Abstract. This work demonstrates a highly nonrandom distribution of specific genes relative to nuclear domains enriched in splicing factors and poly(A) ÷ RNA, and provides evidence for the direct involvement of these in pre-mRNA metabolism. As investigated in hundreds of diploid fibroblasts, human collagen Ietl and 13-actin DNA/RNA showed a very high degree of spatial association with SC-35 domains, whereas three nontranscribed genes, myosin heavy chain, neurotensin, and albumin, showed no such preferential association. Collagen Itxl RNA accumulates within the more central region of the domain, whereas [3-actin RNA localizes at the periphery. A novel approach revealed that collagen RNA tracks are polarized, with the entire gene at one end, on the edge of the domain, and the RNA extending into the domain. Intron 26 is spliced within the RNA track at the domain periphery. Transcriptional inhibition studies show both the structure of the domain and the gene's relationship to it are not dependent upon the continued presence of accumulated collagen RNA, and that domains remaining after inhibition are not just storage sites. Results support a model reconciling light and electron microscopic observations which proposes that transcription of some specific genes occurs at the border of domains, which may also function in the assembly or distribution of RNA metabolic components. In contrast to the apparently random dispersal of total undefined hnRNA synthesis through interdomain space, transcription and splicing for some genes occurs preferentially at specific sites, and a high degree of individual pre-mRNA metabolism is compartmentalized with discrete SC-35 domains.ESPITE the remarkable complexity of critical functions the nucleus performs, a simplified view of the extra-nucleolar nucleoplasm has persisted for many years. The idea that there exists some higher-level organization of the nucleoplasm that facilitates basic nuclear functions has been proposed for some time (see for example Comings, 1980;Blobel, 1985;Jackson, 1991;Lawrence et al., 1993), but the extent to which it exists is still largely unknown. It is known that the distributions of heterochromatin and satellite sequences are cell type specific (Manuelidis, 1984), that chromosomes occupy discrete nuclear "territories" (Cremer, 1982;Lichter, 1988;Pinkel et al., 1988), and that "chromosome position effects" can impact the expression of transgenes for largely unknown reasons (for example AI-Shawi et al., 1990). Relative to overall nuclear space, individual genes do not localize to precise coordinates, but distribute within preferred nuclear regions (Ward, W. S., J. A. McNeil, and J. B. Lawrence, manuscript submitted for publication; Lawrence et al., 1993). However, as explored in this work, greater orAddress all correspondence to Jeanne Bentley Lawrence, Dept. of Cell Biology, University of Massachusetts Medical Center, 55 Lake Ave. North, Worcester, MA 01655.Drs. Xing and Johnson contributed equally to this work.der may become apparent when sequences are local...

Visualization of fibronectin and neurotensin messenger RNAs within mammalian interphase nuclei was achieved by fluorescence hybridization with genomic, complementary DNA, and intron-specific probes. Unspliced transcripts accumulated in one or two sites per nucleus. Fibronectin RNA frequently accumulated in elongated tracks that overlapped and extended well beyond the site of transcription. Splicing appears to occur directly within this RNA track, as evidenced by an unambiguous spatial separation of intron-containing and spliced transcripts. Excised introns for neurotensin RNA appear free to diffuse. The transcription and processing site of the fibronectin gene localized to the nuclear interior and was associated with larger transcript domains in over 88 percent of the cells. These results support a view of nuclear function closely integrated with structure.

Typically, eukaryotic nuclei contain 10–30 prominent domains (referred to here as SC-35 domains) that are concentrated in mRNA metabolic factors. Here, we show that multiple specific genes cluster around a common SC-35 domain, which contains multiple mRNAs. Nonsyntenic genes are capable of associating with a common domain, but domain “choice” appears random, even for two coordinately expressed genes. Active genes widely separated on different chromosome arms associate with the same domain frequently, assorting randomly into the 3–4 subregions of the chromosome periphery that contact a domain. Most importantly, visualization of six individual chromosome bands showed that large genomic segments (∼5 Mb) have striking differences in organization relative to domains. Certain bands showed extensive contact, often aligning with or encircling an SC-35 domain, whereas others did not. All three gene-rich reverse bands showed this more than the gene-poor Giemsa dark bands, and morphometric analyses demonstrated statistically significant differences. Similarly, late-replicating DNA generally avoids SC-35 domains. These findings suggest a functional rationale for gene clustering in chromosomal bands, which relates to nuclear clustering of genes with SC-35 domains. Rather than random reservoirs of splicing factors, or factors accumulated on an individual highly active gene, we propose a model of SC-35 domains as functional centers for a multitude of clustered genes, forming local euchromatic “neighborhoods.”

The coiled bodies are nuclear structures rich in a variety of nuclear and nucleolar components including snRNAs. We have investigated the possibility that coiled bodies may associate with snRNA genes and report here that there is a high degree of association between U2 and U1 genes with a subset of coiled bodies. As investigated in human HeLa cells grown in monolayer culture, about 75% of nuclei had at least one U2 gene associated with a coiled body, and 45% had at least one U1 locus associated. In another suspension-grown HeLa cell strain, 92% of cells showed associated of one or more U2 genes with coiled bodies. In contrast to the U2 and U1 gene associations, a locus closely linked to the U2 gene cluster appeared associated with a coiled body only in 10% of cells. Associated snRNA gene signals were repeatedly positioned at the edge of the coiled body. Thus, this associated was highly nonrandom and spatially precise. Our analysis revealed a much higher frequency of association for closely spaced "doublet" U2 gene signals, with over 80% of paired signals associated as opposed to 35% for single U2 signals. This finding, coupled with the fact that not all genes were associated in all cells, suggested the possibility of a cell-cycle-dependent, possibly S-phase, association. However, an analysis of S- and non-S-phase cells using BrdU incorporation or cell synchronization did not indicate an increased level of association in S-phase. These and other results suggested that a substantial fraction of paired U2 signals represented association of U2 genes on homologous chromosomes rather than only replicated DNA. Furthermore, triple label analysis showed that in a significant fraction of cells U1 and U2 genes were both associated with the same coiled body. U1 and U2 genes were closely paired in approximately 20% of cells, over 60% of which were associated with a readily identifiable coiled body. This finding raises the possibility that multiple genes of a particular class may be in association with each coiled body. Thus, the coiled body may be a dynamic structure which transiently interacts with or is formed by one or more specific genetic loci, possibly carrying out some function related to their expression.

The inactivation of growth suppressor genes appears to play a major role in the malignant process. To assess whether protein phosphotyrosyl phosphatfises (protein-tyrosine-phosphate phosphohydrolase, EC 3.1.3.48) function as growth suppressors, we have isolated a cDNA clone encoding human protein phosphotyrosyl phosphatase 1B for structural and functional characterization. The translation product deduced from the 1305-nucleotide open reading frame predicts a protein containing 435 amino acids and having a molecular mass of 49,966 Da. The amino-terminal 321 amino acids deduced from the cDNA sequence are identical to the empirically determined sequence of protein phosphotyrosyl phosphatase 1B