with acute promyelocytic leukemia [see comments ...

From bloodjournal.hematologylibrary.org by guest on July 26, 2011. For personal use only.

1992 80: 582-586

Correlation of CD2 expression with PML gene breakpoints in patients with acute promyelocytic leukemia [see comments] DF Claxton, CL Reading, L Nagarajan, Y Tsujimoto, BS Andersson, E Estey, A Cork, YO Huh, J Trujillo and AB Deisseroth

Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about subscriptions and ASH membership may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml

Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036. Copyright 2011 by The American Society of Hematology; all rights reserved.


RAPID COMMUNICATION

Correlation of CD2 Expression With PML Gene Breakpoints in Patients With Acute Promyelocytic Leukemia by David F. Claxton, Christopher L. Reading, Lalitha Nagarajan, Yoshihide Tsujimoto, Borje S. Andersson, Elihu Estey, Anne Cork, Yang 0. Huh, Jose Trujillo, and Albert B. Deisseroth The chromosomal translocation t(15;17)(q22:21) of acute promyelocytic leukemia (APL) fuses PML, a novel gene, with RARa, a retinoic acid receptor gene. PML-RAR hybrid transcripts were studied in 18 cases of APL using RNA-PCR. Two forms were noted: one designated 5‘. producing a 439-bp chimeric fragment, and a 3‘ form, producing a pair of fragments of 765 bp and 909 bp. 5’ forms were found in 7 of the 18 cases while the other 11 patients expressed the 3‘ forms. The chromosome 15 specific probes K3 and K2 were used to study genomic breakpoints in 12 APL patients. Comparison of these results with RNA PCR in 11 patients for whom both were available yielded a rearrangement pattern predictive of

whether the hybrid transcript was 5’ or 3’. In this way, an additional three patients in whom DNA but not RNA was available were identified as having 3’ (downstream) breakpoints and, therefore, 3‘ hybridforms. Thus, 21 cases categorized as having 5‘ or 3’ PML-RAR transcripts were analyzed for various phenotypic differences. Surface phenotyping of leukemic promyelocytes demonstrated expression of the CD2 antigen in all cases with the 5’ splice variant. Only 1 of 11 cases with the 3’ form showed CD2 expression. This difference is significant a t P = .001. o 1992by The American Society of Hematology.

T

form of transcript. The fusion type has been determined in this way in 21 cases and is predictive in all but one case of the expression of or lack of the CD2 antigen on leukemic promyelocytes.

HE CYTOGENETIC hallmark of acute promyelocytic leukemia (APL) is the translocation t(15;17)(q22; q21),* found in nearly all leukemic karyotypes of typical cases. Recently, several groups studying APL have described rearrangement of the retinoic acid binding receptor a (RARol) gene with chromosome 17 breakpoints in the second intron of RARa.*s3PML, a novel gene encoded at 15q22, is fused to with resultant hybrid transcripts and gene products apparently contributing to the disease pathogenesis. 5‘ sequences from the PML gene are spliced to the entire DNA and ligand binding domains of RARa at the 3’ end. The amount of PML coding sequences included in the hybrid molecules varies from case to case, falling into at least two groups.46 Biologic studies of two PMGRAR variants suggest quantitative differences in inducible transcription from RARa responsive reporter constructs.4,5 We examined PML-RAR fusion transcripts and genomic breakpoints in a series of fresh or cryopreserved APL cases. A polymerase chain reaction (PCR) assay that amplifies fusion cDNA of APL cells is described. The relationship between the genomic chromosome 15 breakpoints and the form of the resultant chimeric transcripts is studied using novel genomic probes derived from PML sequence. The location of the breakpoint is found to be predictive of the From the Departments of Hematology and Laboratory Medicine, The Universiv of Texas M.D. Anderson Cancer Center, Houston, X ; and Osaka University Medical School, Osaka, Japan. Submitted March 4,1992; accepted May 19,1992. A.B.D. was supported by the Anderson Chairfor Cancer Treatment and Research, the Sid Richardson Foundation, the Ladies Leukemia League of Louisiana, the Bush Leukemia Fund, the Kleberg Foundation, and the National Cancer Institute (PO1 CA55164-OIAI). Address reprint requests to David F. Clarton, MD, Department of Hematology, M.D. Anderson Cancer Center, Box 24, 1515 Holcombe Blvd, Houston, EX 77030. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. section 1734 solely to indicate this fact. 0 1992 by TheAmerican Society of Hematology. 0006-4971/92/8003-0035$3.00/0 582

