BIOLOGICAL OVERVIEW

The SWI/SNF complex of yeast is currently a hot topic because of its ability to activate transcription from DNA covered with histones. SWI/SNF carries out a remodeling process that shifts the histones, giving transcription factors access to their promoter binding sites. Snr1 was isolated on the basis of homology to ini1, a distant human relative of the yeast SNF5 gene (Kalpana, 1994). The INI1 protein activates transcription when it is tethered to DNA via a GAL4 DNA-binding domain, suggesting that INI1 may be involved in transcriptional activation (Kalapana, 1994).

SNR1 shows a high level of homology to INI1 (65% overall). The region of strongest homology, the C-terminal, is sufficient for INI1 interaction with HIV integrase. Both SNR1 and INI1 contain highly charged regions like SNF5, with a strongly acidic C-terminal core. SNF5 contains two regions not found in SNR1 and INI1, including a polyglutamine region at the N terminus of SNF5 and a proline-rich region near the C terminus. SNR1 is found in a high molecular weight complex with the Brahma protein (Tsukiyama, 1995).

Loss-of-function and null mutations in the snr1 gene reveal its essential role in Drosophila development. Mew mutant alleles have been identified, and deleted forms have been ectopically expressed to dissect the specific functions of Snr1. Somatic and germ cell clone analyses confirm its requirement in a continuous and widespread fashion for proper cell fate determination and oogenesis. Expression of Snr1 transgenes reveals unexpected roles in wing patterning, abdomen development, oogenesis, and sustained adult viability. A widespread distribution of Snr1 and BRM on the salivary gland polytene chromosomes shows that the Brm complex associates with many genes, but not always at transcribed loci, consistent with genetic data suggesting roles in both gene activation and repression. Despite essential Brm complex functions in leg development, genetic and protein localization studies have revealed that snr1 is not required or expressed in all tissues dependent on Brm complex activities. Thus, Snr1 is essential for some, but not all Brm functions, and it likely serves as an optional subunit, directing Brm complex activity to specific gene loci or cellular processes (Zraly, 2003).

While the biological functions of Snr1 and its mammalian counterpart INI1 are complex, both are essential for development. Specifically, genetic analyses of both snr1 null and conditional mutants and disrupting Snr1 function with dominant negative transgenes have shown that Snr1 is required at all developmental stages and in adults, though in a restricted set of tissues. Disruption of INI1 function blocks early murine embryogenesis at the peri-implantation stage (E3.5), and chimeric mice harboring an ES cell-derived INI1 knockout develop tumors of the central nervous system and soft tissue sarcomas at high frequency (~30%), suggesting cell lineage specificity. Moreover, mutations in INI1 are strongly (~90%) associated with aggressive childhood cancers, suggesting a critical role in restricting cell growth, a role that has been demonstrated for snr1 as well (Zraly, 2003).

A single snr1 transcript is present at all developmental stages and in adults. Snr1 protein and RNA levels fluctuate during development, with elevated expression in specific tissues. Genetic studies of snr1 reveal that null alleles cause early larval lethality. However, both snr1 germline clones and maternal expression of Snr1-2, lacking the C-terminal 109 amino acids, cause female sterility and late embryonic lethality, implying both a critical role in early development and that a large maternal contribution of snr1 mRNA is sufficient for embryogenesis. An unexpected result of this study was the finding of an important role for Snr1 in sustained adult viability, including males. Consistent with this view, snr1 mRNA was found to be present in male flies; moreover, both mor mRNA and BRM protein have been found in adult males, raising the possibility that the Brm complex may be continuously required (Zraly, 2003).

The snr1 somatic and germline clone phenotypes largely overlap with those observed for brm, and effects caused by expression of interfering transgenes suggest that snr1, brm, and osa are important for proper wing development. Phenotypic differences likely reflect unique tissue-specific functions for snr1. For example, analysis of somatic clones has revealed that snr1, like brm, is required for proper peripheral nervous system development, since both snr1 and brm mutant clones exhibit twinned, missing, stunted, and fused bristles in the abdomen reminiscent of abnormalities associated with mutations in neurogenic genes, such as Notch (Zraly, 2003).

