aHIF: a natural antisense transcript overexpressed in human renal cancer and during hypoxia.

BACKGROUND: Nonpapillary renal carcinoma is the predominant form of human kidney cancer and represents a distinct disease entity, morphologically and molecularly, from papillary renal carcinoma. We have discovered a natural antisense transcript that is complementary to the 3' untranslated region of hypoxia inducible factor alpha (HIF1alpha) messenger RNA (mRNA) and is strikingly overexpressed specifically in nonpapillary kidney cancer. HIF1alpha encodes a protein that is known to have two important functions: 1) to act as a transcription factor for hypoxia inducible genes and 2) to stabilize p53 protein during hypoxia. Because of the importance of HIF1alpha, we have characterized this natural antisense transcript, which we have named "aHIF." METHODS: Differential display, reverse transcription-polymerase chain reaction, ribonuclease protection, and DNA-sequencing methods were used in our analysis. RESULTS AND CONCLUSIONS: We show the following: 1) aHIF is a natural antisense transcript derived from HIF1alpha gene sequences encoding the 3' untranslated region of HIF1alpha mRNA; 2) aHIF is specifically overexpressed in all nonpapillary clear-cell renal carcinomas examined, but not in the papillary renal carcinomas examined; 3) aHIF is overexpressed in an established nonpapillary renal carcinoma cell line under both normoxic (i.e., normal aerobic) and hypoxic conditions; and 4) although aHIF is not further induced by hypoxia in nonpapillary disease, it can be induced in lymphocytes where there is a concomitant decrease in HIF1alpha mRNA. To our knowledge, this is the first case of overexpression of a natural antisense transcript exclusively associated with a specific human malignant disease.

CHD1 is concentrated in interbands and puffed regions of Drosophila polytene chromosomes.

Previously, we reported on the discovery and characterization of a mammalian chromatin-associated protein, CHD1 (chromo-ATPase/helicase-DNA-binding domain), with features that led us to suspect that it might have an important role in the modification of chromatin structure. We now report on the characterization of the Drosophila melanogaster CHD1 homologue (dCHD1) and its localization on polytene chromosomes. A set of overlapping cDNAs encodes an 1883-aa open reading frame that is 50% identical and 68% similar to the mouse CHD1 sequence, including conservation of the three signature domains for which the protein was named. When the chromo and ATPase/helicase domain sequences in various CHD1 homologues were compared with the corresponding sequences in other proteins, certain distinctive features of the CHD1 chromo and ATPase/helicase domains were revealed. The dCHD1 gene was mapped to position 23C-24A on chromosome 2L. Western blot analyses with antibodies raised against a dCHD1 fusion protein specifically recognized an approximately 210-kDa protein in nuclear extracts from Drosophila embryos and cultured cells. Most interestingly, these antibodies revealed that dCHD1 localizes to sites of extended chromatin (interbands) and regions associated with high transcriptional activity (puffs) on polytene chromosomes from salivary glands of third instar larvae. These observations strongly support the idea that CHD1 functions to alter chromatin structure in a way that facilitates gene expression.

Loss of heterozygosity studies indicate that chromosome arm 1p harbors a tumor supressor gene for renal oncocytomas.

We carried out a complete genome scan for loss of heterozygosity (LOH) in four renal oncocytomas by using highly polymorphic CA repeat microsatellite loci. Three of the four tumors exhibited LOH for chromosome arm 1p, and the oncocytomas of both female patients lost Xq. Therefore, these chromosome arms may harbor tumor suppressor genes involved in the etiology of this disease. Although the genomes of ontocytomas are relatively stable, two different microsatellite loci in one tumor were mutated by + or - 2 nt. Similar alterations in CA repeats that are probably due to spontaneous mutation have been observed in renal cell carcinomas.

Genomic alterations and instabilities in renal cell carcinomas and their relationship to tumor pathology.