MATERIALS AND METHODS Patient samples. Bone marrow or peripheral blood specimens were obtained from APL patients at varying disease stages. Specimenswere acquired during routine diagnostic or therapeutic procedures, under protocols approved by The U.T. M. D. Anderson Cancer Center. Mononuclear cells (MNC) were separated on Ficoll-Hypaque (LSM-Organon Technica, Durham, NC) and washed with phosphate-buffered saline (PBS). Some cells were cryopreserved for various periods before DNA and RNA extraction. Samples for RNA were lysed’ and frozen (-70°C) until extraction.‘ DNA samples were embedded in agarose plugs.8 Karyotypes. Bone marrow chromosomes were analyzed by routine methods? Flow cytomew. Immunophenotyping was performed on marrow aspirates collected in EDTA from presentation and relapse APL patients by dual-color flow cytometry. Cells (0.5 to 1 x lo6) were stained with 10 pL of phycoerythrin (PE)-conjugated murine monoclonal antibodies (MoAbs) and fluorescein isothiocyanate (F1TC)-conjugatedMoAbs. The following combinationsof MoAbs were used: IgG2-PE/IgGl-FITC, IgGl-PE/IgG2-FITC, HLADRPE/CDZFITC, CD33-PE/CD8-FITC, CD56-PE/CD3-FITC, IgG2b-FITC/IgM-FITC, and other MoAb-fluorochromecombinations. After 10 minutes at 4°C and erythrocyte lysis with NH&L (10 minutes) cells were washed twice with PBS, fixed in 1% paraformaldehyde, and analyzed on a FACScan using Consort 30 software (BDIS, San Jose, CA). Thawed, cryopreserved APL cells obtained at apheresis, peripheral blood and marrow MNC from healthy donors, and remission marrow MNC from APL patients were examined for coexpression of CD2 and CD33, and coexpressionof CD2 and CD13. Cells were stained with FITC-IgG1, PE-IgG2b, PE-IgG1, FITC-IgGl/PEIgG2b, FITC-IgGl/PE-IgG1, FITC-CD2/PE-CD33, and FITCCDZ/PE-CD13. cDNA synthesis and PCR Total cellular RNA (2 pg in 2 to 3 pL) was heated to 75°C for 3 minutes and cooled on ice. Eighteen microliters of a reverse transcription mix (4 ILLof 5 x RT buffer (BRL, Gaithersburg, MD), 100 pmol of random hexamers (Boehringer-Mannheim. Indianapolis, IN), 16 U RNAse inhibitor (Promega, Madison, WI), 200 U MMLV-reverse transcriptase (BRL), and deoxynucleotides (final concentration 0.5 mM each) Blood, VOI 80,NO 3 (August 1). 1992: pp 582-586


CD2 EXPRESSION AND PML BREAKPOINTS IN APL

were added. After cDNA synthesis at 42°C for 1 hour, 5 p L was added to 25 p L PCR reaction mixtures.'" Primers were: PMLl gcggtaccagcgcgactacgaggagatg,RAR537 gcggaagaagcccttgcagccctcacagg, PMLZ cgcctggatgctgtgctg," PMLASI gggcaggctggaggctccttgtgctct,s RAR504 gtggtagcctgaggacttgt.4 PMLl and RAR5.77 were used for hybrid junction region analysis. PMLl and PMLASI were used to assess normal PML as a control for amplifiable cDNA. Tubes were heated at 93°C for 4.5 minutes before 40 temperature cycles (93°C 30 seconds; 63°C 30 seconds; 72°C 2 minutes). PCR products (12 pL) separated on 2% agarose gels were stained, photographed, and transferred to ny1on.I' Prehybridization and hybridization were at 57°C" with end-labeled oligonucleotide probes (RAR504 and PML2). Final washes were 1 x SSPE (0.15 mol/L NaCI, .01 mol/L Na2H2P04, 1 mmol/L EDTA), 0.1% sodium dodecyl sulphate (SDS) at 57°C for 30 minutes. Genorttic cloning and Soitthem analysis. A genomic library constructed from APL DNA in EMBL3I2 was screened with a 380-bp KpnI/P.vl fragment of the RARa c D N A ' ~from American Type Culture Collection (ATCC; Rockville, MD). Segments of interest were subcloned into pBluescriptllSK+ (Stratagene, La Jolla, CA). DNA specimens digestedX with restriction enzymes were separated on 0.7% agarose gels before Southern transfer and hybridization.'" Probe K2 was pre-annealed with Cot1 DNA (BRL) before hybridization. pH60I4 = TJyl and Jy2 was obtained from ATCC. pSKTJi5l probe, identical to MH6,I5 was a gift from Dr J. Kagan. RESULTS

