SP1 and NFY Regulate the Expression of PNPT1, a Gene Encoding a Mitochondrial Protein Involved in Cancer

Abstract

The Polyribonucleotide nucleotidyltransferase 1 gene (PNPT1) encodes polynucleotide phosphorylase (PNPase), a 3'-5' exoribonuclease involved in mitochondrial RNA degradation and surveillance and RNA import into the mitochondrion. Here, we have characterized the PNPT1 promoter by in silico analysis, luciferase reporter assays, electrophoretic mobility shift assays (EMSA), chromatin immunoprecipitation (ChIP), siRNA-based mRNA silencing and RT-qPCR. We show that the Specificity protein 1 (SP1) transcription factor and Nuclear transcription factor Y (NFY) bind the PNPT1 promoter, and have a relevant role regulating the promoter activity, PNPT1 expression, and mitochondrial activity. We also found in Kaplan-Meier survival curves that a high expression of either PNPase, SP1 or NFY subunit A (NFYA) is associated with a poor prognosis in liver cancer. In summary, our results show the relevance of SP1 and NFY in PNPT1 expression, and point to SP1/NFY and PNPase as possible targets in anti-cancer therapy.

Keywords: NFYA; PNPT1; SP1; liver cancer; mitochondria.

Figure 1

Putative motifs of the human PNPT1 promoter and its conservation across species. (A) Nucleotides 4692 to 5095 of NG_033012.1 NCBI Reference sequence were analyzed with MatInspector (Genomatix) and Promo (Alggen) online applications. ISRE, GC-box and NFY binding site are highlighted with solid boxes, and putative SP1 binding site with a dotted box. NSRE-like sequence is underlined, and initial ATG (Inr) is shown in bold. Previously reported Transcription Start Sites (TSS) are shown. (B) Nucleotides 4984 to 5103 of NG_033012.1 NCBI Reference sequence corresponding to the human (H. sapiens) PNPT1 promoter were aligned with homologous sequences of rabbit (O. cuniculus), cow (B. taurus) and sheep (O. aries), using Clustal Omega software (EMBL-EBI) with default settings. Animal sequences were previously retrieved with the Ensembl Genome Browser by means of a BLAST search using the human sequence as query. Nucleotides matching INRS, SP1 and NFY binding consensus sequences are shown with grey background. NRSE-like sequences are shown boxed, and nucleotides matching the reported NSRE consensus sequences are highlighted with red bold characters. Asterisks indicate nucleotides conserved in the four sequences.

Figure 2

Mutational analysis of the predicted motifs of the human PNPT1 promoter. HeLa cells were transiently transfected with a luciferase reporter plasmid containing 396 bp (nucleotides 4692–5087 of NG_033012.1 NCBI Reference sequence) of the PNPT1 promoter (396bp-PNPT1-prom-luc) 5′ upstream from the ATG (WT, wild-type), or derived plasmids containing deleterious point mutations of the indicated motifs. The activity of the reporter was assayed 40 h after transfection. We represent the relative activity with respect to the native version of the promoter (WT). Statistical analysis: 1 way ANOVA /Tuckey: * p < 0.05; ** p < 0.01; n = 6 (3 assays on duplicated).

Figure 3

Mapping transcription complex binding sites in the 400-bp PNPT1 promoter using EMSA and ChIP. (A) biotinylated double-stranded oligonucleotides representing the indicated cis-elements of 400 bp-PNPT1 promoter were used for EMSA assays. SP1 and NRSE-like motifs overlap and were present in a single oligonucleotide. DNA fragments were incubated in the presence (+) or in the absence (–) of nuclear extracts (NE), and the resulting complexes were analyzed by nondenaturing PAGE and biotin detection. (B,C) the complexes observed in (A) were further competed with, either, growing concentrations of the identical unlabeled oligonucleotide or the corresponding point mutated counterpart, or antibodies to the indicated polypeptides. (C) the arrow points out the complex that is competed with the unlabeled oligonucleotide containing a native SP1 motif or with αSP1 antibodies but not with the native NRSE motif or the presence of αNRSE antibodies. Shown are representative EMSA. (D) SP1 or NFYA chromatin immunoprecipitation in HeLa cells. Immunoprecipitation was performed with antibodies to SP1 or NFYA and the corresponding unspecific control antibody (Cont IgG). Enrichment relative to controls was determined by real-time PCR and expressed as a percentage (fold) over control. Shown are representative ChIP assays performed in triplicate (mean ± SD). Two different pairs of oligonucleotides encompassing the promoter region yield similar results. Statistical analysis: t-test, *** p < 0.001 (n = 6 independent IP).

Figure 4

SP1 and NFYA expression effect on PNPT1 mRNA levels and promoter activity. (A) relative mRNA levels of NFYA, SP1 and PNPT1 were assessed by qPCR in HeLa cells treated with the indicated concentrations of a silencer to SP1 or NFYA (siSP1, siNFYA) or a commercial control silencer (siCONT). (B) HeLa cells were co-transfected with the 396 bp PNPT1 prom-luc plasmid and with the indicated silencer. The luciferase activity was measured in cell lysates and expressed as relative units respect to control cells (mean ± SD). Statistical analysis: 1 way ANOVA/Tuckey: ** p < 0.01, ***, p < 0.001; (A) n = 9 (3 assays on triplicated); (B) n = 8 (2 assays on quadruplicated).

Figure 5

Increased expression of PNPT1, SP1 and NFYA is associated with a poor prognosis in liver cancer. Shown are the Kaplan–Meier survival curves of liver cancer patients, classified as High or Low expression individuals depending on the expression levels of the indicated mRNA species in analyzed cancer tissues. Data were downloaded from Human Protein Atlas (data available from v21.proteinatlas.org searching the indicated genes; all links are shown in Appendix A).