌬ to be at least six cycles lower than the CT values found with The Rn values (yield of generated PCR product) in the any aspecific amplification in normal MNC DNA. Theoreti- reproducible range were lower in all ASO probe cases as cally, the maximal sensitivity should be between 10−4 compared to ASO primer cases. Moreover, the ASO probe (approximately 16 copies of target gene) and 10−5 approach showed a reduction of ⌬Rn values in higher (approximately 1.6 copies of target gene), based on the dilutions (Figure 2). This is probably owing to competition for assumption that a cell contains 6.25 pg DNA and the fact that germline primers during amplification of the patient-specific we used 0.5 ␮g of DNA per tube and two tubes per test (1 clonal IGH gene rearrangement and similar rearrangements in ␮g in total). normal cells. However, the polyclonal IGH gene rearrange- For all experiments, the specificity of amplification and ments are not detected by the ASO probe due to a different detection was tested on four control MNC DNA. For some junctional region. Increase of the primer concentration to 900 cases, an additional eight or 20 MNC DNA were tested nM, to avoid exhaustion of the germline primers in the ASO together with the 10−3 dilution of the diagnosis DNA to check probe approach did not result in higher ⌬Rn values (data for the amplification efficiency. The CT value did not differ not shown). more than one cycle with the original experiments. During the log phase of an optimal amplification, the

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1430 Table 3 Comparison of RQ-PCR data obtained with the ASO probe and the ASO primer approach

Patient ASO Ta Reproducible rangeb,c Maximal sensitivityb,d Aspeciﬁcityb,e

code (°C) (CT) (lowest CT) ⌬ Dilutions CT Rn Slope

6148 probe 60 10−1–10−3 29–37 0.5–0.2 3.9 10−3 (37) 0/4 MNC primer 62 10−1–10−3 29–37 1.0–0.8 3.4 10−3 (37) 1/4 MNC (48) 6154 probe 60 10−1–10−4 24–34 0.8–0.01 3.3 3.16 × 10−5 (36) 0/4 MNC primer 60 10−1–3.16 × 10−5 23–34 1.4–0.9 3.2 3.16 × 10−6 (36) 0/4 MNC 6216 probe 60 10−1–10−3 26–34 0.7–0.09 3.4 3.16 × 10−4 (34 + 36) 0/4 MNC primer 62 10−1–3.16 × 10−4 25–35 1.7–1.0 3.4 3.16 × 10−5 (35) 0/4 MNC 6264 probe 60 10−1–3.16 × 10−5 22–36 1.2–0.1 3.9 10−5 (36 + 42) 0/4 MNC primer 60 10−1–3.16 × 10−5 23–36 1.1–0.8 3.7 10−5 (36) 1/4 MNC (48) 6379 probe 60 10−1–3.16 × 10−4 25–31 0.7–0.03 3.5 3.16 × 10−5 (39) 0/4 MNC primer 62 10−1–3.16 × 10−5 22–34 1.1–0.8 3.4 10−5 (37) 0/4 MNC 6396 probe 60 10−1–10−3 24–32 1.0–0.1 3.7 10−4 (34 + 36) 0/4 MNC primer 60 10−1–10−4 22–35 2.0–1.6 3.7 10−4 (35) 0/4 MNC 6437 probe 60 10−1–10−4 24–34 1.0–0.1 3.6 10−5 (44) 0/4 MNC primer 60 10−1–3.16 × 10−4 23–32 1.2–0.8 3.5 10−5 (36) 0/4 MNC 6480 probe 60 10−1–10−4 23–36 0.7–0.03 3.9 3.16 × 10−5 (40) 0/4 MNC primer 60 10−1–10−5 22–35 2.0–1.5 3.3 10−5 (35) 0/4 MNC 6501 probe 60 10−1–3.16 × 10−5 21–33 1.1–0.2 3.5 3.16 × 10−6 (35) 0/4 MNC primer 60 10−1–3.16 × 10−5 20–32 1.8–0.9 3.4 3.16 × 10−6 (34) 0/4 MNC

aT, annealing/extension temperature. b To be able to compare the CT values of the different primer/probe combinations, the RQ-PCR data have been analyzed using a ﬁxed threshold of 0.01. Only for analyzing the slope of each standard curve optimal thresholds have been used. cThe reproducible range of a primer/probe combination was deﬁned as that part of the standard curve in which the maximal difference

in CT value of the duplicate dilution samples was 1.5 cycles with a maximal CT of 40 cycles. dThe maximal sensitivity of a primer/probe combination was deﬁned as the last dilution of diagnosis DNA in which at least one of the

