Amplified C lambda and c-abl genes are on the same marker chromosome in K562 leukemia cells

Comprehensive, integrated, and phased whole-genome analysis of the primary ENCODE cell line K562

K562 is widely used in biomedical research. It is one of three tier-one cell lines of ENCODE and also most commonly used for large-scale CRISPR/Cas9 screens. Although its functional genomic and epigenomic characteristics have been extensively studied, its genome sequence and genomic structural features have never been comprehensively analyzed. Such information is essential for the correct interpretation and understanding of the vast troves of existing functional genomics and epigenomics data for K562. We performed and integrated deep-coverage whole-genome (short-insert), mate-pair, and linked-read sequencing as well as karyotyping and array CGH analysis to identify a wide spectrum of genome characteristics in K562: copy numbers (CN) of aneuploid chromosome segments at high-resolution, SNVs and indels (both corrected for CN in aneuploid regions), loss of heterozygosity, megabase-scale phased haplotypes often spanning entire chromosome arms, structural variants (SVs), including small and large-scale complex SVs and nonreference retrotransposon insertions. Many SVs were phased, assembled, and experimentally validated. We identified multiple allele-specific deletions and duplications within the tumor suppressor gene FHIT. Taking aneuploidy into account, we reanalyzed K562 RNA-seq and whole-genome bisulfite sequencing data for allele-specific expression and allele-specific DNA methylation. We also show examples of how deeper insights into regulatory complexity are gained by integrating genomic variant information and structural context with functional genomics and epigenomics data. Furthermore, using K562 haplotype information, we produced an allele-specific CRISPR targeting map. This comprehensive whole-genome analysis serves as a resource for future studies that utilize K562 as well as a framework for the analysis of other cancer genomes.

The implication of cancer progenitor cells and the role of epigenetics in the development of novel therapeutic strategies for chronic myeloid leukemia

SIGNIFICANCE

Chronic myeloid leukemia (CML) involves the malignant transformation of hematopoietic stem cells, defined largely by the Philadelphia chromosome and expression of the breakpoint cluster region-Abelson (BCR-ABL) oncoprotein. Pharmacological tyrosine kinase inhibitors (TKIs), including imatinib mesylate, have overcome limitations in conventional treatment for the improved clinical management of CML.

RECENT ADVANCES

Accumulated evidence has led to the identification of a subpopulation of quiescent leukemia progenitor cells with stem-like self renewal properties that may initiate leukemogenesis, which are also shown to be present in residual disease due to their insensitivity to tyrosine kinase inhibition.

CRITICAL ISSUES

The characterization of quiescent leukemia progenitor cells as a unique cell population in CML pathogenesis has become critical with the complete elucidation of mechanisms involved in their survival independent of BCR-ABL that is important in the development of novel anticancer strategies. Understanding of these functional pathways in CML progenitor cells will allow for their selective therapeutic targeting. In addition, disease pathogenesis and drug responsiveness is also thought to be modulated by epigenetic regulatory mechanisms such as DNA methylation, histone acetylation, and microRNA expression, with a capacity to control CML-associated gene transcription.

FUTURE DIRECTIONS

A number of compounds in combination with TKIs are under preclinical and clinical investigation to assess their synergistic potential in targeting leukemic progenitor cells and/or the epigenome in CML. Despite the collective promise, further research is required in order to refine understanding, and, ultimately, advance antileukemic therapeutic strategies.

Detection and Quantification of the Abelson Tyrosine Kinase Domains of the bcr-abl Gene Translocation in Chronic Myeloid Leukaemia Using Genomic Quantitative Real-time Polymerase Chain Reaction

Introduction: Since undetectable BCR-ABL mRNA transcription does not always indicate eradication of the Ph+ CML clone and since transcriptionally silent Ph+ CML cells exist, quantitation by genomic PCR of bcr-abl genes can be clinically useful. Furthermore, hotspot mutations in the Abelson tyrosine kinase (ABLK) domain of the bcr-abl gene translocation in Philadelphia chromosome-positive (Ph+) chronic myeloid leukaemia (CML) cells confer resistance on the specific kinase blocking agent, STI571. Materials and Methods: Genomic DNA from K562, CESS and patient CML cells were amplified using rapid cycle quantitative real-time polymerase chain reaction for the gene regions spanning the mutation hotspots. In assays for ABLK exons 4 or 6, exonic or intronic PCR primers were used. Results: We show that separation of cycle threshold (CT) values for log-fold amplicon quantification was 2.9 cycles for ABLK exon 4, and 3.8 cycles for exon 6 with rapid amplification times. K562 CML cells were found to have a ~2 log-fold ABLK gene amplification. In contrast, patient CML cells had CT differences of 2.2 for both exon, suggesting that there was no significant ABLK gene amplification. DNA sequencing confirmed that neither K562 nor patient CML cells contained ABLK hotspot mutations. Messenger RNA transcription analysis permitted the assessment of BCR-ABL transcription, which was qualitatively correlated to genomic amplification. Conclusions: This novel Q-PCR assay was found to have high fidelity and legitimacy, and potentially useful for monitoring minimal residual disease, transcriptionally silent Ph+ CML cells, and bcr-abl gene amplification. Key words: Drug resistance, Haematologic neoplasms, Molecular diagnostic techniques, Philadelphia chromosome

Complete karyotype characterization of the K562 cell line by combined application of G-banding, multiplex-fluorescence in situ hybridization, fluorescence in situ hybridization, and comparative genomic hybridization

This study combines conventional cytogenetics, fluorescence in situ hybridization (FISH), multiplex-FISH and comparative genomic hybridization (CGH). In applying this multimodal approach on the human leukemia cell line K562, the chromosome composition was refined in detail and compared with data from the literature. A hypotriploid karyotype with a modal chromosome number of 67, and 21 unique marker chromosomes were identified. The classification of six markers was identical to published data and the composition of five further markers from the literature could be fully clarified for the first time. The composition of another five markers, which have been interpreted in divergent ways in different studies, were elucidated without doubt. Finally, five new markers of our study seem to have no equivalents in former studies, very likely due to limitations of conventional cytogenetics. The combinatory application of complementary techniques as shown in this study will be very useful to provide the basis of a refined genotype analysis on the chromosomal level.

