Macromolecular complexes: SMN--the master assembler.

Recent studies indicate that the protein affected in spinal muscular atrophy, SMN, plays a role in the assembly of a number of macromolecular complexes that function in the nucleus, interacting with its partner proteins via their arginine- and glycine-rich domains.

Direct interaction of the spinal muscular atrophy disease protein SMN with the small nucleolar RNA-associated protein fibrillarin.

Disruption of the survival motor neuron (SMN) gene leads to selective loss of spinal motor neurons, resulting in the fatal human neurodegenerative disorder spinal muscular atrophy (SMA). SMN has been shown to function in spliceosomal small nuclear ribonucleoprotein (snRNP) biogenesis and pre-mRNA splicing. We have demonstrated that SMN also interacts with fibrillarin, a highly conserved nucleolar protein that is associated with all Box C/D small nucleolar RNAs and functions in processing and modification of rRNA. Fibrillarin and SMN co-immunoprecipitate from HeLa cell extracts indicating that the proteins exist as a complex in vivo. Furthermore, in vitro binding studies indicate that the interaction between SMN and fibrillarin is direct and salt-stable. We show that the glycine/arginine-rich domain of fibrillarin is necessary and sufficient for SMN binding and that the region of SMN encoded by exon 3, including the Tudor domain, mediates the binding of fibrillarin. Tudor domain missense mutations, including one found in an SMA patient, impair the interaction between SMN and fibrillarin (as well as the common snRNP protein SmB). Our results suggest a function for SMN in small nucleolar RNP biogenesis (akin to its known role as an snRNP assembly factor) and reveal a potential link between small nucleolar RNP biogenesis and SMA.

Internal modification of U2 small nuclear (sn)RNA occurs in nucleoli of Xenopus oocytes.

U2 small nuclear (sn)RNA contains a large number of posttranscriptionally modified nucleotides, including a 5' trimethylated guanosine cap, 13 pseudouridines, and 10 2'-O-methylated residues. Using Xenopus oocytes, we demonstrated previously that at least some of these modified nucleotides are essential for biogenesis of a functional snRNP. Here we address the subcellular site of U2 internal modification. Upon injection into the cytoplasm of oocytes, G-capped U2 that is transported to the nucleus becomes modified, whereas A-capped U2 that remains in the cytoplasm is not modified. Furthermore, by injecting U2 RNA into isolated nuclei or enucleated oocytes, we observe that U2 internal modifications occur exclusively in the nucleus. Analysis of the intranuclear localization of fluorescently labeled RNAs shows that injected wild-type U2 becomes localized to nucleoli and Cajal bodies. Both internal modification and nucleolar localization of U2 are dependent on the Sm binding site. An Sm-mutant U2 is targeted only to Cajal bodies. The Sm binding site can be replaced by a nucleolar localization signal derived from small nucleolar RNAs (the box C/D motif), resulting in rescue of internal modification as well as nucleolar localization. Analysis of additional chimeric U2 RNAs reveals a correlation between internal modification and nucleolar localization. Together, our results suggest that U2 internal modification occurs within the nucleolus.

The snoRNA domain of vertebrate telomerase RNA functions to localize the RNA within the nucleus.

Telomerase RNA is an essential component of the ribonucleoprotein enzyme involved in telomere length maintenance, a process implicated in cellular senescence and cancer. Vertebrate telomerase RNAs contain a box H/ACA snoRNA motif that is not required for telomerase activity in vitro but is essential in vivo. Using the Xenopus oocyte system, we have found that the box H/ACA motif functions in the subcellular localization of telomerase RNA. We have characterized the transport and biogenesis of telomerase RNA by injecting labeled wild-type and variant RNAs into Xenopus oocytes and assaying nucleocytoplasmic distribution, intranuclear localization, modification, and protein binding. Although yeast telomerase RNA shares characteristics of spliceosomal snRNAs, we show that human telomerase RNA is not associated with Sm proteins or efficiently imported into the nucleus. In contrast, the transport properties of vertebrate telomerase RNA resemble those of snoRNAs; telomerase RNA is retained in the nucleus and targeted to nucleoli. Furthermore, both nuclear retention and nucleolar localization depend on the box H/ACA motif. Our findings suggest that the H/ACA motif confers functional localization of vertebrate telomerase RNAs to the nucleus, the compartment where telomeres are synthesized. We have also found that telomerase RNA localizes to Cajal bodies, intranuclear structures where it is thought that assembly of various cellular RNPs takes place. Our results identify the Cajal body as a potential site of telomerase RNP biogenesis.

