14-3-3 isoforms bind directly exon B of the 5'-UTR of human surfactant protein A2 mRNA

Abstract

Human surfactant protein (SP) A (SP-A), an innate immunity molecule, is encoded by two genes, SFTPA1 and SFTPA2. The 5'-untranslated splice variant of SP-A2 (ABD), but not SP-A1 (AD), contains exon B (eB). eB is an enhancer for transcription and translation and contains cis-regulatory elements. Specific trans-acting factors, including 14-3-3, bind eB. The 14-3-3 protein family contains seven isoforms that have been found by mass spectrometry in eB electromobility shift assays (Noutsios et al. Am J Physiol Lung Cell Mol Physiol 304: L722-L735, 2013). We used four different approaches to investigate whether 14-3-3 isoforms bind directly to eB. 1) eB RNA pulldown assays showed that 14-3-3 isoforms specifically bind eB. 2) RNA electromobility shift assay complexes were formed using purified 14-3-3 isoforms β, γ, ε, η, σ, and τ, but not isoform ζ, with wild-type eB RNA. 3 and 4) RNA affinity chromatography assays and surface plasmon resonance analysis showed that 14-3-3 isoforms β, γ, ε, η, σ, and τ, but not isoform ζ, specifically and directly bind eB. Inhibition of 14-3-3 isoforms γ, ε, η, and τ/θ with shRNAs in NCI-H441 cells resulted in downregulation of SP-A2 levels but did not affect SP-A1 levels. However, inhibition of 14-3-3 isoform σ was correlated with lower levels of SP-A1 and SP-A2. Inhibition of 14-3-3 isoform ζ/δ, which does not bind eB, had no effect on expression levels of SP-A1 and SP-A2. In conclusion, the 14-3-3 protein family affects differential regulation of SP-A1 and SP-A2 by binding directly to SP-A2 5'-UTR mRNA.

Keywords: 14-3-3; 5′-untranslated region; human pulmonary surfactant protein A; translation.

Fig. 1.

Surfactant protein A (SP-A) 5′-untranslated region (5′-UTR) exon B (eB) RNA specifically pulls down 14-3-3. eB RNA and random [R (control)] RNA (200 pmol) were 3′-end-biotinylated and used in pulldown assays with NCI-H441 cytoplasmic extract (see materials and methods). RNA pulldown specificity was assessed by Western blotting using input, supernatant (Sup), and eluted proteins from eB (Elution eB) and R (Elution R) RNA. Samples were normalized by volume; bands were detected by rabbit anti-pan-14-3-3 Ab (1:1,000 dilution), secondary anti-rabbit horseradish peroxidase-conjugated Ab (1:2,000 dilution), and the Chemiluminescent Nucleic Acid Detection Module kit with 2-min film exposure. Blots represent results from 2 experiments.

Fig. 2.

Expression of a glutathione S-tranfserase (GST)-tagged 14-3-3 recombinant protein (GST-14-3-3) in pGEX-2TK-14-3-3 vectors, purification by GST-Sepharose, and digestion of the GST tag with thrombin. A: representative silver-stained SDS-polyacrylamide gel used to determine purity of expressed proteins. Lane 1, E. coli cell lysate prior to purification; lane 2, eluate after GST-Sepharose purification (GST-14-3-3); lane 3, digestion of purified protein with thrombin; lanes 4 and 5, flow-throughs from GST-Sepharose columns, which are 14-3-3 proteins without the GST tag. B: Western blot of 14-3-3 proteins with a specific anti-14-3-3 antibody. Lanes 1, 2, 3, 4, 5, 6, and 7: detection of recombinant GST-14-3-3 isoform β, γ, ε, ζ, η, σ, and τ, respectively, eluates after GST-Sepharose purification. Gels and blots represent results from 3 experiments.

Fig. 3.