Both somatic clonal analyses and ectopic expression of Snr1-2 reveal that histoblast proliferation is strongly compromised when functional snr1 is limiting, and the snr1 somatic clone phenotype is enhanced by the presence of a brm mutation, suggesting important functions for the brm complex in histoblasts. Clusters of histoblasts within each segment produce the integument of the adult abdomen on both the dorsal and ventral surfaces. The histoblasts proliferate during pupal development, with division initiated shortly after pupariation (AP) and the commencement of migration at about 14 h AP. Fusion of the different anterior and posterior dorsal histoblast nests that produce the tergite occurs between 18 and 40 h AP. The fusion of the dorsal cuticle is most severely affected by expression of Snr1-2; this phenotype correlates with both the temperature of incubation and the snr1 dosage, suggesting that Snr1 serves to guide brm complex functions in the developing histoblasts. Although the genes regulated by snr1 are unknown, one possibility is the escargot gene, which encodes a zinc-finger protein required for maintaining a diploid state in imaginal cells and normal development of the abdomen. High-level expression alone could not account for the Snr1-2 phenotypes, since Snr1-1 and rescue transgenes had no effect. While formally possible, it appears unlikely that Snr1 has normal functions independent of the brm complex (Zraly, 2003).

In addition to the bristle defects observed in snr1 clones, expression of Snr1-2 disrupts normal wing vein patterning and PNS development. The observed defects all coincide with a developmental period in late larvae and early pupae associated with rapid cell proliferation and differentiation and closely correlate with increased expression of brm complex genes. Of possible significance, snr1 mRNA expression fluctuates coincident with transient pulses of ecdysone in late development, especially as the SWI/SNF complex can assist the activation potential of other steroid hormone-binding transcription factors. Similar to expression of brmK804R that disrupts ATPase activity in vivo but not the formation or stability of the complex, the expression of Snr1-2 results in specific wing vein patterning defects, including disruptions along the L2 vein and ectopic bristles on the L3 vein. Although some of the misexpression phenotypes overlap, the Snr1-2 effects were most striking in anterior wing veins (L2 and L3), while brmK804R phenotypes affected the entire wing, including the L5 vein and posterior crossvein, as well as the wing margin. The expression patterns of Snr1 and brm are nearly identical in the larval and pupal wing disc; thus, Snr1 appears to be important for brm complex activities in restricted regions of the developing wing (Zraly, 2003).

An unexpected finding was an important role for Snr1 in sustained adult viability. Similar to effects caused by expression of Snr1-2, flies rescued with a heat-shock cDNA had a dramatically reduced viability following eclosion (~3 days). The reduced longevity is unlikely due to nonspecific effects, such as accumulation of overexpressed proteins, since a temperature-sensitive snr1 mutant also exhibits shortened lifespans (~7 days), and this effect is rescued by additional copies of the wild-type gene. Interestingly, ash1, another member of the trx-G that genetically interacts with both brm and trx, is also required for adult viability. ASH1 is a component of a large (~2 MDa) complex distinct from the brm complex, but whose composition is unknown (Zraly, 2003).

Since ash1 is important for trx function and Snr1 physically and genetically interacts with Trx, it may be possible that Ash1 can also form transient complexes with Snr1 in vivo. Thus, shortened adult viability may reflect reduced gene transcription due to disruption of brm complex activity or another complex, such as Ash1 (Zraly, 2003).

The Snr1/INI1 subunit copurifies with the fly and mammalian SWI/SNF complexes, and the presence of INI1/hSNF5 helps to reconstitute full in vitro chromatin remodeling activity, suggesting that Snr1/INI1 is a core component (Zraly, 2003).

BLAST database searches and reduced stringency hybridization have confirmed that snr1 is the only SNF5-related gene in flies. Unlike loss-of-function mutations in brm, mor, and osa, that encode other subunits of the brm complex, snr1 null alleles are not significantly dosage-limiting in some sensitized genetic assays. For example, snr1 null alleles do not suppress a Pc mutant phenotype in the male prothoracic leg, an assay that is often used to define members of the trx-G of activators (Zraly, 2003).

Consistent with genetic studies of snr1, most Brm complex components are also not encoded by previously identified trx-G genes. Some of the individual subunits may exist in excess of those required for the formation or function of the complex; they may be required for specific functions of the Brm complex, or they may be components of other complexes, making null phenotypes difficult to interpret. Further, depending on the specific target gene or tissue, a subunit may have different regulatory functions on complex activities (activation or repression); thus, removal of the subunit may have unanticipated effects. Snr1 and Brm efficiently coprecipitate; however, Snr1 is not detectably expressed in all tissues where Brm is found at high levels (Zraly, 2003).

Unlike Brm, Snr1 is not found at elevated levels in the notum portion of the third instar larval wing disc that gives rise to dorsal thorax, and snr1 mutant clones reveal only minor thoracic bristle defects that may represent loss of snr1 function in a restricted set of developing PNS cells. Similarly, a snr1 conditional mutant does not display any significant thoracic defects at any temperature. While it cannot be ruled out that earlier expression of snr1 RNA or protein allows for the completion of normal thoracic development after clone induction, ubiquitous expression of Snr1-2 also does not elicit thoracic phenotypes (Zraly, 2003).

