The ISWI chromatin-remodeling protein is required for gene expression and the maintenance of higher order chromatin structure in vivo.

Drosophila ISWI, a highly conserved member of the SWI2/SNF2 family of ATPases, is the catalytic subunit of three chromatin-remodeling complexes: NURF, CHRAC, and ACF. To clarify the biological functions of ISWI, we generated and characterized null and dominant-negative ISWI mutations. We found that ISWI mutations affect both cell viability and gene expression during Drosophila development. ISWI mutations also cause striking alterations in the structure of the male X chromosome. The ISWI protein does not colocalize with RNA Pol II on salivary gland polytene chromosomes, suggesting a possible role for ISWI in transcriptional repression. These findings reveal novel functions for the ISWI ATPase and underscore its importance in chromatin remodeling in vivo.

Two histone fold proteins, CHRAC-14 and CHRAC-16, are developmentally regulated subunits of chromatin accessibility complex (CHRAC).

The ISWI ATPase of Drosophila is a molecular engine that can drive a range of nucleosome remodelling reactions in vitro. ISWI is important for cell viability, developmental gene expression and chromosome structure. It interacts with other proteins to form several distinct nucleosome remodelling machines. The chromatin accessibility complex (CHRAC) is a biochemical entity containing ISWI in association with several other proteins. Here we report on the identification of the two smallest CHRAC subunits, CHRAC-14 and CHRAC-16. They contain histone fold domains most closely related to those found in sequence-specific transcription factors NF-YB and NF-YC, respectively. CHRAC-14 and CHRAC-16 interact directly with each other as well as with ISWI, and are associated with functionally active CHRAC. The developmental expression profiles of both subunits suggest specialized roles in chromatin remodelling reactions in the early embryo for both histone fold subunits.

The Drosophila kismet gene is related to chromatin-remodeling factors and is required for both segmentation and segment identity.

The Drosophila kismet gene was identified in a screen for dominant suppressors of Polycomb, a repressor of homeotic genes. Here we show that kismet mutations suppress the Polycomb mutant phenotype by blocking the ectopic transcription of homeotic genes. Loss of zygotic kismet function causes homeotic transformations similar to those associated with loss-of-function mutations in the homeotic genes Sex combs reduced and Abdominal-B. kismet is also required for proper larval body segmentation. Loss of maternal kismet function causes segmentation defects similar to those caused by mutations in the pair-rule gene even-skipped. The kismet gene encodes several large nuclear proteins that are ubiquitously expressed along the anterior-posterior axis. The Kismet proteins contain a domain conserved in the trithorax group protein Brahma and related chromatin-remodeling factors, providing further evidence that alterations in chromatin structure are required to maintain the spatially restricted patterns of homeotic gene transcription.

Genetic analysis of brahma: the Drosophila homolog of the yeast chromatin remodeling factor SWI2/SNF2.

The Drosophila brahma (brm) gene encodes an activator of homeotic genes related to the yeast chromatin remodeling factor SWI2/SNF2. Here, we report the phenotype of null and dominant-negative brm mutations. Using mosaic analysis, we found that the complete loss of brm function decreases cell viability and causes defects in the peripheral nervous system of the adult. A dominant-negative brm mutation was generated by replacing a conserved lysine in the ATP-binding site of the BRM protein with an arginine. This mutation eliminates brm function in vivo but does not affect assembly of the 2-MD BRM complex. Expression of the dominant-negative BRM protein caused peripheral nervous system defects, homeotic transformations, and decreased viability. Consistent with these findings, the BRM protein is expressed at relatively high levels in nuclei throughout the developing organism. Site-directed mutagenesis was used to investigate the functions of conserved regions of the BRM protein. Domain II is essential for brm function and is required for the assembly or stability of the BRM complex. In spite of its conservation in numerous eukaryotic regulatory proteins, the deletion of the bromodomain of the BRM protein has no discernible phenotype.

Identification and characterization of Drosophila relatives of the yeast transcriptional activator SNF2/SWI2.