amount of PCR product will duplicate each cycle. In principal, All seven combinations for the rarely used JH1, JH2 and JH3 after 3.3 cycles the amount of PCR product will increase 10 IGH gene rearrangements resulted in sensitive and specific times, resulting in a theoretical slope of the standard curve of amplification and real-time detection under standard con- 3.3. All 18 primer–probe combinations showed slopes ditions, ie an annealing/extension temperature of 60°C (Table between 3.2 and 3.9, indicating good amplification 4). The maximal sensitivity ranged between 10−4 and 3.16 × efficiencies (Table 3). The slopes with the ASO primer 10−6. The reproducible range for accurate quantification approach (3.4 ± 0.16) were slightly better than the slopes with spanned over up to at least four orders of magnitude in all ± the ASO probe approach (3.6 0.22). The correlation coef- cases (Table 4). No positive signals were found with CT values ficients of all 18 standard curves in the reproducible range above 39 cycles. Aspecific amplification of JH1, JH2 and JH3 were excellent (0.96–1.00). rearrangements was not found in MNC DNA. All ⌬Rns (1.4 ± 0.4), slopes (3.2 ± 0.2 ), and correlation coefficients (0.99) of the standard curves indicated robust amplification and PCR Performance of RQ-PCR analysis using ASO primers product detection. The mismatch in the middle of the T-

in 35 IGH gene rearrangements JH1.2.4.5 probe with the JH2 gene apparently did not hamper the detection of the ampliﬁed product, as shown in patients

Next, we tested whether RQ-PCR analysis with germline JH 5771 and 6504. TaqMan probes in combination with an ASO primer is broadly In 15 out of 28 primer/probe combinations for the more

applicable in precursor-B-ALL. For this purpose we tested a frequently occurring JH4, JH5 and JH6 IGH gene rearrange- large series of 35 IGH gene rearrangements from 34 patients, ments, no aspecific amplification with CT values below 45 including the nine rearrangements described above. cycles was found in control MNC DNA using standard con- Before defining the sensitivity, we first tested in pilot experi- ditions with an annealing/extension temperature of 60°C ments whether the primer/probe combinations were able to (Table 4). The maximal sensitivity ranged between 10−4 and amplify and detect the junctional regions in diagnosis DNA 3.16 × 10−6. The reproducible range for accurate quantifi- using normal MNC DNA as negative control. These tests were cation spanned at least four orders of magnitude in all cases.

carried out under standard conditions. All 35 primer/probe No positive signals were found with CT values above 38 combinations resulted in robust amplification. However in 13 cycles. All ⌬Rns (1.7 ± 0.4), slopes (3.3 ± 0.1), and correlation out of 35 cases aspecific amplification and detection of poly- coefficients (0.95–1.00) of the standard curves indicated clonal IGH gene rearrangements in the control MNC DNA robust amplification and PCR product detection.

took place, with CT values below 45 (range 35–45). For these In the other 13 primer/probe combinations for the JH4, JH5 cases, the specificity of the RQ-PCR had first to be increased and JH6 rearrangements, aspecific amplification of polyclonal (see below). IGH gene rearrangements in MNC DNA took place under

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1431

Figure 2 Real-time amplification plots of precursor-B-ALL patient 6480. RQ-PCR analysis of the serial dilutions of diagnosis DNA into MNC DNA were performed in duplo via the ASO probe approach (a) and via the ASO primer approach (b). The ASO primer approach resulted in higher ⌬Rn values, a larger reproducible range, and a higher maximal sensitivity than the ASO probe approach (see also Table 3). standard conditions with an annealing/extension temperature below 45. In the serial dilutions of diagnosis DNA no positive ° of 60 C, resulting in CT values below 45 (range 35–45). More signals were found with CT values above 39, and the positive stringent conditions to avoid this aspecific amplification were signals were at least nine cycles lower than the CT values necessary. In practice two straightforward possibilities exist. found in control MNC DNA. After optimization, almost all 13 Firstly, by increasing the annealing temperature of the ampli- primer/probe combinations showed a maximal sensitivity and fication reaction, and secondly by reducing the length of the a reproducible range for accurate quantification, which was ASO primer, particularly the part of the primer that is comp- comparable to those of the 22 primer/probe combinations lementary to germline gene segments. For cases 5978 and tested under standard conditions. Only for cases 5978 and 6264 shorter ASO primers were developed, both with a Tm 6148 a limited sensitivity of 10−3 was reached (Table 4). of 53°C. For case 6264 optimal annealing with the new primer was achieved at 60°C. In all other cases, we preferred to increase the annealing temperature, because in this way the Specificity of the primers/probe combinations same ASO primer could be used. The annealing temperature was increased in steps of 2 or 3°C, ranging from 62°Cto69°C, For all primer/probe combinations, the specificity of amplifi- till the aspecific amplification of polyclonal IGH gene cation and detection was tested on at least four control MNC rearrangements in control MNC DNA did not reach CT values DNA. For the ASO probe approach, no positive fluorescent

