Functional characterization of minimal residual disease in aml - simultaneous detection of p-glycoprotein function and aberrant phenotypes

10 pages
50 views
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Description
Functional characterization of minimal residual disease in aml - simultaneous detection of p-glycoprotein function and aberrant phenotypes
Transcript
  Leukemia (2001) 15 , 1554–1563 ©  2001 Nature Publishing Group All rights reserved 0887-6924/01 $15.00 www.nature.com/leu Functional characterization of minimal residual disease for P-glycoprotein andmultidrug resistance protein activity in acute myeloid leukemia MA van der Pol 1 , JM Pater 1 , N Feller 1 , AH Westra 1 , A van Stijn 1 , GJ Ossenkoppele 1 , HJ Broxterman 2 and GJ Schuurhuis 1 Departments of   1 Hematology and   2 Medical Oncology, Medical Center Vrije Universiteit, Amsterdam, The Netherlands  Relapse is common in acute myeloid leukemia (AML) due topersistence of residual leukemia cells: minimal residual dis-ease (MRD). In 102 out of 127 patients (80%), cells at diagnosisdisplayed one or more leukemia-associated phenotypes (LAP),ie combinations of cell surface markers which are absent innormal cells and can thus be used to detect MRD at follow-up. Functional characterization of MRD cells for P-glycoprotein(Pgp) and multidrug resistance protein (MRP) activity is essen-tial to investigate the role of these drug transport proteins inmultidrug resistance in AML. A fluorescent probe assay usingSyto16/PSC833 and calcein-AM/probenecid as substrate/ modulator of the Pgp and MRP pump, respectively, and sub-sequent labeling of cells with monoclonal antibodies for LAPdetection allowed simultaneous detection of LAP and Pgp orMRP activity. Validation of this assay is shown for 30 newlydiagnosed AML and 11 MRD situations. In addition, no signifi-cant differences were found when comparing fresh and cryo-preserved  de novo   AML for LAP expression ( n   =  43), Pgp ( n   = 30) and MRP ( n   =  24) function and for MRD samples for simul-taneous LAP expression and Pgp/MRP activity ( n   =  10). Thisapproach enables longitudinal and multicenter studies on thedetection, quantification and functional characterisation ofMRD cells.  Leukemia   (2001)  15,  1554–1563. Keywords:  minimal residual disease; P-glycoprotein; MRP1; acutemyeloid leukemia; leukemia-associated phenotypes Introduction Despite the fact that many patients with acute myeloid leuke-mia (AML) reach complete remission after chemotherapeutictreatment, many will still eventually relapse because of thepersistence and subsequent outgrowth of leukemia cells thatare undetectable by conventional morphological techniques.This condition is known as minimal residual disease (MRD).AML cells at diagnosis frequently display aberrant or uncom-mon phenotypes that allow distinction from normal hemato-poietic cells and therefore can be used as immunophenotyp-ical markers for MRD. 1–8 The detection of MRD withmultiparameter flowcytometry is successful in 80–90% of AML patients, while molecular techniques depending on thepresence of useful chromosomal abnormalities are successfulin less than 30% of AML patients. Multiparameter flow cyto-metry thus has great potential in MRD detection in AML. ForAML it has been shown already in small studies that the pres-ence of threshold numbers of MRD cells or the gradualincrease of the frequency of MRD cells in follow-up samplescorrelates with relapse frequency and relapse-free survival. 7,9–13 In children with acute lymphoblastic leukemia it was shownthat sequential monitoring of MRD by flow cytometry pro-vides significant prognostic information. 14 The advantages of flow cytometric analysis for detecting MRD are its relativelyhigh sensitivity (1:10 3 –1:10 5 ), 1,2,15 rapidity and the possibility Correspondence: GJ Schuurhuis, Medical Center Vrije Universiteit,Dept of Hematology, BR240, PO Box 7057, 1007 MB Amsterdam,The Netherlands; Fax: 31 20 4442601Received 24 April 2001; accepted 13 June 2001 to characterize cells functionally. The strategy for the immun-ophenotypical detection of MRD is based on the identificationduring follow-up, of residual leukemia cells with the same leu-kemia-associated phenotypes (LAPs) as established at diag-nosis. To ensure that a relevant clonogenic leukemia cellpopulation will be followed, whenever possible, primitivemarkers like CD34 or CD117 must be included in the LAP. 16–19 Phenotypic shifts have been reported to occur after treatment. 