EDNRB基因编码区点突变及多态性在浙江地区散发性先天性巨结肠症中的检测与分析

本实验应用聚合酶链反应-单链构象多态性(single strand confbrmation polymorphism analysis of polymerase chain reaction products,PCR-SSCP)和DNA直接测序技术,对75例浙江地区散发性先天性巨结肠病例EDNRB基因编码区的全部7个外显子,进行了点突变与单核苷酸多态性的检测与分析,探讨浙江地区先天性巨结肠患者EDNRB基因的突变特征,阐明EDNRB基因与散发性先天性巨结肠症发病之间可能存在的关系。结果有6例患者在第4外显子上检测到密码子277位点 CTG→CTA的置换,导致亮氨酸的同义突变(L277L),属于单核苷酸多态性,发生率为8％(6/75)。有 2例患者在第2外显子上检测到密码子185位点GTG→ATG的置换,导致缬氨酸到蛋氨酸的错义突变(V185M),此突变型未在国内外文献中报道过,认为是新的基因突变型,突变率2．7％(2/75)。研究结果表明,浙江地区先天性巨结肠群体可发生EDNRB基因的杂合性突变,提示EDNRB基因与先天性巨结肠症的发病存在一定程度的关联。     

Identification of splicing mutations of the last nucleotides of exons, a nonsense mutation, and a missense mutation of the XPAC gene as causes of group A xeroderma pigmentosum.

Four mutations of the XPAC gene were identified as molecular bases of different UV-sensitive subgroups of xeroderma pigmentosum (XP) group A. One was a G to C transversion at the last nucleotide of exon 4 in GM1630/GM2062, a little less hypersensitive subgroup than the most sensitive XP2OS/XP12RO. The second mutation was a G to A transition at the last nucleotide of exon 3 in GM2033/GM2090, an intermediate subgroup. Both mutations caused almost complete inactivation of the canonical 5' splice donor site and aberrant RNA splicing. The third mutation was a nucleotide transition altering the Arg-211 codon (CGA) to a nonsense codon (TGA) in another allele of GM2062. The fourth mutation was a nucleotide transversion altering the His-244 codon (CAT) to an Arg codon (CGT) in XP8LO, an intermediate subgroup. Our results strongly suggest that the clinical heterogeneity in XP-A is due to different mutations in the XPAC gene.     

Identification of missense, nonsense, and deletion mutations in the GRHPR gene in patients with primary hyperoxaluria type II (PH2)

Primary hyperoxaluria type II (PH2) is a rare disease characterized by the absence of an enzyme with glyoxylate reductase, hydroxypyruvate reductase, and D-glycerate dehydrogenase activities. The gene encoding this enzyme (GRHPR) has been characterized, and a single mutation has been detected in four PH2 patients. In this report, we have identified five novel mutations. One nonsense mutation (C295T) results in a premature stop codon at codon 99. A 4-bp deletion mutation has been found in the 5' consensus splice site of intron D, resulting in a predicted splicing error. Three missense mutations have been detected, including a missense transversion (T965G) in exon 9 (Met322Arg), a missense transition (G494A) in the putative co-factor binding site in exon 6 (Gly165Asp), and a substitution of an adenosine for a guanine in the 3' splice site of intron G. The functional consequences of the missense transversion and transition mutations have been investigated by transfection of cDNA encoding the mutated protein into COS cells. Cells transfected with either mutant construct have no enzymatic activity, a finding that is not significantly different from the control (empty) vector (P<0.05). These results further confirm that mutations in the GRHPR gene form the genetic basis of PH2. Ten of the 11 patients that we have genotyped are homozygous for one of the six mutations identified to date. Because of this high proportion of homozygotes, we have used microsatellite markers in close linkage with the GRHPR gene to investigate the possibility that the patients are the offspring of related individuals. Our data suggest that two thirds of our patients are the offspring of either closely or distantly related persons. Furthermore, genotyping has revealed the possible presence of a founder effect for the two most common mutations and the location of the gene near the marker D9S1874.     