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1432 Table 4 RQ-PCR analysis using the ASO primer approach in 35 IGH gene rearrangements

a b,c b,e JH Patient ASO primer T Reproducible range Maximal Aspeciﬁcity b,d gene code (°C) sensitivity (lowest CT) ⌬ (CT) Dilutions CT Rn Slope

−1 −4 −5 JH1 5400 F-VH3-5400a 60 10 –10 24–34 1.2–0.6 2.9 3.16 × 10 (36) 0/4 MNC 6135 F-VH1-6135a 60 10−1–3.16 × 10−5 23–36 1.9–1.0 3.3 10−5 (36 + 36) 0/4 MNC 6513 F-VH3-6513a 60 10−1–10−4 22–34 1.9–0.9 3.3 10−5 (37) 0/4 MNC

−1 −4 −5 JH2 5771 F-VH3-5771a 60 10 –10 26–35 1.4–0.4 3.1 10 (38) 0/4 MNC 6504 F-VH3-6504a 60 10−1–3.16 × 10−5 24–36 0.8–0.4 3.7 10−5 (39) 0/4 MNC −1 × −6 × −6 JH3 5292 F-VH3-5292a 60 10 –3.16 10 18–34 1.7–1.0 3.4 3.6 10 (34) 0/4 MNC 6401 F-VH4-6401a 60 10−1–3.16 × 10−4 25–33 0.9–0.7 3.0 10−4 (35) 0/4 MNC −1 −3 −4 + JH4 5181 F-VH3-5181a 60 10 –10 23–30 1.8–1.1 3.4 10 (32 34) 0/4 MNC 5255 F-VH3-5255a 60 10−1–10−3 24–32 2.0–0.8 3.3 10−5 (38) 0/4 MNC 6154 F-VH3-6154a 60 10−1–3.16 × 10−5 23–34 1.4–0.9 3.2 3.16 × 10−6 (36) 0/4 MNC 6480 F-VH1-6480a 60 10−1–10−5 22–35 2.0–1.5 3.3 10−5 (35) 0/24 MNC 6501 F-VH6-6501a 60 10−1–3.16 × 10−5 20–32 1.8–0.9 3.4 3.16 × 10−6 (34) 0/4 MNC 5538 F-VH3-5538a 62 10−1–10−4 23–34 2.0–0.6 3.6 10−5 (39) 1/12 MNC (48) 5579 F-VH-5579a 64 10−1–10−4 24–35 1.1–0.7 3.3 10−4 (35) 7/12 MNC (42) 5978 F-VH3-5978a 69 10−1–10−5 24–36 1.8–0.5 2.6 10−5 (36) 0/12 MNC 6216 F-VH3-6216a 62 10−1–3.16 × 10−4 25–35 1.7–1.0 3.4 3.16 × 10−5 (36) 1/12 MNC (42)

−1 −5 −5 JH5 3954 F-VH3-3954a 60 10 –3.16 × 10 23–35 1.6–0.5 3.5 3.16 × 10 (35) 0/4 MNC 5282 F-VH1-5282a 60 10−1–10−4 22–33 1.8–0.8 3.4 10−5 (36) 0/4 MNC 5648 F-VH3-5648a 60 10−1–10−5 22–36 1.7–1.0 3.3 10−5 (36) 0/4 MNC 5807 F-VH6-5807a 60 10−1–10−5 23–33 2.7–1.5 3.1 10−5 (35) 0/4 MNC 6396 F-VH3-6396a 60 10−1–10−4 22–34 2.0–1.6 3.7 10−4 (35) 1/24 MNC (46) 5414 F-VH-5414a 64 10−1–10−4 22–33 0.9–0.6 3.4 10−5 (35) 1/12 MNC (48) 5594 F-VH1-5594a 64 10−1–3.16 × 10−5 23–36 1.6–0.7 3.6 3.16 × 10−5 (39) 0/12 MNC 5978 F-VH1-5978bf 66 10−1–10−3 27–34 2.3–1.6 3.6 10−3 (34) 2/12 MNC (45)