A comprehensive genome scan for loss of heterozygosity (LOH) in 33 renal cell carcinomas indicates that mutations of tumor suppressor genes on several different chromosomes are required for malignant transformation in this disease. In the case of nonpapillary renal carcinomas chromosomes 3p, 6q, 8p, 9pq, and 14q exhibit elevated levels of LOH. Although 3p is the most frequently lost chromosome arm, in no case is 3p observed as the sole allelic loss because it always occurs in conjunction with the loss of either 6q, 8p, or 14q. This result indicates that the mutation of a tumor suppressor gene on 3p, most likely von Hippel-Lindau disease (VHL), may be necessary but is not sufficient for the development of nonpapillary renal cell carcinoma. In papillary renal tumors, LOH is observed most often for chromosomes 6pq, 9p, 11q, 14q, and 21q. This suggests that tumor suppressor genes located on chromosomes 6q, 9pq, and 14q may be involved in the development and/or progression of both nonpapillary and papillary renal cell carcinomas. However, LOH in papillary tumors appears to be especially elevated for 11q and 21q and reduced for 3p and 8p indicating that there are also tumor suppressor genes specific to each form of the disease. There is no correlation between stage of disease and the extent of LOH, loss of a particular chromosome, or the number of chromosomes that show allele imbalance. Early and late stage tumors may exhibit either extensive LOH or no apparent allele loss; similarly, allelic imbalances are observed in both early and late stage renal cell carcinomas. This suggests that a gene (or genes) regulating mitotic chromosome stability may be mutated in some renal tumors. Preliminary evidence points to an association between genome instability and LOH of 14q. Finally, a distinct type of microsatellite instability has been detected in 21% of renal cell carcinomas and occurs at a frequency of 4.4 x 10(-4)/locus. The most common mutation is a 2-bp insertion in a CA repeat. This alteration is not restricted to a particular histopathology or clinical stage, and it is not associated with allelic loss of a specific chromosome. The frequency of this event is similar to that which occurs spontaneously in germline microsatellite loci and is probably not the result of a defect in a mismatch repair gene. It is possible that this type of microsatellite instability is general and may occur in most, if not all, carcinomas.

Comprehensive allelotyping of human renal cell carcinomas using microsatellite DNA probes.

The von Hippel-Lindau locus on chromosome 3p is a tumor suppressor gene known to be involved in nonpapillary renal cell carcinoma. A previous loss of heterozygosity (LOH) study aimed at determining the allelotype of kidney tumors has indicated that in addition to 3p, chromosome arms 5q, 6q, 10q, 11q, 17p, and 19p may also harbor tumor suppressor genes. However, cytogenetic studies reveal that chromosomes 3p, 6q, 8p, 9pq, and 14q most frequently undergo karyotypic changes in renal tumors. To resolve these differences, a collection of microsatellite DNA probes has been used to scan for LOH so that 90% of individual tumor genomes were rendered informative for allele loss. The assay is capable of detecting quantitative genomic alterations in tumor cells as well. We find that LOH is most frequent for chromosome arm 3p. However, in no tumor is 3p exclusively affected. LOH for 6q, 8p, 9pq, and 14q is also distinctly elevated for both nonpapillary as well as papillary tumors and suggest that many of the tumor suppressor loci involved may be common to the etiology of both forms of kidney cancer.

Position effect variegation in yeast.

Classically, position effect variegation has been studied in Drosophila and results when a euchromatic gene is placed adjacent to either centromeric heterochromatin or to a telomeric domain. In such a circumstance expression of the locus variegates, being active in some cells and silent in others. Over the last few years a comparable phenomenon in yeast has been discovered. This system promises to tell us much about this curious behavior. Indeed, experiments reported recently(1) indicate that the variegation of a yeast telomeric gene is cell-cycle regulated. The results suggest the following model. During DNA replication there is a disassembly of chromatin that allows a competition between silencing factors and transactivators to take place. Thus, reassembly of the domain may result in either the repression or the expression of the affected gene and, hence, produce a variegating phenotype.

Molecular analysis of cubitus interruptus (ci) mutations suggests an explanation for the unusual ci position effects.

The cubitus interruptus (ci) locus of Drosophila melanogaster is located proximally on chromosome 4. In ci mutants cubital wing veins are interrupted or absent. We have cloned this locus using a gypsy element associated with the ci1 mutation. Analysis of all extant ci mutations reveals that they contain conspicuous molecular alterations within a 13.7 kb region. Of the four homozygous viable mutations, three (ci1, ci361, ciw) have single insertions, while one (ci57g) has a small deletion, all located within a more restricted 1 kb region. The dominant mutations, ciD and Ce2 each contain two insertions within the 13.7 kb region. All these molecular alterations are located upstream of a transcript previously associated with the ciD mutation and thought to derive from a segment polarity gene. We induced revertants of the dominant ci phenotype (wing vein interruption) in ciD and found molecular alterations in this transcript (the ci+ transcript) in two revertant alleles, thereby demonstrating this transcript's involvement in the ci phenotype. The locations of the molecular alterations, together with the results of the ciD reversion experiment, provide a connection between the dominant and recessive ci mutations and argue that all are likely to be alleles of the same complex locus, ci, not two separate loci as previously proposed. The ci phenotype of dominant and recessive mutations can be explained by inappropriate expression of the ci+ transcript in the posterior wing compartment where the cubital vein is affected, while loss of ci+ function generates recessive lethality. Lack of repression of ci+ transcription, through a pairing-dependent, trans-acting silencer element, can explain the unusual position effects associated with ci (the Dubinin effect).

