The telomere-binding protein Tbf1 demarcates snoRNA gene promoters in Saccharomyces cerevisiae.

Small nucleolar RNAs (snoRNAs) play a key role in ribosomal RNA biogenesis, yet factors controlling their expression are unknown. We found that the majority of Saccharomyces snoRNA promoters display an aRCCCTaa sequence motif at the upstream border of a TATA-containing nucleosome-free region. Genome-wide ChIP-seq analysis showed that these motifs are bound by Tbf1, a telomere-binding protein known to recognize mammalian-like T(2)AG(3) repeats at subtelomeric regions. Tbf1 has over 100 additional promoter targets, including several other genes involved in ribosome biogenesis and the TBF1 gene itself. Tbf1 is required for full snoRNA expression, yet it does not influence nucleosome positioning at snoRNA promoters. In contrast, Tbf1 contributes to nucleosome exclusion at non-snoRNA promoters, where it selectively colocalizes with the Tbf1-interacting zinc-finger proteins Vid22 and Ygr071c. Our data show that, besides the ribosomal protein gene regulator Rap1, a second telomere-binding protein also functions as a transcriptional regulator linked to yeast ribosome biogenesis.

Hypoxia Differentially Modulates the Genomic Stability of Clinical-Grade ADSCs and BM-MSCs in Long-Term Culture.

Long-term cultures under hypoxic conditions have been demonstrated to maintain the phenotype of mesenchymal stromal/stem cells (MSCs) and to prevent the emergence of senescence. According to several studies, hypoxia has frequently been reported to drive genomic instability in cancer cells and in MSCs by hindering the DNA damage response and DNA repair. Thus, we evaluated the occurrence of DNA damage and repair events during the ex vivo expansion of clinical-grade adipose-derived stromal cells (ADSCs) and bone marrow (BM)-derived MSCs cultured with platelet lysate under 21% (normoxia) or 1% (hypoxia) O2 conditions. Hypoxia did not impair cell survival after DNA damage, regardless of MSC origin. However, ADSCs, unlike BM-MSCs, displayed altered γH2AX signaling and increased ubiquitylated γH2AX levels under hypoxic conditions, indicating an impaired resolution of DNA damage-induced foci. Moreover, hypoxia specifically promoted BM-MSC DNA integrity, with increased Ku80, TP53BP1, BRCA1, and RAD51 expression levels and more efficient nonhomologous end joining and homologous recombination repair. We further observed that hypoxia favored mtDNA stability and maintenance of differentiation potential after genotoxic stress. We conclude that long-term cultures under 1% O2 were more suitable for BM-MSCs as suggested by improved genomic stability compared with ADSCs.

Gradual processing of the ITS1 from the nucleolus to the cytoplasm during synthesis of the human 18S rRNA.

Defects in ribosome biogenesis trigger stress response pathways, which perturb cell proliferation and differentiation in several genetic diseases. In Diamond-Blackfan anemia (DBA), a congenital erythroblastopenia, mutations in ribosomal protein genes often interfere with the processing of the internal transcribed spacer 1 (ITS1), the mechanism of which remains elusive in human cells. Using loss-of-function experiments and extensive RNA analysis, we have defined the precise position of the endonucleolytic cleavage E in the ITS1, which generates the 18S-E intermediate, the last precursor to the 18S rRNA. Unexpectedly, this cleavage is followed by 3'-5' exonucleolytic trimming of the 18S-E precursor during nuclear export of the pre-40S particle, which sets a new mechanism for 18S rRNA formation clearly different from that established in yeast. In addition, cleavage at site E is also followed by 5'-3' exonucleolytic trimming of the ITS1 by exonuclease XRN2. Perturbation of this step on knockdown of the large subunit ribosomal protein RPL26, which was recently associated to DBA, reveals the putative role of a highly conserved cis-acting sequence in ITS1 processing. These data cast new light on the original mechanism of ITS1 elimination in human cells and provide a mechanistic framework to further study the interplay of DBA-linked ribosomal proteins in this process.

Frameshift mutation in p53 regulator RPL26 is associated with multiple physical abnormalities and a specific pre-ribosomal RNA processing defect in diamond-blackfan anemia.