Cytogenetics of the chronic myeloid leukemia-derived cell line K562: karyotype clarification by multicolor fluorescence in situ hybridization, comparative genomic hybridization, and locus-specific fluorescence in situ hybridization

The transformation of chronic myeloid leukemia (CML) from a chronic phase to an acute phase is frequently accompanied by additional chromosome changes. Extensive chromosome G-banded studies have revealed the secondary changes are nonrandom and frequently include trisomy 8, isochromosome 17q, trisomy 19, or an extra copy of the Philadelphia chromosome. In addition to these secondary chromosome changes, complex structural rearrangements often occur to form marker structures that remain unidentified by conventional G-banded analysis. The CML-derived cell line, K562, has been widely used in research since it was originally established in 1975. The K562 karyotype however, has remained incomplete, and marker structures have never been fully described. Recent advances in fluorescence in situ hybridization (FISH) technology have introduced the possibility of chromosome classification based on 24-color chromosome painting (M-FISH). In this study, we report a clarified karyotype for K562 obtained by a combination of the following molecular cytogenetic techniques: comparative genomic hybridization (CGH), FISH mapping using locus-specific probes, and M-FISH. Multicolor FISH has identified the marker structures in this cell line. The characteristic marker chromosome in K562 has been confirmed by this study to be a der(18)t(1;18). Multicolor FISH confirmed the identity of marker structures partially identified by G-banding as der(6)t(6;6),der(17)t(9;17),der(21)t(1;21),der(5)t(5;6). In addition M-FISH has revealed a deleted 20q and a complex small metacentric marker comprised of material from chromosomes 1, 6, and 20. A cryptic rearrangement was revealed between chromosomes 12 and 21 that produced a structure that looks like a normal chromosome 12 homologue by G-banding analysis. Finally, M-FISH detected regions from chromosome 13 intercalated into two acrocentric markers.

Karyotypic evolution of cells in culture: a new concept

The Chapter summarizes peculiarities of karyotypic variability during establishment and long-term cultivation of permanent cell lines. A new concept on pathways of karyotypic evolution of cells in culture is put forward. A detailed description is presented of the author's original approach of cytogenetic analysis of cell lines provided for a principally new characteristic of the cell line: its generalized reconstructed karyotype (GRK). Its use as a criterion to evaluate authenticity, purity, and stability of cell lines is discussed. Based on analysis of the GRK, two stages of karyotype evolution of cell lines are revealed: establishment and stabilization, different in karyotypic variability of the cell population and in peculiarities of clone selection. Comparison of peculiarities of karyotypic variability of leukemic and tumor cells both in vitro and in vivo was made, and general regularities of their karyotypic evolution have been established, such as nonrandom changes in the number and structure of chromosomes and deletion of one of the sex chromosomes, as well as regularities characteristic only of cells in culture in most human and animal cell lines (at least 85%) of disomy on all autosomes. The rest of the cell lines, 15%, are characterized by either partial or total monosomies on certain autosomes during long-term cultivation. Three main compensatory mechanisms of maintaining viability of cell lines that have lost genetic material are discussed: polyploidization of the initial cell clone, amplification of oncogenes (predominantly of mys family), and extracopying of whole autosomes or of their fragments.

Comparative genomic hybridization reveals previously undescribed amplifications and deletions in the chronic myeloid leukemia-derived K-562 cell line

We used comparative genomic hybridization (CGH) to identify a number of previously undescribed chromosomal imbalances in K-562, a spontaneously transformed cell line originally derived from leukemic cells of a chronic myeloid leukemia (CML) patient in blast crisis. Noteworthy were a discrete amplification in band 13q31, increased copy number of chromosome arms 1q, 5p, 6p, and 16q, and loss of material from 8p, 9p, 10q, and 17p. Amplification within bands 9q34 and 22q11.2 was consistent with previous descriptions of increased copy number of the CML-specific 5'BCR-3'ABL fusion gene in K-562. However, amplification of a large distal segment, 9q31-->9q34, mostly proximal to the ABL locus, was unexpected and is unlikely to be related to BCR-ABL recombination. Previous karyotype studies are reviewed in detail and compared with the CGH findings.

Tissue-specific deoxyribonuclease I-hypersensitive sites in the vicinity of the immunoglobulin C lambda cluster of man

During B cell development, the onset of DNA rearrangements, expression, and somatic hypermutation of Ig genes are regulated through the complex interaction of cis-acting elements with trans-acting factors. Our aim is to identify DNA elements required during activation of the human Ig lambda light chain genes. Determination of deoxyribonuclease (DNase) I-hypersensitive sites in complex regulated genes can lead to the identification of sequence elements which would have been overlooked by employing transient transfection protocols. We have therefore investigated the chromatin structure of human J-C lambda genes and identified three DNase I-hypersensitive sites (HSS-1, -2, and -3) within an 8-kb region downstream of the J-C lambda 7 gene. HSS-2 and HSS-3 are B cell specific. The DNase I-hypersensitive sites are also present in kappa-producing cell lines which have not rearranged the Ig lambda locus and produce germ-line J-C lambda transcripts. We conclude that in mature B cells, both kappa and lambda loci are in an active structure regardless of the type of light chain they produce. This suggests that the chromatin structure of both loci is opened early in B cell development and that the active structure persists in mature B cells. The observed temporal order (first kappa, then lambda) of activation can be explained by consecutive synthesis of the appropriate regulating factors and the tight regulation of the recombination machinery through the products of L chain gene rearrangements.

