江汉油田第一职工医院 (433124) 刘立根 范嘉琏 张志宏 陈润光 张因 男患,60岁,双侧耳垂肿大二年并伴白细胞增多, 于1987年12月12日入院。患者于1985年初发现双测耳垂呈紫红色, 同年底开始肿大伴疼痛。1987年双侧耳垂肿大日见明显且延及耳轮, 伴乏力、盗汗和皮肤搔痒同年11月于本院耳科就诊查血发现白细胞效增多转内科。入院体检:T36.2℃,双侧耳垂及耳轮肿大, 紫红色, 质软, 有压痛,无出血征贫血貌耳后、颈部、腋窝及腹股沟可触及黄豆致蚕豆大小淋巴结数个, 不融合巩膜无黄染胸丹无仄痛。双肺及心脏无异常肝于右肋下4cm,脾大肋下平脐, 质中无压痛。脊柱及神经系统无异常,血象:Hb:80g/L,WBC:80x109 /L,分类中淋巴细胞占85%, 偶见不成熟淋巴细胞:plt:137x109/L。骨髓像:增生极度活跃,淋巴细胞占87.5%。以成熟淋巴细胞为主(81%)。而原始及幼椎淋巴细胞仅占6.5%。Pox(-)。全片见巨核细胞3个, 血小板少见,骨髓报告为CLL,右侧耳垂及右侧腋窝淋巴结活检病理报告为慢性淋巴细胞白血病(CLL)浸润,X线胸片,双肺心隔无异常腹部B超:肝脾大, 肝门及腹膜后淋巴结肿大。诊断:予方案COP化疗三个疗程后, 肿大的耳垂及淋巴结、肝脾均明显缩小, 住院三月余好转出院。此后间断接受COP方案和瘤可宁化疗, 病情稳定, 现仍在治疗观察中。 讨论:CLL皮肤损害的发生率为10%~50%, 表现以结节为主的红皮病。有些类似牛皮癣, 局部浸润,以面部和头皮多见, 结节常常是一单个。本例以双侧耳垂表现首发, 实为罕见,泪腺和(或)唾液腺肿大为CLL浸润少见,且可为首发的表现。因此, 识别的少见表现, 全而详细的物理学和血液学检查对CLL防止的漏诊和误诊十分重要。
以黄疸为首发表现的多发性骨髓瘤一例谢英华, 刘立根, 高武, 赵莉敏, 韩曦瑶( 复旦大学附属上海市第五人民医院血液内科, 上海200240关键词: 黄疸; 多发性骨髓瘤; 肝脏中图分类号: R733.3 文献标识码: B 文章编号: 1671-2870( 2006) 01-0068-01J Diagn Concepts Pract 2006, Vol.5, No.1 病例: 女, 56 岁。皮肤黄染1 周, 以“急性黄疸性肝炎”收住我院传染科。患者1 年前开始出现皮肤巩膜黄染, 伴中上腹胀不适、纳差和乏力, 无发热、腹泻, 自觉尿色加深,大便无异常改变。既往史、个人史和家族史无特殊。入院体检: 神志清, 皮肤、巩膜重度黄染, 无肝掌及蜘蛛痣; 肺脏和心脏未发现异常; 腹平软, 无压痛及反跳痛, 肝上界位于右锁骨中线第5 肋间, 肝脾肋下未触及, 肝区叩击痛阳性, 移动性浊音阴性;双下肢无凹陷性浮肿。入院后血常规:WBC6.03×109/L, 中性粒细胞0.649, RBC 3.55×1012/L, Hb105 g/L,PLT 163×109/L, 网织红细胞百分比为0.01。血清ALT 175 U/L ( 参考值0~40 U/L) , AST 478 U/L ( 参考值4~40 U/L) ,AKP 165 U/L( 参考值15~112 U/L) , γ-GT 268 U/L( 参考值5~54 U/L) , 总胆红素( TB) 229.8 μmol/L, 直接胆红素(DB)161.1 μmol/L, 白蛋白32.2 g/L,球蛋白64.6 g/L, IgG 41.3 g/L,钙2.27 mmol/L, BUN 4.9 mmol/L, Cr 80 μmol/L。酶联免疫吸附法测定甲、乙、丙、丁、戊和型肝炎病毒抗体均阴性。直接和间接Coombs 试验均阴性。腹部B 超和肝脏MRI 检查未见异常。骨髓涂片: 骨髓有核细胞增生较活跃, 浆细胞明显增多, 该类细胞胞体中等偏大, 圆或椭圆型, 胞质量丰富,呈泡沫感, 部分胞质内可见少量细小的紫红色颗粒, 核多偏位、圆形, 染色体呈细网状, 可见一大而明显的核仁, 双核型、三核型可见, 幼浆细胞比例明显增多( 23.5%) , 成熟红细胞呈缗钱状排列。血、尿蛋白免疫固定电泳: IgG 单株峰, κ轻链型。头颅、颈椎、胸椎、肋骨、骨盆X 线平片无异常发现。诊断: 多发性骨髓瘤(MM) IgGκ轻链型( Ⅰ期A 型) 。入院后给予还原型谷胱甘肽、复方甘草酸苷和丁二磺酸腺苷蛋氨酸等保肝退黄治疗, 肝功能逐渐改善。入院半个月后给予酞胺哌啶酮联合大剂量地塞米松( 40.5 mg, 1 次/d, po, d1~4,d 9~12,d17~20) 治疗, 2 疗程后复查球蛋白27.8 g/L, IgG12.2 g/L, 血常规:WBC 17.47×109/L, 中性粒细胞0.824, RBC4.03×1012/L, Hb 121 g/L, PLT 236×109/L, 骨髓涂片: 原幼浆细胞比例3%, 已达完缓解(CR) , 巩固治疗至今。 讨论:MM起病多徐缓, 患者可有数月至10 余年的无症状期, 高血钙、肾脏损害、贫血和骨骼破坏( calcium, renalinsufficiency, anemia, bone lesions; CRAB) 是其突出的临床表现, 以黄疸为首发表现甚为少见。经查找近50 年来文献, 国外约有40 余例MM伴阻塞性黄疸的案例报道, 然而其发病机制一直未阐明。1994 年Terada 等[1]首次从组织学上揭示其原因为M蛋白轻链沉积于肝内外胆管系统, 造成胆管狭窄而引起阻塞性黄疸; 也有因胰腺淀粉样物质沉积, 胰头肿大压迫胆管造成阻塞性黄疸的报道[2]。 MM所致的肝脏损害多由于淀粉样物质在肝脏沉积[1]。淀粉样变性在MM的发生率约10%, 肾脏是MM患者淀粉样变性发生较早且最常见的受累器官, 肝脏是除肾脏外最常见的受累器官。但如不伴有肾脏淀粉样变, 肝功能受影响的程度是极微小的, 阻塞性黄疸更是罕见[3]。本例患者表现为黄疸、肝酶升高, DB/TB>60%, 符合阻塞性黄疸的代谢特点, 肝炎病毒血清学检查阴性, 免疫球蛋白IgG 单克隆升高, 血、尿蛋白免疫固定电泳提示存在M蛋白, 骨髓细胞学检查患者浆细胞比例在15%以上, 患者可诊断为MMIgGκ轻链型。迄今仅有为数不多肝脏淀粉样变而肾脏未明显受累的轻链沉积病( LCDD) 的报道, Michopoulos 等[4]报道了首例确诊LCDD 后不久因阻塞性黄疸及严重的胆汁淤积而死亡却始终无肾功能受累的病例。本例患者经酞胺哌啶酮加大剂量地塞米松辅以复方甘草酸苷治疗取得了CR, 提示部分患者可以通过有效治疗得到控制。[参考文献][1] Terada T, Hirata K, Hisada Y, et al. Obstructive jaundicecaused by the deposition of amyloid-like substances inthe extrahepatic and large intrahepatic bile ducts in apatient with multiple myeloma[J]. Histopathology,1994,24(5):485- 487.[2] Mitchell DG, Hill MC. Obstructive jaundice due to multiplemyeloma of the pancreatic head: CT evaluation[J]. JComput Assist Tomogr,1985,9(6):1118- 1119.[3] MaciasRoblesMD,Navia-OsorioGarcia-Braga JM,MenendezCaro JL, et al. Jaundice secondary to intrahepatic depositof light chains as a presenting form of multiple myeloma[J].An Med Interna,1994,11(2):74- 76.[4] Michopoulos S, Petraki K, Petraki C, et al. Light chaindeposition disease of the liver without renal involvementin a patient with multiple myeloma related to liver failureand rapid fatal outcome[J]. Dig Dis Sci,2002,47(4):730-734.( 收稿日期: 2005-11-01)
Identification of three F5 gene mutations associatedwith inherited coagulation factor V deficiency in two ChinesepedigreesQ.-H. FU,*1 R.-F. ZHOU,*1 L.-G. LIU, W.-B. WANG,* W.-M. WU,* Q.-L. DING,* Y.-Q. HU,*X.-F. WANG,* Z.-Y. WANG* and H.-L. WANG**Division of Thrombosis and Hemostasis, Shanghai Institute of Hematology, Ruijin Hospital, Shanghai Second MedicalUniversity, Shanghai; Blood Centre of Zhejiang Province, Hangzhou, Zhejiang; and Department of Hematology,Shanghai Fifth People乫s Hospital affiliated to Fudan University, Shanghai, ChinaSummary. To investigate the molecular defects intwo Chinese pedigrees with inherited factor V (FV)deficiency. A 37-year-old male (proband 1) and an18-month-old boy (proband 2) were diagnosed asinherited coagulation FV deficiency by severelyreduced plasma levels of FV activity and antigen.All 25 exons and their flanking sequence of F5 genewere amplified by polymerase chain reaction (PCR)for both probands and the PCR products weredirectly sequenced. Total RNA was extracted fromthe peripheral lymphocytes of proband 1 for detectingthe changes at mRNA level.The homozygousdeletion IVS8 )2A>G was identified in the F5 gene ofproband 1 and complementary DNA (cDNA) analysisrevealed the abolishment of the canonical splicingsite by the mutation and the activation of the crypticacceptor site 24 bp upstream instead. The insertionintroduced eight additional amino acids (AA) intothe FV protein. Two heterozygous mutations of F5gene were discovered in proband 2. The 2238-9delAG in exon 13 introduced a premature terminationcode at 689 AA and the substitution of G6410 byT in exon 23 lead to the missense mutationGly2079Val.Three F5 gene mutations, IVS8)2A>G, 2238-9del AG and G6410T, have beenidentified in two Chinese pedigree with congenitalFV deficiency, respectively.Keywords: coagulation factor V, deficiency, genemutation, splice siteIntroductionHuman coagulation factor V (FV), a single chainglycoprotein with molecular weight ratio (MWR)about 330 kDa, is an essential pro-cofactor in bloodcoagulation cascade. FV protein is comprised of2224 AA residues including 28 AA signal peptidesand six domains that is orientated as A1-A2-B-A3-C1-C2. FV is activated to its active form (FVa) bythrombin or activated FXa which removed the Bdomain, generating a heavy chain (1.709 AAresidues, A1-A2 domain) and a light chain (1546.2196 AA residues, A3-C1-C2 domain) that arelinked together in the presence of calcium ions.FVa binds to FXa and serves as its cofactor in theprothrombinase complex that convert prothrombinto thrombin [1.3].The F5 gene is mapped to human chromosome1q21.25 [2], and spans more than 80 kb consistingof 25 exons and 24 introns. There is a 35.40%homology existing in A and C domains of FV andFVIII. Conversely, extensive divergence exists in theB domains of the two genes [4].Congenital FV deficiency is a rare bleeding disorder,inherited as an autosomal recessive trait, withan incidence of about 1 in 1 million [5,6]. Commonsymptoms are occasional nose bleeding, easy bruising,bleeding following surgery and excessive menstrualbleeding in female. Haemarthroses andhaematomas occur in only one-quarter of thepatients, whereas life-threatening bleeding episodes1Qi-Hua Fu and Rong-Fu Zhou contribute equally to this workand should be considered as co-first author.Correspondence: Prof. Hong-Li Wang, Vice Director, ShanghaiInstitute of Hematology, Ruijin Hospital, Shanghai Second MedicalUniversity, No. 197 Ruijin II Road, Shanghai 200025, China.Tel.: +86-21-64370045 (extn 610602); fax: +86-21-64743206;e-mail: wanghongli602@163.comAccepted after revision 19 February 2004Haemophilia (2004), 10, 264.270 DOI: 10.1111/j.1365-2516.2004.00896.x264 2004 Blackwell Publishing Ltdin the gastrointestinal tract and in the central nervoussystem are rare [7]. Up to now, over 200 cases of FVdeficiency have been reported, but the molecularbasis for FV deficiency has been established in only afew cases [4,7.9,11.27].In this study, we reported three F5 gene mutationsin two unrelated Chinese families with FV deficiency.In proband 1, a homozygous IVS8 )2A>G transitionin genomic DNA and an in-frame insertion of 24 bpin cDNA were identified in F5 gene. These resultssuggested that the splice mutation IVS8 )2A>Gabolished the acceptor splice site of intron 8 thuscausing the activation of a cryptic acceptor splice site24 bp upstream of the physiological one within theintron. In proband 2, novel compound heterozygousmutations, 2238-9del AG in exon 13 and G6410Tmissense mutation in exon 23 were found.Materials and methodsCase historyProband 1, a 37-year-old male, was admitted tohospital because of severe headache and vomiting for2 days. During childhood, he mainly suffered fromgingival bleeding. At the age of 16 years, he wastreated with fresh frozen plasma infusion for apersistent haemorrhage after having had appendectomy.Computerized tomography (CT) scan of thebrain revealed that there was a haematoma of50 丒 29 丒 60 mm in size in frontal lobe of thecerebrum. Laboratory tests suggested that he had anormal platelet count, liver function and was negativefor antibodies to hepatitis B and C. He was treatedwith transfusion of fresh frozen plasma and medicationfor decreasing intracranial pressure. Threemonths later, a repeated CT scan of the brain showedthat the haematoma was completely absorbed. Hisparents were known to be consanguineous. Hisgrandmother died of severe gastrointestinal bleeding.Proband 2 is an 18-month-old boy. At the age of6 months, he was treated with transfusion of freshfrozen plasma for gastrointestinal bleeding. Subsequently,he presented with recurrent epitaxis, easybruising and gum bleeding. No history of bleeding isreported in the other family members and his parentsare not consanguineous.The pedigrees of these two families are illustratedin Fig. 1a and b.Sample collection for coagulation testsFollowing informed consent, peripheral venousblood from the probands and family members wascollected gently into 0.109 mol L)1 sodium citrate(9:1 v/v). Platelet poor plasma was obtained bycentrifugation at 2000 g for 10 min and aliquotswere stored at )80 C until use.Coagulation factor activities and FV antigen assaysThe plasma coagulation factors activities weremeasured following the protocol of one-stageclotting assay using ACL 3000 Plus automatedcoagulation apparatus (Instrumentation Laboratory,Milan, Italy); plasma FV antigen levels were measuredby using paired antibodies for FV (CedarlaneLaboratories limited, Ontario, Canada) in anenzyme-linked immunosorbent assay (ELISA). FVantigen level was expressed as percentage of controlplasma pooled from 30 normal individuals, set as100%. The linear range of the functional andimmunological tests was 1.200% and 0.78.200%,respectively.DNA and RNA isolationGenomic DNA was extracted from whole bloodaccording to a standard phenol-chloroform protocol.Total RNA from peripheral lymphocytes of proband1 was prepared for further studies [8].Genomic DNA amplification and sequencingThe primers for polymerase chain reaction (PCR,primers sequences are available on request) weredesigned according to the published F5 genesequence (GenBank accession number Z99275).The amplification was performed in final volume of100 lL, including 10 lL 10X PCR buffer with(a)(b)Fig. 1. Two pedigrees of factor V (FV) deficiency. (a) Pedigreewith IVS8 )2A>G mutation. (b) Pedigree with compoundheterozygous mutation G6410T and 2238-9del AG. The probandsare indicated by arrows.THREE F5 GENE MUTATIONS 265 2004 Blackwell Publishing Ltd Haemophilia (2004), 10, 264.270MgCl2 (containing 100 mmol L)1 Tris-HCl (pH8.3), 500 mmol L)1 KCl and 15 mmol L)1 MgCl2),8 lL 2.5 mmol L)1 dNTP, 5 lL of each primer(10 