−1 −5 −5 JH6 5566 F-VH2-5566a 60 10 –10 23–35 1.5–1.1 3.2 10 (35) 0/4 MNC 5627 F-VH3-5627b 60 10−1–10−4 23–34 1.4–0.8 3.5 10−4 (34) 0/4 MNC 5772 F-VH-5772a 60 10−1–3.16 × 10−5 24–37 1.5–0.9 3.4 10−5 (36) 1/4 MNC (46) 6198 F-VH6-6198a 60 10−1–10−4 25–36 1.2–0.4 3.2 10−4 (36) 0/4 MNC 6437 F-VH3-6437b 60 10−1–3.16 × 10−4 23–32 1.2–0.8 3.5 10−5 (36) 0/24 MNC 5221 F-VH4-5221a 64 10−1–3.16 × 10−5 24–36 1.2–0.7 3.6 3.16 × 10−5 (37) 2/12 MNC (48) 5902 F-VH-5902a 64 10−1–10−5 22–37 1.4–0.3 4.1 10−5 (37) 3/12 MNC (43) 6110 F-VH-6110a 62 10−1–10−5 22–35 1.8–0.6 3.6 10−5 (35) 0/12 MNC 6148 F-VH2-6148a 62 10−1–10−3 29–37 1.0–0.8 3.4 10−3 (37) 1/12 MNC (48) 6264 F-VH6-6264bf 60 10−1–3.16 × 10−5 23–36 1.1–0.8 3.7 10−5 (36 + 39) 1/12 MNC (48) 6379 F-VH3-6379a 62 10−1–3.16 × 10−5 22–34 1.0–0.8 3.4 10−5 (37) 0/12 MNC

aT, annealing/extension temperature. Optimized temperatures are indicated in bold. b To be able to compare the CT values of the different primer/probe combinations, the RQ-PCR data have been analyzed using a ﬁxed threshold of 0.01. Only for analyzing the slope of each standard curve optimal thresholds have been used. cThe reproducible range of a primer/probe combination was deﬁned as that part of the standard curve in which the maximal difference

duplicate dilution samples resulted in a positive fluorescent signal, with a CT value of maximal 40 cycles and at least 6 cycles lower than the CT values found for the aspecific amplification for this primer/probe combination in normal MNC DNA. eFor all experiments, the specificity of amplification and detection was tested on at least four control MNC DNA, which was in some cases extended to 12 (8 samples extra) or 24 (20 samples extra). fASO-primers F-VH1-5978b and F-VH6-6264b had a Tm of 53°C (nearest neighbour method).

signal was ever found in control MNC DNA samples (Table optimized conditions, eight extra control MNC DNA samples 3). For the ASO primer approach under standard conditions, were tested in addition to the four control MNC DNA samples at first instance 13 out of 35 (37%) primer/probe combinations tested before. Of the 16 primer/probe combinations tested, showed fluorescent signals in control MNC DNA, indicating incidental aspecific amplification of polyclonal IGH gene aspecific amplification of normal polyclonal IGH rearrange- rearrangements in MNC DNA was found in 10 cases (Table ments. After optimization of the RQ-PCR (mainly by increas- 4). In nine out of 10 this concerned one to three MNC samples ing the annealing temperature), only five primer/probe combi- positive out of the total of 12 to 24 control samples, but for

nations still showed aspecific amplification, but with high CT one case (5579) seven out of 12 samples were positive. In values above 45. seven cases, positive signals were found after 45 cycles, all To test whether the specificity of the ASO primer approach at least nine cycles later than the maximal sensitivity of the could be reproduced consistently, we tested an additional 20 primer/probe combinations. In three cases (5579, 5902 and

control MNC DNA samples for three primer/probe combi- 6216) positive signals in MNC DNA had CT values of 42–43 nations (patients 6396, 6437 and 6480) under standard con- cycles, but still six to eight cycles above that of the maximal ditions. Also for all 13 primer/probe combinations under the sensitivity (Table 4).

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1433 Discussion get. For example, in patient 5978 another IGH gene rearrange- −5 ment (VH3–JH4) reached a 100-fold higher sensitivity (10 In a recent publication we have shown that the RQ-PCR tech- instead of 10−3). nique using patient-specific Ig/TCR gene rearrangements as The applicability of the ASO primer approach depends on PCR targets is a reliable method for the quantification of MRD its sensitivity and specificity. Specificity of the primer/probe in precursor-B-ALL.18 Patient-specific (ASO) TaqMan probes combinations depends on one patient-specific forward primer. complementary to the junctional region, were combined with Factors that influence the specificity of RQ-PCR analysis using two germline primers. In the present study, we show that at an ASO primer are: (1) the gene segments used in the VH–DH least equally sensitive detection and quantification of IGH and DH–JH junctional regions; (2) the size and sequence of the gene rearrangements can be reached by positioning the Taq- junctional region; (3) and the background of normal cells with Man probe and one of the primers at germline sequences of comparable gene rearrangements. By positioning the 3Ј-end