Niemann-Pick type C disease: mutations of NPC1 gene and evidence of abnormal expression of some mutant alleles in fibroblasts

We analyzed Niemann-Pick type C disease 1 (NPC1) gene in 12 patients with Niemann-Pick type C disease by sequencing both cDNA obtained from fibroblasts and genomic DNA. All the patients were compound heterozygotes. We found 15 mutations, eight of which previously unreported. The comparison of cDNA and genomic DNA revealed discrepancies in some subjects. In two unrelated patients carrying the same mutations (P474L and nt 2972del2) only one mutant allele (P474L), was expressed in fibroblasts. The mRNA corresponding to the other allele was not detected even in cells incubated with cycloheximide. The promoter variants (-1026T/G and -1186T/C or -238 C/G), found to be in linkage with 2972del2 allele do not explain the lack of expression of this allele, as they were also found in control subjects. In another patient, (N1156S/Q922X) the N1156S allele was expressed in fibroblasts while the expression of the other allele was hardly detectable. In a fourth patient cDNA analysis revealed a point mutation in exon 20 (P1007A) and a 56 nt deletion in exon 22 leading to a frameshift and a premature stop codon. The first mutation was confirmed in genomic DNA; the second turned out to be a T-->G transversion in exon 22, predicted to cause a missense mutation (V1141G). In fact, this transversion generates a donor splice site in exon 22, which causes an abnormal pre-mRNA splicing leading to a partial deletion of this exon. In some NPC patients, therefore, the comparison between cDNA and genomic DNA may reveal an unexpected expression of some mutant alleles of NPC1 gene.     

Novelty of Axin 2 and lack of Axin 1 gene mutation in colorectal cancer: a study in Kashmiri population

Colorectal cancer is (CRC) one of the leading causes of mortality and morbidity. Various genetic factors have been reported to be involved in the development of colorectal cancers including Axin gene. Axin, a major scaffold protein, plays an important role in various bio signaling pathways. We aim to study mutational pattern of Axin gene in colorectal cancer patients of Kashmiri population. The paired tumor and adjacent normal tissue specimens of 50 consecutive patients with CRC were used in our study. The DNA preparations were evaluated for the occurrence of Axin 1 and Axin 2 gene mutations by direct DNA sequencing. We analyzed exon 1a, 1b, 1c, 2, 4, 6, and 10 of Axin 1 and exon 7 of Axin 2. In this study, we found a novel mutation of G>T (GCT>TCT) transversion in exon 7 of Axin 2 gene at codon G695T (p.alanine >?serine) at a frequency of 6% (3/50). In the same exon of Axin 2 gene a single nucleotide polymorphism (SNP) was detected in codon L688L (CCT>CTT) at a frequency of 36% (18/50). In exon 1c of Axin 1 a SNP was detected at codon D726D (GAT>GAC) at a frequency of 62.5% (31/50). Both the SNPs were synonymous hence do not lead to change of amino acid. Although Axin 1 and Axin 2 gene mutations have been found to be involved in the development of colorectal cancers, it seems to be a relatively rare event in Kashmiri population. However, an interesting finding of this study is the novelty of Axin 2 gene mutations which may be a predisposing factor in ethnic Kashmiri population to CRC.     

Identification of genetic mutations in Japanese patients with fructose-1,6-bisphosphatase deficiency.

Fructose-1,6-bisphosphatase (FBPase) deficiency is an autosomal recessive inherited disorder and may cause sudden unexpected infant death. We reported the first case of molecular diagnosis of FBPase deficiency, using cultured monocytes as a source for FBPase mRNA. In the present study, we confirmed the presence of the same genetic mutation in this patient by amplifying genomic DNA. Molecular analysis was also performed to diagnose another 12 Japanese patients with FBPase deficiency. Four mutations responsible for FBPase deficiency were identified in 10 patients from 8 unrelated families among a total of 13 patients from 11 unrelated families; no mutation was found in the remaining 3 patients from 3 unrelated families. The identified mutations included the mutation reported earlier, with an insertion of one G residue at base 961 in exon 7 (960/961insG) (10 alleles, including 2 alleles in the Japanese family from our previous report [46% of the 22 mutant alleles]), and three novel mutations--a G-->A transition at base 490 in exon 4 (G164S) (3 alleles [14%]), a C-->A transversion at base 530 in exon 4 (A177D) (1 allele [4%]), and a G-->T transversion at base 88 in exon 1 (E30X) (2 alleles [9%]). FBPase proteins with G164S or A177D mutations were enzymatically inactive when purified from E. coli. Another new mutation, a T-->C transition at base 974 in exon 7 (V325A), was found in the same allele with the G164S mutation in one family (one allele) but was not responsible for FBPase deficiency. Our results indicate that the insertion of one G residue at base 961 was associated with a preferential disease-causing alternation in 13 Japanese patients. Our results also indicate accurate carrier detection in eight families (73%) of 11 Japanese patients with FBPase deficiency, in whom mutations in both alleles were identified.     