lmol L)1), and 2.5 U Taq DNA polymerase[Takara Biotechnology (Dalian) Co., Ltd, Dalian,China], 500 ng genomic DNA. After denatured at95 C for 5 min, 30 cycles of 95 C for 30 s,55.60 C for 30 s, 72 C for 30 s were performed,then extended at 72 C for another 10 min. The PCRproducts were purified from agarose gel usingQIAquick Gel Extraction Kit (Qiagen GmbH,Hilde, Germany) and subsequently sequenced byusing the ABI 377 sequencer (Applied Biosystems,Foster City, CA, USA). All the 25 exons and theirintron.exon boundaries of F5 gene from the probandsgenomic DNA were amplified and sequenced,whereas only corresponding sequences were amplifiedand sequenced for other family members of thetwo pedigrees.cDNA synthesis and sequencingThe first strand of cDNA was synthesized by reversetranscriptionPCRfrom totalRNAof proband 1, usingrandom hexamers as primers and MMLV-RT (GibcoBRL, Life Technologies, Rockville, MD, USA). Thereverse transcription product was then amplified byPCR with the primers from exon 8 (5亼-ACAGGTCTAGCATTTGGAT-3亼) and exon 12 (5亼-TCCTCATGCCTCTTTCCATA-3亼) and sequencedthereafter.Restriction fragment length polymorphism analysisRestriction fragment length polymorphism (RFLP)analysis was performed to validate the G6410Tmutation. The PCR products of exon 23 and itsflanking regions of proband 2, his parents and 100random normal individuals unrelated to the patientwere digested with restriction enzyme RsaI (MBI,Fermentas, Vilnius, Lithuania) per the conditionssuggested by the manufacturer. The digested productswere examined by agarose electrophoresis.ResultsPhenotype identificationThe patients had a significantly prolonged activatedpartial thromboplastin time (APTT) and prothrombintime (PT). FV activity and antigen level of proband 1were 1.6% and 7.2% of the normal control, respectively;whereas for proband 2, they were <1% and1.5%, respectively. The activities of other coagulationfactors, including FII, FVII, FVIII, FIX, FX, FXIand FXII were in normal ranges for both patients.Other heterozygous family members also haddecreased FV activities and antigens (see Table 1).Molecular defects of these two pedigreesProband 1 Direct genomic DNA sequencing of F5gene 25 exons and their boundaries sequence ofproband 1 revealed three homozygous variations,including one silent mutation in exon 2 (A327G),one polymorphism in exon 13 (C3930A), whichintroduced Leu to Ile substitution at amino acid (AA)position 1257, and a previously reported homozygousputative causal splicing mutation IVS8 )2A>G[9] (Fig. 2a). Further analysis of DNA samples fromthe family showed that all of the members but theprobands wife were heterozygous of this splicingmutation, in agreement with the autosomal recessivepattern of inheritance.Table 1. Some laboratory test resultsof pedigrees.Activated partialthromboplastintime (APTT, S)Prothrombintime (PT, S) FV:C (%) FV:Ag (%)Pedigree 1I1 37.2 12.1 66.3 47.6I2 35.9 12.4 68 49.6II1 33.4 11.0 128.6 89.7II2 (proband 1) 123 43.4 1.6 7.2II3 36.6 11.3 68 47.4III1 37.0 11.5 55.1 37.9Pedigree 2I1 38.5 12.6 54.3 49.5I2 37.8 13.1 54.3 56.1II1 (proband 2) 249.2 46.6 <1 1.5Control 33.44(n . 30)11.2.13.6(n . 30)50.0.150.0(n . 30)100266 Q.-H. FU et al.Haemophilia (2004), 10, 264.270 2004 Blackwell Publishing LtdTo clarify where the cryptic acceptor splice sitecaused by this splicing mutation was, further analysisof FV cDNA sequence of proband 1 was performedand a 24 bp insertion between exon 8 and exon 9was identified (Fig. 2b). The insertion sequencescould be traced back to the 3亼-flanking sequence ofintron 8, which indicated that the IVS8 )2A>Gmutation abolished the acceptor splice site of intron8 thus causing the activation of a cryptic acceptorsplice site 24 bp upstream of the physiological one(Fig. 2a). The sequence of the amplified cDNA didnot reveal the presence of normal mRNA of F5 gene.Proband 2 TheDNAanalysis disclosed four variationsin the proband乫s F5 gene, including one silent mutationsin exon 4 (G642T), one polymorphism in exon10 (G1628A), which introduced Arg to Lys substitutionat AA position 485. This polymorphism has beenreported previously and the frequency of Lys485 inChinese population was 0.15 [10]. Two heterozygousputative mutations, 2 bp deletion (2238-9del AG) inexon 13 and G6410T in exon 23 were identified(Fig. 3a and b), the 2238-9del AG mutation introducinga frameshift and a premature stop at codon 689,while the G6410T missense mutation causing thesubstitution of Gly by Val at codon 2079. Theprobands father was heterozygous for the G6410Tmutation, whereas the probands mother was heterozygousfor the 2238-9del AG mutation.Restriction enzyme analysisThe nucleotide substitution G6410T creates a newrecognition site for restriction enzyme RsaI, whichwill cleave the 438 bp PCR product of exon 23 andits flanking regions with the mutation into twofragments at the length of 194 and 244 bp (Fig. 4).The digested products from proband 2 and his fatherhad all three fragments 194, 244 and 438 bp,indicating both of them were heterozygous for thegene variation, while that from his mother and 100normal controls had only one 438 bp fragment,precluding the possibility of the gene variation as acommon polymorphism.DiscussionSince Murray et al. [27] reported the molecularmechanism of FV deficiency for the first time in1995, 29 F5 gene mutations associated with FVdeficiency have been reported [4,8,9,11.27]. Mostof mutations (18 of 29) lead to the productionof truncated FV protein, and three of them aresplice site mutations [9,17,18]. In this study, we(a) Intron8 Start exon9(b) Exon8 Insertion sequence Exon9Fig. 2. DNA and cDNA analysis of proband 1 with mutation IVS8 )2A>G. (a) The genomic DNA analysis:Oindicates the A>G transition.The cryptic accept splice site in intron 8 is underlined. (b) The cDNA analysis. The sequence between two arrows is inserted into thepresumed junction of exon 8 and 9. The insertion will introduce eight additional amino acids, listed below, into the factor V (FV) peptides.AAT AAA TTT GAT TTA ACT TTG TGGAsn Lys Phe Asp Leu Thr Leu TrpTHREE F5 GENE MUTATIONS 267 2004 Blackwell Publishing Ltd Haemophilia (2004), 10, 264.270investigated two patients with FV deficiency fromunrelated Chinese families, and identified three F5gene mutations associated with FV deficiency.The IVS8 )2A>G transition identified in proband1 has been detected in another non-related Chinesepatient [9]. In mammalian, 98.71% splice sitesequences have canonical GT-AG pairs [28]. AllF5 exons and introns are canonical GT-AG junctionsexcept for intron 6 [1], so the mutation mightinterfere with the process of pre-mRNA modification.With the analysis of mRNA extracted from theperipheral blood cells, we found that the mutationabolishes the normal acceptor site of intron 8 andthe cryptic acceptor site 24 bp upstream is conscriptedinstead, causing an eight AA (Asn-Lys-Phe-Asp-Leu-Thr-Leu-Trp) in-frame insertionbetween 404 AA (encoded by exon 8) and 405AA (encoded by exon 9) of FV protein (Fig. 2b).Both exon 8 and exon 9 code part of A2 domain ofFV protein. Normally, A2 domain has 316 AA andis the only region where FV reacts to FIIa [29]. Theinsertion of 8 AA into the A2 domain would affectthe stereotype of FV protein and might thus impairthe stability and secretion of the protein. It is veryinteresting to note that the plasma FV antigen levelof the proband 1 was about 7% of normal whileprocoagulation activity only 1.6%. So there mightbe some FV mutant exists in the circulation,however, the eight AA insertion might impair theprocoagulant functions of FV by disturbing itsinteraction with FIIa. It is of course possible thatsome normally spliced mRNA and hence somenormal FV protein is produced, but sequence of thePCR products from the amplified cDNA of proband1 did not reveal the presence of any normal mRNA(Fig. 2b).The exon 13 codes the B domain of FV and eight of29 mutations occur in this exon. We also discovereda 2-nucleotide deletion (2238-9del AG) in exon 13 in1 2 3 4 52000 bp1000 bp750 bp500 bp250 bp100 bp438 bp 仺244 bp 仺194 bp 仺Fig. 4. Restriction fragment length polymorphism (RFLP) analysisof missense mutation Gly2079Val. The polymerase chain reaction(PCR) products of exon 23 were digested by restriction enzymeRsaI. Lanes 1.4 stands for the normal control, family member I2,I1 and II1 of pedigree 2, respectively. Lane 5 is the DNA marker(DL-2000, Takara).(a)(b)Fig. 3. DNA sequencing for proband 2 (a) and (b) was part of the sequence of exon 13 and exon 23, respectively. Arrows indicated are themutation sites.268 Q.-H. FU et al.Haemophilia (2004), 10, 264.270 2004 Blackwell Publishing Ltdproband 2. The deletion shifts the reading frame andleads to the termination of translation at codon 689.This would predict the synthesis of a truncated FVmolecular, lacking 98% part of the B domain and thecomplete light chain. The putative mutation mightlead to severe FV deficiency by promoting theselective degradation of the corresponding mRNAbecause of the nonsense-mediated mRNA decaypathway and/or by the quality control system ofsecretive proteins to retain and intracellular degradethe truncated FV protein [12,16,30].The FV C2 domain is encoded by exon 23.25 andcomposed of eight antiparallel strands arranged in ajelly-roll structure, whose lower surface exhibitsthree adjacent loops (spikes 1, 2 and 3) [3,31]. Thethree spikes are linked to one another and to threeshorter loops by an intricate hydrogen-bondingnetwork, which extends to residues at the bottomof the b-barrel. They have been proposed to mediatebinding of activated FV to phospholipid membranes[31]. Another putative causal mutation, G6410T,also identified in proband 2, lead to the substitutionof Gly by Val at codon 2079, which is strictlyconserved in species (human, bovine, mouse) and inFVIII C2 domain. The Gly2079 is involved in theformation of the second spike and the second specificphospholipid-binding site. The Gly2079Val mutationmight not only affect the formation of phospholipid-binding site but also change the overalltertiary structure of the domain, causing the mutantprotein destability. Among 10 missense mutationsreported so far with FV deficiency, three mutationswere characterized by expression studies [4,19], andall resulted in a secretion defect with a rapidintracellular degradation. The probands father washeterozygous with this mutation and had levels of FVof approximately 50% (Table 1). Similarly, putativeGly2079Val mutation is likely to cause the secretiondefect with a rapid intracellular degradation of themutant FV protein.In conclusion, three mutations have been identifiedin two unrelated Chinese pedigree with FV deficiency.The splicing mutation (IVS8 )2A>G) abolishesthe acceptor splice site of intron 8 thus causing theactivation of a cryptic acceptor splice site 24 bpupstream of the physiological one. The compoundheterozygous 2238-9del AG and G6410T missencemutations are identified for the first time.AcknowledgementsAuthors thank all family members for their participationin this study, and Dr Jining Lu (MEDPulmonary Center, Boston University) for criticalreading of the manuscript. The financial supportfrom Natural Science Foundation of Shanghai (grantno. 