JH gene segments in combination with one ASO primer. For of the ASO primer at the junctional region, no aspecific ampli- ASO primer design we used both VH–DH and DH–JH junctional fication was found in 22 out of 35 primer/probe combinations regions, although for MRD studies it may be preferred to use (63%) under standard RQ-PCR conditions. Particularly ampli- mainly DH–JH junctional regions, in order to prevent false- fication of the JH1, JH2 and JH3 rearrangements did not result negative results due to ongoing IGH gene rearrangements in aspecific background signals, which can probably be 5,33–35 involving pre-existing DH–JH joinings. explained by the fact that these JH gene segments are rarely 36–39 The sequence variability between the different JH gene seg- used in pediatric and adult peripheral B lymphocytes. ments allowed us to design three different JH TaqMan probes, This is in contrast to the commonly used JH4, JH5 and JH6 one specific for JH3, one specific for JH6 and one specific for rearrangements, where in 13 out of 28 primer/probe combi- JH1, JH2, JH4 and JH5 gene segments. To obtain higher speci- nations (46%) aspecific amplification was found. In these ficity, these probes were used in combination with reverse cases additional stringent conditions were necessary to dimin- primers that were made specific for each of the six JH gene ish this aspecific amplification. In virtually all cases, the strin- segments by positioning them at the downstream introns. gency of amplification was successfully improved by increas- First, we compared the alternative ASO primer approach ing the annealing/extension temperature (up to 69°C). This with our initial ASO probe approach for nine IGH gene resulted in specific amplification and did not affect the rearrangements (four JH4, one JH5 and four JH6 efficiency of amplification and detection, except for cases rearrangements) obtained from nine randomly chosen precur- 5978 and 6148 as described above. We preferred to increase sor-B-ALL. Both approaches resulted in successful amplifi- the annealing temperature instead of shortening the ASO cation and real-time detection for all IGH rearrangements. primer, because the former approach is easy to perform and Specific amplification using the ASO primers resulted for five allows the use of the existing ASO primer, thereby saving time out of the nine IGH rearrangements in a longer reproducible and costs. range for accurate quantification and in higher maximum sen- We evaluated whether positioning of the ASO primer and sitivities, compared to the results obtained with the ASO probe the lengths of the junctional region were related to the occur- approach. In the remaining four IGH rearrangements similar rence of aspecific amplification. Although aspecific amplifi- results were obtained for both approaches. Representative cation indeed tended to be absent in the case of IGH results of both approaches are given in Figure 2, which illus- rearrangements with high nucleotide insertion in the junc- trates that a disadvantage of the ASO probe approach was the tional region, several exceptions were found, such as case low plateau level of fluorescent signal (maximum ⌬Rn) at 5594 with 20 N-nucleotides in the ASO primer, which needed higher dilutions. We assume that this frequent finding was stringent conditions to remove background signals. caused by competition for germline primers between patient- To gain more insight into the characteristics of aspecific specific IGH gene rearrangements and similar polyclonal IGH amplification, we tested all primer/probe combinations that gene rearrangements in normal cells. In the higher dilutions were at least once positive in MNC-derived DNA in separate ⌬ the maximum Rn is even so low that it is hard to distinguish RQ-PCR reactions using normal MNC DNA as input. The CT between specific signals and background signals. So, follow- values of ‘false-positive’ signals were in three out of 16 cases up samples with low MRD levels will just like the higher below 45, but always above 42 and at least six to eight cycles ⌬ dilutions also show very low Rn values and will be difficult above the CT value found at the detection limit of the dilution to evaluate. This is in contrast to the ASO primer approach, series. In conclusion, specific amplification could be easily where specific amplification results in high plateau levels, discriminated from incidental aspecific amplification, even in the higher dilutions. implying that sensitive quantification of MRD levels is The RQ-PCR analysis by the ASO primer approach of 35 possible.

IGH gene rearrangements showed robust amplification. The The germline JH TaqMan probes in combination with JH- sensitivities obtained were higher (10−4–10−5) than previously specific reverse primers might also be useful for quantification reported with ASO probes.18 The improved sensitivity is prob- of IGH gene rearrangements in mature B cell malignancies. ably due to the fact that in the ASO primer approach no com- Although in most of these cases the presence of somatic hyp- petition exists for primers between the amplification of the ermutation allow the design of two patient-specific primers,29 leukemic and normal IGH gene rearrangements. The repro- our good results justify screening for the applicability of our ducible range for accurate quantification spanned at least four set of germline JH TaqMan probes and primers in future stud- orders of magnitude in almost all cases (31/35). In only two ies. In addition, our JH TaqMan probes and primers might also out of 35 (6%) was the maximum sensitivity below the theor- be useful for MRD detection via chromosomal translocations −4 etical sensitivity threshold of 10 (approximately 16 copies in which JH gene segments are involved, like t(14;18) and of the target gene). In both patients this was found under extra t(11;14). stringent amplification conditions to avoid aspecific amplifi- Our results demonstrate that RQ-PCR analysis of IGH gene cation of polyclonal IGH gene rearrangements in MNC DNA. rearrangements with one primer at the patient-specific VH–DH In such patients, it is probably better to use another PCR tar- or DH–JH junctional region combined with a germline JH Taq-