HLA-B*2736等位基因的序列分析

使用FLOW-SSO、PCR-SSP以及测序等分型技术, 发现一个与HLA-B*270401基因相关的未知基因。设计基因特异性引物单独扩增B*27基因的外显子2-5, 包括内含子2-4, 并进行双向测序, 分析与B*270401基因序列的差异。该基因的扩增产物为1 815 bp。与B*270401相比在外显子3和4共有10个碱基的改变, 从而使相应氨基酸发生错义或同义突变。碱基634 A→C (密码子130丝氨酸→精氨酸); 670 A→T (密码子142苏氨酸→丝氨酸); 683 G→T (密码子146色氨酸→亮氨酸); 698 A→T (密码子151谷氨酸→缬氨酸); 774 G→C (密码子176谷氨酸→天冬氨酸); 776 C→A (密码子177苏氨酸→赖氨酸); 781 C→G (密码子179谷氨酰胺→谷氨酸); 789 G→T (密码子181丙氨酸同义突变); 1 438 C→T (密码子206甘氨酸同义突变); 1 449 G→C (密码子210甘氨酸→丙氨酸)。在IMGT/HLA数据库中B*27组只有3个基因（B*270502 / 2706 / 2732）提交了内含子序列。该未知基因的内含子2序列与B*2706相同, 显示了与B*27组基因的同源性, 但其同源性在内含子3、4均未得到支持, 与B*27组基因相比, 内含子3的第106个碱基C→G, 碱基168缺失, 碱基179 G→A, 碱基536 G→A; 内含子4中碱基82 T→C。但其内含子3、4序列却与B*070201完全相同。该基因序列已提交GenBank, 编号为被DQ915176, 被WHO确认为HLA-B*2736等位基因。     

III型神经中丝蛋白基因与中国高度近视人群相关性的研究

为了检测peripherin基因（PRPH）的突变与高度近视的病因有无相关关系,采用PCR-SSCP检测180例中国人高度近视先证者及60例正常人中PRPH基因所有外显子有无突变;对有突变的外显子区域进行克隆测序。结果表明,分析180例高度近视先证者PRPH基因编码区9个外显子及其邻近内含子,分别发现有下列核苷酸改变：密码子21TTC→TTT(Phe21Phe、4/180),nt2138C→G（IVS3、1/180）,密码子277GCC→ACC（Ala277Thr、8/180）,密码子237CCA→TCA（Arg237stop、1/180）,密码子292GCG→GCA(Ala292Ala,1/180),密码子361CUG→CUC（Leu361Leu,12/180）,密码子369AAA→AAG（Lys369Lys,12/180）,nt3331G→C（IVS7、3/180）,其中GCC277ACC为错义突变（Ala277Thr）;CCA237TCA为无义突变（Arg237stop）;密码子361CUG→CUC,密码子369AAA→AAG属于同义突变并且相连锁。Ala277Thr突变尚存在于正常人群中（1/60）,亦存在于患者正常亲属中;Arg237stop仅见于一个常染色体隐性遗传家系的患者中,为杂合性突变。分析180例高度近视先证者PRPH基因,未发现致病突变,可排除PRPH基因与高度近视病因的相关性。在中国人群中PRPH基因有多种变异。 Variation of the Peripherin Gene in Chinese with or Without High Myopia LI Jiang1,ZHANG Qing-jiong1,FU Rong2,XIAO Xue-shan1,LI Jia-zhang3,ZHANG Feng-sheng4, LI Shi-qiang1,LI Wei5,LI Tuo3,JIA Xiao-yun1,GUO Li1,GUO Xiang-ming 1.Zhongshan Ophthalmic Center,Sun Yat-sen University,Guangzhou 510060,China; 2.Shenzhen Municipal People's Hospital,Shenzhen 518000,China; 3.Department of Opthalmolgy,The people's Hospital of Enshi Autonomous Prefecture,Enshi 445000,Hubei,China; 4.Chaoju Eye Hospital,Baotou,Inner Mongolia 014000,China; 5.Shenzhen 2nd People's Hopital,Shenzhen 518000,China Abstract:To analyze the relationship of the peripherin gene(PRPH,OMIM17071) mutations with high myopia,genomic DNA was collected from 180 probands with high myopia (≤-6.0 dipoters) and 60 unrelated persons without high myopia.The coding sequences of PRPH gene in 240 subjects were analyzed using exon-by-exon PCR-heteroduplex-SSCP analysis and sequencing.Variations at codon21TTC→TTT(Phe21Phe、4/180),nt2138C→G（IVS3、1/180）,codon277 GCC→ACC（Ala277Thr、8/180）,codon237 CCA→TCA （Arg237stop、1/180）,codon292CCG→CCA (Ala292Ala,1/180),codon361CUG→CUC（Leu361Leu,12/180）,codon369 AAA→AAG（Lys369Lys,12/180）,nt3331G→C（IVS7、3/180）were detected in a number of probands as indicated in the blanket.Of the 8 variations one( codon 277,G→A,Ala277Thr) is a missense mutation identified in 8 of the 180patients and one of 60 controls;The mutation of codon361 and codon 369were synonymous one and linkage each other;Another one(codon237,CCA→TCA,Arg237stop) is a heterozygous nonsense mutation identified in one patient with autosomal recessive inheritance mode population but not in the 60 normal controls.The others were synonymous mutations.Eight nucleotide variations were found in the PRPH gene.We found no evidence that mutations in the PRPH gene are responsible for the high myopia in Chinese. Key words：high myopia; peripherin gene; PCR-SSCP     