02ZB14043) and part of Clyde Wu foundation ofSIH is gratefully acknowledged.References1 Cripe LD, Moore KD, Kane WH. Structure of the genefor human coagulation factor V. Biochemistry 1992;31: 3777.85.2 Koeleman BPC, Reitsma PH, Bakker E, Bertina RM.Location on the human genetic linkage map of 26genes involved in blood coagulation. Thromb Haemost1997; 77: 873.8.3 Jenny RJ, Pittman DD, Toole JJ et al. Complete cDNAand derived amino acid sequence of human factor V.Proc Natl Acad Sci USA 1987; 84: 4846.50.4 Duga S, Montefusco MC, Asselta R et al. R2074Cmissense mutation in the C2-domain of factor Vcausing moderately severe factor V deficiency:molecular characterization by expression of therecombinant protein. Blood 2003; 101: 173.7.5 Peyvandi E, Mannucci PM. Rare coagulation disorders.Thromb Haemost 1999; 82: 1207.14.6 Kane WH. Factor V. In: Colman RW, Hirsh J, MarderVJ, Clowes AW, George JN, eds. Hemostasis andThrombosis: Basic Principles and Clinical Practice, 4thedn. Philadelphia, PA: Lippincott Williams and Wilkins,2001: 157.69.7 Lak M, Sharifian R, Peyvandi F, Mannucci PM.Symptoms of inherited factor V deficiency in 35 Iranianpatients. Br J Haematol 1998; 103: 1067.9.8 Van Wijk R, Nieuwenhuis K, van de Berg M et al.Five novel mutations in the gene for human bloodcoagulation factor V associated with type I factor Vdeficiency. Blood 2001; 98: 358.67.9 Fu WJ, Hou J, Wang DX, Yu RQ. A novel molecularmechanism of congenital FV deficiency: mutation in theintron acceptor splice site of human blood coagulationFV gene. Zhonghua Yi Xue Za Zhi 2003; 83: 24.6.10 Le W, Yu J, Lu L et al. Arg485Lys polymorphism offactor V increases the risk of coronary artery disease ina Chinese population. Chin Med J (Engl) 2000; 113:963.6.11 Guasch JF, Cannegieter S, Reitsma PH, van乫tVeer-Korthof ET, Bertina RM. Severe coagulationfactor V deficiency caused by a 4 bp deletion in thefactor V gene. Br J Haematol 1998; 101: 32.9.12 Van Wijk R, Montefusco MC, Duga S et al. Coexistenceof a novel homozygous nonsense mutation in exon13 of the factor V gene with the homozygous Leidenmutation in two unrelated patients with severe factor Vdeficiency. Br J Haematol 2001; 114: 871.4.13 Castold E, Simioni P, Kalafatis M et al. Combinationsof 4 mutations (FV R506Q, FV H1299R,FVY1702C,PT 20210G/A) affecting the prothrombinasecomplex in a thrombophilic family. Blood 2000; 96:1443.8.THREE F5 GENE MUTATIONS 269 2004 Blackwell Publishing Ltd Haemophilia (2004), 10, 264.27014 Bossone A, D乫Angelo F, Santacroce R, De Lucia D,Margaglione M. Factor V Arg2074Cys: a novel missensemutation in the C2 domain of factor V. ThrombHaemost 2002; 87: 923.4.15 Ajzner EE, Balogh I, Szabo T, Marosi A, Haramura G,Muszbek L. Severe coagulation factor V deficiencycaused by 2 novel frameshift mutations: 2952delT inexon 13 and 5493insG in exon 16 of factor 5 gene.Blood 2002; 99: 702.5.16 Montefusco MC, Duga S, Asselta R et al. A novel twobase pair deletion in the factor V gene associated withsevere factor V deficiency. Br J Haematol 2000; 111:1240.6.17 Schrijver I, Koerper MA. Jones CD, Zehnder JL.Homozygous factor V splice site mutation associatedwith severe factorVdeficiency. Blood 2002; 99: 3063.5.18 Asselta R, Montefusco MC, Duga S et al. Severe factorV deficiency: exon skipping in the factor V genecausing a partial deletion of the C1 domain.J Thromb Haemost 2003; 1: 1237.44.19 Montefusco MC, Duga S, Asselta R et al. Clinical andmolecular characterization of 6 patients affected bysevere deficiency of coagulation factor V: broadeningof the mutational spectrum of factor V gene andin vitro analysis of the newly identified missensemutations. Blood 2003; 102: 3210.6.20 Fu Q, Wu W, Ding Q et al. Type I coagulation factor Vdeficiency caused by compound heterozygous mutationof F5 gene. Haemophilia 2003; 9: 646.9.21 Castoldi E, Lunghi B, Mingozzi F, Muleo G et al. Amissense mutation (Y1702C) in the coagulation factorV gene is a frequent cause of factor V deficiency in theItalian population. Haematologica 2001; 86: 629.33.22 Xie F, Cheng F, Zhu X. Studies on hereditary deficiencyof coagulation factor V. Zhonghua Xue Ye XueZa Zhi 2001; 22: 453.6.23 Schrijver I, Houissa-Kastally R, Jones CD, Garcia KC,Zehnder JL. Novel factor V C2-domain mutation(R2074H) in two families with factor V deficiency andbleeding. Thromb Haemost 2002; 87: 294.9.24 Asselta R, Tenchini ML, Holme R, Brosstad F, StormorkenH. The discovery of Mary乫s mutation.J Thromb Haemost 2003; 1: 397.8.25 Fu WJ, Hou J, Wang DX, Yu RQ. Identification of anovel mutation of human blood coagulation FV geneassociated with congenital FV deficiency. ZhonghuaXue Ye Xue Za Zhi 2003; 24: 119.21.26 Hou LH, Xie F, Liu XE et al. A novel mutation causescongenital factor V deficiency. Zhonghua Xue Ye XueZa Zhi 2003; 24: 455.9.27 Murray JM, Rand MD, Egan JO, Murphy S, Kim HC,Mann KG. Factor Vnew Brunswick: Ala221-to-Valsubstitution results in reduced cofactor activity. Blood1995; 86: 1820.7.28 Burset M, Seledtsov IA, Solovyev VV. Splice DB:database of canonical and non-canonical mammaliansplice sites. Nucleic acids Res 2001; 29: 255.9.29 Villoutreix BO, Dahlback B. Structural investigation ofthe A domains of human blood coagulation factor V bymolecular modeling. Protein Sci 1998; 7: 1317.25.30 Frischmeyer PA, Dietz HC. Nonsense-mediated mRNAdecay in health and disease. Hum Mol Genet 1999; 8:1893.900.31 Macedo-Ribeiro S, Bode W, Huber R et al. Crystalstructures of the membrane-binding C2 domain ofhuman coagulation factor V. Nature 1999; 402: 434.9.270 Q.-H. FU et al.Haemophilia (2004), 10, 264.270 2004 Blackwell Publishing Ltd
Chronic Myelogenous Leukemia with e13a3 (b2a3)Type of BCR-ABL Transcript Having a DNA Breakpointbetween ABL exons a2 and a3Li-Gen Liu,1,2 Hideo Tanaka,1* Kinro Ito,1 Taiichi Kyo,3 Takuo Ito,1 and Akiro Kimura11Department of Hematology & Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan2Hematological Department, Shanghai Fifth People’s Hospital affiliated with Fudan University, Shanghai, China3Fourth Department of Internal Medicine, Hiroshima Red Cross & Atomic Bomb Survivors Hospital, Hiroshima, JapanWe describe a patient with chronic myelogenous leukemia (CML), in whom the DNAbreakpoint in the BCR-ABL fusion gene was determined to result in a rare e13a3 (b2a3)transcript. The breakpoint in BCR was intron 13, which was 30 bp downstream from exon13, and the breakpoint in ABL was intron 2, and was 46 bp downstream from exon a2.This case conforms to the mechanism of DNA breakage occurring within ABL intron 2,but not at 50 to ABL exon a2. With our review of this case and the literature, it seems thatCML with the BCR-a3 fusion product is associated with a low proportion of circulatingimmature cells, mild or lack of splenomegaly, slow progressiveness, rather resistance toIFN-a, and good response to imatinib mesylate. This is the first report of BCR-a3-typeCML in which the exact DNA breakpoint was identified and located between exons a2 anda3 of the ABL gene. Am. J. Hematol. 74:268–272, 2003. 2003 Wiley-Liss, Inc.Key words: chronic myelogenous leukemia; BCR; ABL; DNA breakpointINTRODUCTIONThe hallmark of chronic myelogenous leukemia(CML) is the chimeric BCR-ABL fusion gene, whichis usually formed as a result of the t(9;22) translocation(Philadelphia chromosome, Ph). Most BCR-ABL fusiontranscripts are e13a2 (b2a2), e14a2 (b3a2), and lesscommonly, e1a2, e19a2 [1]. CML with an atypicalhybrid transcript, in which BCR sequences are fusedto ABL exon a3 rather than exon a2, is very rare, and itscharacteristics are poorly understood [2–11]. Untilnow, the position of the DNA breakpoint in the ABLgene with BCR-a3-type CML has been reported in onlythree cases. ABL exon a2 encodes 58 amino acids, thelast 17 of which formpart of a stretch of 50 amino acidsof Src homology 3 (SH3) domain of the ABL protein.The SH3 is believed to negatively regulate the kinasedomain (SH1). Thus, in theory, deletion of ABL exona2 should result in increased tyrosine kinase activityand, therefore, increased transforming activity. However,recent data suggest that SH3 does not necessarilycontribute to aggressive phenotype. Details of the clinicalphenotype of CML with this type of transcript areunclear because of its very low incidence.Here, we report a case of CML with the e13a3 transcriptand the DNA breakpoints in the BCR and ABLgenes. This case was rather resistant to interferon a(IFN-a) and showed rapid response to imatinib mesylate(imatinib). This is the first report of the exact DNAbreakpoint in the ABL gene determined by nucleotidesequencing in BCR-a3–type CML.Contract grant sponsor: Tsuchiya Foundation*Correspondence to: Hideo Tanaka, M.D., Ph.D., Department ofHematology & Oncology, Research Institute for Radiation Biologyand Medicine, Hiroshima University, Kasumi 1-2-3, Minami-ku,Hiroshima 743-8553, Japan. E-mail: dtanaka@hiroshima-u.ac.jpReceived for publication 21 October 2002; Accepted 15 July 2003Published online inWiley InterScience (www.interscience.wiley.com).DOI: 10.1002/ajh.10429American Journal of Hematology 74:268–272 (2003) 2003 Wiley-Liss, Inc.MATERIALS AND METHODSRNA and DNA were extracted from the cells of abone marrow aspirate from a patient with CML beforetreatment. Informed and written consent was obtainedbefore bone marrow aspiration. Reverse transcriptionpolymerase chain reaction (RT-PCR) for specific fusiontranscript of major BCR-ABL was performed usingsets of primers for sequences within BCR exon 13 andABL exon a3. The primer sequences were as follows(The primer sequences were derived from GenBankaccession numbers U07000 and U07563):BCR1: 50-GCTTCTCCCTGACATCCGTG-30ABL1: 50-GGCCCATGGTACCAGGAGTG-30BCR2: 50-GGAGCTGCAGATGCTGACCAAC-30ABL2: 50-GTTTCTCCAGACTGTTGACTG-30Nested PCRwas performed with the set of primers offirst, BCR1 and ABL1, and second, BCR2 and ABL2.GenomicDNA PCR was also performed with the samesets of primers. The PCR product was cloned into avector using TA cloning kit (Invitrogen Corp, Carlsbad,CA) and followed by sequencing in ABI PRISM 310genetic analyzer (Perkin Elmer, Foster City, CA).CASEA 49-year-old man was diagnosed with chronicphase(CP) CML. He had mild splenomegaly at diagnosisin August of 2000. Peripheral blood analysisresults were as follows: white blood cell (WBC) count,87 109/L; hemoglobin, 11.5 g/dL; platelet count,331 109/L. TheWBC differential count was as follows:0.5% blasts; 3.5% promyelocytes; 20.5% myelocytes;11% metamyelocytes; 16.5% band neutrophils; 29.5%segmented neutrophils; 6.0% basophils; 4.5% eosinophils;1.5% monocytes; 7.0% lymphocytes. The bonemarrow aspirate showed hypercellularity, and the karyotypewas 46, XY, t(9;22)(q34;q11) [20/20]. The positiverate of BCR-ABL fusion in peripheral blood was97.2% as determined by fluorescence in situ hybridization(FISH). Starting in September 2000, the patientwas treated with IFN-a and hydroxyurea. The peripheralWBC count was 6.8 109/L in March 2001, butthe abnormal white blood cell differential (1% promyelocyte,17% myelocyte, and 10% metamyelocyte) indicatedonly a partial hematologic response (PHR).Chromosome analysis in August 2001 showed thatPh-positive clone remained at 95%, indicating poorresponse. FISH analysis at November 8 2001 showedthat the positive rate of the BCR-ABL fusion gene was75.8%. Thus, IFN-a with hydroxyurea did not inducegood response. Then, from December 2001, imatinibwas administered. In April 2002, the karyotype wasnormalized (46, XY [20/20]), and FISH for BCR-ABLdropped to 0%, indicating complete