The box C/D motif directs snoRNA 5'-cap hypermethylation.

The 5'-cap structure of most spliceosomal small nuclear RNAs (snRNAs) and certain small nucleolar RNAs (snoRNAs) undergoes hypermethylation from a 7-methylguanosine to a 2,2, 7-trimethylguanosine structure. 5'-Cap hypermethylation of snRNAs is dependent upon a conserved sequence element known as the Sm site common to most snRNAs. Here we have performed a mutational analysis of U3 and U14 to determine the cis-acting sequences required for 5'-cap hypermethylation of Box C/D snoRNAs. We have found that both the conserved sequence elements Box C (termed C' in U3) and Box D are necessary for cap hypermethylation. Furthermore, the terminal stem structure that is formed by sequences that flank Box C (C' in U3) and Box D is also required. However, mutation of other conserved sequences has no effect on hypermethylation of the cap. Finally, the analysis of fragments of U3 and U14 RNAs indicates that the Box C/D motif, including Box C (C' in U3), Box D and the terminal stem, is capable of directing cap hypermethylation. Thus, the Box C/D motif, which is important for snoRNA processing, stability, nuclear retention, protein binding, nucleolar localization and function, is also necessary and sufficient for cap hypermethylation of these RNAs.

Interaction of the U3-55k protein with U3 snoRNA is mediated by the box B/C motif of U3 and the WD repeats of U3-55k.

U3 small nucleolar RNA (snoRNA) is a member of the Box C/D family of snoRNAs which functions in ribosomal RNA processing. U3-55k is a protein that has been found to interact with U3 but not other members of the Box C/D snoRNA family. We have found that interaction of the U3-55k protein with U3 RNA in vivo is mediated by the conserved Box B/C motif which is unique to U3 snoRNA. Mutation of Box B and Box C, but not of other conserved sequence elements, disrupted interaction of U3-55k with U3 RNA. Furthermore, a fragment of U3 containing only these two conserved elements was bound by U3-55k in vivo. RNA binding assays performed in vitro indicate that Box C may be the primary determinant of the interaction. We have cloned the cDNA encoding the Xenopus laevis U3-55k protein and find strong homology to the human sequence, including six WD repeats. Deletion of WD repeats or sequences near the C-terminus of U3-55k resulted in loss of association with U3 RNA and also loss of localization of U3-55k to the nucleolus, suggesting that protein-protein interactions contribute to the localization and RNA binding of U3-55k in vivo.

Nuclear retention elements of U3 small nucleolar RNA.

The processing and methylation of precursor rRNA is mediated by the box C/D small nucleolar RNAs (snoRNAs). These snoRNAs differ from most cellular RNAs in that they are not exported to the cytoplasm. Instead, these RNAs are actively retained in the nucleus where they assemble with proteins into mature small nucleolar ribonucleoprotein particles and are targeted to their intranuclear site of action, the nucleolus. In this study, we have identified the cis-acting sequences responsible for the nuclear retention of U3 box C/D snoRNA by analyzing the nucleocytoplasmic distributions of an extensive panel of U3 RNA variants after injection of the RNAs into Xenopus oocyte nuclei. Our data indicate the importance of two conserved sequence motifs in retaining U3 RNA in the nucleus. The first motif is comprised of the conserved box C' and box D sequences that characterize the box C/D family. The second motif contains conserved box sequences B and C. Either motif is sufficient for nuclear retention, but disruption of both motifs leads to mislocalization of the RNAs to the cytoplasm. Variant RNAs that are not retained also lack 5' cap hypermethylation and fail to associate with fibrillarin. Furthermore, our results indicate that nuclear retention of U3 RNA does not simply reflect its nucleolar localization. A fragment of U3 containing the box B/C motif is not localized to nucleoli but retained in coiled bodies. Thus, nuclear retention and nucleolar localization are distinct processes with differing sequence requirements.