cDNA amplification. Amplification of PML-RAR hybrid cDNA and normal PML cDNA produced results consistent with published sequenccs. PCR of normal PML (primers PMLl and PMLASl) gave rise to two products of 826 bp and 970 bp. The difference in size of these two fragments results from inclusion of a 144-bp altcrnativc exon in the 970-bp fragment.s PML-RAR PCR of 3' fusion transcripts using primers PMLl and RAR537 yiclded two products of 765 bp and 909 bp, rcspectivcly,s thc difference again seen because of the 144-bp exon included in the larger band. 5' hybrid cDNAs produced 439-bp fragments because of omission of 3' PML sequences. Diagnosis or relapse bone marrow RNA samples wcre available from 18 patients with t(15;17). Each cDNA was amplified for both the normal PML transcript (primers PMLl and PMLASI) and the chimcric transcript (PML1 and RAR537). APL cases, normal bone marrow, cell lines K562 and Hela, and several other acute myelogenous leukemia (AML) specimens demonstrated the normal PML

Fig 1. (A) Ethidium bromidestained gel showing amplification of PML (primers P M L l and PMLAS1) in upper panels and PML-RAR fusion cDNAs (using PMLl and RAR537) in lower panels. Lanes 1 through 5, patients 5.6.9.13,and 14; ianes6through 8. other AML cases; lane 9, K562 RNA; lane 10, no RNA; lane 11, PhrXIHaelll markers. ( 8 )RAR504 hybridization. (C) PMLZ hybridization.

583

product. Results of a represcntativc expcriment are shown in Fig 1A. In thc uppcr panel some normal PML amplification is sccn in all lanes except the no RNA blank (lanc 10). The lower pancl shows PML-RAR chimeric cDNAs amplificd using primcrs PMLl and RAR537. Lanes 1 and 5 have a promincnt amplified band of 439 bp, the expected fragment of thc 5' junctional hybrid cDNA.4 In lancs 2 through 4, thc 3' fragments described above,s two bands of 909 and 765 bp arc sccn. Lanes 6 through 8, AMLs without t(15;17), and lane 9 (K562 cells), show no PML-RAR products. To confirm that PCR products were segments of the previously reported C D N A S ~they . ~ were hybridizcd to probes specific for intervening PML (PML2) and RAR (RAR504) sequcncc. RAR504 hybridized to both hybrid cDNAs but not to normal PML (Fig lB), while PMLZ hybridized to both the hybrid forms and to normal PML bands (Fig 1C). Five hybrid cDNAs amplified from diagnosis or rclapsc bone marrow arc shown in lancs 2 through 6. In the thrcc samplcs with 3' fusions (lanes 2 through 4) an additional (third) fragmcnt is secn at approximatcly 500 bp. As it hybridizes to both PML2 and RAR504, it may represent a novel altcrnativc splicing form. Results of RNA-PCR are given in Table 1 for 18 patients. In addition, five AML leukemic, K562, and Hela cell RNAs havc bccn studied. No typical PML-RAR bands wcre secn in AMLF,but four of five demonstrated normal PML bands (one did not show any, possibly bccausc of tcchnical factors). Genomic cloning and study of rearrangements. Three RARa hybridizing cloncs werc isolated from the APL gcnomic library. Figure 2 shows restriction maps of the phagc. Clonc 5b provcd to bc rcarrangcd in the second intron of RARa. Table 1 gives results of Southern analysis. K3, a probe from the novcl portion of 5b, was used to study chromosomc 15 brcakpoints in 12 APL DNA specimcns. Cases 2,7, and 8 had rearrangement with BamHI digestion but not HindIII, and paticnt 5 showed rcarrangemcnt with Hind111 and EcoRI but not EamHI. Thereforc, these four patients must have breakpoints upstream (to the K3 sidc) of the chromosome 15 HindIII site (see Fig 1). APLs 1,9, 10, and 13 had rearrangements with EcoRI but not BamHI or HindIII; thus breakpoints must be downstream of the HindIII site. As cases 3, 4, and 15 had no rearrangements with K3, an additional probe K2, located downstream of K3, was used. With probe K2 cases 4 and 15 showed