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1434 Man probe and a germline primer, is applicable for MRD dardized detection of minimal residual disease in acute lympho- analysis in precursor-B-ALL. However, not all junctional blastic leukemia using immunoglobulin and T cell receptor gene regions might be suitable for designing a specific ASO primer, rearrangements and TAL1 deletions as PCR targets: report of the BIOMED-1 Concerted Action: investigation of minimal residual particularly in case of a limited insertion of N-nucleotides. In disease in acute leukemia. Leukemia 1999; 13: 110–118. those cases where aspecific amplification still occurs after 9 Szczepanski T, Langerak AW, Wolvers-Tettero ILM, Ossenkoppele optimization, it is frequently possible to use another PCR tar- GJ, Verhoef G, Stul M, Petersen EJ, de Bruijn MAC, van’t Veer get.5,9,10 Currently, we are testing the ASO primer approach MB, van Dongen JJM. Immunoglobulin and T cell receptor gene for other Ig/TCR gene rearrangements suitable for MRD PCR rearrangement patterns in acute lymphoblastic leukemia are less analysis, ie TCRD, TCRG and IGK-Kde.5,8 We expect that stan- mature in adults than in children: implications for selection of PCR targets for detection of minimal residual disease. Leukemia 1998; dardized RQ-PCR analysis for MRD detection in ALL will soon 12: 1081–1088. be available for large-scale clinical studies. 10 Szczepanski T, Beishuizen A, Pongers-Willemse MJ, Ha¨hlen K, van Wering ER, Wijkhuijs JM, Tibbe GJM, De Bruijn MAC, van Dongen JJM. Cross-lineage T-cell receptor gene rearrangements Acknowledgements occur in more than ninety percent of childhood precursor-B-acute lymphoblastic leukemias: alternative PCR targets for detection of minimal residual disease. Leukemia 1999; 13: 196–205. This work was supported by the Dutch Cancer Foundation/ 11 Cross NC. Quantitative PCR techniques and applications. Br J Koningin Wilhelmina Fonds (grant SNWLK 97–1567) and PE Haematol 1995; 89: 693–697. Biosystems (Nieuwerkerk a/d IJssel, The Netherlands). We 12 Cave H, Guidal C, Rohrlich P, Delfau MH, Broyart A, Lescoeur thank Dr T Szczepanski for support in the design of the Taq- B, Rahimy C, Fenneteau O, Monplaisir N, d’Auriol L. Prospective Man probes and for critical reading of the manuscript. We monitoring and quantitation of residual blasts in childhood acute lymphoblastic leukemia by polymerase chain reaction study of thank the Dutch Childhood Leukemia Study Group (DCLSG) delta and gamma T-cell receptor genes. Blood 1994; 83: 1892– for kindly providing ALL cell samples. Board members of the 1902. DCLSG are IM Appel, H van den Berg, JPM Bo¨kkerink, MCA 13 Sykes PJ, Neoh SH, Brisco MJ, Hughes E, Condon J, Morley AA. Bruin, JJ Groot-Loonen, SSN de Graaf, K Ha¨hlen, PM Hooger- Quantitation of targets for PCR by use of limiting dilution. Biotech- brugge, WA Kamps, FAE Nabben, JA Rammeloo, T Revesz, niques 1992; 13: 444–449. AYN Schouten-van Meeteren, AJP Veerman, M van Weel- 14 Ouspenskaia MV, Johnston DA, Roberts WM, Estrov Z, Zipf TF. Accurate quantitation of residual B-precursor acute lymphoblastic Sipman and RS Weening. leukemia by limiting dilution and a PCR-based detection system: a description of the method and the principles involved. Leukemia 1995; 9: 321–328. References 15 Neale GAM, Coustan Smith E, Pan Q, Chen X, Gruhn B, Stow P, Behm FG, Pui CH, Campana D. Tandem application of flow 1 Coustan-Smith E, Behm FG, Sanchez J, Boyett JM, Hancock ML, cytometry and polymerase chain reaction for comprehensive Raimondi SC, Rubnitz JE, Rivera GK, Sandlund JT, Pui CH, Cam- detection of minimal residual disease in childhood acute lym- pana D. Immunological detection of minimal residual disease in phoblastic leukemia. Leukemia 1999; 13: 1221–1226. children with acute lymphoblastic leukaemia. Lancet 1998; 351: 16 Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative 550–554. PCR. Genome Res 1996; 6: 986–994. 2 Cave H, van der Werff ten Bosch, Suciu S, Guidal C, Waterkeyn 17 Holland PM, Abramson RD, Watson R, Gelfand DH. Detection of specific polymerase chain reaction product by utilizing the 5Ј– C, Otten J, Bakkus M, Thielemans K, Grandchamp B, Vilmer E. Ј Clinical significance of minimal residual disease in childhood 3 exonuclease activity of Thermus aquaticus DNA polymerase. acute lymphoblastic leukemia. European Organization for Proc Natl Acad Sci USA 1991; 88: 7276–7280. Research and Treatment of Cancer – Childhood Leukemia Cooper- 18 Pongers-Willemse MJ, Verhagen OJ, Tibbe GJ, Wijkhuijs AJ, de ative Group. New Engl J Med 1998; 339: 591–598. Haas V, Roovers E, van der Schoot CE, van Dongen JJM. Real- 3 van Dongen JJM, Seriu T, Panzer-Grumayer ER, Biondi A, Pongers- time quantitative PCR for the detection of minimal residual disease Willemse MJ, Corral L, Stolz F, Schrappe M, Masera G, Kamps in acute lymphoblastic leukemia using junctional region specific WA, Gadner H, van Wering ER, Ludwig WD, Basso G, De Bruijn TaqMan probes. Leukemia 1998; 12: 2006–2014. MA, Cazzaniga G, Hettinger K, van der Does-van den Berg A, Hop 19 Mensink E, van de Locht A, Schattenberg A, Linders E, Schaap N, WC, Riehm H, Bartram CR. Prognostic value of minimal residual Geurts vK, De Witte T. Quantitation of minimal residual disease disease in acute lymphoblastic leukaemia in childhood. Lancet in Philadelphia chromosome positive chronic myeloid leukaemia 1998; 352: 1731–1738. patients using real-time quantitative RT-PCR. Br J Haematol 1998; 4 Gruhn B, Hongeng S, Yi H, Hancock ML, Rubnitz JE, Neale GA, 102: 768–774. Ј Kitchingman GR. Minimal residual disease after intensive induc- 20 Luthra R, McBride JA, Cabanillas F, Sarris A. Novel 5 exonucle- tion therapy in childhood acute lymphoblastic leukemia predicts ase-based real-time PCR assay for the detection of outcome. Leukemia 1998; 12: 675–681. t(14;18)(q32;q21) in patients with follicular lymphoma. Am J 5 van Dongen JJ, Szczepanski T, De Bruijn MA, van den Beemd Pathol 1998; 153: 63–68. MW, de Bruin-Versteeg S, Wijkhuijs JM, Tibbe GJ, van Gastel- 21 Marcucci G, Livak KJ, Bi W, Strout MP, Bloomfield CD, Caligiuri Mol EJ, Groeneveld K, Hooijkaas H. Detection of minimal residual MA. Detection of minimal residual disease in patients with disease in acute leukemia patients. Cytokines Mol Ther 1996; 2: AML1/ETO-associated acute myeloid leukemia using a novel 121–133. quantitative reverse transcription polymerase chain reaction assay. 6 Seriu T, Hansen-Hagge TE, Erz DH, Bartram CR. Improved detec- Leukemia 1998; 12: 1482–1489. tion of minimal residual leukemia through modifications of poly- 22 Pallisgaard N, Clausen N, Schroder H, Hokland P. Rapid and merase chain reaction analyses based on clonospecific T cell sensitive minimal residual disease detection in acute leukemia by receptor junctions. Leukemia 1995; 9: 316–320. quantitative real-time RT-PCR exemplified by t(12;21) TEL-AML1 7 Steenbergen EJ, Verhagen OJ, van Leeuwen EF, van den Berg H, fusion transcript. Genes Chromosomes Cancer 1999; 26: 355– Behrendt H, von dem Borne AE, van der Schoot CE. IgH/TCR delta 365. PCR oligonucleotide liquid hybridization, a fast and sensitive 23 Emig M, Saussele S, Wittor H, Weisser A, Reiter A, Willer A, assay for monitoring minimal residual disease in childhood B-pre- Berger U, Hehlmann R, Cross NC, Hochhaus A. Accurate and cursor ALL. Leukemia 1995; 9: 216–222. rapid analysis of residual disease in patients with CML using spe- 8 Pongers-Willemse MJ, Seriu T, Stolz F, d’Aniello E, Gameiro P, cific fluorescent hybridization probes for real time quantitative RT- Pisa P, Gonzalez M, Bartram CR, Panzer-Grumayer ER, Biondi A, PCR. Leukemia 1999; 13: 1825–1832. San Miguel JF, van Dongen JJM. Primers and protocols for stan- 24 Verhagen OJHM, Wijkhuijs AJM, van der Sluijs-Gelling AJ,