8 Since blasts frequently display more than one LAP at diag-nosis 4 as many as possible of these are applied in the MRDsituation to ensure that all MRD cells are detected at follow-up.The survival of leukemia cells could be due to the persist-ence of resistant cell subpopulations or the induction of resist-ant cell subpopulations during chemotherapy. Putativeresistance mechanisms include (1) the presence of drug effluxpumps (P-glycoprotein (Pgp) or multidrug resistance protein(MRP) or (2) resistance due to changes in apoptotic pathways.Of these, the role of Pgp, which is an ATP-dependent mem-brane efflux pump, is most intensively studied. Expression andfunction of Pgp in  de novo   AML patients has been shown tobe an independent adverse prognostic factor for response andsurvival. 20–26 Furthermore, in some studies higher Pgpexpression was found in refractory and relapsed AML. 22,24,27 The use of rhodamine123 as Pgp probe, combined withPSC833 as Pgp inhibitor in an accumulation assay representsa sensitive and reproducible functional assay in AML both inefflux 28 and accumulation 29,30 assay and is also applicable forfrozen/thawed samples. 29 More recently, we have shown thatthe nucleic acid stain Syto16 combined with PSC833 is moresensitive than rhodamine123 to assess Pgp function. 31 Apart from Pgp, MRP1 (further referred to as MRP) mightcontribute to the multidrug resistance (MDR) phenotype inAML. 24,25,32–36 It has been shown that among others the com-bination of calcein-AM and probenecid can be used to detectMRP function. 37,38 To investigate the role of Pgp and MRP in drug resistance inAML, the most informative approach would be to functionallycharacterize MRD cells at different stages of disease of thepatient. For that purpose, MRD cells need to be identified,while in addition one should be able to determine Pgp and/orMRP function in these cells. In the present study, wedeveloped a sensitive and reproducible four-color FACS assaythat allows the simultaneous detection of Pgp or MRP functionand the LAP. An accumulation assay was used with the com-bination Syto16/PSC833 or calcein-AM/probenecid for detec-tion of Pgp and MRP function, respectively. The feasibility of the simultaneous detection of Pgp or MRP function with LAPis shown both in  de novo   AML, as well as in MRD situations.Since the use of cryopreserved samples would (1) enableenlargement of patient groups, eg in multicenter studies onMRD and (2) improve the flexibility in the laboratory, theeffect of freezing in liquid nitrogen on LAP expression andPgp/MRP function was also studied both in  de novo   AML andMRD material. The ability to study drug resistance at different  Pgp and MRP function in minimal residual disease in AML MA van der Pol  et al  1555 stages of disease of an AML patient will lead to a better under-standing of the role of multidrug resistance proteins in theemergence of MRD and might have consequences for chemo-therapeutic strategies like multidrug resistance modulation.Further the approach presented can be extended to othermechanisms putatively involved in drug resistance in AML,like apoptotic resistance mechanisms. Materials and methods Reagents  RMPI 1640, Iscove’s modified Dulbecco’s medium and fetalcalf serum (FCS) were obtained from Life Technologies(Paisley, UK), bovine serum albumin (BSA) and Hepes fromICN Biomedicals (Aurora, OH, USA), Dulbecco’s minimalessential medium (DMEM) without bicarbonate and phenolred was from Flow Laboratories (Irvine, UK), phosphate buff-ered saline from ICN (Costa Mesa, CA, USA), Ficoll–Paquefrom Pharmacia Biotech (Uppsala, Sweden), 7-amino acti-nomycin D (7-AAD) (Via-Probe) from Pharmingen (San Diego,CA, USA), FACS lysing solution from Becton Dickinson (San Jose, CA, USA), Dimethyl sulphoxide (DMSO) and