The genetic basis of Cowden’s syndrome: three novel mutations in PTEN/MMAC1/TEP1

Cowden’s syndrome (CS) is an autosomal dominant disorder associated with an increased risk of developing benign and malignant tumors in a variety of tissues, including the skin, thyroid, breast and brain. Women with CS are felt to have an increased risk of developing breast cancer, and virtually all women with CS develop bilateral fibrocystic disease of the breast. Recently, a series of germline mutations have been identified from CS families in a gene known as PTEN/MMAC1/TEP1. In this study, we used heteroduplex analysis and direct sequencing analysis and identified three novel germline mutations in the PTEN/MMAC1/TEP1 coding sequence from unrelated individuals with CS. We report a de novo transition (T→C) at nucleotide 335 in exon 5. This missense mutation resulted in a leucine to proline (CTA to CCA) change at codon 112. We also describe a novel splice site mutation (801+2T→G) in intron 7 that caused exon skipping in PTEN/MMAC1/TEP1 mRNA. The third mutation we report is a missense mutation, consisting of a transition (T→C) at nucleotide 202 in exon 3, resulting in a tyrosine to histidine (TAC to CAC) change at codon 68. Finally, we also detected a rare polymorphism in exon 7 of the PTEN/MMAC1/TEP1 coding sequence. These data confirm the observation that mutations of the PTEN/MMAC1/ TEP1 coding sequence are responsible for at least some cases of CS, and further define the spectrum of mutations in this autosomal dominant disorder. Received: 21 October 1997 / Accepted: 15 December 1997     

Identification of one novel causative mutation in exon 4 of WFS1 gene in two Italian siblings with classical DIDMOAD syndrome phenotype

The aim of the present paper is to describe a novel missense mutation (G107R) of WFS1 gene that was unexpectedly detected, in two siblings from Southern Italy, outside exon 8; a very unusual finding which has previously been reported only twice in Italian patients with Wolfram syndrome (WS). Although in Spanish pedigrees' WFS1 mutations are frequently located in exon 4, this finding is very infrequent in other pedigrees, particularly in Italian patients. Conclusions: a) our report of two siblings with one novel WSF1 mutation (G107R) expands the molecular spectrum of WS; b) this is the 3rd report of Italian patients harbouring one mutation outside exon 8 and the 2nd with one mutation in exon 4; c) on the basis of the present observations, and literature data we can infer that mutation locations outside exon 8 do not seem to be clearly associated with peculiar phenotype expressions of WFS1 gene.     

Mucopolysaccharidosis IVA: four new exonic mutations in patients with N-acetylgalactosamine-6-sulfate sulfatase deficiency.