Leukemia RQ-PCR for detection of MRD in ALL OJHM Verhagen et al 1435 Szczepanski T, van der Linden-Schrever BEM, Pongers-Willemse locus: a systematic analysis provides no evidence for the use of MJ, van Wering ER, van Dongen JJM, van der Schoot CE. Suitable DIR segments, inverted D segments, ‘minor’ D segments or D–D DNA isolation method for the detection of minimal residual dis- recombination. J Mol Biol 1997; 270: 587–597. ease by PCR techniques. Leukemia 1999; 13: 1298–1299. 33 Steenbergen EJ, Verhagen OJ, van Leeuwen EF, von dem Borne 25 van Dongen JJM, Wolvers-Tettero IL. Analysis of immunoglobulin AE, van der Schoot CE. Distinct ongoing Ig heavy chain rearrange- and T cell receptor genes. Part I: Basic and technical aspects. Clin ment processes in childhood B-precursor acute lymphoblastic leu- Chim Acta 1991; 198: 1–91. kemia. Blood 1993; 82: 581–589. 26 Beishuizen A, Verhoeven M-AJ, Mol EJ, Breit TM, Wolvers-Tettero 34 Beishuizen A, Verhoeven MA, van Wering ER, Hahlen K, ILM, Van Dongen JJM. Detection of immunoglobulin heavy-chain Hooijkaas H, van Dongen JJM. Analysis of Ig and T-cell receptor gene rearrangements by Southern blot analysis. Recommendations genes in 40 childhood acute lymphoblastic leukemias at diagnosis for optimal results. Leukemia 1993; 7: 2045–2053. and subsequent relapse: implications for the detection of minimal 27 Steenbergen EJ, Verhagen OJ, van den Berg H, van Leeuwen EF, residual disease by polymerase chain reaction analysis. Blood Behrendt H, Slater RR, von dem Borne AE, van der Schoot CE. 1994; 83: 2238–2247. Rearrangement status of the malignant cell determines type of sec- 35 Steward CG, Goulden NJ, Katz F, Baines D, Martin PG, Langlands ondary IgH rearrangement (V-replacement or V to DJ joining) in K, Potter MN, Chessells JM, Oakhill A. A polymerase chain reac- childhood B precursor acute lymphoblastic leukemia. Leukemia tion study of the stability of Ig heavy-chain and T-cell receptor 1997; 11: 1258–1265. delta gene rearrangements between presentation and relapse of 28 Szczepanski T, Pongers-Willemse MJ, Langerak AW, Harts WA, childhood B-lineage acute lymphoblastic leukemia. Blood 1994; Wijkhuijs AJM, van Wering ER, van Dongen JJM. Ig heavy chain 83: 1355–1362. gene rearrangements in T-cell acute lymphoblastic leukemia exhi- 36 Sanz I. Multiple mechanisms participate in the generation of diver-