Complete nucleotide sequence of the gene for human heparin cofactor II and mapping to chromosomal band 22q11

Heparin cofactor II (HCII) is a 66-kDa plasma glycoprotein that inhibits thrombin rapidly in the presence of dermatan sulfate or heparin. Clones comprising the entire HCII gene were isolated from a human leukocyte genomic library in EMBL-3 lambda phage. The sequence of the gene was determined on both strands of DNA (15,849 bp) and included 1749 bp of 5'-flanking sequence, five exons, four introns, and 476 bp of DNA 3' to the polyadenylation site. Ten complete and one partial Alu repeats were identified in the introns and 5'-flanking region. The HCII gene was regionally mapped on chromosome 22 using rodent-human somatic cell hybrids, carrying only parts of human chromosome 22, and the chronic myelogenous leukemia cell line K562. With the cDNA probe HCII7.2, containing the entire coding region of the gene, the HCII gene was shown to be amplified 10-20-fold in K562 cells by Southern analysis and in situ hybridization. From these data, we concluded that the HCII gene is localized on the chromosomal band 22q11 proximal to the breakpoint cluster region (BCR). Analysis by pulsed-field gel electrophoresis indicated that the amplified HCII gene in K562 cells maps at least 2 Mbp proximal to BCR-1. Furthermore, the HCII7.2 cDNA probe detected two frequent restriction fragment length polymorphisms with the restriction enzymes BamHI and HindIII.

The (6;9) chromosome translocation, associated with a specific subtype of acute nonlymphocytic leukemia, leads to aberrant transcription of a target gene on 9q34

The specific (6;9)(p23;q34) chromosomal translocation is associated with a defined subtype of acute nonlymphocytic leukemia (ANLL). The 9q34 breakpoint is located at the telomeric side of the c-abl gene. Through a combination of chromosome jumping, long-range mapping, and chromosome walking, the chromosome 9 breakpoints of several t(6;9) ANLL patients were localized within a defined region of 8 kilobases (kb), 360 kb telomeric of c-abl. Subsequent cDNA cloning revealed that this region represented an intron in the middle of a gene, called Cain (can), encoding a 7.5-kb transcript. Disruption of the can gene by the translocation resulted in the expression of a new 5.5-kb can mRNA from the 6p- chromosome. Isolation of chromosome 6 sequences showed that breakpoints on 6p23 also clustered within a limited stretch of DNA. These data strongly suggest a direct involvement of the translocation in the leukemic process of t(6;9) ANLL.

The (6;9) chromosome translocation, associated with a specific subtype of acute nonlymphocytic leukemia, leads to aberrant transcription of a target gene on 9q34

The specific (6;9)(p23;q34) chromosomal translocation is associated with a defined subtype of acute nonlymphocytic leukemia (ANLL). The 9q34 breakpoint is located at the telomeric side of the c-abl gene. Through a combination of chromosome jumping, long-range mapping, and chromosome walking, the chromosome 9 breakpoints of several t(6;9) ANLL patients were localized within a defined region of 8 kilobases (kb), 360 kb telomeric of c-abl. Subsequent cDNA cloning revealed that this region represented an intron in the middle of a gene, called Cain (can), encoding a 7.5-kb transcript. Disruption of the can gene by the translocation resulted in the expression of a new 5.5-kb can mRNA from the 6p- chromosome. Isolation of chromosome 6 sequences showed that breakpoints on 6p23 also clustered within a limited stretch of DNA. These data strongly suggest a direct involvement of the translocation in the leukemic process of t(6;9) ANLL.

Comparative mapping of the constitutional and tumor-associated 11;22 translocations

The reciprocal t(11;22)(q23;q11) is the most common non-Robertsonian constitutional translocation in humans. The tumor-associated 11;22 rearrangement of Ewing sarcoma (ES) and peripheral neuroepithelioma (NE) is cytologically very similar to the 11;22 constitutional rearrangement. Using immunoglobulin light-chain constant region, ETS1 probes, and the technique of in situ hybridization, we previously were able to show that the constitutional and ES/NE breakpoints are different. As a first step toward isolating these translocation junctions and to further distinguish between them, we have made somatic cell hybrids. Cells from a constitutional 46,XX,inv(9),t(11;22) carrier and from an ES cell line with a t(11;22) were separately fused to a hypoxanthine-guanine phosphoribosyltransferase-deficient Chinese hamster cell line (RJK88). Resulting clones were screened with G-banding and Southern hybridization. Hybrid clones derived from the constitutional t(11;22) were established which contained the der(22) and both the der(22) and the der(11). Hybrid clones derived from the ES cell line containing the der(11) were isolated. Using the technique of Southern hybridization we have sublocalized the loci; ApoA1/C3, CD3D, ETS1, PBGD, THY1, D11S29, D11S34, and D11S147 to the region between the two breakpoints on chromosome 11 and V lambda I, V lambda VI, V lambda VII, and D22S10 to the region between the breakpoints on chromosome 22. Using anonymous DNA probes, we found that D22S9 and D22S24 map proximal to the constitutional breakpoint and that D22S15 and D22S32 map distal to the ES breakpoint on chromosome 22.