The SH2-like Akt homology (AH) domain of c-akt is present in multiple copies in the genome of vertebrate and invertebrate eucaryotes. Cloning and characterization of the Drosophila melanogaster c-akt homolog Dakt1.

The Akt proto-oncogene encodes a serine-threonine protein kinase whose carboxyterminal catalytic domain is closely related to the catalytic domains of all the known members of the protein kinase C (PKC) family. Akt, however, differs from PKC in its N-terminal region which contains a domain related distantly to the SH2 domain of cytoplasmic tyrosine kinases and other signalling proteins, which we have named Akt homology (AH) domain. Low stringency hybridization of a c-akt AH probe to a panel of genomic DNAs from vertebrate and invertebrate eucaryotes detected multiple DNA bands (perhaps multiple genes) in all tested species. Drosophila DNA contains at least three hybridizing DNA bands. One of them was cloned, and found by sequence analysis, to define an Akt related gene (Dakt1). Comparison between the coding regions of c-akt and Dakt1 revealed 64.6% identity at the nucleotide level and 76.5% similarity at the amino acid level. The highest degree of homology was detected in the AH domain (68.3% similarity at the amino acid level) and the catalytic domain (83.3% similarity). Additional sequence comparisons revealed that the amino acid similarity between the catalytic domains of Dkt1 and the three known members of the Drosophila protein kinase C (PKC) family, Dpkc1, Dpkc2 and Dpkc3, is 68%, 63.6% and 67.1%, respectively. Dakt1 was mapped to Drosophila chromosome 3R 89BC. Its expression is subject to developmental regulation with the highest levels detected within the fourth hour of embryonic development. These results confirm that the AH domain of Akt defines new protein families in both vertebrate and invertebrate eucaryotes. The high degree of homology between the catalytic domains of Dkt1 and the three known members of the Drosophila PKC family suggests an evolutionarily conserved functional relationship between the members of the two families.

The Exo-gap method employing the phage f1 endonuclease generates a nested set of unidirectional deletions.

An Exo-gap method for producing a nested set of unidirectional deletions in a piece of cloned DNA is described. The protein (pII) encoded by gene II of phage f1 makes a single-stranded (ss) nick at the f1 origin of replication (ori) in supercoiled DNA. Many phagemids, such as pBluescriptSK+ contain this ori on one side of the multiple cloning site, thereby permitting purified pII endonuclease to create a nick at one end of a cloned DNA insert. The nick may be expanded into gaps of increasing size by the timed 3' to 5' exonuclease (Exo) activity of the Vent DNA polymerase. Double-stranded deletions are produced by subsequent treatment with ss-specific mung bean nuclease. After size fractionation by agarose-gel electrophoresis, the DNA from the melted gel slices is ligated and transfected into host cells to produce a set of plasmids that contain a unidirectional nested set of deletions. This deletion method is independent of restriction sites, requires only one universal DNA primer to sequence a cloned insert, and may be applied to virtually any cloned segment given the unique nature of the 46-bp recognition site for pII endonuclease.

Spontaneous and radiation-induced renal tumors in the Eker rat model of dominantly inherited cancer.

Hereditary renal carcinoma (RC) in the rat, originally reported by R. Eker in 1954, is an example of a Mendelian dominant predisposition to a specific cancer in an experimental animal. At the histologic level, RCs develop through multiple stages from early preneoplastic lesions (e.g., atypical tubules) to adenomas in virtually all heterozygotes by the age of 1 year. The homozygous mutant condition is lethal at approximately 10 days of fetal life. Ionizing radiation induces additional tumors in a linear dose-response relationship, suggesting that in heterozygotes two events (one inherited, one somatic) are necessary to produce tumors, and that the predisposing gene is a tumor suppressor gene. No genetic linkage has yet been found between the Eker mutation and rat DNA sequences homologous to those in human chromosome 3p, the presumed site of the putative tumor suppressor gene responsible for human RC. Nonrandom loss of rat chromosome 5 in RC-derived cell lines is sometimes associated with homozygous deletion of the interferon gene loci at rat chromosome bands 5q31-q33. Since this locus is not linked with the predisposing inherited gene in the Eker rat, it probably represents a second tumor suppressor gene involved in tumor progression.