Nucleolar localization signals of box H/ACA small nucleolar RNAs.

The two major families of small nucleolar RNAs (snoRNAs), Box C/D and Box H/ACA, are generated in the nucleoplasm and transported to the nucleolus where they function in rRNA processing and modification. We have investigated the sequences involved in the intranuclear transport of Box H/ACA snoRNAs by assaying the localization of injected fluorescent RNAs in Xenopus oocyte nuclear spreads. Our analysis of U17, U64 and U65 has revealed that disruption of either of the conserved sequence elements, Box H or Box ACA, eliminates nucleolar localization. In addition, the stem present at the base of the 3' hairpin is required for efficient nucleolar localization of U65. Fragments or rearrangements of U65 that consist of Box H and Box ACA flanking either the 5' or 3' hairpin are targeted to the nucleolus. The targeting is dependent on the presence of the Box sequences, but not on their orientation. Our results indicate that in each of the two major families of snoRNAs, a motif composed of the signature conserved sequences and an adjacent structural element that tethers the sequence elements directs the nucleolar localization of the RNAs. We demonstrate that telomerase RNA is also targeted to the nucleolus by a Box ACA-dependent mechanism.

Role of the box C/D motif in localization of small nucleolar RNAs to coiled bodies and nucleoli.

Small nucleolar RNAs (snoRNAs) are a large family of eukaryotic RNAs that function within the nucleolus in the biogenesis of ribosomes. One major class of snoRNAs is the box C/D snoRNAs named for their conserved box C and box D sequence elements. We have investigated the involvement of cis-acting sequences and intranuclear structures in the localization of box C/D snoRNAs to the nucleolus by assaying the intranuclear distribution of fluorescently labeled U3, U8, and U14 snoRNAs injected into Xenopus oocyte nuclei. Analysis of an extensive panel of U3 RNA variants showed that the box C/D motif, comprised of box C', box D, and the 3' terminal stem of U3, is necessary and sufficient for the nucleolar localization of U3 snoRNA. Disruption of the elements of the box C/D motif of U8 and U14 snoRNAs also prevented nucleolar localization, indicating that all box C/D snoRNAs use a common nucleolar-targeting mechanism. Finally, we found that wild-type box C/D snoRNAs transiently associate with coiled bodies before they localize to nucleoli and that variant RNAs that lack an intact box C/D motif are detained within coiled bodies. These results suggest that coiled bodies play a role in the biogenesis and/or intranuclear transport of box C/D snoRNAs.

A common maturation pathway for small nucleolar RNAs.

We have shown that precursors of U3, U8 and U14 small nucleolar RNAs (snoRNAs) are not exported to the cytoplasm after injection into Xenopus oocyte nuclei but are selectively retained and matured in the nucleus, where they function in pre-rRNA processing. Our results demonstrate that Box D, a conserved sequence element found in these and most other snoRNAs, plays a key role in their nuclear retention, 5' cap hypermethylation and stability. Retention of U3 and U8 RNAs in the nucleus is saturable and relies on one or more common factors. Hypermethylation of the 5' caps of U3 RNA occurs efficiently in oocyte nuclear extracts lacking nucleoli, suggesting that precursor snoRNAs are matured in the nucleoplasm before they are localized to the nucleolus. Surprisingly, m7G-capped precursors of spliceosomal small nuclear RNAs (snRNAs) such as pre-U1 and U2, can be hypermethylated in nuclei if the RNAs are complexed with Sm proteins. This raises the possibility that a single nuclear hypermethylase activity may act on both nucleolar and spliceosomal snRNPs.

Retention and 5' cap trimethylation of U3 snRNA in the nucleus.