We report four new mutations in Japanese patients with mucopolysaccharidosis IVA (MPSIVA) who were heterozygous for a common double gene deletion. A nonsense mutation of CAG to TAG at codon 148 in exon 4 was identified, resulting in a change of Q to a stop codon and three missense mutations. V (GTC) to A (GCC) at codon 138 in exon 4, P (CCC) to S (TCC) at codon 151 in exon 5, and P (CCC) to L (CTC) at codon 151 in exon 5. Introduction of these mutations into the normal GALNS cDNA and transient expression in cultured fibroblasts resulted in a significant decrease in the enzyme activity. V138A and Q148X mutations result in changes of restriction site, which were analyzed by restriction-enzyme assay. P151S and P151L mutations that did not alter the restriction site were detected by direct sequencing or allele specific oligohybridization. Detection of the double gene deletion was initially done using Southern blots and was confirmed by PCR. Haplotypes were determined using seven polymorphisms to the GALNS locus in families with the double gene deletion. Haplotype analysis showed that the common double gene deletion occurred on a single haplotype, except for some variation in a VNTR-like polymorphism. This finding is consistent with a common founder for all individuals with this mutation.     

Two mutations within the coding sequence of the phenylalanine hydroxylase gene

Summary Two previously unidentified mutations at the phenylalanine hydroxylase locus were found during a study of the relationship between genotype and phenotype in phenylketonuria and hyperphenylalaninemia. One mutation eliminates the BamHI site in exon 7 and the other eliminates the HindIII site in exon 11 of the phenylalanine hydroxylase gene. They were suspected because of deviating restriction fragment patterns and confirmed by amplification, via the polymerase chain reaction, of exon 7 and exon 11, respectively, followed by digestion with the appropriate restriction enzyme. Direct sequencing of amplified mutant exon 7 revealed a G/C to T/A transversion at the first base of codon 272, substituting a GGA glycine codon for a UGA stop codon. Direct sequencing of amplified mutant exon 11 revealed a deletion of codon 364, a CTT leucine codon. The exon 7 mutation can be expected to result in a truncated protein and the exon 11 mutation in the elimination of an amino acid in the catalytic region of the enzyme. A patient who is a compound heterozygote for these two mutations has classical phenylketonuria. It is concluded that each of the two mutations leads to a profound loss of enzymatic activity. The segregation of these mutations with disease alleles in 4 and 2 families, respectively, supports the hypothesis that multiple mutations at the phenylalanine hydroxylase locus explain the variable phenylalanine tolerance in patients with phenylalanine hydroxylase deficiency.     

Identification and expression of five mutations in the human acid sphingomyelinase gene causing types A and B Niemann-Pick disease. Molecular evidence for genetic heterogeneity in the neuronopathic and non-neuronopathic forms.

The deficient activity of the human lysosomal hydrolase, acid sphingomyelinase (ASM, EC 3.1.4.12), results in the neuronopathic (Type A) and non-neuronopathic (Type B) forms of Niemann-Pick disease (NPD). To investigate the genetic basis of the phenotypic heterogeneity in NPD, the molecular lesions in the ASM gene were determined from three unrelated NPD patients and evaluated by transient expression in COS-1 cells. A Type A NPD patient of Asian Indian ancestry (proband 1) was homoallelic for a T to A transversion in exon 2 of the ASM gene which predicted a premature stop at codon 261 of the ASM polypeptide (designated L261X). In contrast, an unrelated Type A patient of European ancestry (proband 2) was heteroallelic for a two-base (TT) deletion in exon 2 which caused a frame-shift mutation at ASM codon 178 (designated fsL178), leading to a premature stop at codon 190, and a G to A transition in exon 3 which caused a methionine to isoleucine substitution at codon 382 (designated M382I). Transient expression of the fsL178, L261X, and M382I mutations in COS-1 cells demonstrated that these lesions did not produce catalytically active ASM, consistent with the severe neuronopathic Type A NPD phenotype. In contrast, an unrelated Type B patient of European descent (proband 3) was heteroallelic for two missense mutations, a G to A transition in exon 2 which predicted a glycine to arginine substitution at ASM codon 242 (designated G242R), and an A to G transition in exon 3 which resulted in an asparagine to serine substitution at codon 383 (designated N383S). Interestingly, the G242R allele produced ASM activity in COS-1 cells at levels about 40% of that expressed by the normal allele, thereby explaining the mild Type B phenotype of proband 3 and the high residual activity (i.e. approximately 15% of normal) in cultured lymphoblasts. In contrast, the N383S allele did not produce catalytically active enzyme. None of these five ASM mutations was detected in over 60 other unrelated NPD patients analyzed, nor were these mutations found in over 100 normal ASM alleles. Thus, small deletions or nonsense mutations which trunctated the ASM polypeptide, or missense mutations that rendered the enzyme noncatalytic, resulted in Type A NPD disease, whereas a missense mutation that produced a defective enzyme with residual catalytic activity caused the milder nonneuronopathic Type B phenotype. These findings have facilitated genotype/phenotype correlations for this lysosomal storage disease and provided insights into the functional organization of the ASM polypeptide.     