A superfamily of Drosophila satellite related (SR) DNA repeats restricted to the X chromosome euchromatin.

The 1.688 g/cm3 class of Drosophila satellite DNA is predominantly localized to the centromeric heterochromatin of the X chromosome. We report here the existence of 1.688 satellite related (SR) DNA arrays present at numerous locations throughout the euchromatic portion of the X. Unlike their heterochromatic counterparts, euchromatic SRs consist of a small number of repeating units (usually 2-4), each of which is 63-81% identical to the 359-bp monomer of the 1.688 satellite. Although it appears that SR DNA arrays are not transcribed, in at least two cases, they are located adjacent to transcriptionally active genes. SR sequences also have significant similarity to a previously described Drosophila middle repeat found almost exclusively in the X euchromatin. It seems likely that these X linked sequences are required for sex chromosome specific functions.

Mechanisms for the construction and developmental control of heterochromatin formation and imprinted chromosome domains.

The study of variegating position effects in Drosophila provides a model system to explore the mechanism and material basis for the construction and developmental control of heterochromatin domains and the imprinted genomic structures that they may create. The results of our experiments in this regard have implications for a diverse assortment of long-range chromosome phenomena related to gene and chromosome inactivation. Specifically, as a consequence of our studies on position effect variegation, we propose a simple mechanism of X-chromosome inactivation, suggest a purpose for genomic imprinting, and postulate a general means for regulating the time in development at which certain genes become heterochromatically repressed.

Towards an understanding of position effect variegation.

Most variegating position effects are a consequence of placing a euchromatic gene adjacent to alpha-heterochromatin. In such rearrangements, the affected locus is inactivated in some cells, but not others, thereby giving rise to a mosaic tissue of mutant and wild-type cells. A detailed examination of the molecular structure of three variegating white mottled mutations of Drosophila melanogaster, all of which are inversions of the X chromosome, reveals that their euchromatic breakpoints are clustered and located approximately 25 kb downstream of the white promoter and that the heterochromatic sequences to which the white locus is adjoined are transposons. An analysis of three revertants of the wm4 mutation, created by relocating white to another euchromatic site, demonstrates that they also carry some heterochromatically derived sequences with them upon restoration of the wild-type phenotype. This suggests that variegation is not controlled from a heterochromatic sequence immediately adjacent to the variegating gene but rather from some site more internal to the heterochromatic domain itself. As a consequence of this observation we have proposed a boundary model for understanding how heterochromatic domains may be formed. It has been recognized for many years that the phenotype of variegating position effects may be altered by the presence of trans-acting dominant mutations that act to either enhance or suppress variegation. Using P-element mutagenesis, we have induced and examined 12 dominant enhancers of variegation that represent four loci on the second and third chromosomes. Most of these mutations are cytologically visible duplications or deficiencies. They exert their dominant effects through changes in the copy number of wild-type genes and can be divided into two reciprocally acting classes. Class I modifiers are genes that act as enhancers of variegation when duplicated and as suppressors when mutated or deficient. Conversely, class II modifiers are genes that enhance when mutated or deleted and suppress when duplicated. The available data indicate that, in Drosophila, there are 20-30 loci capable of dominantly modifying variegation. Of these, most appear to be of the class I type whereas only two class II modifiers have been identified so far.(ABSTRACT TRUNCATED AT 400 WORDS)

Dosage-dependent modifiers of position effect variegation in Drosophila and a mass action model that explains their effect.