The first BCR gene intron contains breakpoints in Philadelphia chromosome positive leukemia

The hallmark of chronic myelogenous leukemia (CML) is a translocation between chromosomes 9 and 22 - the Philadelphia (Ph') translocation. The translocation is also found in acute lymphocytic leukemia (ALL) albeit in a lower percentage of patients. The breakpoint on chromosome 22 is located within the BCR gene: in CML, breakpoints are clustered within 5.8 kb of DNA, the major breakpoint cluster region (Mbcr). In ALL, breakpoints have been reported within the Mbcr but also in more 5' regions encompassing the BCR gene. To characterize the latter breakpoints, we have molecularly cloned and mapped the entire gene, which encompasses approximately 130 kb of DNA. Mbcr negative, Ph'-positive ALL breakpoints were not distributed at random within the gene but rather were found exclusively within the 3' half of the first BCR gene intron. In contrast to the Mbcr, which is limited to a region of 5.8 kb, this part of the intron has a size of 35 kb. Translocation breakpoints in this region appear to be specific for ALL, since it was not rearranged in clinically well-defined CML specimens nor in any other tumor DNA samples examined.

The human Vpre B gene is located on chromosome 22 near a cluster of V lambda gene segments

The chromosomal location of the human VpreB gene was determined by Southern blotting analysis of restriction enzyme-digested DNAs from a panel of 17 mouse-human somatic cell hybrids. The pattern of hybridization of a Vpre B-specific probe in conjunction with earlier analysis of several marker genes allowed the following conclusions: 1) Vpre B is on human chromosome 22 within band 22q11.2 distal to the bcr-like gene, bcr-2 and proximal to the bcr-like gene, bcr-4. 2) Vpre B has been localized relative to several constitutional and tumor-specific breakpoints within 22q11.2, segregates in hybrids retaining 22q- chromosomes with some but not with all members of the V lambda 1 subgroup of the V lambda genes, and is amplified with these genes in K562 cells. 3) The order of the loci on chromosome 22 is centromere----bcr-2, Vpre B, V lambda 1----bcr-4----C lambda----bcr-1----bcr-3----sis.

Duplication of the bcr and gamma-glutamyl transpeptidase genes

The Philadelphia (Ph') translocation involves rearrangement of the bcr gene located on chromosome 22. Hybridization experiments revealed the presence of multiple bcr gene-related loci within the human genome. Two of these were molecularly cloned and characterized. Both loci contain exons and introns corresponding to the 3' region of the bcr gene. Restriction enzyme and DNA sequence analysis indicate a very high degree of conservation between bcr and the two related genomic sequences. Both bcr-related loci are located on chromosome 22, one centromeric, the other telomeric, of the bcr gene. Within the two bcr related genomic sequences, fragments or the complete coding sequences of an unrelated gene were found to be present. This gene was identified; it encodes gamma-glutamyl transferase, an enzyme involved in the glutathione metabolism.

Localization of gelsolin proximal to ABL on chromosome 9

Gelsolin is a plasma and cytoskeletal protein that severs actin filaments and is regulated by both Ca+2 and polyphosphoinositides. The two forms of gelsolin are encoded by a single gene and derived through alternative message splicing. By Southern blot analysis of somatic cell hybrids and in situ chromosomal localization, we demonstrate that the gelsolin gene is present on human chromosome 9 in bands q32-q34. In situ hybridization of gelsolin to cells containing a Philadelphia chromosome [(9;22)(q34;q11)], as well as Southern blot analysis of K562 cell DNA, indicates that gelsolin is centromeric to the ABL locus in 9q34. Southern blot analysis of NotI-digested, pulsed-field gel electrophoresis-separated DNA indicates the gelsolin gene is greater than or equal to 40 kb centromeric to ABL. These studies and standard Southern blot analysis of digested DNA also indicate that the NotI restriction site contained in the gelsolin gene is uncleavable in DNA from white blood cells and hematopoietic cell lines.

Oncogenes in human leukemias

Eukaryotic cells contain a family of genes termed cellular oncogenes or proto-oncogenes thought to regulate normal cell growth and development. In some abnormal circumstances, such as following transduction by retroviruses, activation of these genes causes leukemias in animals. Possible mechanisms of activation of cellular oncogenes include: point mutation, deletion, or insertion; amplification; activation by internal rearrangement, chromosomal translocation, or promoter insertion; recombinatorial events resulting in the formation of novel chimeric genes; among others. In this review, we consider data implicating activation of cellular oncogenes in the pathogenesis of leukemia in humans. We discuss possible mechanisms whereby oncogene activation may induce leukemias, as well as potential diagnostic and therapeutic implications.

Oncogenes and human leukemias

Eukaryotic cells contain a family of genes termed "cellular oncogenes" or "proto-oncogenes," thought to regulate normal cell growth and development. In some circumstances, such as following transduction by retroviruses, activation of these genes causes tumors and leukemias in animals. Possible mechanisms of cellular oncogene activation include: 1) DNA point mutation, deletion or insertion, 2) gene amplification, 3) gene activation by internal rearrangement, chromosomal translocation or promoter insertion, 4) recombinative events resulting in the formation of novel chimeric genes, and others. In this review, we consider data which implicates cellular oncogene activation in the pathogenesis of leukemia in humans. We discuss possible mechanisms by which oncogene activation may induce leukemias, as well as potential diagnostic and therapeutic implications.

The first intron in the human c-abl gene is at least 200 kilobases long and is a target for translocations in chronic myelogenous leukemia

The c-abl protooncogene is unusual in two respects; it has multiple, widely space N-terminal coding exons transcribed by different promoters, and it is the target of the translocations that form the Philadelphia chromosome found in cells of chronic myelogenous leukemia patients. To understand the organization of the gene in normal and chronic myelogenous leukemia patient DNA we have mapped c-abl by pulsed field gradient gel electrophoresis. We find that one of the alternative 5' exons of the gene lies at least 200 kilobases upstream of the remaining c-abl exons, posing formidable transcription and splicing problems. The 5'-most c-abl exon includes an unusually long 1,276-base-pair segment that contains 15 ATG codons and multiple short open reading frames, upstream of the abl initiator codon. Its peculiar structure suggests that c-abl may be decapitated in most chronic myelogenous leukemia patients, and we demonstrate that this is the case in the chronic myelogenous leukemia cell line K562.