cytogenetic response(CCR).RT-PCR was performed to detect BCR-ABL transcript.The product (Fig. 1A, lane 2) showed one bandthat was smaller than the common e13a2 (lanes 3) ore14a2 (lane 4) transcripts. Sequencing of the PCRproduct revealed the e13a3-type BCR-ABL transcript,which was 174 bp shorter than that of the usual e13a2type, meaning that the ABL exon a2 transcript wascompletely gone.To determine the DNA breakpoints of the BCRABLfusion gene, genomic DNA was amplified withthe preceding sets of primers. The PCR product showedone band (lane 5) at the position of approximately 800bp. Sequencing of the product revealed that the DNAbreakpoint of BCR was in intron 13, 30 bp downstreamfrom exon 13, and that the breakpoint of ABLwas in intron 2 and was 46 bp downstream from exona2 (Fig. 1B).Fig. 1. (A)Agarose gel electrophoresis ofRT-PCRproduct ofBCR-ABL mRNA transcripts and DNA PCR product of BCRABLfusion gene fromgenomic DNA. Lane 1, 1-kb DNA laddermarker. Lane 2, RT-PCR product of our patient’s BCR-ABLtranscript. Lane 3, RT-PCR product of e13a2 (b2a2) as apositive control. Lane 4, RT-PCR product of e14a2 (b3a2) as apositive control. Lane 5, DNA PCR product of BCR-ABL fusiongene from our patient’s genomic DNA. (B) Sequencing resultof thePCRproduct fromgenomicDNAof our patient. Theverticalarrowindicates theDNAbreakpoint and BCR-ABL junctionof the fusion gene. [Color figure can be viewed in the onlineissue, which is available at www.interscience.wiley.com.]Case Report: CML Lacking ABL exon a2 269DISCUSSIONTo our knowledge, this is the first report of CMLwith the BCR-a3 type, in which the exact DNA breakpointof the ABL gene was identified. We found thatthe breakpoint of the ABL gene was between exons a2and a3. This is consistent with earlier reports of twocases of Ph-positive acute lymphocytic leukemia (ALL)[12] and 1 case of CML with the e13a3 transcript [2].However, in these cases, the breakpoints were determinedby Southern blotting. DNA breakpoint determinedby sequencing was reported only in one of thosetwo patients with ALL [12], in which the breakpointwas in ABL intron 2 just like our case. Their breakpointwas 22 bp downstream from exon a2, at a position24 bp upstream from the breakpoint that we identifiedin our case. In contrast, there has been one detailed reportand one short report of CML patients with the BCRa3-type, in which no DNA rearrangement was detectedin ABL intron 2 by Southern blotting, suggesting thatdeletion of exon a2 from the transcript was due to thesplicing mechanism in these cases [3,4]. Thus, at leasttwo different mechanisms seem to be involved in theformation of the BCR-a3 fusion transcript, one isDNAbreakage within ABL intron 2, and the other is DNAbreakage at 50 to ABL exon a2 and thereafter transcriptfor exon a2 is spliced out.The primary factors that determine preferentialbreakage sites in the BCR and ABL genes are presentlyunknown. The low incidence of BCR-a3 fusion productmay be due to the relatively short length of ABL intron2 (0.6 kb) compared with the introns at the beginning ofthe ABL gene (>200 kb). The theoretically predictedfrequency of occurrence of a DNA breakpoint betweenABL exons a2 and a3 is 0.3%of BCR-ABL–rearrangedpatients, assuming that breakpoints in ABL are randomlydistributed. This theoretical frequency may holdtrue, because we found this case during the RT-PCRstudy of 110CML cases. Alu element is thought to be astrong candidate involved in DNA breakage and chimericgene rearrangement, because it is the most abundantrepeat sequence in both of these genes. However,in our case, we could not find Alu element at the breakpointsof both BCR and ABL genes. A lack of Aluelements was observed across the major and minorbreakpoint cluster regions of BCR and across a 25-kbregion with a high frequency of breakage in ABL1 [13].Thus, the incidence and DNA breakage of BCR-a3type may not be determined only by the size of intronsor by Alu element. Thus, mechanisms of DNA breakagein BCR-ABL gene seem more complicated.A review of the literature revealed 13 cases of BCRa3–type CML including our case and 6 cases of BCRa3–type ALL. The clinical data of BCR-a3–type CMLare summarized in Table I. There does not seem to beany particular feature with regard to age or genderdistribution, WBC count, or platelet count. However,there seemed to be an association with low proportionof circulating immature cells and mild or no splenomegaly.All but 3 of the 13 reported cases had a benignprognosis. The longest lived patient has survived for atleast 126 months and was in CP at the last follow-up.Two patients did not require any treatment andremained in the CP for at least 60 and 66 months.This supports the notion that BCR-a3–type CMLpatients may have a benign prognosis. In eight cases,IFN-a has been administered. Among the five cases inwhich hematologic responses were assessed, our caseachieved PHR, two cases achieved complete hematologicresponse (CHR), and two cases achieved remissionwithout precise statements from points of responsecriteria. However, three of these latter four cases lostresponsiveness to IFN-a within 2 years. Among thefour cases in which cytogenetic responses were assessed,two showed partial cytogenetic response (PCR), oneshowed no cytogenetic response (NCR), and our caseshowed MCR, but the Ph chromosome was 95% positive,meaning close to NCR. Imatinib was used only inthree cases including our case. All of them achievedCCR. Thus, imatinib treatment seems very promisingeven for BCR-a3–type CML, in addition to commonBCR-a2–type CML.A CML patient with BCR-a3 transcript is a nicemodel in which to investigate the roles of SH3 domainin BCR-ABL protein in vivo. Bcr-Abl mutants lackingSH3 resulted in reduced tissue invasiveness, and proliferationof leukemic cells slowed down in vivo [14].Another study demonstrated that both SH3 and SH2are required for STAT5 activation by BCR-ABL protein[15]. In murine models of CML, e14a3-type of Bcr-Abl/p210 induced CML-like myeloproliferative diseasesbut showed small delay of the increase in peripheralblood WBC counts and showed small, but consistent,increase in survival compared with the common e14a2-type [16]. These findings seem to agree with the generallyless progressive clinical histories of BCR-a3–type CMLpatients.In conclusion, CML with the BCR-a3 transcriptseems less progressive but rather resistant to IFN-aand sensitive to imatinib as is seen in the commonBCR-a2 type. Therefore, diagnosis of this type ofCML is important, because it may have a differentclinical expression and may affect treatment strategies.ACKNOWLEDGMENTSWe thank Ryoko Yamaguchi, Sachiko Hidani, andNanae Nakaju for their excellent technical support.270 Case Report: Liu et al.TABLE I. Clinical and Laboratory Characteristics of CML Patients with BCR-ABL Transcripts Lacking ABL Exon a2Case No. 1 2 3 4 5 6 7 8 9 10 11 12 13Gender Male Male Female Female Male Male NA Male Female Male Female Male MaleAge (y) 49 41 64 75 69 51 NA 23 68 19 39 27 59Diagnostic phase CP CP CP CP CP CP CP CP CP CP CP CP CPSplenomegaly Mild Yes No No No No NA No No NA No NA YesKaryotype Ph Ph Ph Ph t(4:9:22) Ph Ph Ph Ph Ph Ph/13q Ph PhBCR-ABL transcript e13a3 e1a3/e1a2 e1a3 e1a3 e14a3 e14a3 e13a3 e14a3 e1a3/e13a3 e14a3 e14a3 NA e13a3Hemoglobin level (g/dL) 11.5 NA NA 12.4 13.3 15.1 NA 12.7 12.9 NA NA NA (6.0 mmol/L)WBC count (109/L) 87 189.5 53.2 18.5 18 19.9 NA 95.8 38.4 42 9 NA 254WBC differentiation NA NA NABlast (%) 0.5 2 1 0 0 0 0 0 0 0Promyelocyte (%) 3.5 0 0 0 0 1 1 1 0 10Myelocyte (%) 20.5 0 0 0 0 0 0 14 3 42Metamyelocyte (%) 11 32a 22a 0 0 0 0 11 9 5Band +segmented (%)46 60 72 76 82 78 88 53 46 38Basophils (%) 6 0 0 1 4 1 1 6 24 0Eosinophils (%) 4.5 4 0 0 8 0 0 7 1 2Monocytes (%) 1.5 0 2 6 0 7 7 0 2 0Lymphocytes (%) 7 2 3 17 6 13 3 8 15 0Normoblasts (%) 3Platelet count (109/L) 331 184 156 257 527 566 NA 485 629 381 NA NA 180Initial treatment IFN/Hy IFN/Hy IFN/Hy No No!Hy IFN NA IFN/AraC IFN IFN/Hy No Bu/IFN HyIFN response(hematologic)PR CR NA NA Remission!progressionRemission!progressionCR!progressionIFN response(cytogenetic)MR NR PR PRSecond-line treatment Imatinib Imatinib Imatinib Bu/6MP AutoBMT Hy BMTImatinib response(cytogenetic)CR CR CRDuration of CPbefore imatinib (mos)16 56 62Duration offollow-up* (mos)24 56 62 66 36 126 96 60 20 34 60 24Clinical outcome CP CP CP CP CP CP CP CP BC CP BC APDNA breakpoint ofABL geneIntron 2 Upstreamof a2Upstreamof a2Intron 2Authors, publishedyearPresentcaseAl-Ali,2002Al-Ali,2002Roman,2001Tribelli,2000Tribelli,2000Wilson,2000Amabile,1999Martinelli,1999Pola k,1998Iwata,1994Pa ldi-Haris, 1994van derPlas, 1991*Between the diagnosis and the last follow-up.aAll immature neutrophils included.NA, not applicable; CP, chronic phase; AP, accelerated phase; BC, blastic crisis; Ph, t(9:22)(q34:q11); IFN, interferon; Hy, hydroxyurea; Bu, busulfan; 6-MP, 6-mercaptopurine: auto,autologous; BMT, bone marrow transplantation; AraC, cytarabine; CR, complete response; PR, partial response; MR, minor response; NR, no response (by criteria of Talpaz).Case Report: CML Lacking ABL exon a2 271REFERENCES1. Melo JV. BCR-ABL gene variants. Baillie`re Clin Haematol 1997;10:203–222.2. van der Plas DC, Soekarman D, van Gent AM, Grosveld G,Hagemeijer A. bcr-abl mRNA lacking abl exon a2 detected by polymerasechain reaction in a chronic myelogenous leukemia patient.Leukemia 1991;5:457–461.3. Pa ldi-Haris P, Barta A, Lengyel L, et al. Molecular background ofa new case of chronic myelogenous leukemia with bcr-abl chimeramRNA lacking the A2 exon. Leukemia 1994;8:1791.4. Iwata S, Mizutani S, Nakazawa S, Yata J. Heterogeneity of thebreakpoint in the ABL gene in cases with BCR/ABL transcriptlacking ABL exon a2. Leukemia 1994;8:1696–1702.5. Polak J, Zemanova Z, Michalova K, Klamova H, Cˇ ermak J, Hasˇ-kovec C. A new case of chronic myeloid leukemia (CML) in myeloidblast crisis with an atypical (b3/a3) junction of the BCR/ABLgene. Leukemia 1998;12:250.6. Amabile M, Martinelli G, Terragna C, Montefusco V, Tabilio A,Tura S. An atypical (b3a3) junction of the bcr/abl gene lacking ablexon a2 in a patient with chronic myeloid leukemia. Haematologica1999;84:573–575.7. Martinelli G, AmabileM, Terragna C, et al. Concomitant expressionof the rare E1/A3 and B2/A3 types of BCR/ABL transcript ina chronic myeloid leukemia (CML) patient. Leukemia 1999;13:1463–1464.8. Tiribelli M, Tonso A, Ferrero D, et al. Lack of SH3 domain doesnot imply a more severe clinical course in Ph+ chronic myeloidleukemia patients. Blood 2000;95:4019–4020.9. Wilson GA, Vandenberghe EA, Pollitt RC, et al. Are aberrantBCR-ABL transcripts more common than previously thought?Br J Haematol 2000;111:1109–1111.10. Roman J, Jimenez A, Barrios M, Castillejo JA, Maldonado J,Torres A. E1A3 as a unique, naturally occurring BCR-ABL transcriptin an indolent case of chronic myeloid leukaemia. Br J Haematol2001;114:635–637.11. Al-Ali H-K, Leiblein S, Kovacs I, Henning E, Niederwieser D,Deininger MWN. CML with an e1a3 BCR-ABL fusion: rare, benign,and a potential diagnostic pitfall. Blood 2002;100:1092–1093.12. Soekarman D, van Denderen J, Hoefsloot L, et al. A novel variantof the bcr-abl fusion product in Philadelphia chromosome-positiveacute lymphoblastic leukemia. Leukemia 1990;4:397–403.13. Jeffs AR, Wells E, Morris CM Nonrandom distribution of interspersedrepeat elements in the BCR and ABL1 genes and itsrelation to breakpoint cluster regions. Genes ChromosomesCancer 2001;32:144–154.14. Skorski T, Nieborowska-Skorska M, Wlodarski P, et al. The SH3domain contributes to BCR/ABL-dependent leukemogenesis in vivo:role in adhesion, invasion, and homing. Blood 1998;91:406–418.15. Nieborowska-Skorska M, Wasik MA, Slupianek A, et al. Signaltransducer and activator of transcription (STAT) 5 activation byBCR/ABL is dependent on intact Src homology (SH) 3 and SH2domains of BCR/ABL and is required for leukemogenesis. J ExpMed 1999;189:1229–1242.16. Gross AW, Zhang X, Ren R. Bcr-Abl with an SH3 deletion retainsthe ability to induce a myeloproliferative disease in mice, yet c-Ablactivated by an SH3 deletion induces only lymphoid malignancy.Mol Cell Biol 1999;19:6918–6928.272 Case Report: Liu et al.
