Richard J. Maraia, MD, Head, Section on Molecular and Cell Biology
Robert V. Intine, PhD, Staff Scientist
Ying Huang, PhD, Visiting Associate
Mark A. Bayfield, PhD, Visiting Fellow
Jung-Min Park, PhD, Visiting Fellow
Liqiang Zhang, PhD, Visiting Fellow
Nathan H. Blewett, BS, Postbaccalaureate Fellow
Monique W. Bruinsma, BS, Postbaccalaureate Fellow
Trish E. Kaiser, BS, Technician
Amanda M. Day, BS, Technical Training Fellow

The La antigen is a target of auto-antibodies in patients who suffer from systemic lupus, neonatal lupus, and Sjögren's syndrome; however, a causal relationship, if any, between autoimmunity and La remains unknown. We study the activities of the multifunctional La protein and its regulation in normal cells. La is a sequence-specific RNA binding protein whose capacity to bind to many non-coding and messenger RNAs lends itself to a diversity of activities. The best-established ligands of La are transcripts synthesized by RNA polymerase (Pol) III, which include tRNAs, 5S rRNA, and U6 snRNA as well as other non-coding RNAs. High-affinity binding occurs via recognition of UUU-OH, the RNAs' common 3′ terminal motif, which results from transcriptional termination by Pol III. We developed an La-dependent tRNA suppressor that drives a red-white reporter system in fission yeast. Because La is conserved, the system lends itself to the study of La proteins of human, mouse, yeasts, and other species. Our goal is to understand how the functions of the La protein are integrated by cells and organisms during growth and development. We use genetics, cell and structural biology, and biochemistry in model systems that include yeast, human tissue culture cells, and gene-altered mice.

Functions of the La antigen in RNA expression

Maraia, Intine, Huang, Bayfield, Kaiser

The La proteins of yeast and humans are the most extensively characterized. The large share of phosphorylated La resides in the nucleus, where it promotes the maturation of its UUU-OH-containing RNA ligands, in part by protecting them from 3′ exonucleolytic digestion and for some, e.g., pre-tRNAs, by directing the order in which the 5′ and 3′ ends and introns are processed. The tRNAs undergo the most complex maturation of all of the Pol III transcripts, requiring 5′- and 3′-end processing, base and ribose modifications, splicing for those pre-tRNAs that have introns, addition of CCA to the processed 3′ end, aminoacylation, and nuclear export. La's effects on pre-tRNAs have been studied more extensively than those associated with other RNAs. La also functions to retain precursor RNAs in the nucleus, where many RNA processing enzymes reside, until early maturation steps are completed. For the Pol III RNAs, whose post-transcriptional maturation is relatively minimal, La's involvement is limited.

Accumulating evidence indicates that La also affects the metabolism of mRNAs (synthesized by Pol II) and rRNA (synthesized by Pol I). We mapped the major phosphorylation site of La to S366 and developed two sets of antibodies that differ in whether they recognize phosphorylated (pLa) or nonphosphorylated (npLa). The antibodies demonstrated that npLa and pLa exhibit different subnuclear localizations and are differentially associated with certain RNAs in vivo. In the cytoplasm, npLa binds to and facilitates the translation of specific sets of mRNAs. For example, npLa is associated with a family of mRNAs that bear a common 5′-terminal oligopyrimidine (5′TOP) motif and that coordinately produce ribosomal proteins and translation factors in response to nutritional status. The introns of some non-coding 5′TOP RNAs are processed to produce several snoRNAs involved in rRNA modification in the nucleolus. A cis element known as the short basic motif (SBM) located adjacent to the major phosphorylation site of La, serine 366 (S366) in the carboxyl-terminal domain, directs npLa-specific interaction with nucleolin, a factor involved in rRNA processing in the nucleolus. As such, La appears to be involved in the metabolism of transcripts synthesized by Pols I, II, and III and, moreover, in a manner that contributes to maintenance of the cell's translational capacity, which is a major central activity.

Our data point to several trafficking signals in La proteins that control nuclear, nucleolar, and cytoplasmic localization. We uncovered a nuclear export activity that is conserved in La proteins from yeast to human and is therefore intriguing because the known functions of La in Pol III transcript biogenesis are entirely intranuclear. Features of La's nuclear export activity are reminiscent of proteins that carry mRNAs to the cytoplasm. Accordingly, we have begun to look for La-associated mRNAs whose expression is critical to growth and development.