bit predominant DH6–19 and DH7–27 gene usage, can result in sity of human H chain CDR3 regions. J Immunol 1991; 147: complete V–D–J rearrangements. and are rare in T-cell receptor 1720–1729. ␣␤ lineage. Blood 1999; 93: 4079–4085. 37 Wasserman R, Galili N, Ito Y, Reichard BA, Shane S, Rovera G. 29 Willems PMW, Verhagen OJHM, Segeren C, Veenhuizen P, Gui- Predominance of fetal type DJH joining in young children with kema J, Wiemer E, Groothuis L, de Jong T, Kok H, Bloem A, Bos B precursor lymphoblastic leukemia as evidence for an in utero N, Vellenga E, Mensink EJBM, Sonneveld P, Lokhorst H, van der transforming event. J Exp Med 1992; 176: 1577–1581. Schoot CE, Raymakers RAP. A consensus strategy to quantitate 38 Yamada M, Wasserman R, Reichard BA, Shane S, Caton AJ, Rov- malignant cells in Myeloma patients validated in a multicentre era G. Preferential utilization of speciﬁc immunoglobulin heavy study. Blood (in press). chain diversity and joining segments in adult human peripheral 30 Cook GP, Tomlinson IM. The human immunoglobulin VH reper- blood B lymphocytes. J Exp Med 1991; 173: 395–407. toire. Immunol Today 1995; 16: 237–242. 39 Steenbergen EJ, Verhagen OJ, van Leeuwen EF, Behrendt H, Merle 31 Tomlinson IM, Cook GP, Walter G, Carter NP, Riethman H, PA, Wester MR, von dem Borne AE, van der Schoot CE. B precur- Buluwela L, Rabbitts TH, Winter G. A complete map of the human sor acute lymphoblastic leukemia third complementarity- immunoglobulin VH locus. Ann NY Acad Sci 1995; 764: 43–46. determining regions predominantly represent an unbiased recom- 32 Corbett SJ, Tomlinson IM, Sonnhammer ELL, Buck D, Winter G. bination repertoire: leukemic transformation frequently occurs in Sequence of the human immunoglobulin diversity (D) segment fetal life. Eur J Immunol 1994; 24: 900–908.

Leukemia