Mapping of four distinct BCR-related loci to chromosome region 22q11: order of BCR loci relative to chronic myelogenous leukemia and acute lymphoblastic leukemia breakpoints

A probe derived from the 3' region of the BCR gene (breakpoint cluster region gene) detects four distinct loci in the human genome. One of the loci corresponds to the complete BCR gene, whereas the others contain a 3' segment of the gene. After HindIII cleavage of human DNA, these four loci are detected as 23-, 19-, 13-, and 9-kilobase-pair fragments, designated BCR4, BCR3, BCR2, and BCR1, respectively, with BCR1 deriving from the original complete BCR gene. All four BCR loci segregate 100% concordantly with human chromosome 22 in a rodent-human somatic cell hybrid panel and are located at chromosome region 22q11.2 by chromosomal in situ hybridization. The BCR2 and BCR4 loci are amplified in leukemia cell line K562 cells, indicating that they fall within the amplification unit that includes immunoglobulin lambda light chain locus (IGL) and ABL locus on the K562 Philadelphia chromosome (Ph1); additionally, in chronic myelogenous leukemia-derived mouse-human hybrids retaining a Ph1 chromosome in the absence of the 9q+ and normal chromosome 22, BCR2 and BCR4 loci are retained, whereas the 3' region of BCR1 and the BCR3 locus are lost, indicating that BCR3 is distal to BCR1 on chromosome 22. Similarly, in mouse-human hybrids retaining a Ph1 chromosome derived from an acute lymphoblastic leukemia-in the absence of the 9q+ and 22, only BCR2 and BCR4 loci are retained, indicating that the breakpoint in this acute lymphoblastic leukemia, as in chronic myelogenous leukemia, is proximal to the BCR1 3' region, but distal to the IGLC locus and the BCR2 and BCR4 3' loci. Thus, the order of loci on chromosome 22 is centromere----BCR2, BCR4, and IGL----BCR1----BCR3----SIS, possibly eliminating BCR2 and BCR4 loci as candidate targets for juxtaposition to the ABL gene in the acute lymphoblastic leukemia Ph1 chromosome.

The first intron in the human c-abl gene is at least 200 kilobases long and is a target for translocations in chronic myelogenous leukemia

The c-abl protooncogene is unusual in two respects; it has multiple, widely space N-terminal coding exons transcribed by different promoters, and it is the target of the translocations that form the Philadelphia chromosome found in cells of chronic myelogenous leukemia patients. To understand the organization of the gene in normal and chronic myelogenous leukemia patient DNA we have mapped c-abl by pulsed field gradient gel electrophoresis. We find that one of the alternative 5' exons of the gene lies at least 200 kilobases upstream of the remaining c-abl exons, posing formidable transcription and splicing problems. The 5'-most c-abl exon includes an unusually long 1,276-base-pair segment that contains 15 ATG codons and multiple short open reading frames, upstream of the abl initiator codon. Its peculiar structure suggests that c-abl may be decapitated in most chronic myelogenous leukemia patients, and we demonstrate that this is the case in the chronic myelogenous leukemia cell line K562.

Protooncogene abnormalities in colon cancers and adenomatous polyps

To determine the frequency and clinical significance of oncogene abnormalities in colon cancer, deoxyribonucleic acids from 45 colon carcinomas and 15 benign adenomas were hybridized with 14 different protooncogene probes. Abnormalities of oncogenes were found in 22% of cancers at the time of resection. Amplification of c-myc or c-erbB-2 and allelic deletion of c-ras-Ha or c-myb were the most frequent abnormalities. The presence of altered oncogenes did not correlate with Dukes' stage, tumor progression, or patient survival after resection. One adenoma had an allelic deletion of the c-myb oncogene which was not seen in either the normal colon or an adjacent carcinoma. These data indicate that the spectrum of altered protooncogenes in colon carcinoma is similar to that of other adenocarcinomas, and that unstable oncogenes can be found before overt malignancy develops.

Human chromosome 22

The acrocentric chromosome 22, one of the shortest human chromosomes, carries about 52 000 kb of DNA. The short arm is made up essentially of heterochromatin and, as in other acrocentric chromosomes, it contains ribosomal RNA genes. Ten identified genes have been assigned to the long arm, of which four have already been cloned and documented (the cluster of lambda immunoglobulin genes, myoglobin, the proto-oncogene c-sis, bcr). In addition, about 10 anonymous DNA segments have been cloned from chromosome 22 specific DNA libraries. About a dozen diseases, including at least four different malignancies, are related to an inherited or acquired pathology of chromosome 22. They have been characterised at the phenotypic or chromosome level or both. In chronic myelogenous leukaemia, with the Ph1 chromosome, and Burkitt's lymphoma, with the t(8;22) variant translocation, the molecular pathology is being studied at the DNA level, bridging for the first time the gap between cytogenetics and molecular genetics.

The chronic myelocytic cell line K562 contains a breakpoint in bcr and produces a chimeric bcr/c-abl transcript

In the DNAs of all Ph1-positive chronic myelocytic leukemia patients studied to date, a breakpoint on chromosome 22 (the Ph1 chromosome) can be demonstrated with a probe from the bcr (breakpoint cluster region). Although the K562 cell line was established from cells of a chronic myelocytic leukemia patient, we have been unable to detect the Ph1 chromosome by cytogenetic means. Employing a probe from the 5' region of bcr, we have cloned an amplified Ph1 breakpoint fragment from K562. This demonstrates that K562 contains multiple remnants of a Ph1 chromosome with a breakpoint within bcr and thus may serve as a model system for the study of Ph1-positive chronic myelocytic leukemia at a molecular level. The isolation of bcr cDNA sequences shows that parts of bcr encode a protein. Employing K562, we demonstrate the presence of an abnormally sized mRNA species hybridizing to c-abl and to a bcr cDNA probe, indicating the possible consequence of the Ph1 translocation on a transcriptional level in chronic myelocytic leukemia. The isolation and sequencing of a cDNA containing the breakpoint area of this mRNA provide further evidence for its chimeric structure. Cloning of large stretches of chromosomal DNA flanking bcr and c-abl sequences in K562 and identification of the exons participating in the formation of the chimeric mRNA shows that a splice of at least 99 kilobases is made to fuse the 3' bcr exon to the 5' c-abl exon. Furthermore two chimeric cDNAs were isolated containing chromosome 9 sequences that map 43.5 kilobases downstream from the K562 breakpoint. These chromosome 9 sequences neither hybridize to the 8.5-kilobase chimeric c-abl mRNA nor to normal c-abl mRNAs in Hela cells and probably represent incorrect splicing products present in the K562 cell line.