Our recent structure-function-related work focused on the conserved N-terminal domain of La protein, which is composed of an La motif and RNA recognition motif (RRM, one of the most numerous domains encoded in the human genome), both known to be required for UUU-OH recognition. Although RRM proteins play several roles in RNA metabolism, their RRMs all engage RNA by a common mechanism lying across the surface of their four-stranded beta-sheets. Surprisingly, an La-RNA co-crystal indicated an unusual use of the RRM in UUU-OH recognition that leaves the typical surface of the RRM unoccupied. We analyzed several suppressor tRNA alleles in vivo and found differential sensitivities to La and Rrp6, a 3′-exonuclease component of pre-tRNA decay. Mutations in the La motif but not in the RRM surface compromised 3′ end protection in vivo and in vitro. We characterized a second activity of La that requires an intact RRM surface. The La motif mediates UUU-OH recognition while the RRM surface is required for a distinct activity, indicating that separate RNA-binding surfaces on La mediate distinguishable activities in tRNA maturation and suggesting a modular model that has implications for La's other ligands. Our observations provide a framework for appreciating the various activities attributed to this multifunctional phosphoprotein, which include its principal function in the form of snRNA 3′ end protection as well as mRNA-related and RNA chaperone-like activities and DNA- and chromatin-associated activities.

Activities of RNA polymerase III and associated factors

Maraia, Huang; in collaboration with Fairley, Grewal, Noma, White

The Pol III enzyme consists of 17 subunits, several with strong homology to subunits of Pol I and Pol II. In addition, TFIIIC, composed of six subunits, binds to the A- and B-box promoters and recruits TFIIIB to direct Pol III to the correct start site. Supporting many cycles of initiation, termination, and reinitiation with great productivity, Pol III complexes are highly stable. For example, each 5S rRNA gene in human cells must produce approximately 104-105 transcripts per cell division to provide sufficient 5S rRNA for ribosomes. While Pol I, Pol II, and Pol III are homologous, their properties are distinct in accordance with the unique functions related to the different types of genes they transcribe. Given that some mRNA genes can be hundreds of kilobase pairs long, Pol II must be very processive and avoid premature termination. Pol II terminates in response to complex termination/RNA-processing signals requiring endonucleolytic cleavage of the RNA upstream of the elongating polymerase. By contrast, formation of the UUU-OH 3′ terminus of nascent Pol III transcripts appears to occur in the Pol III active center. The dT(\n) tracts at the ends of class III genes directly signal pausing and release by Pol III such that termination and RNA 3′ end formation are coincident and efficient. Some data suggest that TFIIIC may participate in termination.

We developed a model Pol III system in S. pombe that uses an opal suppressor tRNA that suppresses a stop codon in the mRNA encoding a purine-synthetic enzyme (Ade6-704) whose activity can be monitored by an in vivo colorimetric colony assay. In wild-type cells, suppressor tRNA expression is dependent on accurate and efficient termination by Pol III, which requires five or more dT residues. We cloned several of the subunits comprising S. pombe TFIIIB, TFIIIC, and Pol III and showed that many could affect Pol III transcription in vivo. The system responds to factors that modulate Pol III initiation as well as termination. We also developed in vitro transcription from S. pombe.

We examined RNA UUU-OH 3′ end formation by the Pol III subunit Rpc11p (homologous to TFIIS), which was known to mediate 3′ cleavage of the growing RNA associated with the template DNA in the transcription bubble. Mutations in Rpc11p that were shown to decrease catalytic RNA 3′ cleavage activity led to increased 3′ oligo-U length and La-dependent tRNA processing in fission yeast. We believe that this system represents a eukaryotic RNA polymerase for which the catalysis of termination can be studied. Our Rpc11p mutants produce pre-tRNAs with elongated 3′ oligo-U tracts that are better substrates for La and its associated processing pathway. Our proof-of-principal project established that this system can be used in a genetic screen for termination-altering mutants. We are planning to test certain TFIIIC subunits as well as additional Pol III subunits for termination-altering activities.