sodiumazide from Merck (Darmstadt, Germany). Probenecid wasobtained from Sigma (St Louis, MO, USA), Syto16 and calceinacetoxymethylester (calcein-AM) from Molecular Probes(Eugene, OR, USA), doxorubicin from Pharmacia and UpjohnBV (Woerden, The Netherlands) and colchicine from ServaFeinbiochemica (Heidelberg, Germany), while PSC833 was agift from Novartis (Basel, Switzerland). The source of mono-clonal antibodies (MoAbs) conjugated with fluorescein isothi-ocyanate (FITC) was as follows: CD2 and CD45 from Coulter(Coultronics, Mergency, France), CD5, CD11c, CD13 andHLADR from Dako (Glostrup, Denmark), CD7 from CLB(Amsterdam, The Netherlands), CD15, CD34 and CD61 fromBecton Dickinson, CD33 from Caltag (San Francisco, CA),CD38 from Immunotech (Marseille, France), CD65 from And-erGrub (Kaumberg, Austria) and CD90 from Pharmingen. ForMoAbs conjugated with phycoerythrin (PE) the source was:CD2, CD13, CD15 and CD33 from Coulter, CD7 from Immu-notech, CD4, CD11b, CD34, CD38, CD56 and CD117 fromBecton Dickinson, CD45 from Dako and AC133 from Mil-tenyi (CLB, Amsterdam, The Netherlands). The MoAbs conju-gated with peridinyl chlorophyllin (PerCP) were CD45 andHLADR (Becton Dickinson) and the MoAbs conjugated withallophyocyanin (APC) were CD14, CD19, CD33, CD34,CD38, CD45 and CD117 (all Becton Dickinson). Cell lines  The human epidermoid carcinoma cell line KB3–1 and its Pgpmultidrug resistant sublines, KB8 and KB8–5, were obtainedfrom Dr I Roninson (University of Illinois, Chicago, IL, USA;KB3–1 and KB8) or via ATCC (KB8–5). The cells were culturedin DMEM supplemented with 7.5% heat-inactivated FCS at37 ° C in a humid atmosphere with 5% CO 2 . The KB8 andKB8–5 cells were cultured in the presence of 5 and 10 ng/mlcolchicine, respectively. The human small-cell lung cancercell line GLC4 and its MRP overexpressing subline GLC4-ADRwere cultured in RMPI 1640 supplemented with 10% FCS.The resistant cells were cultured in the presence of 1.2   M doxorubicin. The human ovarian cancer cell line A2008 andits stable MRP transfectants MRP1–4, MRP1–6 (MRP1 clones) Leukemia and cMOAT (MRP2) (gift from Dr M Kool, Netherlands CancerInstitute, Amsterdam, The Netherlands) were cultured in RPMI1640 supplemented with 10% FCS. All resistant cells werecultured for 7 days without drugs before experiments wereperformed. The cell lines were mycoplasma negative, as wastested monthly using a PCR-mediated EIA with 16S ribosomalRNA as a target. Patients and controls  Normal bone marrow samples were obtained after informedconsent from patients undergoing cardiovascular surgery.Control peripheral blood stem cells were obtained after mobil-ization with G-CSF, from healthy donors involved in an allo-geneic peripheral blood stem cell transplantation procedureor from non-Hodgkin’s lymphoma patients in completeremission. Bone marrow or peripheral blood samples fromAML patients were obtained at the time of diagnosis and aftertreatment (BM after first, second and third course of chemo-therapy, peripheral blood stem cell transplants, BM at differenttime points after peripheral blood stem cell transplantation),following informed consent. In the MRD study we included127 patients thus far (73 female, 54 male) with a median ageof 55 years, ranging from 15 to 75. The French–American–British (FAB) classification of the AML patients was as follows:M0 (6), M1 (12), M2 (26), M3 (7), M4 (24), M5 (26), M6 (7),RA (1), RAEB (2), RAEB-T (11), hypoplastic MDS (1), whilefour were not classified as caused by inadequate samples.Immunophenotypical analysis on bone marrow or peripheralblood samples from  de novo   AML was performed within 24 hafter sampling (see establishment of leukemia-associatedphenotypes). These  de novo   AML samples showed a medianof 74% (range 24–94) blasts. For functional studies, mono-nuclear cells were isolated via a Ficoll gradient (1.077 g/ml)from the  de novo   AML sample. Erythrocytes were lysed after-wards by a 10-min incubation on ice with ammonium chlor-ide. After this procedure, a median of 92% (range 19–100)blasts was present in these samples. The fresh mononuclearcells were tested for Pgp and MRP activity and the