Twelve dominant enhancers of position effect variegation, representing four loci on the second and third chromosomes of Drosophila melanogaster, have been induced by P-element mutagenesis. Instead of simple transposon insertions, seven of these mutations are cytologically visible duplications and three are deficiencies. The duplications define two distinct regions, each coinciding with a locus that also behaves as a dominant haplo-dependent suppressor of variegation. Conversely, two of the deficiencies overlap with a region that contains a haplo-dependent enhancer of variegation while duplications of this same region act to suppress variegation. The third deficiency defines another haplo-dependent enhancer. These data indicate that loci capable of modifying variegation do so in an antipodal fashion through changes in the wild-type gene copy number and may be divided into two reciprocally acting classes. Class I modifiers enhance variegation when duplicated or suppress variegation when deficient. Class II modifiers enhance when deficient but suppress when duplicated. From our data, and those of others, we propose that in Drosophila there are about 20 to 30 dominant loci that modify variegation. Most appear to be of the class I type whereas only two class II modifiers have been identified so far. From these observations we put forth a model, based on the law of mass action, for understanding how such suppressor-enhancer loci function. We propose that each class I modifier codes for a structural protein component of heterochromatin and their effects on variegation are a consequence of their dosage dependent influence on the extent of the assembly of heterochromatin at the chromosomal site of the position effect. It is further proposed that class II modifiers may inhibit the class I products directly, bind to hypothetical termination sites that define heterochromatin boundaries or promote euchromatin formation. Consistent with our mass action model we find that combining two enhancers together produce additive and not epistatic effects. Also, since different enhancers have different relative strengths on different variegating mutants, we suggest that heterochromatic domains are constructed by a combinatorial association of proteins. The mass action model proposed here is of general significance for any assembly driven reaction and has implications for understanding a wide variety of biological phenomena.(ABSTRACT TRUNCATED AT 400 WORDS)

New cloning vectors and techniques for easy and rapid restriction mapping.

We have modified plasmid, phage lambda and cosmid cloning vectors to be of general use for easily and unambiguously determining restriction maps of recombinant DNA molecules. Each vector is constructed so that it contains the rarely found NotI restriction site joined to a short synthetic linker sequence that is followed by a multiple cloning site. DNA cloned into these vectors may be restriction-mapped by either of two methods. In one technique, the cloned DNA is completely digested with NotI, followed by partial digestion with any other restriction enzyme. After electrophoresis and transfer to a nylon membrane, the fragments are hybridized to a labeled probe complementary to the NotI linker. In the second technique, referred to as recession hybridization detection, cloned DNA is digested with NotI and then briefly treated with exonuclease III to recess the 3' ends. After hybridizing a labeled complementary oligodeoxynucleotide to the single-stranded 5' end containing the linker sequence, the DNA is partially digested with another restriction enzyme, electrophoresed and the gel is exposed to x-ray film. With either method the size of each labeled fragment corresponds directly to the distance that a restriction site is located from the NotI linker terminus. Methods for obtaining partial restriction enzyme digests have been devised so that as many as 20 different enzymes may be conveniently mapped on a single gel in little more than a day. The vectors and techniques described may also be adapted to automated or semi-automated devices that read fragment lengths and calculate the resulting restriction map.(ABSTRACT TRUNCATED AT 250 WORDS)

A two-stage model for the control of rDNA magnification.

Males of the genotype bb/Ybb- have been shown to produce both magnified (bbm+) and, less frequently, reduced (bbrl) X chromosomes. An analysis of the progeny of single magnifying bb/Ybb- males reveals that bbm+ revertants may be recovered either as rare single events or, more frequently, in large clusters. To analyze the role of the bb phenotype in the induction of rDNA magnification we have constructed a series of bb and bb+ derivatives of Ybb-. Males carrying an X chromosomal bb allele and one of these derivatives (bb/bbYbb- or bb/bb+Ybb-) produce small numbers (one to two) of bbm+ progeny at a frequency similar to that observed for bb/Ybb- males but do not produce large clusters of bbm+ revertants. In addition, bb/bb+Ybb- males produce essentially equal numbers of magnified (bbm+) and reduced (bbrl) X chromosomes. These data, together with a consideration of the growth properties of the male germline in Drosophila, suggest that magnification/reduction may occur at two different times during development. Those events that give rise to large clusters, and, thus, necessarily arise early in germ cell development, appear to be dependent on the bb phenotype. However, those events that give rise to single bbm+ chromosomes arise late in spermatogenesis, probably at meiosis, and are independent of the bb phenotype.

A structural basis for variegating position effects.

Variegating position effects in Drosophila result from chromosome rearrangements where normal genes, having been placed next to heterochromatin, are inactivated in some cells but not in others, thereby producing a variegated tissue. We have determined that the euchromatic breakpoints for three variegating white mutants are clustered and lie approximately 25 kb downstream of the white structural gene. In each case the white locus is adjoined in the heterochromatin to a mobile genetic element. Satellite sequences are not involved. We also demonstrate that revertants of the variegating mutant, wm4, are reinversions that leave the initial wm4-heterochromatic junction intact so that some heterochromatin-derived sequences remain joined to white at its new location. These results suggest a simple model for understanding the structure of heterochromatic domains and how variegating position effects may arise.