Original PaperActa Haematol 2005;113:113–123DOI: 10.1159/000083449Absence of Gene Mutation in TRAIL Receptor 1(TRAIL-R1) and TRAIL Receptor 2 (TRAIL-R2) inChronic Myelogenous Leukemia andMyelodysplastic Syndrome, and Analysis ofmRNA Expressions of TRAIL and TRAIL-RelatedGenes in Chronic Myelogenous LeukemiaLi-Gen Liua, cHideo TanakaaKinro ItoaTakuo ItoaTanvira A. SultanaaTaiichi KyobAkiro KimuraaaDepartment of Hematology and Oncology, Research Institute for Radiation Biology and Medicine,Hiroshima University and bFourth Department of Internal Medicine, Hiroshima Red Cross Hospital andAtomic Bomb Survivors Hospital, Hiroshima, Japan; cHematological Department, Shanghai Fifth People’s Hospitalaffiliated with Fudan University, Shanghai, ChinaReceived: January 5, 2004Accepted after revision: May 26, 2004Hideo Tanaka, MD, PhDDepartment of Hematology and OncologyResearch Institute for Radiation Biology and Medicine, Hiroshima University1-2-3 Kasumi, Minami-ku, Hiroshima JP-734-8553 (Japan)Tel. +81 82 257 5861, Fax +81 82 256 7108, E-Mail dtanaka@hiroshima-u.ac.jpABCFax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com 2005 S. Karger AG, Basel0001–5792/05/1132–0113$22.00/0Accessible online at:www.karger.com/ahaKey WordsChronic myelogenous leukemia W Interferon W MutationWMyelodysplastic syndromeW TRAILAbstractTumor necrosis factor-related apoptosis-inducing ligand(TRAIL) is an interferon (IFN)-induced molecule withapoptotic activity. We examined gene mutations in thedeath domains of TRAIL receptor 1 (TRAIL-R1) and TRAILreceptor 2 (TRAIL-R2) , and in the TRAIL gene promoter in46 chronic myelogenous leukemia (CML) patients. In 23of the 46 patients, all the coding regions of TRAIL-R2were also examined. However, no mutation or loss ofheterozygosity was found. Furthermore, no mutation inthe death domains of TRAIL-R1 and TRAIL-R2 genes,which causes amino acid change, was found in 18 my-elodysplastic syndrome (MDS) patients. Ribonucleaseprotection assay (RPA) and real-time quantitative poly-merase chain reaction using polymorphonuclear neutro-phils of five new CML patients showed that the TRAILmRNA expression was very low before in vitro IFN-·stimulation and markedly upregulated after IFN-· stimu-lation. FAS mRNA was also upregulated with IFN- · stim-ulation but the fold induction was far lower than that ofTRAIL mRNA. In addition, RPA revealed that the ratio of(TRAIL-R1 plus TRAIL-R2) to TRAIL-R3 was also in-creased after IFN- · stimulation. Taken together, genemutations of TRAIL-R1, TRAIL-R2 are infrequent in pa-tients with CML and MDS. And so is the TRAIL promoterfor CML. These mutations seem unrelated to tumorigen-esis, disease progression, and response to IFN- · therapyin CML. A markedly high induction of TRAIL mRNA byIFN- · may have some relevance to IFN-· action in CMLpatients.Copyright 2005 S. Karger AG, BaselIntroductionTumor necrosis factor (TNF)-related apoptosis-induc-ing ligand (TRAIL/Apo2L) is a proapoptotic member ofthe TNF superfamily involved in cell homeostasis, im-mune tumor surveillance, and also potential cancer thera-114 Acta Haematol 2005;113:113–123 Liu/Tanaka/Ito/Ito/Sultana/Kyo/Kimurapy [1–5]. Interferon alpha (IFN- ·) enhances the TRAILexpression in anti-CD3-stimulated peripheral blood T cells[6], and TRAIL may participate in IFN- ·-induced anti-proliferative effects on malignant tumors. TRAIL is consti-tutively expressed in a variety of tissues and cell types, andhas relatively specific cytotoxic effects on transformed cellsand a variety of tumor cells but not on many normal cells.More positively, it is thought to be nontoxic when adminis-tered systemically in animals [7]. There are at least fivetypes of TRAIL receptors, including TRAIL receptor 1(TRAIL-R1/DR4), TRAIL receptor 2 (TRAIL-R2/DR5),TRAIL receptor 3 (TRAIL-R3/DcR1), TRAIL receptor 4(TRAIL-R4/DcR2), and the soluble receptor osteoprote-gerin [4]. TRAIL-R3 and TRAIL-R4 are known to act asdecoy receptors that interfere with agonistic receptors,TRAIL-R1 and TRAIL-R2. The differential expression ofthese receptors was initially considered as the source ofselective TRAIL-induced apoptosis [8].In TRAIL receptor expressions in hematological ma-lignancies, a previous report demonstrated that the ratioof death receptors TRAIL-R1 and TRAIL-R2 (R1 + R2)to decoy receptors TRAIL-R3 and TRAIL-R4 (R3 + R4)in bone marrow (BM) cells in myelodysplastic syndrome(MDS) patients is significantly higher than in normalindividuals, which may explain the higher susceptibilityto TRAIL-induced death observed in BM cells in MDSpatients [9]. In Daudi B lymphoma cells, in addition tothe TRAIL expression, expressions of TRAIL-R1 andTRAIL-R2 were also upregulated by IFN- · stimulation,and were considered to be involved in the mechanism ofIFN-·-induced apoptosis [10]. However, several recentreports found no correlation between the expression levelsof these TRAIL receptors and the sensitivity of TRAIL-induced apoptosis in acute leukemia cells [11], and mela-noma cells [12]. Thus, the significance of the differentialexpression of the four TRAIL receptors is still unclear, atleast in hematological malignancies.Gene mutations of TRAIL-R1 and TRAIL-R2 havebeen reported in some solid tumors such as head and neckcancer [13, 14], colorectal carcinoma [15], lung cancers[13, 16, 17], breast cancers [18, 19], and hepatocellularcarcinoma [20], although the frequency of the mutationsvaries widely. For example, the mutation rate of theTRAIL-R2 gene in non-small cell lung cancer is as high as10.6% [17], whereas some other malignancies showed acomplete absence of these mutations, indicating that thefrequency of the mutation seems to depend mainly on thetumor cell type. Among these mutations, the most com-mon are missense mutations either in the death domainor the ligand-binding domain, indicating that the muta-tions may be involved in the development of tumors. Inhematological malignancies, however, mutation analysisis very limited, and up to now, there has been only onereport about non-Hodgkin’s lymphoma [21]. IFNs areconsidered apoptosis-inducing cytokines [22] and theIFN-stimulated responsive element (ISRE) has alreadybeen identified in the TRAIL gene promoter [23]. It seemspossible that mutations in the TRAIL gene promoter maymodify the expression of TRAIL, which may further mod-ulate the clinical response to IFN-· in chronic myeloge-nous leukemia (CML) patients.In this paper, we examined gene mutations of TRAIL-R1, TRAIL-R2 and the TRAIL promoter in CML pa-tients, and gene mutations of TRAIL-R1 and TRAIL-R2in MDS patients, by polymerase chain reaction single-strand conformation polymorphism (PCR-SSCP). As faras we know, this is the first report of gene mutations ofTRAIL-R1 and TRAIL-R2 and the TRAIL promoter inCML or MDS patients. We also examined stable andIFN-·-stimulated mRNA expressions of TRAIL, FAS,TRAIL receptors, and other TRAIL-related genes by ribo-nuclease protection assay (RPA), and by real-time quanti-tative polymerase chain reaction (RQ-PCR) for expres-sions of TRAIL and FAS.Patients, Materials, and MethodsPatientsForty-six adult CML patients were enrolled in the PCR-SSCPanalysis for gene mutations of TRAIL-R1, TRAIL-R2, and theTRAIL promoter. Thirty-eight patients were in the chronic phase(CP) and eight in blastic crisis (BC). The mean age was 51.6. Thirty-three were males and 13 were females. For RPA, five newly diag-nosed untreated patients in CP were studied. They were subsequent-ly treated with IFN- ·, and 2 achieved a minor cytogenetic response(MCR), 1 a complete cytogenetic response (CCR), and 2 no cytoge-netic response (NCR) after 12 months of follow-up. The responsecriteria of Talpaz’s group [24] were adopted to evaluate the effective-ness of the IFN-· therapy. In addition to CML, 18 MDS patients [5RA, 4 RAEB-I, 2 RAEB-II, and 7 acute myelocytic leukemia (AML)derived from MDS, 14 males and 4 females] were analyzed for muta-tion analysis. Written informed consent was obtained before BMaspiration in each patient. We used polymorphonuclear neutrophils(PMNs) instead of mononuclear cells as the source of DNA and RNAthroughout this study, because it is known that, in most CMLpatients, mature T cells and NK cells do not have the Philadelphia(Ph) chromosome, and mature B cells are a mixture of Ph-positiveand Ph-negative cells, whereas over 90% of neutrophils are Ph-posi-tive [25]. CD34-positive cells were thought to be one of the cellsource candidates, but we could not obtain CD34-positive cells fromthese patients, and if it had been possible, the number of CD34-posi-tive cells was very small and the amount of RNA would not havebeen sufficient to perform the RPA.Gene Mutation of TRAIL Receptors inCML and MDSActa Haematol 2005;113:113–123 115Table 1. Primers used for PCR amplification of individual exons of TRAIL-R1 and TRAIL-R2 , and the TRAIL promoterName Exon Forward primer sequence (5)-3 )’) Reverse primer sequence (5 )-3 )) Size, bpTRAIL-R1E9a 9 ACCCCACTGAGACTCTGATGCTGT CGTCCAGTTTTGTTGACCCATTTC 187E9b 9 AATGAGATCGATGTGGTCAGAGC ACTTTCCAGAGTCCACCAAGAGG 187TRAIL-R2aE1a 1 TATAAGAGCGTTCCCTACCGC AGCACAAGGGTCTTGGGGAC 141E1b 1 ACCCAGGGAGGCGCGGGGAG ccagggaccgcggcgggact 105E2 2 catgcccatccctgctttgtccc caaggattagagacccatcttgaac 163E3 3 tcactggcttttctctcccctcc taggcatggggtccatatgttctg 167E4 4 ccaaaaccttatgctctgtcgtca gggagcccccggctcctgtc 161E5a 5 cctctccctccgtgtgtgtacc TGTGCTTTGTACCTGATTCTTTGTGG 111E5b 5 GATTGTACACCCTGGAGTGACATCG CCACAGTAAAGACTTGCAAACAAACAC 207E5c 5 TCCTGCCTCTCCCTGTTCTC cccccgccctcagccagcac 161E6 6 cttcctgctaaggagactgcttc aagagttcctgaacccctgagcc 112E7a 7 tcggctttttgccttcccaatgtc CATTTCCTGCTCAGGGACCTGGG 120E7b 7 TCAATGAGATCGTGAGTATCTTGCAG gaaacaaaatgatctgtcccccact 150E8 8 ccaggacattccccactatgtg ttcctgctgcatctccaggag 133E9a 9 ctgtctctgtggctttctccac CCCGCTGCCTCAGCTTTAGC 146E9b 9 CCTCATGGACAATGAGATAAAGG CAATCTTCTGCTTGGCAAGTCTCTCTC 170E9c 9 CACCCTGCTGGATGCCTTGG TGACTTCCTGAAGAGAATCACAC 147TRAIL promoterbgagtgcagataaggggtgca gcccttccttctctattctt 230Upper case letters correspond to exons and lower case letters correspond to introns for TRAIL-R1 and TRAIL-R2 .a Most primers for TRAIL-R2 are based on a report by Arai et al. [15] except for E1b and E4.b Primers for the TRAIL promoter were designed to encompass the ISRE in the human TRAIL gene promoter based on the report by Gongand Almasan [23].PCR-SSCP and DNA SequencingIn all 46 CML cases, mutations of the TRAIL gene promoter,exon 9 of the TRAIL-R1 and exon 9 of the TRAIL-R2 , were exam-ined as the death domain falls into exon 9 for both genes, and mostmutations have been found in this region in other malignancies.Additionally, the remaining eight exons of the TRAIL-R2 gene werealso examined in 23 of the 46 CML patients, of whom 1 patient wasin BC. Only exon 9 was examined for TRAIL-R1 because the organi-zation of other exons of this gene was unclear. PCR-SSCP was per-formed for exon 9 in both the TRAIL-R1 and TRAIL-R2 genes in the18 MDS patients. Genomic DNA was extracted from PMNs of BMaspirate according to the routine method [26]. PCR amplificationwas performed using a thermal cycler (Perkin Elmer, Branchburg,N.J., USA). The PCR condition was one cycle at 94° C for 2 min,followed by 35 cycles at 94 ° C for 2 min, 56 ° C for 1 min and 72° Cfor 2 min. The extension time at 72 ° C was extended to 10 min in thelast cycle. The primer sequences are listed in table 1. Nonradioiso-topic SSCP was essentially performed according to the methoddescribed previously [27]. Electrophoresis was carried out usingGeneGel Excel 12.5/24 kit according to the manufacturer’s proce-dure (Pharmacia Biotech, Uppsala, Sweden). Gel was stained using aDNA silver staining kit (Pharmacia Biotech). When PCR-SSCP anal-ysis showed additional bands on the gels, the PCR product wascloned into a vector using a TOPO T/A cloning kit (Invitrogen, Carls-bad, Calif., USA) and DNA sequencing was carried out using an ABI371 Automated Sequencer with the ABI PRISMTM Dye TerminatorCycle Sequencing Ready Reaction Kit (Perkin Elmer, Branchburg,N.J., USA).RNA ExtractionBM PMNs from the 5 newly diagnosed patients with CML-CPwere cultured in RPMI 1640 medium plus 10% fetal calf serum(FCS) with or without 1 ! 