Nuclease sensitive element binding protein 1 associates with the selenocysteine insertion sequence and functions in mammalian selenoprotein translation

Abstract

Biosynthesis of selenium-containing proteins requires insertion of the unusual amino acid selenocysteine by alternative translation of a UGA codon, which ordinarily serves as a stop codon. In eukaryotes, selenoprotein translation depends upon one or more selenocysteine insertion sequence (SECIS) elements located in the 3'-untranslated region of the mRNA, as well as several SECIS-binding proteins. Our laboratory has previously identified nuclease sensitive element binding protein 1 (NSEP1) as another SECIS-binding protein, but evidence has been presented both for and against its role in SECIS binding in vivo and in selenoprotein translation. Our current studies sought to resolve this controversy, first by investigating whether NSEP1 interacts closely with SECIS elements within intact cells. After reversible in vivo cross-linking and ribonucleoprotein immunoprecipitation, mRNAs encoding two glutathione peroxidase family members co-precipitated with NSEP1 in both human and rat cell lines. Co-immunoprecipitation of an epitope-tagged GPX1 construct depended upon an intact SECIS element in its 3'-untranslated region. To test the functional importance of this interaction on selenoprotein translation, we used small inhibitory RNAs to reduce the NSEP1 content of tissue culture cells and then examined the effect of that reduction on the activity of a SECIS-dependent luciferase reporter gene for which expression depends upon readthrough of a UGA codon. Co-transfection of small inhibitory RNAs directed against NSEP1 decreased its expression by approximately 50% and significantly reduced luciferase activity. These studies demonstrate that NSEP1 is an authentic SECIS binding protein that is structurally associated with the selenoprotein translation complex and functionally involved in the translation of selenoproteins in mammalian cells.

Fig. 1

Co-immunoprecipitation of selenoprotein mRNAs bound to NSEP1 in human and rat cell lines. Upper parts: RT-PCR detection of mRNA encoding human cytosolic glutathione peroxidase (GPx1) or phospholipid hydroperoxide glutathione peroxidase (GPx4), as indicated in the left margin, from human HeLa cells. Lower parts: RT-PCR detection of mRNA encoding rat GPx1 or GPx4, as indicated in the left margin, from rat McArdle 7777 cells. Intact cells were treated with formaldehyde to crosslink mRNA to associated binding proteins, then lysed and the lysates immunoprecipitated with the antibodies listed below and protein A-sepharose beads; the crosslinking was reversed and the released RNA species identified by RT-PCR. RT-PCR products were detected by ethidium bromide staining of agarose gels. Lane 1, normal rabbit serum; Lane 2, purified rabbit IgG against human NSEP1; Lane 3, rabbit antiserum against human poly (A) binding protein. Lane 4, RT-PCR using total cellular RNA prior to protein A-sepharose bead extractions.

Fig. 2

Schematic diagram of the SECIS element in the 3′UTR of mRNA encoding human GPx1. Regions of short conserved sequences are indicated by bold face type. Shading indicates regions of deletions, as described in the text. Light shading indicates the extent of the distal stem and terminal loop deletion; dark shading indicates the extent of the 5′ stem and conserved sequence deletion The oval area of dark shading indicates a conserved sequence region deleted in both constructs. The total SECIS deletion construct removed the entire sequence.

Fig. 3

Co-immunoprecipitation of epitope-tagged human GPX1 transcripts bound to NSEP1. mRNA from Saos-2 cells was analyzed as described in the legend to Figure 1. Upper part: RT-PCR products were detected by ethidium bromide staining of agarose gels. Lane 1: no-template control; Lanes 2, 4, 6, 8, 10: RT-PCR using total cellular RNA prior to protein A-sepharose bead extractions from cells transfected with the constructs indicated in the upper margin; Lanes 3, 5, 7, 9, 11: co-immunoprecipitated epitope-tagged GPX1 mRNA from cells transfected with the constructs indicated in the upper margin. Lower part: co-immunoprecipitated endogenous GPX1 mRNA from the same preparations as the corresponding lanes from the main part of the figure.

Fig. 4

Co-immunoprecipitation of epitope-tagged and endogenous human GPX1 transcripts bound to NSEP1. mRNA from Saos-2 cells was analyzed as described in the legend to Figure 1. RT-PCR products were detected by ethidium bromide staining of agarose gels. The upper part shows the products amplified by PCR primers directed to the epitope-tagged transcripts (“GPx1-HA”); the lower part, to endogenous transcripts (“GPx1”). Lane 1: no-RT control; Lane 2: total cellular RNA; Lane 3: co-immunoprecipitated RNA from cells transfected with a construct containing the SECIS element; Lane 4: co-immunopreci-pitated RNA from cells transfected with a construct with the distal stem and loop of the SECIS element deleted (as indicated by light shading in Fig. 2).

Fig. 5

Biosynthetic incorporation of ⁷⁵Se into GPx1 selenoprotein. Saos-2 cells were transfected with epitope-tagged GPx1 constructs as indicated in the top margin, incubated with ⁷⁵Se, lysed and immunoprecipitated with antibody to human GPx1 or to the epitope tag as indicated in the lower margin, then analyzed by SDS–PAGE and autoradiography. The autoradiogram shows Lanes 1, 2: immunoprecipitated selenoprotein from non-transfected cells; Lanes 3, 4: from cells transfected with an epitope-tagged GPx1 construct with an intact SECIS element; Lanes 5, 6: from cells transfected with an epitope-tagged GPx1 construct with the SECIS element deleted. Lanes 1, 3, 5: immunoprecipitated with anti-GPx1 antibody; Lanes 2, 4, 6: immunoprecipitated with antibody to the epitope tag (“Epi”).

Fig. 6

Inhibition of NSEP1 protein expression by siRNAs. Western blot assay for detection of NSEP1 (top part) and β-actin (bottom part) from lysates of HEK 293 cells transfected with siRNA duplexes as indicated in the bottom margin: none, no siRNA; GFP, GFP-silencing siRNA duplex; 1, NSEP1 siRNA duplex 1; 2, NSEP1 siRNA duplex 2; 3, NSEP1 siRNA duplex 3. HEK 293 cells were transfected with the indicated siRNA constructs, incubated 48 h, then lysed and analyzed by Western blotting for the proteins indicated in the left margin.

Fig. 7

Effect of NSEP1 siRNA duplexes on SECIS-dependent luciferase activity in HEK 293 cells. Cells were cotransfected with reporter plasmid pBPHsec (light bars) or pH9A (dark bars) plus siRNA duplexes as indicated in the bottom margin: None, no siRNA; GFP, GFP-silencing siRNA duplex; 1, NSEP1 siRNA duplex 1; 2, NSEP1 siRNA duplex 2; 3, NSEP1 siRNA duplex 3. The ordinate indicates luciferase activity as a percent of no-siRNA controls; bar heights and error bars indicate the mean and standard deviations of triplicate measurements, normalized for transfection efficiency by Renilla luciferase activity derived by co-transfection of plasmid pRL-TK.