RNA electromobility shift assays (REMSAs) of eB biotinylated RNA with purified 14-3-3 isoforms β, γ, ε, ζ, η, σ, and τ. Lane 1, free probe (negative control); lane 2, eB RNA incubated with purified 14-3-3 isoform ζ does not form complexes; lane 3, eB RNA incubated with 14-3-3 isoform β forms weak eB/14-3-3 isoform β complexes; lanes 4, 5, 6, 7, and 8, eB RNA incubated with purified 14-3-3 isoforms γ, ε, η, σ, τ, respectively, forms eB/14-3-3 complexes; lane 9, eB incubated with a mixture of all 14-3-3 isoforms (3 μg of total protein) results in eB/14-3-3 complexes. S, specific eB ribonucleoprotein complexes; NS, nonspecific shift. In all cases, 3 μg of purified protein were incubated with 200 nM eB riboprobe. The eB/14-3-3 protein complexes were excised from the gel and subjected to mass spectrometry. Gels represent results from 1 experiment.

Fig. 4.

RNA affinity chromatography assays show direct binding of 14-3-3 isoforms to eB RNA. Purified 14-3-3 isoforms β, γ, ε, η, σ, τ, and ζ were incubated with streptavidin magnetic beads coated with 3′-end eB-biotinylated RNA or R RNA. Interacting proteins were eluted and separated on SDS-polyacrylamide gel, transferred to nitrocellulose membrane, and immunoblotted with specific anti-14-3-3 Ab. Input, supernatant (sup.), and eluted fractions from eB (elution eB) and R (elution R) RNA-coated magnetic beads are shown. R RNA served as a negative control. Blots represent results from 2 experiments.

Fig. 5.

Surface plasmon resonance studies of 14-3-3 isoforms binding to SP-A2 5′-UTR eB RNA. Biotinylated 3′-end-labeled eB RNA was immobilized on a streptavidin-coated sensor chip, resulting in capture of 100 RU of eB RNA, and 14-3-3 protein isoforms were injected at 200 nM over the surface. Injections were performed at a flow rate of 5 μl/min for 2 min, followed by 5 min of buffer flow and regeneration of the surface before the next injection. Injections were repeated twice and were aligned at time 0. Sensograms show that 14-3-3 isoforms β, γ, ε, η, σ, and τ bind to eB RNA and form complexes. Sensograms also show that the eB/14-3-3 isoform σ complex is the most stable, whereas the eB/14-3-3 isoform β complex is the least stable. 14-3-3 isoform ζ did not interact with eB RNA. These results are consistent with those from REMSAs and RNA affinity chromatography assays. Traces represent results from 3 experiments.

Fig. 6.

Knockdown of 14-3-3 isoforms in the NCI-H441 cell line and correlation with SP-A2 expression. Three or 4 different shRNAs (1–4) were used to knock down expression of 14-3-3 isoforms γ, ζ/δ, ε, η, σ, and τ by RNA interference. A scrambled RNA that does not interact with any human, mouse, or rat gene was used as negative control. Western blot analysis was used to monitor expression of specific 14-3-3 isoforms (28 kDa), SP-A2 (35 kDa), SP-A1 (35 kDa), and actin (42 kDa). Experimental shRNAs inhibited 14-3-3 isoforms to a varying degree, with some shRNAs being more efficient in silencing 14-3-3 isoforms, and this correlated with levels of SP-A2. SP-A1 levels remained unaffected. A: all 4 shRNAs knocked down 14-3-3 isoform γ and resulted in downregulated expression of SP-A2, but not SP-A1. B: 3 shRNAs knocked down 14-3-3 isoform τ expression and resulted in downregulated expression of SP-A2, but not SP-A1. C: knockdown of 14-3-3 isoform ε correlates with low levels of SP-A2 but does not affect SP-A1. D: inhibition of 14-3-3 isoform η was the highest with shRNA 3, and this correlates with very low levels of SP-A2, although levels of SP-A1 remain unaffected. E: shRNA 3 and 4 knock down 14-3-3 isoform σ, and this correlates with low levels of SP-A1 and SP-A2. F: 3 shRNAs were used to silence YWHAZ gene expressing 14-3-3 isoform ζ. Of these, shRNA 3 showed the most interference; however, shRNA 3 had no effect on SP-A2 or SP-A1 expression levels. Isoform-specific Abs were used to detect 14-3-3 isoforms (see materials and methods). A polyclonal anti-rabbit Ab that recognized all 14-3-3 isoforms (pan Ab) was used to show that shRNAs specifically target only 14-3-3 isoform ζ. Blots represent results from 2 experiments.