Replication of proto-oncogenes early during the S phase in mammalian cell lines

Members of several classes of proto-oncogenes replicate during the first third of S-phase in two human (K562 erythroleukemia and HeLa), one Chinese hamster (CHO) and eight mouse cell lines. These cell lines exhibit a variety of specialized functions characteristic of pre-B and B cells, T cells and erythroid cells. The proto-oncogenes studied include fos, myc, myb, abl, fes, fms, mos, raf, rel, sis, Ha-ras, Ki-ras, and N-ras. In K562 cells, amplified and rearranged c-abl genes show a pattern of temporal replication during S that is similar to the pattern observed for the 5' breakpoint cluster region (bcr) and the amplified C lambda light chain immunoglobulin genes. The c-Ki-ras related sequences in CHO cells provide one example of late replicating proto-oncogene sequences that are present in multiple copies. The cellular gene N-myc replicates late during S in some of these cell lines. In three pre-B cell lines in which N-myc specific transcripts have been detected, N-myc replicates earlier in the S phase than in the other cell lines studied here.

Carcinogen-mediated methotrexate resistance and dihydrofolate reductase amplification in Chinese hamster cells

We have investigated different parameters characterizing carcinogen-mediated enhancement of methotrexate resistance in Chinese hamster ovary (CHO) cells and in simian virus 40-transformed Chinese hamster embryo (C060) cells. We show that this enhancement reflects dihydrofolate reductase (dhfr) gene amplification. The carcinogens used in this work are alkylating agents and UV irradiation. Both types of carcinogens induce a transient enhancement of methotrexate resistance which increases gradually from the time of treatment to 72 to 96 h later and decreases thereafter. Increasing doses of carcinogens decrease cell survival and increase the enhancement of methotrexate resistance. Enhancement was observed when cells were treated at different stages in the cell cycle, and it was maximal when cells were treated during the early S phase. These studies of carcinogen-mediated dhfr gene amplification coupled with our earlier studies on viral DNA amplification in simian virus 40-transformed cells demonstrate that the same parameters characterize the amplification of both genes. Possible cellular mechanisms responsible for the carcinogen-mediated gene amplification phenomenon are discussed.

Breakpoints on chromosomes 9 and 22 in Philadelphia chromosome-positive chronic myelogenous leukemia (CML). Amplification of rearranged c-abl oncogenes in CML blast crisis

We surveyed 20 Philadelphia chromosome (Ph1) positive chronic myelogenous leukemia (CML) samples by Southern blot hybridization to determine the location of the breakpoints that occur on chromosomes 9 and 22 in the Ph1 translocation. Only 3 of 20 samples exhibited breakpoints on chromosome 9 within 18 kilobases (kb) of the v-abl homologous sequences. Mapping of these three chromosome 9 breakpoints indicates that each is at a separate location within this 18-kb region, indicating that there are no breakpoint "hot spots" in this area. In contrast, all 20 CML samples exhibited breaks on chromosome 22 within a 5.0-kb Bgl II fragment that lies within the previously described breakpoint cluster region (bcr). Several patients with CML blast crisis exhibiting multiple Ph1 chromosomes/metaphase exhibited amplified and rearranged c-abl-related fragments. These additional Ph1 chromosomes in blast crisis cells do not arise from a second, independent 9:22 translocation but rather result from a duplication of the preexisting Ph1 chromosome.

Clinical implications of oncogene activation in human neuroblastomas

Amplification of the oncogene N-myc has been identified in almost all human neuroblastoma cell lines tested. Eighty-nine primary neuroblastomas from untreated patients were studied to determine the frequency and clinical significance of N-myc amplification. Tumor DNA was analyzed by hybridization with the radiolabeled probe pNB-1 for N-myc. Amplification (3-300 copies) of the N-myc gene was found in 34 of the 89 tumors (38%). Amplification was not found in 8 Stage I or 5 Stage IV-S tumors, but it was found in 2 of 16 with Stage II, 13 of 20 with Stage III, and 19 of 40 with Stage IV tumors (P less than 0.01). Correlation of N-myc amplification with progression-free survival (PFS) indicated that N-myc amplification was associated with a worse prognosis (P less than 0.0001). The PFS at 18 months was 70%, 30%, and 5% for patients whose tumors had 1, 3-10, and more than 10 copies, respectively. Even within individual stages, the presence of N-myc amplification correlated with rapid progression. For instance, of 16 patients with Stage II disease, the 2 with N-myc amplification developed progressive disease rapidly, whereas only 1 of 14 without amplification progressed (P = 0.03). Similarly, those with Stage III and IV disease whose tumors have multiple copies of N-myc had a substantially worse prognosis. The correlation between N-myc amplification and age at diagnosis was also analyzed. Although N-myc amplification was detected in only 4 of 28 infants less than 1 year of age, compared to 30 of 61 older patients (P less than 0.005), this difference disappeared when corrected for disease stage. The results suggest that N-myc amplification is a powerful prognostic indicator, and that this gene may play an important role in the progression of certain neuroblastomas.