It is shown here that maturation of the m7G-capped precursors of U3 small nuclear RNA (snRNA) occurs by a previously unknown pathway. In contrast to the 5' m7G-capped precursors of other snRNAs, this RNA is not exported to the cytoplasm but is retained in the nuclei of Xenopus laevis oocytes, where it undergoes trimethylation of its 5' cap. The m7G caps of most snRNA precursors are trimethylated only after transport of the RNAs to the cytoplasm. The nuclear retention and maturation of this nucleolar RNA raises the possibility that other m7G-capped RNAs are also retained and modified in the nucleus.

Multiple cis-acting signals for export of pre-U1 snRNA from the nucleus.

We have identified cis-acting sequences that promote nuclear export of pre-U1 RNA injected into Xenopus oocyte nuclei. At least three elements, the 5' m7G cap, the 3'-terminal stem-loop structure, and sequences in the 5'-terminal 124 nucleotides, contribute to efficient export of this RNA. Both the 5' and 3' export signals can function separately and do so independently of the cap structure. Experiments using hybrid RNAs indicate that the 5' and 3' export sequences of U1 RNA are sufficient to direct export of the heterologous, otherwise nonexportable, U6 RNA. The absence of comparable export signals in U6 RNA appears to be responsible for its retention in the nucleus. Stability of the pre-snRNAs in the nucleus depends on the presence of both a 5' cap structure and a 3' base-paired stem. The 5' m7G cap is neither sufficient nor necessary for nuclear export. The m7G cap by itself did not promote export of U6 RNA or nonspecific small RNAs. Moreover, substitution of this cap with either an AppG cap or gamma-mppG cap did not eliminate export of either full-length or a "minimal" U1 RNA (lacking most of the internal U1 RNA sequences), but it reduced the rate of export by about two to threefold. However, in the absence of the 3' stem-loop, substitution of the m7G cap led to a greater decrease in export rate, underscoring the cooperative action of the three different export elements of pre-U1 RNA. The m7G cap analog, m7GpppG, selectively destabilized pre-U1 RNA within the nucleus. Thus, nuclear components that recognize the 5' m7G cap may be important for both the stability and the export of pre-U1 RNA.

A pre-export U1 snRNP in Xenopus laevis oocyte nuclei.

We demonstrate that precursors of U1 snRNA are associated with nuclear proteins prior to export to the cytoplasm. The approximately 15S complexes containing pre-U1 RNA, which we call pre-export U1 snRNPs, were identified in extracts of Xenopus laevis oocyte nuclei that were synthesizing U1 RNAs from injected U1 genes. The U1 snRNP-specific A protein was associated with nuclear pre-U1 RNA since both this protein and the RNA were co-precipitated by antibodies directed against either the m7G-cap of the precursor RNA or the U1-A protein. The interaction of the U1-A protein with pre-U1 RNA required sequences in the loop II region although this region of U1 RNA was not necessary for the association of U1 A protein with mature U1 snRNPs. The U1 A protein helps protect pre-U1 RNA against degradation in the nucleus.

3'-end-dependent formation of U6 small nuclear ribonucleoprotein particles in Xenopus laevis oocyte nuclei.

We have identified and characterized a U6 small nuclear (sn) ribonucleoprotein particle (RNP) present in the nuclei of Xenopus laevis oocytes. The structure of this U6 snRNP was investigated by native gel shift analysis and a combination of RNA-protein UV cross-linking, RNase T1 fingerprinting, and immunoprecipitation assays. These analyses demonstrate that certain forms of U6 snRNA associate with the 50-kDa nuclear antigen La both in vivo and in vitro. The La protein binds the stretch of uridylates at the 3' hydroxyl end of newly synthesized U6 snRNA. La does not bind to mature U6 snRNAs that have 2',3'-cyclic phosphate (greater than p) groups at their 3' ends (E. Lund and J. E. Dahlberg, Science 255:327-330, 1992) or to U6 snRNAs in anti-Sm-precipitable U4/U6 snRNPs. We propose that 3'-end modification, including posttranscriptional UMP addition, modulates the binding of La protein to U6 snRNA which, in turn, may affect the function of this RNA.

Potential role of poly(A) polymerase in the assembly of polyadenylation-specific RNP complexes.