Incidence of fibroblast growth factor receptor 3 gene (FGFR3) A248C, S249C, G372C, and T375C mutations in bladder cancer

Bladder cancer is the most frequent cancer of the urinary system. Fibroblast growth factor receptors (FGFR) belong to the tyrosine kinase family and have important roles in cell differentiation and proliferation and embryogenesis. FGFR3 is located on chromosome 4p16.3, and missense mutations of FGFR3 are associated with autosomal dominant human skeletal disorders and have some oncogenic effects. We examined the incidence of FGFR3 thanatophoric dysplasia mutations located in exon 7, A248C and S249C, and in exon 10, G372C and T375C, and their correlation with clinical-pathological parameters in bladder carcinoma patients. Fifty-six paraffin-embedded specimens of transitional cell carcinoma of the urinary bladder were included in this study. Analysis of FGFR3 thanatophoric dysplasia mutations located in exon 7, A248C and S249C, and in exon 10, G372C and T375C, was performed by PCR-restriction fragment length polymorphism (RFLP) analysis and DNA sequencing. FGFR3 thanatophoric dysplasia mutations located in exon 7, A248C and S249C, and in exon 10, G372C and T375C, were detected in 33 of the 56 patients (heterozygous mutant). Among the 56 transitional cell carcinomas, missense point mutations were detected in seven of them at codon A248C, 28 of them at codon S249C, and three of them at codon T375C, similar to data from previous reports. When the results of the FGFR3 thanatophoric dysplasia mutations located in exon 7, A248C and S249C and in exon 10, G372C and T375C, were analyzed one by one or as a group, despite the findings of previous research reports, our data suggest that these mutations are detected homogenously regardless of the tumor classification and tumor grade.     

Screening for nonsense mutations in patients with severe hemophilia A can provide rapid,direct carrier detection

Summary Despite marked genetic heterogeneity in families with hemophilic patients, transition mutations in CG dinucleotides occur frequently. Of 71 CG dinucleotides in the factor VIII cDNA, a C-to-T transition in 12 would lead to a new Stop codon (CGA to TGA). Using restriction enzyme digestion of 11 amplified DNA fragments, seven point mutations were localized among 60 patients with severe hemophilia A. Five were detected as loss of a natural or introduced TaqI site at codons -5, 583, 1941, 2116, and 2209 and were confirmed as CGA (Arg) to TGA (Stop) nonsense mutations by DNA sequencing. A novel C-to-T nonsense mutation was detected as loss of the RsaI site at codon 1966 and confirmed by sequence in two unrelated individuals. Two partial gene deletions were detected as selective failure to amplify exon 1 and exons 15–22, respectively. In an additional (61st) patient who was subsequently found to have mild (instead of severe) hemophilia, digests suggested a mutation in codon 1696. Upon sequencing, this codon contained a novel missense mutation, a C-to-G transversion changing CGA (Arg 1696) to GGA (Gly). In four families with women available for testing, carrier status was rapidly determined by direct screening for the point mutation. In two of three with sporadic occurrences, the mother was a carrier as were two of four sisters. In the other family, the mother and a sister were homozygous for the TaqI cleavage site in their amplified exon 24 fragment, indicating a de novo C-to-T transition in codon 2209 in the patient's factor VIII gene. This final patient's sister was a noncarrier even though by linkage analysis she inherited the same factor VIII gene as her brother.These results have already been published in part in abstract form: Reiner AP, Thompson AR (1990) Circulation Research 82:304