The Drosophila brahma (brm) gene encodes an activator of homeotic genes that is highly related to the yeast transcriptional activator SWI2 (SNF2), a potential helicase. To determine whether brm is a functional homolog of SWI2 or merely a member of a family of SWI2-related genes, we searched for additional Drosophila genes related to SWI2 and examined their function in yeast cells. In addition to brm, we identified one other Drosophila relative of SWI2: the closely related ISWI gene. The 1,027-residue ISWI protein contains the DNA-dependent ATPase domain characteristic of the SWI2 protein family but lacks the three other domains common to brm and SWI2. In contrast, the ISWI protein is highly related (70% identical) to the human hSNF2L protein over its entire length, suggesting that they may be functional homologs. The DNA-dependent ATPase domains of brm and SWI2, but not ISWI, are functionally interchangeable; a chimeric SWI2-brm protein partially rescued the slow growth of swi2- cells and supported transcriptional activation mediated by the glucocorticoid receptor in vivo in yeast cells. These findings indicate that brm is the closest Drosophila relative of SWI2 and suggest that brm and SWI2 play similar roles in transcriptional activation.

brahma: a regulator of Drosophila homeotic genes structurally related to the yeast transcriptional activator SNF2/SWI2.

The brahma (brm) gene is required for the activation of multiple homeotic genes in Drosophila. Loss-of-function brm mutations suppress mutations in Polycomb, a repressor of homeotic genes, and cause developmental defects similar to those arising from insufficient expression of the homeotic genes of the Antennapedia and Bithorax complexes. The brm gene encodes a 1638 residue protein that is similar to SNF2/SWI2, a protein involved in transcriptional activation in yeast, suggesting possible models for the role of brm in the transcriptional activation of homeotic genes. In addition, both brm and SNF2 contain a 77 amino acid motif that is found in other Drosophila, yeast, and human regulatory proteins and may be characteristic of a new family of regulatory proteins.

Interacting proteins identified by genetic interactions: a missense mutation in alpha-tubulin fails to complement alleles of the testis-specific beta-tubulin gene of Drosophila melanogaster.

In this paper we demonstrate that failure to complement between mutations at separate loci can be used to identify genes that encode interacting structural proteins. A mutation (nc33) identified because it failed to complement mutant alleles of the gene encoding the testis-specific beta 2-tubulin of Drosophila melanogaster (B2t) did not map to the B2t locus. We show that this second-site noncomplementing mutation is a missense mutation in alpha-tubulin that results in substitution of methionine in place of valine at amino acid 177. Because alpha- and beta-tubulin form a heterodimer, our results suggest that the genetic interaction, failure to complement, is based on the structural interaction between the protein products of the two genes. Although the nc33 mutation failed to complement a null allele of B2t (B2tn), a deletion of the alpha-tubulin gene to which nc33 mapped complemented B2tn. Thus, the failure to complement appears to require the presence of the altered alpha-tubulin encoded by the nc33 allele, which may act as a structural poison when incorporated into either the tubulin heterodimer or microtubules.

Proof of recombination between viral and cellular genomes in human KB cells productively infected by adenovirus type 12: structure of the junction site in a symmetric recombinant (SYREC).

In previous work we have described a symmetric recombinant (SYREC1) between Ad12 DNA and human KB cell DNA. This recombinant DNA molecule has been generated during productive infection and is encapsidated into virions. From the DNA of a similar symmetric recombinant (termed SYREC2) between the left terminus of Ad12 DNA and human KB cellular DNA, the site of linkage between the two DNAs was cloned and sequenced. It was demonstrated that the first 2081 Ad12 nucleotides counting from the left viral terminus are conserved and linked to a sequence of GC-rich (70.4% G + C) KB cell DNA which occurs about 20 times per cellular genome. Except for a common 5'-CTGGC-3' pentanucleotide between the Ad12 DNA and KB cell DNA sequences, extensive patch homologies were not apparent at the site of junction. Similarly, comparisons of the deleted Ad12 DNA sequence and the cellular sequence replacing it did not reveal patch homologies. The 304 bp abutting the Ad12 terminus were shown to hybridize to KB cell DNA. These results provided definitive proof for the occurrence of recombinants between viral and cellular DNAs in human cells productively infected by Ad12 as previously shown by less direct experiments (Burger and Doerfler, 1974; Schick et al., 1976). Across the site of junction, an open reading frame exists which extends the truncated 54-kDal protein of the E1b region of Ad12 DNA for another 66 amino acids encoded by KB cellular DNA. This sequence is terminated by two UGA translational termination signals. The hypothetical protein has not yet been isolated.