103 U/ml of IFN-· (HLBI; kindly suppliedby Sumitomo Pharmaceutical Co, Tokyo, Japan) for 12 h, and totalRNA was extracted using Trizol reagent (Invitrogen). A BM samplefrom a normal subject was also included in this assay with informedconsent.Ribonuclease Protection Assay RPA was performed using RiboQuantTM with an hAPO-3d mul-tiprobe template set (PharMingen, San Diego, Calif., USA) accordingto the manufacturer’s procedure. The probe set contains 14 probes,i.e., caspase 8, FAS ligand (FASL), FAS, TRAIL, TRAIL-R1,TRAIL-R2, TRAIL-R3, TRAIL-R4, DR3, TNF receptor 1(TNFR1), TRADD, RIP, and two housekeeping genes, glyceralde-hyde-3-phosphate dehydrogenase (GAPDH) and L32, to confirmequal RNA loading. The hAPO-3d template set was used for the T7RNA polymerase-directed synthesis of [32P]-UTP-labeled antisenseRNA probes. The probe (4 ! 105 cpm) was hybridized with eachtotal RNA (10 g) sample overnight at 56 ° C. RNA samples were116 Acta Haematol 2005;113:113–123 Liu/Tanaka/Ito/Ito/Sultana/Kyo/KimuraFig. 1. EMSA. a Nucleotide sequence of the5)-flanking region of the human TRAILgene, reported by Gong and Almasan [23].Two horizontal arrows show the positions ofprimers used for PCR-SSCP in this study.The region indicated by § is the position ofthe probe used for EMSA. The position ofthe ISRE sequence is indicated. Nucleotidenumbering is referenced to the +1 nucleotideC identified as the major transcription startsite (bent arrow) [23]. b Nuclear extracts ofDaudi lymphoblastoid cells were preparedwith (+; lanes 3–5) or without (–; lane 2)100 U/ml of IFN-· treatment for 30 min.Oligonucleotides of ISRE either for ISRE wt(lane 4) or for ISRE mt (lane 5) were addedas a competitor at 100-molar excess to thebinding buffer. Lane 1 shows the probe with-out the nuclear extract.then digested with RNase A and T1, purified, and resolved on 6%denaturing polyacrylamide gels. After exposure to an X-ray film, theintensity of each band was determined by quantification software,Quantity One version 4.2 (Bio-Rad Laboratories, Hercules, Calif.,USA) according to the manufacturer’s procedure, and divided by itsown GAPDH intensity to get the relative intensity value. For statisti-cal analysis, Student’s paired t test was performed to compare therelative intensity between values with and without IFN- · stimula-tion.Real-Time Quantitative Polymerase Chain Reaction cDNA synthesis was performed, using a First-Strand cDNA Syn-thesis Kit (Amersham Biosciences), with 1 g of the total RNA ofeach sample as a template. RQ-PCR was performed using an ABIPRISM 7700 Sequence Detection System (Perkin Elmer Biosystems,Foster City, Calif., USA). PCR was performed using TaqMan Uni-versal PCR Master Mix (Applied Biosystems, Branchburg, N.J.,USA) in a 25-l reaction containing 1 l of cDNA (2 ng/ l), 900 n Mforward and reverse primers and 250 n M TaqMan probe. TheTRAIL primers and probe were: TRAIL forward (5 )-AATCAT-CAAGGAGTGGGCATTC-3)), TRAIL reverse (5 )-ATGACCAG-TTCACCATTCCTCAA-3 )), and TRAIL probe (5 )-FAM-TCCTG-AGCAACTTGCA-MGB-3 )). The FAS primers and probe were: FASforward (5 )-TGACATCAACTCCAAGGGATTG-3 )), FAS reverse(5 )-ACCTGGAGGACAGGGCTTATG-3 )), and FAS probe (5 )-FAM-ACTCAGAACTTGGAAGGC-MGB-3)). The thermal cyclingconditions were: 2 min at 50 ° C, 10 min at 95 ° C, 50 cycles at 95 ° Cfor 15 s and 60° C for 1 min. The estimated size of the amplifiedfragment (a single band at electrophoresis on a 3% ethidium bro-mide-stained gel) matched the calculated size (64 bp for TRAIL and115 bp for FAS). To construct standard curves, cDNA was synthe-sized from total RNA of Daudi cells stimulated with 100 U/ml ofGene Mutation of TRAIL Receptors inCML and MDSActa Haematol 2005;113:113–123 117IFN-· for 12 h. The cDNA was serially diluted (1:5) in water to makefive points. GAPDH was used as an internal reference. All sampleswere tested in triplicate. Each GAPDH, TRAIL, and FAS expressionin the samples was related to each corresponding standard curve.Each expression level of the TRAIL and FAS transcript was normal-ized by the level of the GAPDH transcript.Nuclear Protein Extraction from Daudi CellsDaudi lymphoblastoid cells were cultured in RPMI 1640 me-dium with 10% FCS, then nuclear extracts were prepared. After100 U/ml of IFN- · treatment for 30 min, cells were washed with PBSby centrifugation, and swelled and homogenized with Dounce ho-mogenizer in buffer A [20 mM HEPES (pH 7.0), 10 mM KCl, 1 mMMgCl2, 10% glycerol, 0.1% NP-40, 0.5 m M dithiothreitol (DTT), and0.25 mM phenylmethylsulfonyl fluoride (PMSF)]. Nuclei were col-lected by centrifugation, washed in buffer A without NP-40, andresuspended at 1 ! 108 nuclei per milliliter in buffer B (buffer A plus0.4 M NaCl), and stirred gently for 30 min at 4° C. Insoluble materialwas removed by centrifugation, then the clarified supernatant wasdialyzed for about 12 h at 4 ° C against 20 mM HEPES (pH 7.9),40 mM KCl, 1 mM MgCl2, 0.1 mM EGTA, 10% glycerol, 0.5 mMDTT, and 0.25 mM PMSF. The dialysate was clarified by centrifuga-tion at 10,000 g at 4 ° C.Electrophoretic Mobility Shift AssayThe probe used was a 23-bp-long synthesized oligonucleotidecontaining ISRE in the TRAIL gene promoter (see fig. 1) [23]. Theoligonucleotide for each strand was end-labeled (1 ! 108 cpm/g)with [-32P]ATP and T4 polynucleotide kinase, then annealed to-gether to make a double-stranded probe. For the electrophoreticmobility shift assay (EMSA), nuclear protein extracts (3 g of pro-tein) from Daudi cells were incubated with 2 g of poly(dI-dC)W (dI-dC) (Pharmacia) and F1 ng of end-labeled probe DNA ( F1 !104cpm) in binding buffer [20 mM HEPES (pH 7.9), 40 mM KCl,1 m M MgCl2, 0.1 m M EGTA, 10% glycerol, 0.5 mM DTT] for20 min at room temperature. Samples were run on a 4% polyacryl-amide gel in 0.5 ! TBE. Gels were dried and visualized by autora-diography. Double-stranded oligonucleotides for wild-type ISRE(ISRE wt) in the Pkr gene promoter [28, 29] and mutant-type ISRE(ISRE mt) were synthesized and used for competition assays. Thenucleotide sequences were as follows (ISRE are underlined, mutatednucleotides are in bold type):ISRE wt: 5 )-CCGCCGGGAAAACGAAACA GAAGA-3)3)-GGCGGCCCTTTTGCTTTGT CTTCT-5)ISRE mt: 5 )-CCGCCGGGAAAACCACACAGAAGA-3)3)-GGCGGCCCTTTTGGTGTGTCTTCT-5)The unlabeled double-stranded competitor was added to thebinding buffer at 100-molar excess to the labeled probe.ResultsElectrophoretic Mobility Shift Assay We first tried to confirm that the region amplified byour primers for PCR-SSCP contained authentic ISRE forthe TRAIL gene promoter. Figure 1a shows the nucleotidesequence of the 5 )-flanking region of the human TRAILgene including the ISRE motif, which was previouslyshown to have IFN-inducible promoter activity [23]. Weperformed EMSA using a probe containing this ISREregion and nuclear extract from Daudi cells, and con-firmed that IFN-·-induced protein(s) actually binds to theISRE in this region of the TRAIL gene (fig. 1b, lane 3),and that the competition assay showed that the shiftedband was blocked with the addition of ISRE wt oligonu-cleotide (lane 4), but not with the addition of mutated-ISRE oligonucleotide (lane 5).Mutation SearchDespite extensive searches using PCR-SSCP analysis,we found no gene mutations of TRAIL-R1 , TRAIL-R2and the TRAIL promoter in CML patients. The disap-pearance of any band implicating the possibility of loss ofheterozygosity was also not identified. Only 1 patient (IDNo. 829) in CP showed a shifted band in exon 5 of theTRAIL-R2 gene (fig. 2a, lane 1). Sequence analysis of thisband revealed a C to T nucleotide change at nucleotideposition 572 (note: the first nucleotide ‘A’ of the startcodon is taken as position 1) (fig. 2b), resulting in anAla191Val amino acid change. In order to determinewhether this change was a mutation or polymorphism,PCR-SSCP analysis was performed using DNA from thebuccal cells of this patient (fig. 2a, lane 2). The band pat-tern was identical to that using DNA from BM (fig. 2a,lane 1), indicating that this nucleotide change was a poly-morphism. This polymorphism has already been reportedin other studies [15, 20]. In addition to CML, in 18 MDSpatients, we performed a PCR-SSCP analysis of exon 9for both TRAIL-R1 and TRAIL-R2 genes. Single muta-tion was found in exon 9 in the TRAIL-R1 gene of 1patient (AML derived from MDS) with C to T change atnucleotide position 1,209, which, however, resulted in noamino acid change (403 Gly to Gly). Thus, overall, wefound no significant mutation in CML and MDS pa-tients.Expressions of TRAIL-Related Genes Examined byRPARPA disclosed a lot of information for both stableexpressions and IFN-·-stimulated expressions in 14 genes(fig. 3). TRAIL mRNA expressions were strongly inducedafter in vitro IFN-· stimulation in all 5 CML patients,while they were only barely expressed before the stimula-tion, indicating that the fold inductions were very high.Fold induction of the TRAIL mRNA by IFN-· appearedvariable among the 5 patients. Patient 825 (lanes 5 and 6),who showed a rather weaker induction of TRAIL mRNA118 Acta Haematol 2005;113:113–123 Liu/Tanaka/Ito/Ito/Sultana/Kyo/KimuraFig. 2. A polymorphism in exon 5 of the TRAIL-R2 gene. a PCR-SSCP analysis using BM PMN cells (lane 1), buccal cells (lane 2) fromone CML-CP patient (ID No. 829) and BM PMN cells from a normalcontrol (lane 3) is shown. b Nucleotide sequence of the shifted bandobserved in the BM sample (lane 1 in a) from the patient. C to Tchange in the sense strand was observed at nucleotide position 572.The bottom section shows the raw sequencing result of the antisensestrand.Fig. 3. RPA. BM PMNs from 5 newly diagnosed patients with CML-CP (lanes 3–12) and 1 normal subject (lanes 1 and 2) were stimulatedwith (+) or without (–) 1 ! 