remainingcells were frozen in medium with 20% FCS and 10% DMSOusing controlled freezing with a Kryo 10 (Planer Biomed, Sun-bury, UK) and stored in liquid nitrogen. Follow-up sampleswere tested within 24 h after sampling for the presence of MRD cells (see MRD detection). For functional studies,erythrocytes were lysed with ammonium chloride and thefresh cells were tested for Pgp and MRP activity and theremaining cells were frozen and stored in liquid nitrogen.Frozen  de novo   AML or follow-up samples were rapidlythawed in a waterbath (37 ° C) and immediately diluted in15 ml RPMI 1640 with 40% FCS, prewarmed to 37 ° C. Thecells were centrifuged and resuspended in RPMI 1640 with10% FCS for further experiments. When thawed cells wereused for functional flow cytometric assays, they were preincu-bated for 60 min in accumulation medium (DMEM withoutbicarbonate and phenol red) with 40% FCS at 37 ° C to allowrecovery of metabolic activity, 29 before resuspension inaccumulation medium with 10% FCS. These cells had amedian trypan blue viability of 86%. Flow cytometry  Flow cytometry was performed using a FACScalibur (BectonDickinson) equipped with an argon and red diode laser. When  Pgp and MRP function in minimal residual disease in AML MA van der Pol  et al  1556 Leukemia functional Pgp or MRP activity was combined with detectionof leukemia-associated phenotypes, special attention had tobe paid with respect to compensation of spectral overlap. 1n M  calcein-AM (fluorescence in FL1) was used for the detec-tion of MRP function without causing problems in the FL2channel.Syto16 (fluorescence in FL1) showed considerable spectraloverlap in FL2, which was a problem at concentrationsexceeding 5 n M . Therefore, a low concentration of Syto16 (0.6n M ) was used and attention was paid to compensate FL2 flu-orescence (used for PE labelled MoAbs) for the Syto16 fluor-escence present in FL1. Data acquisition and analysis wasperformed using Cellquest software (Becton Dickinson). Establishment of leukemia-associated phenotypes  The phenotypic analysis of   de novo   AML was performed onwhole bone marrow or peripheral blood samples upon stain-ing with FITC, PE, PerCP and APC conjugated MoAbs. Briefly,0.3  ×  10 6 cells in PBS with 0.1% BSA and 0.05% sodiumazide (FACS buffer) were incubated in a final volume of 50  l with the appropriate combinations of MoAbs for 15 min atroom temperature. Afterwards, 1 ml of FACS lysing solutionwas added and another 7.5 min incubation was performed atroom temperature. Finally, cells were washed twice with 3 mlFACS buffer and resuspended in 0.2 ml FACS buffer.For detailed immunophenotypical analysis of leukemiablasts, a double-step procedure was used. First, a panel of 15quadruple MoAb combinations was used to define the immu-nophenotype of the blasts. In a second step, further MoAbcombinations were used to assess the precise characteristicsof different blast cell subpopulations. With every sample, anegative control (IgG1 isotype) and a positive control (CD45)conjugated with different fluorochromes was included tocheck the compensation settings of the flow cytometer. Dataanalysis was performed on blast cells gated with CD45dimstaining and low side scatter. MRD detection The detection of MRD cells in follow-up material was perfor-med on whole bone marrow or peripheral blood samplesupon staining with FITC, PE, PerCP and APC conjugatedMoAbs.Cells were incubated with appropriate MoAb combinationsto be able to detect cells with expression of the leukemia-associated phenotypes found in the corresponding  de novo  AML. Briefly, 1.0–2.0  ×  10 6 cells were incubated at a finalconcentration of 6  ×  10 6 cells/ml FACS buffer with the appro-priate combinations of MoAbs for 15 min at room tempera-ture. Afterwards, 1 ml of FACS lysing solution was added andanother 7.5 min incubation was performed at room tempera-ture. Finally, cells were washed twice with 3 ml FACS bufferand resuspended in 0.6 ml FACS buffer. Corresponding iso-type controls were included for every LAP to determine nega-tivity and positivity for the markers tested. Data analysis wasperformed on blasts with