From bloodjournal.hematologylibrary.org by guest on July 26, 2011. For personal use only. CLAXTON ET AL

584 Table 1. RNA-PCR and Southern Analysis of APL Cases Case No.

PML-RAR RNA-PCR

K3 BamHl (12 kb)

K3 Hindlll (7 kb)

K3 FcoRl (8kb)

G

G G G G

G G G

K2 BamHl (8kb)

~

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

3' 5'

R

-

G G G G

5' 3' 5' 5' 3' 3' 3' 5' 3' 5'

R R

G G G R

G

-

-

G -

3' 3' 5' 3' 3' 3'

-

-

R

R

R

G G G G G G G G G -

G G G R R R

G R

G -

-

-

-

~

Correlation of RNA-PCR results with Southern analysis. Probes are given above the enzymes used. Sizes of germline restrictionfragments are given below each enzyme. G, germline; R, rearranged. K3 hybridization to EcoRV digested specimens gave identical results as for 8amHl and are not shown.

rearrangements after BamHI digestion. No rearrangement has been shown in case 3, but a 3' breakpoint seems likely given germline K3 BamHI and HindIII fragments. Correlation of cDNA splice point with genomic rearrangement. Results of PCR hybrid transcript analysis and Southern blots are shown in Table 1. DNA samples were studied for chromosome 15 breakpoints in 11cases in which RNA was analyzed. As seen in Table 1, rearrangement upstream of the chromosome 15HindIII site shown in Fig 1 (ie, K3-BamHI or K3-HindIII) predicts the 5' chimeric cDNA. Breakpoints downstream of this site (rearranged K3-EcoRI or K2-BamHI fragments with germline K3HindIII) occur in 3' fusion cases.

K

K

E

I

I

-

KV

I I

K3

BK K H K

I I I

I

0

K2

I

BEB H

I l l I

Patient samples and cytogenetic studies. Hematologic values are listed in Table 2. Data are subdivided into cases with 3' breakpoints and those demonstrating 5' fusions. Routine parameters showed no significant differences between the two groups. Bone marrow cytogenetics were available for all 21 patients, and 19 patients had the t(15;17)(q21;22). Patients 2 and 12 had no analyzable metaphases. Flow cytomeqv. Bone marrow surface phenotype was examined at referral to this institution in all but six cases. Patient 15's specimens were studied at first and second relapses. There was no evidence for expression of pan-T markers CD3, CD5, or CD7 on any APL but there was a striking correlation of CD2 expression with the 5' fusion cDNA. For this reason cryopreserved specimens of blasts and promyelocytes from cases 5, 10, and 12 were thawed and studied for coexpressionof CD2 and CD33. Results are given for all cases in Table 2. Most 3' fusion APLs had uniformly low numbers of cells in the blast region staining for CD2. 3' cases 6 and 19 showed some CD2 expression but had similar numbers of CD3 staining cells, suggesting that these were T cells. Only 1 of the 11 3' cases (no. 20) showed significant CD2 expression (48%). In marked contrast, all seven of the 5' leukemias showed significant CD2 marking with few CD3 staining cells. Given the numerical overlap between CD2 and CD33 it may be concluded for all 5' cases except no. 7 that there must have been coexpression of these two antigens on the same cells. Although this cannot be proven for patient 7, whose blast gated cells showed 9% CD2 and 82% CD33, such coexpression was likely given the absence of CD3 and CD8 staining populations. Differences in CD2 expressionbetween 5' and 3' cases were significant at P = .001. Cases 5, 10, and 12 were studied for CD2 coexpression with CD33. APLs 5 and 12, which had 5' PML fusion junctions, showed coexpression of CD2 and CD33 on 34.3% and 22.9% of cells, respectively. Case 10 (3' fusion) had no significant coexpression (0.2%) and few (2.5%) CD2 staining cells. CD2 and CD33 coexpression was also demonstrated in case 8 during remission (Fig 3). Collected 5 weeks before relapse, this specimen showed 3.9% CD2 and CD33 coexpression. T-cell receptors. To further assess the lymphoid differen-

B

I

K

I

GL

K

BEB H

B

K

I

I I I I

I

I

2 Kb

RR

Fig 2. Genomic clones and probes. Phage clones are fine lines. Heavy bars are genomic fragmentswith chromosome 17derived sequences (solid) and chromosome 15 portions (open). Restriction sites: K, Kpnl; E, EcoRI; B, B8mHI; H, Hindlll; V, EcoRV (partial map). The chromosome 15 Hindlll site separated cases into 5' (upstream) and 3' (downstream).


