Biosynthesis of selenocysteine on its tRNA in eukaryotes

A recoding element that stimulates decoding of UGA codons by Sec tRNA[Ser]Sec

Selenocysteine insertion during decoding of eukaryotic selenoprotein mRNA requires several trans-acting factors and a cis-acting selenocysteine insertion sequence (SECIS) usually located in the 3' UTR. A second cis-acting selenocysteine codon redefinition element (SRE) has recently been described that resides near the UGA-Sec codon of selenoprotein N (SEPN1). Similar phylogenetically conserved elements can be predicted in a subset of eukaryotic selenoprotein mRNAs. Previous experimental analysis of the SEPN1 SRE revealed it to have a stimulatory effect on readthrough of the UGA-Sec codon, which was not dependent upon the presence of a SECIS element in the 3' UTR; although, as expected, readthrough efficiency was further elevated by inclusion of a SECIS. In order to examine the nature of the redefinition event stimulated by the SEPN1 SRE, we have modified an experimentally tractable in vitro translation system that recapitulates efficient selenocysteine insertion. The results presented here illustrate that the SRE element has a stimulatory effect on decoding of the UGA-Sec codon by both the methylated and unmethylated isoforms of Sec tRNA([Ser]Sec), and confirm that efficient selenocysteine insertion is dependent on the presence of a 3'-UTR SECIS. The variation in recoding elements predicted near UGA-Sec codons implies that these elements may play a differential role in determining the amount of selenoprotein produced by acting as controllers of UGA decoding efficiency.

A new strategy for assessing selenoprotein function: siRNA knockdown/knock-in targeting the 3'-UTR

Selenocysteine insertion into protein in mammalian cells requires RNA elements in the 3'-untranslated regions (3'-UTRs) of selenoprotein genes. The occurrence of these conserved sequences should make selenoproteins particularly amenable for knockdown/knock-in strategies to examine selenoprotein functions. Herein, we utilized the 3'-UTR of various selenoproteins to knock down their expression using siRNAs and then knock in expression using constructs containing mutations within the target region. Thioredoxin reductase 1 (TR1) knockdown in a mouse kidney cell line resulted in the cells growing about 10% more slowly, being more sensitive to UV radiation, and having increased apoptosis in response to UV than control cells. The knockdown cells transfected with a construct encoding the wild-type TR1 gene and having mutations in the sequences targeted by siRNA restored TR1 expression and catalytic activity, rendered the knockdown cells less sensitive to UV, and protected the cells against apoptosis. We also applied this technique to other selenoproteins, selenophosphate synthetase 2 and glutathione peroxidase 1, and found that mRNA and protein levels were restored following transfection of knockdown cells with the corresponding knock-in constructs. In addition to important new insights into the functions of key mammalian selenoproteins, the data suggest that the RNAi-based knock-in technology could distinguish phenotypes due to off-targeting and provide a new method for examining many of the subtleties of selenoprotein function not available using RNAi technology alone.

Selenophosphate synthetase 2 is essential for selenoprotein biosynthesis

Selenophosphate synthetase (SelD) generates the selenium donor for selenocysteine biosynthesis in eubacteria. One homologue of SelD in eukaryotes is SPS1 (selenophosphate synthetase 1) and a second one, SPS2, was identified as a selenoprotein in mammals. Earlier in vitro studies showed SPS2, but not SPS1, synthesized selenophosphate from selenide, whereas SPS1 may utilize a different substrate. The roles of these enzymes in selenoprotein synthesis in vivo remain unknown. To address their function in vivo, we knocked down SPS2 in NIH3T3 cells using small interfering RNA and found that selenoprotein biosynthesis was severely impaired, whereas knockdown of SPS1 had no effect. Transfection of SPS2 into SPS2 knockdown cells restored selenoprotein biosynthesis, but SPS1 did not, indicating that SPS1 cannot complement SPS2 function. These in vivo studies indicate that SPS2 is essential for generating the selenium donor for selenocysteine biosynthesis in mammals, whereas SPS1 probably has a more specialized, non-essential role in selenoprotein metabolism.

Novel tRNA aminoacylation mechanisms

In nature, ribosomally synthesized proteins can contain at least 22 different amino acids: the 20 common amino acids as well as selenocysteine and pyrrolysine. Each of these amino acids is inserted into proteins codon-specifically via an aminoacyl-transfer RNA (aa-tRNA). In most cases, these aa-tRNAs are biosynthesized directly by a set of highly specific and accurate aminoacyl-tRNA synthetases (aaRSs). However, in some cases aaRSs with relaxed or novel substrate specificities cooperate with other enzymes to generate specific canonical and non-canonical aminoacyl-tRNAs.

Identification of genes encoding tRNA modification enzymes by comparative genomics

As the molecular adapters between codons and amino acids, transfer-RNAs are pivotal molecules of the genetic code. The coding properties of a tRNA molecule do not reside only in its primary sequence. Posttranscriptional nucleoside modifications, particularly in the anticodon loop, can modify cognate codon recognition, affect aminoacylation properties, or stabilize the codon-anticodon wobble base pairing to prevent ribosomal frameshifting. Despite a wealth of biophysical and structural knowledge of the tRNA modifications themselves, their pathways of biosynthesis had been until recently only partially characterized. This discrepancy was mainly due to the lack of obvious phenotypes for tRNA modification-deficient strains and to the difficulty of the biochemical assays used to detect tRNA modifications. However, the availability of hundreds of whole-genome sequences has allowed the identification of many of these missing tRNA-modification genes. This chapter reviews the methods that were used to identify these genes with a special emphasis on the comparative genomic approaches. Methods that link gene and function but do not rely on sequence homology will be detailed, with examples taken from the tRNA modification field.