Evidence from our laboratory and others suggests that human La may be associated with a Pol III holoenzyme, from which La is transferred to newly completed Pol III transcripts. Data from two other laboratories suggest that human La participates in the termination and reinitiation phases of Pol III transcription. NpLa was active for transcription while pLa was not. Recent chromatin immunoprecipitation (ChIP) data show that only npLa is associated with Pol III-transcribed genes in vivo. While a role for La in transcription remains questionable and its association with Pol III-transcribed genes is intriguing but as yet unexplained, the conserved activity of La proteins, namely, 3′ UUU-OH binding, which links Pol III termination and post-transcriptional processing, is well established.

We discovered a surprising activity in heterochromatic gene silencing and genome organization for S. pombe TFIIIC. In a collaborative effort, we showed that boundary elements flanking the fission yeast mating-type heterochromatin contain B-box sequences that prevent heterochromatin from spreading into nearby euchromatin by recruiting TFIIIC without Pol III. Genome-wide analysis further revealed high enrichment of TFIIIC at many sites located between divergent Pol II promoters. The sites appear to tether distant loci to the nuclear periphery, at which TFIIIC is concentrated into distinct bodies. Many intriguing questions regarding this new activity of TFIIIC in genome organization and gene regulation remain to be answered.

Role of La antigen in mouse development

Maraia, Park, Kaiser, Bruinsma; in collaboration with DePamphilis, Grinberg, Ko, Kohn, Love

Ubiquitous in eukaryotic nuclei, the La phosphoprotein promotes the maturation of tRNA precursors and other small non-coding RNAs synthesized by Pol III. As alluded to above, binding to these transcripts occurs via recognition of UUU-OH, their common 3′ terminal motif, which results from transcription termination by Pol III. In the cytoplasm, the non-phosphorylated isoform of human La is associated with a family of mRNAs that bear a common 5′-terminal oligopryimidine (5′TOP) motif and coordinately produce ribosomal proteins and translation factors in response to nutritional status. Thus, it was surprising that La is dispensable in yeasts, the organisms in which it has been characterized most extensively. To determine if La is essential in mammals and, if so, at which developmental stage, we created mice with a disrupted La gene and analyzed the offspring from La+/- X La+/- intercrosses. We detected La-/- offspring at the expected frequency among blastocysts before uterine implantation but observed no nullizygotes after implantation, indicating that La is required early in development. Blastocysts derived from La+/- x La+/- intercrosses yielded 38 La+/+ and La+/- embryonic stem (ES) cell lines but no La-/- ES cell lines, suggesting that La contributes a critical function toward the establishment of ES cells. Consistent with our observations, blastocyst outgrowth assays revealed loss of the inner cell mass specifically from La-/- embryos. The results indicate that, in contrast to yeasts, La is essential in mammals and is one of a limited number of genes required as early as development of the inner cell mass.

In addition, La+/- X La+/- crosses revealed that female La+/- offspring were obtained at a 25 percent lower frequency than expected, suggesting that La haploinsufficiency may be detrimental in some genetic backgrounds. Congenic breeding should yield insight into this observation. In a separate project, we have developed conditional knockout alleles of La; the resulting mice are being bred and are under investigation by using the appropriate Cre-mater mice. One goal is to determine at what developmental stage, if any, La can be deleted with minimal effects on development. Another goal is to obtain ES cell lines that can be used to examine the effects of loss of the conditional La knockout allele.

COLLABORATORS

Mel DePamphilis, PhD, Laboratory of Molecular Growth Regulation, NICHD, Bethesda, MD
Jennifer Fairley, PhD, University of Glasgow, Glasgow, UK
Shiv Grewal, PhD, Laboratory of Molecular Cell Biology, NCI, Bethesda, MD
Alex Grinberg, DVM, Laboratory of Mammalian Genes and Development, NICHD, Bethesda, MD
Minoru Ko, PhD, Laboratory of Genetics, NIA, Baltimore, MD
Matthew Kohn, PhD, Laboratory of Molecular Growth Regulation, NICHD, Bethesda, MD
Paul Love, MD, PhD, Laboratory of Mammalian Genes and Development, NICHD, Bethesda, MD
Kenichi Noma, PhD, Laboratory of Molecular Cell Biology, NCI, Bethesda, MD
Alex Vassilev, PhD, Laboratory of Molecular Growth Regulation, NICHD, Bethesda, MD
Robert White, PhD, FRSE, FMedSci, University of Glasgow, Glasgow, UK

For further information, contact maraiar@mail.nih.gov.