103 U/ml of IFN-· for 12 h. The positionsof the 14 genes are indicated on the right. Clinical responses of thepatients are shown at the bottom. Fig. 4. Quantification of IFN- · inducibility of each gene in the RPA.Gene expressions and the ratio of (TRAIL-R1 plus TRAIL-R2) toTRAIL-R3 [(R1 + R2)/R3] were compared between values with andwithout stimulation of IFN-· in 5 CML patients studied by RPA.The vertical axis shows the relative intensity to GAPDH. The meanvalue B SD for each value is shown. The mRNA expressions ofTRAIL (lane 1), FAS (lane 6), RIP (lane 10), and the ratio of (R1 +R2)/R3 (lane 5) were enhanced by IFN-· stimulation.Gene Mutation of TRAIL Receptors inCML and MDSActa Haematol 2005;113:113–123 119Table 2. In vitro effects of IFN- · on the gene expression in each CML patient, grouped on the basis of the clinical response determined by theRPATRAILIFN– IFN+ ratioFASIFN– IFN+ ratio(R1 + R2)/R3IFN– IFN+ ratioRIPIFN– IFN+ ratioNormal 0.36 0.81 2.30 0.43 0.44 1.02 1.41 1.42 1.01 0.47 0.59 1.26No. 869 (NCR) 0.56 0.97 1.751.84 B0.090.79 0.87 1.111.11 B0.011.58 1.72 1.091.11 B0.020.71 0.99 1.381.31 B0.08No. 825 (NCR) 0.23 0.45 1.930.25 0.27 1.101.88 2.14 1.140.31 0.38 1.23No. 863 (MCR) 0.35 0.81 2.29 0.38 0.40 1.08 1.33 1.40 1.06 0.48 0.58 1.20No. 868 (MCR) 0.27 0.76 2.85 2.55B0.20 0.36 0.38 1.05 1.11 B0.07 1.66 1.70 1.02 1.04 B0.02 0.33 0.47 1.41 1.33 B0.09No. 892 (CCR) 0.24 0.61 2.500.31 0.37 1.211.27 1.34 1.050.36 0.49 1.37Each figure shows the quantified value of RPA. The mean B SD of the ratio (IFN+/IFN–) for each group is shown.compared to the other patients, was found to be NCRafter subsequent IFN-· therapy. Regarding the fourTRAIL receptors, TRAIL-R3 was strongly, and TRAIL-R1 and TRAIL-R2 were rather weakly expressed withoutIFN-· stimulation, in all 5 CML patients. In addition tothese genes, some other genes were constitutively ex-pressed, and some appeared to be modified with IFN- ·stimulation.The quantification of each band was then performed inthese patients. FASL and TRAIL-R4 were excluded in thequantification analysis, since clear bands were not seenfor these genes in RPA, indicating that these genes wereneither expressed constitutively nor transcriptionally in-duced by IFN-· stimulation. As shown in figure 4, tran-scripts of three genes, TRAIL, FAS, and RIP, were signifi-cantly induced after in vitro stimulation of IFN-· (IFN+)compared to those without stimulation (IFN–) (p = 0.001,p = 0.03, p = 0.01, respectively). The expressions weremore or less identical after IFN-· stimulation for the oth-er genes, TRAIL-R1, TRAIL-R2, TRAIL-R3, TNFR1,DR3, TRADD, and caspase 8. The normal subject alsoshowed similar results to the CML patients, i.e., expres-sions of the TRAIL, FAS, and RIP genes were alsoinduced after IFN- · stimulation (quantification data forthe normal subject is not shown in fig. 4), implicating thatpatterns of IFN-·-induced expressions for these genes didnot differ between CML patients and the normal subject.We tried to calculate the ratio of TRAIL-R1 plus TRAIL-R2 to TRAIL-R3 plus TRAIL-R4, (R1 + R2)/(R3 + R4).As mentioned above, there was no clear band of TRAIL-R4 in RPA, so we calculated the ratio of TRAIL-R1 plusTARIL-R2 to TRAIL-R3 [(R1 + R2)/R3], and comparedthe values between samples with and without IFN-· stim-ulation. Although this result may be cautiously inter-preted because the band intensities for both TRAIL-R1and TRAIL-R2 were rather weak in the RPA, statisticalanalysis showed that the ratio of (R1 + R2)/R3 was slight-ly increased after IFN- · stimulation (p = 0.04; lane 5).Table 2 shows the raw quantification results with orwithout IFN- · stimulation for the three genes, TRAIL,FAS, and RIP, and for the (R1 + R2)/R3 ratio. The 5patients were grouped according to the clinical response,i.e., 2 NCR patients and 3 MCR or CCR patients. Themean fold induction after IFN-· stimulation of theTRAIL mRNA expression was 2.55 in the MCR or CCRpatients, and 1.84 in NCR patients. Thus, good-responderpatients tended to show a higher induction rate of TRAILmRNA with in vitro IFN-· stimulation than nonrespon-der patients, although this tendency should be cautiouslyinterpreted because the number of patients was small.This tendency was not found in the gene expressions ofFAS, RIP, and in the (R1 + R2)/R3 ratio.Expressions of TRAIL and FAS mRNAs Examined byRQ-PCRFrom the results of RPA, in which TRAIL mRNA wasinduced by IFN- · more strongly than FAS mRNA, wetried to confirm the expressions of these two genes usingthe RQ-PCR method (fig. 5). TRAIL mRNA expressionswere very low before in vitro IFN- · stimulation, and werethen strongly induced with IFN- · in all 5 patients and alsoin 1 normal individual (fig. 5a). FAS mRNA expressionswere already detectable before IFN-· stimulation, andwere also inducible by IFN-· (fig. 5b). Fold inductions ofthe TRAIL mRNA expressions by IFN- · stimulationwere as high as 8- to 34-fold (fig. 5c). Good-responder(MCR and CCR) patients (No. 863, 868 and 892) tendedto show higher TRAIL mRNA induction after IFN- ·stimulation than nonresponder (NCR) patients (No. 869and 825). These results were similar to the RPA results. In120 Acta Haematol 2005;113:113–123 Liu/Tanaka/Ito/Ito/Sultana/Kyo/KimuraFig. 5. Expressions of TRAIL and FASmRNAs examined by RQ-PCR. Total RNAextracted from PMN cultured with (+) orwithout (–) 100 U/ml of IFN-· stimulationfor 12 h was reverse-transcribed and sub-jected to RQ-PCR for TRAIL and FASmRNA expressions in the 5 CML patients.a, b Vertical axes show the values of mRNAexpression corrected by the expression ofGAPDH. c , d Vertical axes show the foldinduction after IFN-· stimulation. Themean values B SD of triplicate results areshown. Clinical responses to IFN- · therapieswere: NCR for No. 869 and 825, MCR forNo. 863 and 868, and CCR for No. 892.a TRAIL mRNA expression. b FAS mRNAexpression. c Fold induction of TRAILmRNA expression after IFN-· stimulation.d Fold induction of FAS mRNA expressionafter IFN-· stimulation.contrast, fold inductions of FAS mRNA expressions withIFN-· stimulation were between 1- and 2-fold, and nocorrelation with the clinical response was evident(fig. 5d).DiscussionUntil now, mutation analyses of TRAIL-R1 or TRAIL-R2 genes have been reported mainly in solid tumors, suchas head and neck cancer, colorectal cancer, lung cancer,breast caner, gastric cancer, and hepatocellular carcinoma[13–20]. Gene mutations of these genes in hematologicalmalignancies are almost unknown. Only one study of non-Hodgkin’s lymphoma showed 1.7% mutation (2/117) inexon 9 of the TRAIL-R1 gene, and 5.1% mutation (6/117)in exon 9 of the TRAIL-R2 gene [21]. In our study, weshowed that mutations in the death domains of bothTRAIL-R1 and TRAIL-R2 genes were absent in 46 CMLpatients, including 8 BC patients, and that mutations inall other exons in the TRAIL-R2 gene were also absent in23 CML patients. One patient with CML-CP had a poly-morphism in exon 5 of the TRAIL-R2 gene. The positionof this amino acid change was not located within thetransmembrane or death domain region, and its biologicalsignificance is uncertain. It also seems unlikely that themutations of TRAIL-R1 and TRAIL-R2 genes are in-volved in disease progression since 8 patients with CML-BC also showed no mutation or loss of heterozygosity.These results are reasonable, as all four membrane-typeTRAIL receptors, TRAIL-R1, TRAIL-R2, TRAIL-R3,and TRAIL-R4 genes, are clustered on the 8p21–p22region of chromosome 8 [4]. Chromosomal aberration asa deletion in this region in CML is relatively uncommon,and abnormalities of chromosome 8 are more frequentlyfound as trisomy 8 upon disease progression to BC [30,31]. In addition, we found no mutation in the TRAIL genepromoter in 46 CML patients. Taken together, we consid-er that gene mutations of TRAIL-R1 , TRAIL-R2 and theTRAIL promoter are absent or at least infrequent inCML, and seem to be unrelated to tumorigenesis, diseaseprogression, and response to IFN-· therapy in CML.Mutations in other hematological malignancies are also ofinterest. In this study, 18 MDS patients were also studied,and one mutation was found in exon 9 of the TRAIL-R1gene, which, however, caused no amino acid change.Gene Mutation of TRAIL Receptors inCML and MDSActa Haematol 2005;113:113–123 121Thus, mutations of these genes seem uncommon in MDSpatients. Further studies of more patients and mutationsearches in other hematological malignancies, such asacute leukemia, are needed in the future.Our results showed that TRAIL mRNA expressions inPMN in CML patients were barely detectable withoutIFN-· stimulation; however, upregulation of the TRAILmRNA expression by IFN-· was marked compared toother genes studied in RPA. Regarding FAS expression,Selleri et al. [32, 33] initially reported that IFN· enhancedthe FAS expression on CD34-positive cells in CML pa-tients, and that responsiveness to FAS-induced apoptosisfollowing stimulation with IFN- · correlated with the clin-ical effects of IFN-· therapy. We also confirmed that FASmRNA was upregulated by IFN-·, but the induction ratewas far lower than that of TRAIL mRNA. In addition, ourstudy suggested that the clinical response to IFN- · thera-py tended to correlate with the induction rate of TRAILmRNA after in vitro IFN-· stimulation, although thenumber of patients studied was small, and the relationwith clinical response should be cautiously interpreted.We actually tried to detect cell surface expressions ofTRAIL protein on PMNs in CML by flow cytometry anal-ysis. The stable expression was detectable but at a verylow level, and IFN- ·-induced expressions were very weak(data not shown), which was puzzling in light of themarked IFN-· inducibility of TRAIL mRNA in PMNcells in our study. This discrepancy may make sense, sincea very recent report showed that IFN-·-stimulated neu-trophils and monocytes release a soluble form of TRAILrather than a membrane-bound form in leukemia cells,including CML [34].We found that the mRNA ratio of (R1 + R2)/R3 inTRAIL receptors was increased after in vitro stimulationof IFN-·. This implies that IFN- · may augment TRAIL-induced apoptosis by increasing the ratio of authenticreceptor expression to decoy receptors. In line with this, aprevious study of a Daudi cell line showed that IFN- ·increased cell surface TRAIL-R1 and TRAIL-R2 expres-sions, which may be one mechanism to augment the anti-tumor action of TRAIL [10]. However, the signaling path-way at the intracellular level is complex and we suspectthat cell sensitivity to TRAIL or to IFN- · may not besimply determined at the receptor level. In fact, manymolecules play important roles in the TRAIL-induced sig-nal [5], for example, a recent study showed that IFN-sensitized cells to TRAIL through the IFN-induced en-hancement of the X-linked inhibitor of apoptosis-associ-ated factor-1 (XAF1), at least in melanoma cells [35]. Arecent DNA microarray study also indicated that TRAILand IFN synergistically induce apoptosis and caspase acti-vation with multiple levels of possible molecular cross-talk between the two diverse cytokine pathways [36].During the RPA experiment in this study, we foundthat RIP mRNA was induced by IFN- ·, which may havesome importance for the IFN-related apoptotic pathway.Death domain kinase RIP is known to play a key role inTNF signaling [37], in which RIP mediates the TNF-induced NF- B signal [38], or protects thymocytes fromTNF receptor type 2-induced cell death [39]. This mole-cule may be a candidate to be investigated for the antitu-mor action of IFN-· in the future.From the viewpoint of a therapeutic approach, TRAILalone or TRAIL plus other reagents such as IFN-· may begood candidates for novel anticancer therapy for manytumors, as IFNs sensitize a variety of solid tumor cells toTRAIL-induced apoptosis [40–42]. TRAIL-inducedapoptosis has also been reported in hematological malig-nancies such as malignant lymphoma, multiple myeloma[43–45], and Philadelphia chromosome-positive leuke-mia cells even in imatinib-resistant cells [46], implyingthat TRAIL alone or in combination with other drugscould be a novel choice for imatinib-resistant patientswith CML.In summary, gene mutations of TRAIL-R1 , TRAIL-R2and TRAIL promoter are absent or at least infrequent inCML. In addition, mutations of the TRAIL-R1 andTRAIL-R2 genes seem rare in MDS patients. However,the markedly high induction of TRAIL mRNA by IFN- ·may have some relevance to IFN-· action in CMLpatients.AcknowledgmentsWe thank Sachiko Hidani, Ryoko Matsumoto, and Nanae Naka-ju for their excellent technical support. This work was supported inpart by the Tsuchiya Foundation.122 Acta Haematol 2005;113:113–123 Liu/Tanaka/Ito/Ito/Sultana/Kyo/KimuraReferences1 Wiley SR, Schooley K, Smolak PJ, Din WS,Huang CP, Nicholl JK, Sutherland GR, SmithTD, Rauch C, Smith CA, Goodwin RG: Identi-fication and characterization of a new memberof the TNF family that induces apoptosis. Im-munity 1995;3:673–682.2 Wajant K, Pfizenmaier K, Scheurich P: TNF-related apoptosis inducing ligand (TRAIL) andits receptors in tumor surveillance and cancertherapy. Apoptosis 2002;7:449–459.3 Pitti RM, Marsters SA, Ruppert S, DonahueCJ, Moore A, Ashkenazi A: Induction of apop-tosis by Apo-2 ligand, a new member of thetumor necrosis factor cytokine family. J BiolChem 1996;271:12687–12690.4 Almasan A, Ashkenazi A: Apo2L/TRAIL:Apoptosis signaling, biology, and potential forcancer therapy. Cytokine Growth Factor Rev2003;14:337–348.5 MacFarlane M: TRAIL-induced signaling andapoptosis. Toxicol Lett 2003;139:89–97.6 Kayagaki N, Yamaguchi N, Nakayama M, EtoH, Okumura K, Yagita H: Type I interferons(IFNs) regulate tumor necrosis factor-relatedapoptosis-inducing ligand (TRAIL) expressionon human T cells: A novel mechanism for theantitumor effects of type I IFNs. J Exp Med1999;189:1454–1460.7 Lawrence D, Shahrokh Z, Marsters S, AchillesK, Shih D, Mounho B, Hillan K, Totpal K,DeForge L, Schow P, Hooley J, Sherwood S,Pai R, Leung S, Khan L, Gliniak B, Bussiere J,Smith CA, Strom SS, Kelley S, Fox JA, ThomasD, Ashkenazi A: Differential hepatocyte toxici-ty of recombinant Apo2L/TRAIL versions. NatMed 2001;7:383–385.8 Sheridan JP, Marsters SA, Pitti RM, Gurney A,Skubatch M, Baldwin D, Ramakrishnan L,Gray CL, Barker K, Wood WI, Goddard AD,Godowski P, Ashkenazi A: Control of TRAIL-induced apoptosis by a family of signaling anddecoy receptors. Science 1997;277:818–821.9 Zang DY, Goodwin RG, Loken MR, Bryant E,Deeg HJ: Expression of tumor necrosis factor-related apoptosis-inducing ligand, Apo2L, andits receptors in myelodysplastic syndrome: Ef-fects on in vitro hemopoiesis. Blood 2001;98:3058–3065.10 Oshima K, Yanase N, Ibukiyama C, Yamashi-na A, Kayagaki N, Yagita H, Muzoguchi J:Involvement of TRAIL/TRAIL-R interactionin IFN· -induced apoptosis of Daudi B lym-phoma cells. Cytokine 2001;14:193–201.11 Wuchter C, Krappmann D, Cai Z, Ruppert V,Scheidereit C, Drken B, Ludwig W-D, Kara-wajew L: In vitro susceptibility to TRAIL-induced apoptosis of acute leukemia cells in thecontext of TRAIL receptor gene expression andconstitutive NF-B activity. Leukemia 2001;15:921–928.12 Griffith TS, Chin WA, Jackson GC, LynchDH, Kubin MZ: Intracellular regulation ofTRAIL-induced apoptosis in human melano-ma cells. J Immunol 1998;161:2833–2840.13 Fisher MJ, Virmani AK, Wu L, Aplenc R,Harper JC, Powell SM, Rebbeck TR, SidranskyD, Gazdar AF, El-Deiry WS: Nucleotide sub-stitution in the ectodomain of TRAIL receptorDR4 is associated with lung cancer and headand neck cancer. Clin Cancer Res 2001;7:1688–1697.14 Pai SI, Wu GS, Ozoren N, Wu L, Jen J Sidrans-ky D, El-Deiry WS: Rare loss-of function muta-tion of a death receptor gene in head and neckcancer. Cancer Res 1998;58:3513–3518.15 Arai T, Akiyama Y, Okabe S, Saito K, Iwai T,Yuasa Y: Genomic organization and mutationanalyses of the DR5/TRAIL receptor 2 gene incolorectal carcinomas. Cancer Lett 1998;133:197–204.16 Wu WG, Soria JC, Wang L, Kemp BL, Mao L:TRAIL-R2 is not correlated with p53 status andis rarely mutated in non-small cell lung cancer.Anticancer Res 2000;20:4525–4529.17 Lee SH, Shin MS, Kim HS, Lee HK, Park WS,Kim SY, Lee JH, Han SY, Park JY, Oh RR,Jang JJ, Han JY, Lee JY, Yoo NJ: Alterationsof the DR5/TRAIL receptor 2 gene in non-small cell lung cancers. Cancer Res 1999;59:5683–5686.18 Shin MS, Kim HS, Lee SH, Park WS, Kim SY,Park JY, Lee JH, Lee SK, Lee SN, Jung SS,Han JY, Kim H, Lee JY, Yoo NJ: Mutations oftumor necrosis factor-related apoptosis-induc-ing ligand receptor 1 (TRAIL-R1) and receptor2 (TRAIL-R2) genes in metastatic breast can-cers. Cancer Res 2001;61:4942–4946.19 Seitz S, Wassmuth P, Fischer J Nothnagel A,Jandrig B, Schlag PM, Scherneck S: Mutationanalysis and mRNA expression of trail-recep-tors in human breast cancer. Int J Cancer 2002;102:117–128.20 Jeng YM, Hsu HC: Mutation of the DR5/TRAIL receptor 2 gene is infrequent in hepato-cellular carcinoma. Cancer Lett 2002;181:205–208.21 Lee SH, Shin MS, Kim HS, Lee HK, Park WS,Kim SY, Lee JH, Han SY, Park JY, Oh RR,Kang CS, Kim KM, Jang JJ, Nam SW, Lee JY,Yoo NJ: Somatic mutations of TRAIL-receptor1 and TRAIL-receptor 2 genes in non-Hodg-kin’s lymphoma. Oncogene 2001;20:399–403.22 Chawla-Sarkar M, Lindner DJ, Liu Y-F, Wil-liams BR, Sen GC, Silerman RH, Borden EC:Apoptosis and interferons: Role of interferon-stimulated genes as mediators of apoptosis.Apoptosis 2003;8:237–249.23 Gong B, Almasan A: Genomic organizationand transcriptional regulation of human Apo2/TRAIL gene. Biochem Biophys Res Commun2000;278:747–752.24 Talpaz M, Kantarjian HM, McCredie K, Tru-jillo JM, Keating MJ, Gutterman JU: Hemato-logic remission and cytogenetic improvementinduced by recombinant human interferon al-pha A in chronic myelogenous leukemia. NEngl J Med 1986;314:1065–1069.25 Takahashi N, Miura I, Saitoh K, Miura AB:Lineage involvement of stem cells bearing thePhiladelphia chromosome in chronic myeloidleukemia in the chronic phase as shown by acombination of fluorescence-activated cell sort-ing and fluorescence in situ hybridization.Blood 1998;92:4758–4763.26 Sambrook J, Fritsch EF, Maniatis T: MolecularCloning: A Laboratory Manual, ed 2. NewYork, Cold Spring Harbor Laboratory Press,1989.27 Oto M, Miyake Y, Yuasa Y: Optimization ofnonradioisotopic single strand conformationpolymorphism analysis with a conventionalminislab gel electrophoresis apparatus. AnalBiochem 1993;213:19–22.28 Tanaka H, Samuel CE: Mechanism of interfer-on action: Structure of the mouse PKR geneencoding the interferon-inducible RNA-depen-dent protein kinase. Proc Natl Acad Sci USA1994;91:7995–7999.29 Tanaka H, Samuel CE: Mouse interferon-indu-cible RNA-dependent protein kinase Pkr gene:Cloning and sequence of the 5)-flanking regionand functional identification of the minimalinducible promoter. Gene 2000;246:373–382.30 Johansson B, Fioretos T, Mitelman F: Cytoge-netic and molecular genetic evolution ofchronic myeloid leukemia. Acta Haematol2002;107:76–94.31 Tanaka H, Tanaka K, Oguma N, Ito K, Ito T,Kyo T, Dohy H, Kimura A: Effect of interferonalpha on chromosome abnormalities in treatedchronic myelocytic leukemia patients. CancerGenet Cytogenet, in press.32 Selleri C, Sato T, Del Vecchio L, Luciano L,Barrett AJ, Rotoli B, Young NS, MaciejewskiJP: Involvement of Fas-mediated apoptosis inthe inhibitory effects of interferon-· in chronicmyelogenous leukemia. Blood 1997;89:957–964.33 Selleri C, Maciejewski JP, Pane F, Luciano L,Raiola AM, Mostarda I, Salvatore F, Rotoli B:Fas-mediated modulation of Bcr/Abl inchronic myelogenous leukemia results in differ-ential effects on apoptosis. Blood 1998;92:981–989.34 Tecchio C, Huber V, Scapini P, Calzetti F,Margotto D, Todeschini G, Pilla L, MartinelliG, Pizzolo G, Rivoltini L, Cassatella MA:IFN ·-stimulated neutrophils and monocytesrelease a soluble form of TNF-related apopto-sis-inducing ligand (TRAIL/Apo-2 ligand) dis-playing apoptotic activity on leukemic cells.Blood 2004;103:3837–3844.35 Leaman DW, Chawla-Sarker M, Vyas K, Rehe-man M, Tamai K, Toji S, Borden EC: Identifi-cation of X-linked inhibitor of apoptosis-asso-ciated factor-1 as an interferon-stimulated genethat augments TRAIL Apo2L-induced apopto-sis. J Biol Chem 2002;277:28504–28511.36 Kumar-Sinha C, Varambally S, Sreekumar A,Chinnaiyan AM: Molecular cross-talk betweenthe TRAIL and interferon signaling pathway. JBiol Chem 2002;277:575–585.Gene Mutation of TRAIL Receptors inCML and MDSActa Haematol 2005;113:113–123 12337 Stanger BZ, Leder P, Lee TH, Kim E, Seed B:RIP: A novel protein containing a death do-main that interacts with Fas/APO-1 (CD95) inyeast and caused cell death. Cell 1995;81:513–523.38 Kelliher MA, Grimm S, Ishida Y, Kuo F, Stan-ger BZ, Leder P: The death domain kinase RIPmediates the TNF-induced NF-B signal. Im-munity 1998;8:297–303.39 Cusson N, Oikemus S, Kilpatrick ED, Cun-ningham L, Kelliher M: The death domainkinase RIP protects thymocytes from tumornecrosis factor receptor type 2-induced celldeath. J Exp Med 2002;196:15–26.40 Chawla-Sarkar M, Leaman DW, Jacobs BS,Borden EC: IFN- pretreatment sensitizes hu-man melanoma cells to TRAIL/Apo2 ligand-induced apoptosis. J Immunol 2002;169:847–855.41 Kim KB, Choi YH, Kim IK, Chung CW, KimBJ, Park YM, Jung YK: Potentiation of Fas-and TRAIL-mediated apoptosis by IFN- inA549 lung epithelial cells: Enhancement of cas-pase-8 expression through IFN-response ele-ment. Cytokine 2002;20:283–288.42 Shigeno M, Nakao K, Ichikawa T, Suzuki K,Kawakami A, Abiru S, Miyazoe S, NakagawaY, Ishikawa H, Hamasaki K, Nakata K, IshiiN, Eguchi K: Interferon-· sensitizes humanhepatoma cells to TRAIL-induced apoptosisthrough DR5 upregulation and NF-B inacti-vation. Oncogene 2003;22:1653–1662.43 Toomey NL, Deyev VV, Wood C, Boise LH,Scott D, Liu LH, Cabral L, Podack ER, BarberGN, Harrington WJ Jr: Induction of a TRAILmediated suicide program by interferon alphain primary effusion lymphoma. Oncogene2001;20:7029–7040.44 Mitsiades CS, Treon SP, Mitsiades N, ShimaY, Richardson P, Schlossman R, Hideshima T,Anderson KC: TRAIL/Apo2L ligand selective-ly induces apoptosis and overcomes drug resis-tance in multiple myeloma: Therapeutic appli-cations. Blood 2001;98:795–804.45 Mitsiades N, Mitsiades CS, Poulaki V, Ander-son KC, Treon SP: Intracellular regulation oftumor necrosis factor-related apoptosis-induc-ing ligand-induced apoptosis in human multi-ple myeloma cells. Blood 2002;99:2162–2171.46 Uno K, Inukai T, Kayagaki N, Goi K, Sato H,Nemoto A, Takahashi K, Kagami K, Yamagu-chi N, Yagita H, Okumura K, Koyama-Okaza-ki T, Suzuki T, Sugita K, Nakazawa S: TNF-related apoptosis-inducing ligand (TRAIL) fre-quently induces apoptosis in Philadelphiachromosome-positive leukemia cells. Blood2003;101:3658–3667.