An unusual symmetric recombinant between adenovirus type 12 DNA and human cell DNA.

On purification of human adenovirus type 12 (Ad12) by equilibrium sedimentation in CsCl density gradients, two bands of particles, Ad12-3 and Ad12-3a, are observed. The particles from band Ad12-3a contain a recombinant of human host cell DNA and of Ad12 DNA. The human cell DNA sequences contain repetitive DNA recurring 200 to 500 times in cellular DNA. Ad12 DNA and the recombinant genomes exhibit the same or similar lengths. This finding suggests that a constant amount of DNA is packaged into complete Ad12 particles. On cleavage of KB cellular DNA with EcoRI, BamHI, HinfI, Msp I, Mbo I Pst I, or Bgl II, the (32)P-labeled cellular DNA from Ad12-3a particles hybridizes on Southern blots to distinct bands of KB DNA. There is also less-specific background hybridization that is not observed in the control. The cellular DNA from Ad12-3a particles is not methylated, whereas the same cellular sequences in KB cell DNA appear to be extensively methylated. On denaturation and renaturation, the recombinant DNA molecules are converted to molecules half as long as Ad12 DNA, as determined by gel electrophoresis and electron microscopy. The recombinant DNA molecules were terminally labeled by exonuclease III treatment and subsequent refilling of the depleted segments with [(32)P]dNTPs by using DNA polymerase I (Klenow fragment). When these molecules were cleaved with EcoRI, BamHI, Msp I, or Pst I, only one terminal DNA fragment was found to be labeled. The results of partial digestion experiments using Msp I, HinfI, or Mbo I are consistent with a model in which 700-1150 base pairs from the left terminus of Ad12 DNA are linked to host cell DNA containing repetitious sequences, and this structure is symmetrically duplicated as a large inverted repeat of the type ABCDD'C'B'A'. The Ad12 DNA sequences are flanking the entire molecule, which consists mainly of human KB cell DNA. The recombinants appear to be stable on serial passage of the virus preparation for many years, although variations in the sequence of the recombinants occur. These symmetric recombinant (SYREC) molecules suggest a way to use adenovirus DNA as a eukaryotic vector. Their occurrence provides further evidence for the generation of virus-host DNA recombinants and may help elucidate the role this interaction may have in adenovirus replication and oncogenesis.

Viruses as tools for studies on the molecular biology of mammalian cells.

1. Multiple copies of intact adenovirus type 12 (Ad12) DNA are integrated into the DNA of Ad12-transformed hamster and Ad12-induced rat brain tumor cells. Free viral DNA is not present in the Ad12-transformed lines investigated. 2. Only few sites of integration are found in Al2-transformed hamster and Ad12-induced rat brain tumor cells. Integration may have occurred into repetitive cellular sequences at selective sites. These sites may be different in different cell lines, however, none of these sites has been analyzed in sufficient detail. 3. Three lines of Ad12-induced rat brain tumor cells exhibit identical patterns of integration. These lines have been derived from three brain tumors in one animal and may have evolved from the same transformed cell. 4. In Ad12-induced rat brain tumor cells early and late segments of the viral genome are expressed as polysome-associated messenger RNA. 5. In human cells productively infected with adenovirus type 2 (Ad2), a large number of viral genome copies are linked to cellular DNA early postinfection. There are only a few sites of recombination (illegitimate?) which probably lie in repetitive cellular DNA sequences. The functional significance of this frequent recombination is unknown.