Strikingly, Snr1 is not present at detectable levels in the leg discs; there are no Snr1-2 phenotypes in the legs, nor is there any evidence for a genetic role of snr1 in the development of the legs, despite convincing requirements for other brm complex genes. In addition to an absence of detectable snr1 clone or dominant-negative expression phenotypes in legs, genetic interaction tests using a snr1 null allele have failed to reveal any function for snr1 in suppressing the ectopic sex comb phenotype associated with mutations in several Pc-G genes, including Pc, Pcl and E(z). Furthermore, an unusual allele of the Pc-G gene E(z) that mimics trithorax mutant phenotypes, including similar genetic interactions, shows enhanced mutant phenotypes in the abdomen and legs in combination with alleles of brm, mor, and osa; however, only the abdomen phenotype is enhanced by snr1. Thus, Snr1 is not required for brm complex functions in leg development and, perhaps similar to what has been proposed for Osa, it appears to be an optional component in a subset of brm complexes (Zraly, 2003).

The finding that Snr1-2 expression phenotypes are sensitive to snr1 dosage and that the truncated protein is found predominantly in the cytoplasm raises the possibility that Snr1-2 interfers with normal Snr1 function by antagonizing specific interactions between the brm complex and cellular proteins. It appears unlikely that Snr1-2 phenotypes result from mislocalization of the brm complex, since no cytoplasmic redistribution of brm was observed following induction of Snr1-2 (Zraly, 2003).

Expression of large Snr1 deletions have no phenotype, suggesting that expression of the N-terminal residues alone is not deleterious, a view supported by analyses of snr1 mutant alleles that had similar truncations. The Repeat regions in Snr1 and INI1 have been shown to mediate specific protein contacts, acting independently or in concert to promote the recruitment of the metazoan SWI/SNF complexes to target loci. For example, HIV integrase, the Epstein-Barr viral protein EBNA2, c-MYC, and the human papillomavirus (HPV) E1 protein required for replication of the viral genome, all interact with INI1 through the repeat regions (Zraly, 2003).

Interactions have been identified between Snr1/INI1 and the SET domains of the Drosophila Trithorax (Trx) and human trithorax (HRX) proteins that function as transcriptional regulators of the homeotic genes. Trx physically interacts with Snr1 through conserved residues within the Repeat 2 region, and their interaction is important for proper development. Since Snr1-2 can assemble into brm complexes and Trx does not associate with Snr1-2 in vivo, the observed phenotypes might reflect diminished functions of Trx or other factors (Zraly, 2003).

The activities of other cellular proteins are also likely to require Snr1, including transcription factors such as Bicoid (unpublished data) and the cell cycle regulators Cyclin E/CDK2, and these may be candidates for functional disruption by Snr1-2. One possibility for a protein shown to associate directly with Repeat 1 is c-MYC, whose transactivation activity depends on interaction with INI1. Although a functional homolog of c-MYC exists in flies, it is unknown whether a similar relationship exists between the Drosophila counterparts or whether Drosophila c-MYC activity is affected by expression of Snr1-2. Another possibility is that Snr1-2 protein in the cytoplasm associates with unknown proteins and blocks their activity or transport into the nucleus, reflecting novel functions of the Snr1-2 protein. Further tests are necessary to resolve among these and other possibilities (Zraly, 2003).

Mounting evidence suggests that Brm complex subunits have specific roles in defining the range of targets and developmental functions of the complex. Mutations in mor result in midgut abnormalities and reduced eye size. Ectopic expression of Osa affects wing vein patterning and disrupts eye development, and mutations in brm, mor, and snr1 can modulate those phenotypes. Genetically, osa functions in concert with brm in wing tissues, but with the opposite effect in eyes, and osa has been shown to have genetic functions distinct from both brm and mor in oogenesis. Similarly, some snr1 mutant phenotypes overlap with those of brm, and brm can suppress many of the snr1 conditional mutant phenotypes (Zraly, 2003).

GENE STRUCTURE

Bases in 5' UTR -126

Exons - two

Bases in 3' UTR - 88

PROTEIN STRUCTURE

Amino Acids - 370

Structural Domains

The SNR1 and BRM proteins are present in a large (> 2 x 10(6) Da) complex, and they co-immunoprecipitate from Drosophila extracts (Dingwall, 1995).

date revised: 25 April 2024

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.