Variable expression of the translocated c-abl oncogene in Philadelphia-chromosome-positive B-lymphoid cell lines from chronic myelogenous leukemia patients

The consistent cytogenetic translocation of chronic myelogenous leukemia (the Philadelphia chromosome, Ph1) has been observed in cells of multiple hematopoietic lineages. This translocation creates a chimeric gene composed of breakpoint-cluster-region (bcr) sequences from chromosome 22 fused to a portion of the abl oncogene on chromosome 9. The resulting gene product (P210c-abl) resembles the transforming protein of the Abelson murine leukemia virus in its structure and tyrosine kinase activity. P210c-abl is expressed in Ph1-positive cell lines of myeloid lineage and in clinical specimens with myeloid predominance. We show here that Epstein-Barr virus-transformed B-lymphocyte lines that retain Ph1 can express P210c-abl. The level of expression in these B-cell lines is generally lower and more variable than that observed for myeloid lines. Protein expression is not related to amplification of the abl gene but to variation in the level of bcr-abl mRNA produced from a single Ph1 template.

Carcinogen-mediated methotrexate resistance and dihydrofolate reductase amplification in Chinese hamster cells

We have investigated different parameters characterizing carcinogen-mediated enhancement of methotrexate resistance in Chinese hamster ovary (CHO) cells and in simian virus 40-transformed Chinese hamster embryo (C060) cells. We show that this enhancement reflects dihydrofolate reductase (dhfr) gene amplification. The carcinogens used in this work are alkylating agents and UV irradiation. Both types of carcinogens induce a transient enhancement of methotrexate resistance which increases gradually from the time of treatment to 72 to 96 h later and decreases thereafter. Increasing doses of carcinogens decrease cell survival and increase the enhancement of methotrexate resistance. Enhancement was observed when cells were treated at different stages in the cell cycle, and it was maximal when cells were treated during the early S phase. These studies of carcinogen-mediated dhfr gene amplification coupled with our earlier studies on viral DNA amplification in simian virus 40-transformed cells demonstrate that the same parameters characterize the amplification of both genes. Possible cellular mechanisms responsible for the carcinogen-mediated gene amplification phenomenon are discussed.

Heterogeneity of chromosome 22 breakpoint in Philadelphia-positive (Ph+) acute lymphocytic leukemia

In chronic myelogenous leukemias (CML) with the t(9;22)(q34;q11) chromosome translocation the breakpoints on chromosome 22 occur within a 5.8-kilobase segment of DNA referred to as "breakpoint cluster region" (bcr). The same cytogenetically indistinguishable translocation occurs in approximately 10% of patients with acute lymphocytic leukemias (ALL). In this study we have investigated the chromosome breakpoints in several cases of ALL carrying the t(9;22) translocation. In three of five cases of ALL we found that the bcr region was not involved in the chromosome rearrangement and that the 22q11 chromosome breakpoints were proximal (5') to the bcr region at band 22q11. In addition, we observed normal size bcr and c-abl transcripts in an ALL cell line carrying the t(9;22) translocation. We conclude, therefore, that if c-abl is inappropriately expressed in ALL cells without bcr rearrangements, the genetic mechanism of activation must be different from that reported for CML.

The chronic myelocytic cell line K562 contains a breakpoint in bcr and produces a chimeric bcr/c-abl transcript

In the DNAs of all Ph1-positive chronic myelocytic leukemia patients studied to date, a breakpoint on chromosome 22 (the Ph1 chromosome) can be demonstrated with a probe from the bcr (breakpoint cluster region). Although the K562 cell line was established from cells of a chronic myelocytic leukemia patient, we have been unable to detect the Ph1 chromosome by cytogenetic means. Employing a probe from the 5' region of bcr, we have cloned an amplified Ph1 breakpoint fragment from K562. This demonstrates that K562 contains multiple remnants of a Ph1 chromosome with a breakpoint within bcr and thus may serve as a model system for the study of Ph1-positive chronic myelocytic leukemia at a molecular level. The isolation of bcr cDNA sequences shows that parts of bcr encode a protein. Employing K562, we demonstrate the presence of an abnormally sized mRNA species hybridizing to c-abl and to a bcr cDNA probe, indicating the possible consequence of the Ph1 translocation on a transcriptional level in chronic myelocytic leukemia. The isolation and sequencing of a cDNA containing the breakpoint area of this mRNA provide further evidence for its chimeric structure. Cloning of large stretches of chromosomal DNA flanking bcr and c-abl sequences in K562 and identification of the exons participating in the formation of the chimeric mRNA shows that a splice of at least 99 kilobases is made to fuse the 3' bcr exon to the 5' c-abl exon. Furthermore two chimeric cDNAs were isolated containing chromosome 9 sequences that map 43.5 kilobases downstream from the K562 breakpoint. These chromosome 9 sequences neither hybridize to the 8.5-kilobase chimeric c-abl mRNA nor to normal c-abl mRNAs in Hela cells and probably represent incorrect splicing products present in the K562 cell line.

Amplification units containing human N-myc and c-myc genes

The amplification units in human tumors containing amplified myc genes were examined. The amplification unit in all cases consisted of a large genomic region coamplified with the coding region of the myc genes themselves. In eight independent neuroblastomas containing N-myc amplifications, the amplification unit was estimated to be 290 to 430 kilobases. This amplification unit was highly conserved among the different neuroblastomas, with some neuroblastomas containing almost identical units. In contrast, five tumor cell lines containing c-myc amplifications exhibited amplification units that were more variable in size (90 to 300 kilobases) and sequence content; at least three different patterns of c-myc amplification units could be discerned.