To elucidate the mechanism by which poly(A) polymerase functions in the 3'-end processing of pre-mRNAs, polyadenylation-specific RNP complexes were isolated by sedimentation in sucrose density gradients and the fractions were analyzed for the presence of the enzyme. At early stages of the reaction, the RNP complexes were resolved into distinct peaks which sedimented at approximately 18S and 25S. When reactions were carried out under conditions which support cleavage or polyadenylation, the pre-mRNA was specifically assembled into the larger 25S RNP complexes. Polyclonal antibodies raised against the enzyme purified from a rat hepatoma, which have been shown to inhibit cleavage and polyadenylation (Terns, M., and Jacob, S. T., Mol. Cell. Biol. 9:1435-1444, 1989) also prevented assembly of the 25S polyadenylation-specific RNP complexes. Furthermore, formation of these complexes required the presence of a chromatographic fraction containing poly(A) polymerase. UV cross-linking analysis indicated that the purified enzyme could be readily cross-linked to pre-mRNA but in an apparent sequence-independent manner. Reconstitution studies with the fractionated components showed that formation of the 25S RNP complex required the poly(A) polymerase fraction. Although the enzyme has not been directly localized to the specific complexes, the data presented in this report supports the role of poly(A) polymerase as an essential component of polyadenylation-specific complexes which functions both as a structural and enzymatic constituent.

Polyadenylate polymerases from normal and cancer cells and their potential role in messenger RNA processing: a review.

Two structurally and immunologically distinct species of nuclear polyadenylate [poly(A)] polymerases have been characterized. One of these enzymes is relatively absent in normal tissues but is predominant in primary and transplanted tumors and transformed cell lines. The presence of the tumor type enzyme in fetal liver, but not in regenerating liver, suggests that it is an oncofetal protein. Antibodies against the tumor-type poly(A) polymerases are present in the sera of rats bearing tumors and in some cancer patients. These antibodies are also found in the sera of rats fed hepatocarcinogen even before preneoplastic nodules were visible, which suggests that elicitation of these antibodies is an early event in neoplastic transformation. Autoantibodies against both liver-type and tumor-type poly(A) polymerase are also present in some rheumatic autoimmune sera. Polyclonal antibodies against purified enzyme from a rat hepatoma, which exhibit a single band upon immunoblot analysis, were used in cell-free extracts to study the role of poly(A) polymerase in the 3'-end processing of pre-mRNA. These studies showed that the antibodies blocked both endonucleolytic cleavage and poly(A) addition at the cleavage site and complex formation between factors in the extract and pre-mRNA. Independent studies in other laboratories have demonstrated that both the cleavage and poly(A) polymerase activities require the same component for their function. These observations suggest that both cleavage and polyadenylation reactions are tightly coupled in a functional complex.

Role of poly(A) polymerase in the cleavage and polyadenylation of mRNA precursor.

To determine the role of poly(A) polymerase in 3'-end processing of mRNA, the effect of purified poly(A) polymerase antibodies on endonucleolytic cleavage and polyadenylation was studied in HeLa nuclear extracts, using adenovirus L3 pre-mRNA as the substrate. Both Mg2+- and Mn2+-dependent reactions catalyzing addition of 200 to 250 and 400 to 800 adenylic acid residues, respectively, were inhibited by the antibodies, which suggested that the two reactions were catalyzed by the same enzyme. Anti-poly(A) polymerase antibodies also inhibited the cleavage reaction when the reaction was coupled or chemically uncoupled with polyadenylation. These antibodies also prevented formation of specific complexes between the RNA substrate and components of nuclear extracts during cleavage or polyadenylation, with the concurrent appearance of another, antibody-specific complex. These studies demonstrate that (i) previously characterized poly(A) polymerase is the enzyme responsible for addition of the poly(A) tract at the correct cleavage site and probably for the elongation of poly(A) chains and (ii) the coupling of these two 3'-end processing reactions appears to result from the potential requirement of poly(A) polymerase for the cleavage reaction. The results suggest that the specific endonuclease is associated with poly(A) polymerase in a functional complex.