CD45dim staining and SSC lowproperties. Functional Pgp and MRP assay  Cells were washed in accumulation medium with 10% FCS.For Pgp function cells were incubated at a concentration of 0.3  ×  10 6 cells/ml accumulation medium with 0.6 n M  Syto16with or without the Pgp inhibitor PSC833 (1   M ) in a 37 ° Cwater bath and carefully shaken. After 45 min of incubation,after which for Syto16 a steady state cellular accumulation isreached, 31 the cells were immediately diluted with ice coldaccumulation medium and centrifuged. For MRP function asimilar protocol was followed using 1 n M  calcein-AM withor without the MRP inhibitor probenecid (3 m M ) in a 60-minincubation period. In both protocols the cells wereresuspended in FACS buffer and kept on ice.Pgp and MRP activities are expressed as ratios of drug flu-orescence with modulator and drug fluorescence withoutmodulator after subtraction of the fluorescence of the control(cells in accumulation medium without drugs present). Thedescribed parental and Pgp and MRP expressing sublines wereused to optimise the functional Pgp and MRP assay. Whenusing Syto16/PSC833 for detection of Pgp function in theKB3–1, KB8 and KB8–5 cells the following ratios (means  ± s.d. of four experiments) were found: 0.99  ±  0.09, 2.21  ±  0.08and 8.62  ±  0.93, respectively. In 79  de novo   AML blasts testedfor Pgp function a median ratio of 1.45 (range 1.0–31) wasfound. From CD34 + cells present in normal bone marrowsamples ( n  =  6) a median ratio of 2.30 (range 1.73–3.73) wasfound while CD34 + cells, present in peripheral blood stem celltransplants from normal allogeneic donors or non-Hodgkin’slymphoma patients in complete remission ( n  =  15), showed amedian ratio of 6.0 (range 2.8–33).The functional MRP assay was optimized using the MRPexpressing cell lines. A 60 min accumulation assay with 1 n M calcein-AM and 3 m M  probenecid showed maximal ratios. Anefflux assay for 30–120 min, recommended in the literature, 38–40 in our hands did not improve the sensitivity. Moreover,since we aim at the possibility of studying small samples, aswell as frozen/thawed samples with sometimes low cellyields, any procedure that omits further cell loss, such as theuse of short incubation periods without extra washing steps,is preferable. For the A2008, A2008/cMOAT, A2008/MRP1–4, A2008/MRP1–6, GLC4 and GLC4/ADR the following ratios(means  ±  s.d.;  n  =  3–5) were found: 1.87  ±  0.35, 3.23  ±  0.96,13.03  ±  1.94, 9.57  ±  0.80, 0.97  ±  0.06 and 6.18  ±  0.95,respectively. In 47  de novo   AML blasts tested for MRP func-tion a median ratio of 1.25 (range 1.0–2.29) was found. FromCD34 + cells present in normal bone marrow samples ( n  =  3)or normal peripheral blood stem cell transplants ( n  =  8) amedian ratio of 1.48 (range 1.44–1.52) and 1.86 (range 1.59–2.09) were found, respectively. All cell lines were used regu-larly to control the overall performance of the Pgp and MRPassay. Combined detection of leukemia-associated phenotype and Pgp or MRP function To detect Pgp or MRP function in subpopulations of blasts,cells were first loaded with the appropriate substrate/blockercombination in an accumulation assay, followed, after awashing step, by incubation with 7-AAD and appropriate PE-and APC-conjugated MoAbs for detection of LAP for 30 min.The cells were kept on ice to prevent efflux of drugs. 7-AAD(10   l/10 6 cells) was included to exclude dead cells in thefinal analysis. The cells were washed once with 3 ml FACSbuffer and resuspended in 0.2 ml FACS buffer and kept on iceuntil acquisition. Gating of blast cells in  de novo   AML andfollow-up samples was performed using the LAPs that weredetected at diagnosis. In our assay two channels (FL2 and FL4)  Pgp and MRP function in minimal residual disease in AML MA van der Pol  et al  1557 were available for LAP detection, since one channel (FL1) isused for the functional Pgp or MRP assay and one channel(FL3) is used for viability staining with 7-AAD. Data analysisof control peripheral blood stem cells was performed on gatedCD34 +  /SSC low cells. When cells undergo a freeze–thawingprocedure, a subpopulation can become