Expression of a translocated c-abl gene in hybrids of mouse fibroblasts and chronic myelogenous leukaemia cells

Chronic myelogenous leukaemia (CML) is a clonal disease arising from malignant transformation of pluripotent hematopoietic stem cells. In most cases, it is characterized by the presence of the Philadelphia (Ph1) chromosome (22q-) which results from a reciprocal translocation between chromosomes 9 and 22 (refs 1-3). In this translocation, the human homologue of the Abelson virus oncogene, c-abl, normally on chromosome 9, is moved to chromosome 22, while c-sis, the cellular homologue of the simian sarcoma virus oncogene, is moved from chromosome 22 to chromosome 9 (refs 4-6). CML cells carrying the t(9;22) chromosomal translocation are known to produce an 8-kilobase (kb) c-abl transcript in addition to the normal 6- and 7-kb transcripts and to express the normal p145 abl protein and a p210 c-abl protein possessing a tyrosine kinase activity not detected in the p145 species. Results of our analyses using somatic cell hybrids between a mouse fibroblast line and two human CML-derived cell lines which carry the Ph1 chromosome and are phenotypically identical to the fibroblast parent indicate that only the hybrid cells containing Ph1 chromosome express both the 8-kb c-abl RNA and the p210 protein. Thus, expression of the altered c-abl transcripts and protein depends on the presence of the Ph1 chromosome and is not myeloid-specific.

Association of multiple copies of the N-myc oncogene with rapid progression of neuroblastomas

Eighty-nine patients with untreated primary neuroblastomas were studied to determine the relation between the number of copies of the N-myc oncogene and survival without disease progression. Genomic amplification (3 to 300 copies) of N-myc was detected in 2 of 16 tumors in Stage II, 13 of 20 in Stage III, and 19 of 40 in Stage IV; in contrast, 8 Stage I and 5 Stage IV-S tumors all had 1 copy of the gene (P less than 0.01). Analysis of progression-free survival in all patients revealed that amplification of N-myc was associated with the worst prognosis (P less than 0.0001); the estimated progression-free survival at 18 months was 70 per cent, 30 per cent, and 5 per cent for patients whose tumors had 1, 3 to 10, or more than 10 N-myc copies, respectively. Of 16 Stage II tumors, 2 with amplification metastasized, whereas only 1 of 14 without amplification did so (P = 0.03). Stage IV tumors with amplification progressed most rapidly: nine months after diagnosis the estimated progression-free survival was 61 per cent, 47 per cent, and 0 per cent in patients whose tumors had 1, 3 to 10, or more than 10 copies, respectively (P less than 0.0001). These results suggest that genomic amplification of N-myc may have a key role in determining the aggressiveness of neuroblastomas.

Regional mapping of two human immunoglobulin V lambda genes and analysis of the V lambda locus in chronic myeloid leukaemia

The human immunoglobulin V lambda locus has been studied in relation to chromosomal translocations involving chromosome 22. DNA probes for two V lambda genes which belong to different subgroups and do not cross hybridize, were used to show that both V lambda genes are located on the Philadelphia chromosome in chronic myeloid leukaemia (CML). Both genes map in band 22q11 to a region that is bounded on the distal side by the breakpoints for CML 9:22 translocations and on the proximal side by the breakpoint for an X:22 translocation. We have found no evidence for rearrangements or amplification of either V lambda gene in CML, in either the chronic or acute phases of the disease. In K562 cells which are derived from the pleural effusion of a patient with Ph1-positive CML, there appears to be no rearrangement of the V lambda genes, but they are both amplified about four times. We have estimated that the minimum size for the amplification unit in K562 cells is 186 kb.

Sequence rearrangements and genome instability. A possible step in carcinogenesis

A substantial part of the mammalian genome is composed of sequences that do not contain structural genes. These sequences may constitute the major target for physical, chemical and biological DNA-damaging agents and can be involved in carcinogenesis. DNA-damaging agents contribute to the instability of the genome by introducing recombination-prone sites at DNA; these agents lead to extensive chromosomal lesions and rearrangements of genes and their regulatory sequences. Movable sequences that exist and operate in certain bacteria, yeast, and the fruit fly are responsible for sequence rearrangements and contribute to the majority of mutations. Their presence and role in higher animals is not well established. Extensive chromosomal rearrangements were identified in numerous malignancies in man and animals and definitely seem to represent a characteristic of malignancy. Vast chromosomal damage and sequence reshuffling may be of no less importance in the malignant transformation than the point mutation of a particular gene.

"Masked" Ph1-chromosome in chronic myelogenous leukaemia (CML)

The CML patients with so called masked Ph1-chromosome have been reviewed. Although the importance of c-sis and c-abl oncogenes is gaining popularity yet their role in the genesis of CML remain obscure. Patients with masked Ph1-chromosomes where chromosome 9 is not involved in the translocation(s) will provide a clue to the role of c-abl and/or c-sis in oncogenesis.

Analysis of gene expression during hematopoiesis: present and future applications

Recombinant DNA technology now provides the strategies required to identify genes whose expression controls the development of normal and pathologic blood cells. Characterization of the gene families responsible for synthesis of hemoglobins, immunoglobulins, histocompatibility antigens, and cellular enzymes have already, or are about to, provide major insights into the mechanisms producing normal erythroid cells, immunocytes, and immune surface features. Hemoglobinopathies, leukemias, and autoimmune diseases of the bone marrow can now be examined to a degree of detail previously inaccessible to investigators. Oncogene translocation analysis is shedding new light on the pathogenesis of leukemias and lymphomas. Recent basic advances now permit direct cloning and identification of genes in host organisms which express their protein products, thus allowing isolation of genes coding for the hematopoietic surface markers and growth factors which characterize and regulate blood cell progenitors. This review summarizes the molecular genetic approach to analysis of normal and pathologic hematopoiesis, surveys major findings which have resulted, and examines the potential use of refined gene cloning strategies for improved understanding of blood cell development.