585

CD2 EXPRESSION AND PML BREAKPOINTS IN APL Table 2. Clinical Features and Flow Cytometry WBC

Hg

Plls

CD2

CD3

COB

1 46 3 32 4 54 6 41 9 55 10 62 11 32 13 27 15 54 16 27 17 40 19 21 20 17 21 47 Mean: 39.6

8.9 3.6 1.4 1.3 0.5 17.5 2.3 12.7 1.1 1.3 0.8 36.6* 6.6 6.9 7.3

9.2 86 10.0 2 95 11.3 60 10.2 10.3 47 9.0 5 9.4 118 10 10.0 21 8.2 7.0 4 10.8 84 6.5t 63* 27 7.6 24 11 9.3 46

NA 2 1 4 1 3' 1 NA It 4 3 5t 48

NA 2 5 4 1 NA NA 2t 7 1 6t 6 1 3.3

NA 99 0 97 1 4 8 0 97 0 79 NA 92' 0 96 NA NA O t 97t 1 72 0 89 7 t 91t 0 94 0 90 6q22. Proc Natl Acad Sci USA 83: t(15;17) translocation of APL fuses the retinoic acid receptor a 6568,1986 gene to a novel transcribed locus. Nature 347558,1990 13. Petkovch M, Brand NJ, Krust A, Chambon P: A human 4. Kakizuka A, Miller WH Jr, Umesono K, Warrel RP Jr, RAR belonging to the family of nuclear receptors. Nature 330:444, Frankel SR, Murty VS, Dmitrovsky E, Evans R M Chromosomal 1987 translocation t(15;17) in human acute promyelocytic leukemia 14. Lefranc MP, Rabbitts TH: Two tandemly organized human fuses RARa with a novel putative transcription factor, PML. Cell genes encoding the T-cell y constant region show multiple rear66:663,1991 rangement in different cell types. Nature 316:464,1985 5. de The H, Lavau C, Marchio A, Chomienne C, Degos L, 15. Hara J, Benedict SH, Chamagne E, Takihara Y, Mak TW, Dejean A The PML-RARa fusion mRNA generated by the Minden M, Gelfand EW T cell receptor 6 gene rearrangements in t(15;17) translocation in APL encodes a functionallyaltered RAR. acute lymphoblasticleukemia. J Clin Invest 821974,1988 Cell 66:675,1991 16. Dalla-Favera R, Lombardi L, Pelicci PG, Lanfrancone L, 6. Pandolfi PP, Grignani F, Alcalay M, Mencarelli A, Biondi A, Cesarman E, Neri A: Mechanism of activation of c-myc oncogene LoCoco F, Grignani F, Pelicci PG. Structure and origin of the APL in B-cell lymphomagenesis. Ann NY Acad Sci 511:207,1987 myl/RARa cDNA and characterization of its retinoid-binding and 17. Mills K, MacKenzie ED, Birnie GD: The site of the transactivation properties. Oncogene 61285,1991 breakpoint within the bcr is prognostic in CML. Blood 72:1237, 7. Chomczynski P, Sacchi N Single step method of RNA 1988 isolation. Anal Biochem 162:156,1987 18. Przepiorka D: Breakpoint of bcr in CML does not correlate 8. Feddersen RM, Van Ness B G Single copy gene detection with disease phase or prognosis. Cancer Genet Cytogenet 36:117, requiring minimal cell numbers. BioTechniques 744,1989 9. Trujillo JM, Cork A, Ahern MJ, Youness EL, McCredie KB: 1988 19. Springer T A Adhesion receptors of the immune system. Hematologic and cytogenetic characterization of the 8/21 transloNature 346425,1990 cation acute granulocyticleukemia. Blood 53:695,1989