apoptotic. We haveshown previously, 41 that the combination of Syto16  +  PSC833and 7-AAD can be used to distinguish viable, dead, as wellas early and late apoptotic CD34-positive cells. Since apop-totic and dead cells do not contain the capacity to effluxdrugs, 41 these cells were always excluded from the analysis(MA van der Pol  et al  , manuscript in preparation). Statistical analysis  A Spearman’s rank correlation test was used to determine thecorrelation coefficient. Linear regression was used to deter-mine the slope and intercept of the relationship. The Wil-coxon signed-rank test for paired samples was used with a95% confidence interval to compare results obtained for freshand frozen/thawed samples. Level of significance:  P     0.05. Results Functional characterization of MRD for Pgp and MRP functionis important to explain the role of multidrug resistance inAML. In order to perform such a study, detection of leukemia-associated phenotype (LAP) and Pgp or MRP function must becombined preferably on fresh, but if possible also on cryopre-served samples. To check the feasibility of this approach, thefollowing questions have been addressed: (1) Which part of  de novo   AML samples can be characterized by LAPs for sub-sequent MRD detection? (2) Can Pgp and MRP activitymeasurements be combined with LAP detection both in  de novo   AML and in MRD situations? (3) What is the effect of freeze–thawing on LAP determination both in  de novo   AMLand in MRD situations? and (4) What is the effect of freeze–thawing on Pgp/MRP activity combined with LAP determi-nation both in  de novo   AML and in MRD situations? Leukemia-associated phenotypes  We first examined which percentage of   de novo   AML samplescan be characterized by one or more LAP. Blasts from 80%(102 out of 127) of AML patients displayed one or more LAPat diagnosis. Thirty-seven patients displayed one LAP, 24patients had two, 20 patients had three and in 21 patientsmore than three LAPs could be established. In 67% (68 outof 102) of the patients with a LAP present on the blasts, thestem cell marker CD34 was included. If CD34 could not bedetected, CD117 was used as a progenitor marker in 71% (24out of 34) of the patients. Cross-lineage antigen expressionwas found in 57%, asynchronous antigen expression in 39%and clear antigen overexpression in 4% of all the LAPs found.The most frequent cross-lineage phenotype is characterizedby CD7 expression on myeloid CD34 + blasts, followed byCD2- and CD19-positive phenotypes. Coexpression of CD34and CD56 offered the most frequent asynchronous phenotype,followed by coexpression of CD34 and CD11b. Antigen over-expression was most frequently found with CD33. LAPs withCD117 as early progenitor marker were frequently of a cross-lineage antigen expression type with coexpression of CD7. In Leukemia patients with LAPs present at diagnosis, the above-mentionedLAPs were used to detect MRD cells in different follow-upsamples. Simultaneous detection of Pgp or MRP function and LAP  In order to be able to determine Pgp and MRP function inMRD cells we developed a four-color FACS assay for the sim-ultaneous detection of Pgp or MRP activity and LAP. First itwas examined if the functional Pgp and MRP assay, whichwas performed prior to immunophenotyping, did not affectLAP expression. For 15  de novo   AML samples direct immuno-phenotyping or immunophenotyping following incubation forPgp and MRP activity were compared and no significant dif-ferences were found in the percentage of blasts showing a LAP(Figure 1), indicating that the activity assays did not alter LAPexpression of   de novo   AML blasts.In a similar way, 24 MRD samples were studied. As for  de novo   AML samples, the correlation between LAP percentages Figure 1  LAP expression of   de novo   AML blasts determined bydirect immunophenotyping or immunophenotyping following thefunctional Pgp or MRP assay.  De novo   AML cells were either directlylabelled for the relevant LAP or after previous incubation for Pgp (a)or MRP (b) activity. (a) Effect of the Pgp assay. The percentage of blastspositive for LAP was not significantly different ( P   =  0.624) betweenboth types of incubation: y  =  0.92 × +  5.01. (b) Effect of the MRP assay.The percentage of blasts positive for LAP was not significantly different( P   =  0.919) between both types of incubation: y  =  1.05 × −  3.30.  