Diamond-Blackfan anemia (DBA) is an inherited form of pure red cell aplasia that usually presents in infancy or early childhood and is associated with congenital malformations in ∼30-50% of patients. DBA has been associated with mutations in nine ribosomal protein (RP) genes in about 53% of patients. We completed a large-scale screen of 79 RP genes by sequencing 16 RP genes (RPL3, RPL7, RPL8, RPL10, RPL14, RPL17, RPL19, RPL23A, RPL26, RPL27, RPL35, RPL36A, RPL39, RPS4X, RPS4Y1, and RPS21) in 96 DBA probands. We identified a de novo two-nucleotide deletion in RPL26 in one proband associated with multiple severe physical abnormalities. This mutation gives rise to a remarkable ribosome biogenesis defect that affects maturation of both the small and the large subunits. We also found a deletion in RPL19 and missense mutations in RPL3 and RPL23A, which may be variants of unknown significance. Together with RPL5, RPL11, and RPS7, RPL26 is the fourth RP regulating p53 activity that is linked to DBA.

Eukaryotic snoRNAs: a paradigm for gene expression flexibility.

Small nucleolar RNAs (snoRNAs) are one of the most ancient and numerous families of non-protein-coding RNAs (ncRNAs). The main function of snoRNAs - to guide site-specific rRNA modification - is the same in Archaea and all eukaryotic lineages. In contrast, as revealed by recent genomic and RNomic studies, their genomic organization and expression strategies are the most varied. Seemingly snoRNA coding units have adopted, in the course of evolution, all the possible ways of being transcribed, thus providing a unique paradigm of gene expression flexibility. By focusing on representative fungal, plant and animal genomes, we review here all the documented types of snoRNA gene organization and expression, and we provide a comprehensive account of snoRNA expressional freedom by precisely estimating the frequency, in each genome, of each type of genomic organization. We finally discuss the relevance of snoRNA genomic studies for our general understanding of ncRNA family evolution and expression in eukaryotes.

Assembly into snoRNP controls 5'-end maturation of a box C/D snoRNA in Saccharomyces cerevisiae.

The SNR52 gene, coding for a box C/D snoRNA, is the only snoRNA gene transcribed by RNA polymerase (Pol) III in Saccharomyces cerevisiae. Pol III transcription generates a precisely terminated primary transcript that undergoes extensive 5'-end processing. Here, we show that mutations of the box C/D core motif required for snoRNP assembly compromise 5'-end maturation of the SNR52 snoRNA. Upstream processing was also impaired by specific depletion of either Nop1p or Nop58p snoRNP proteins. We further show that the nuclear exosome is required for 3'-end maturation of SNR52 snoRNA, at variance with all the other known Pol III transcripts. Our data suggest a functional coupling between snoRNP assembly and 5'-end maturation of independently transcribed box C/D snoRNAs.

A minimal promoter for TFIIIC-dependent in vitro transcription of snoRNA and tRNA genes by RNA polymerase III.

The Saccharomyces cerevisiae SNR52 gene is unique among the snoRNA coding genes in being transcribed by RNA polymerase III. The primary transcript of SNR52 is a 250-nucleotide precursor RNA from which a long leader sequence is cleaved to generate the mature snR52 RNA. We found that the box A and box B sequence elements in the leader region are both required for the in vivo accumulation of the snoRNA. As expected box B, but not box A, was absolutely required for stable TFIIIC, yet in vitro. Surprisingly, however, the box B was found to be largely dispensable for in vitro transcription of SNR52, whereas the box A-mutated template effectively recruited TFIIIB; yet it was transcriptionally inactive. Even in the complete absence of box B and both upstream TATA-like and T-rich elements, the box A still directed efficient, TFIIIC-dependent transcription. Box B-independent transcription was also observed for two members of the tRNA(Asn)(GTT) gene family, but not for two tRNA(Pro)(AGG) gene copies. Fully recombinant TFIIIC supported box B-independent transcription of both SNR52 and tRNA(Asn) genes, but only in the presence of TFIIIB reconstituted with a crude B'' fraction. Non-TFIIIB component(s) in this fraction were also required for transcription of wild-type SNR52. Transcription of the box B-less tRNA(Asn) genes was strongly influenced by their 5'-flanking regions, and it was stimulated by TBP and Brf1 proteins synergistically. The box A can thus be viewed as a core TFIIIC-interacting element that, assisted by upstream TFIIIB-DNA contacts, is sufficient to promote class III gene transcription.