Pgp and MRP function in minimal residual disease in AML MA van der Pol  et al  1558 Leukemia Figure 2  Number of MRD cells in follow-up samples determinedby direct immunophenotyping or immunophenotyping following thefunctional Pgp or MRP assay. Follow-up samples were either directlylabelled for the relevant LAP or after previous incubation for Pgp (a)or MRP (b) activity. (a) Effect of the Pgp assay. The number of MRDcells was not significantly different ( P   =  0.164) between both types of incubation: y  =  0.78 × +  0.19. (b) Effect of the MRP assay. The numberof MRD cells was not significantly different ( P   =  0.330) between bothtypes of incubation: y  =  0.74 × +  0.29. determined with and without applying the functional assaywas good (Figure 2). Two representative examples of the sim-ultaneous detection of Pgp activity and LAP in MRD situationsare shown in Figure 3. In the first example, a large fraction of the CD34-positive cells showed the CD2 expression (Figure3b), while in the second example only a minor fraction of theCD34-positive cells was positive for CD2 (Figure 3f). In bothcases gating has been performed for the CD34-positive cellswith a normal phenotype (CD34 + CD2 − , R6 and R11) and forthe CD34-positive cells with an aberrant phenotype(CD34 + CD2 + , R7 and R12). The accompanying histograms (cand g for the CD2 − fraction) and (d and h for the CD2 + fraction) show differences in Pgp function between these twodifferent CD34-positive fractions. In both samples tested, Pgpfunction of CD34 + CD2 + cells in a MRD situation (d: 1.0 andh: 1.4) was similar to the Pgp activity found in  de novo   situ-ation (shown in a: 1.1 and e: 1.2). Likewise the Pgp functionfound in the CD34 + CD2 − cells (c: 2.3 and g: 4.1) in both MRDsamples is close to the range of values found in CD34 positivecells obtained from six normal bone marrow samples (Pgpfunction: mean  ±  s.d.: 2.3  ±  0.8, range: 1.7–3.7), for whichone example is shown in Figure 3j. Figure 3j also shows the Figure 3  Pgp function in minimal residual disease. Two represen-tative examples of bone marrow obtained from AML patients afterchemotherapeutic treatment, one with a high and one with a low per-centage of aberrant CD34 + cells. The filled curve represents the auto-fluorescence of the cells, the dotted line represents the fluorescenceof cells incubated with Syto16 alone and the solid line represents thefluorescence of cells incubated with Syto16 + PSC833. a–d Patient 1,64% of the CD34 + cells showed CD2 expression (b). Regions wereset around the CD34 + CD2 − (R6) and the CD34 + CD2 + (R7) cell popu-lation. Pgp function of the CD34 + CD2 − and CD34 + CD2 + cells isshown in c (2.3) and d (1.0), respectively. (a) 97% of the CD34 + cellsin  de novo   AML sample showed CD2 expression (not shown) and Pgpactivity of these CD34 + CD2 + de novo   blasts was 1.1 which is closeto the activity found in AML MRD cells (see d). (e–h) Patient 2, 3%of the CD34 + cells showed CD2 expression (f). Regions were setaround the CD34 + CD2 − (R11) and CD34 + CD2 + (R12) cell population.Pgp function of the CD34 + CD2 − and CD34 + CD2 + cells is shown in g(4.1) and h (1.4), respectively. (b) 50% of the CD34 + cells in  de novo  AML sample showed CD2 expression (not shown) and Pgp activity of these CD34 + CD2 + was 1.2 which is close to the activity found in AMLMRD cells (see h). Representative example of Pgp activity (j) in normalCD34 + bone marrow cells (i); compare with the presumed normalCD34 + CD2 − cells in c and g. characteristic bimodal distribution of Pgp activity in normalCD34 + bone marrow cells, which is also found in theCD34 + CD2 − MRD fractions (c and g). Taken together, theseresults indicate that the simultaneous detection of Pgp or MRPfunction and LAP can be performed without interference bothin  de novo   AML and in MRD situations.
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks