Biotin Regulates the Expression of Holocarboxylase Synthetase in the miR-539 Pathway in HEK-293 Cells¹²³

Cell culture.

Human Embryonic Kidney-293 (HEK-293) cells, Jurkat human lymphoblast cells, Caco-2 human colon cells, and IMR-90 primary human lung fibroblasts were purchased from ATCC. In some experiments, transgenic HEK-293 cells that overexpress HCS fused to green fluorescent protein (GFP) were used. Cells were cultured using routine procedures as recommended by ATCC (22). IMR-90 primary cells were used before senescence. For most experiments, cells were cultured in biotin-defined media containing 25 pmol/L biotin (“deficient”), 250 pmol/L biotin (“physiological”), or 10,000 pmol/L biotin (“pharmacological”). These concentrations represent levels typically seen in serum from biotin-deficient, biotin-normal, and biotin-supplemented adults (23, 24). Biotin-defined media were prepared as described previously (2). Carboxylase-bound biotin was routinely monitored using streptavidin blots as marker for cellular biotin (2).

Ecd-miR539-CMV-Luc-HCS3UTR.

The Ecd-miR539-CMV-Luc-HCS3UTR plasmid and other plasmids were built using standard molecular biology techniques (Supplemental Methods) (25–27).

Identification of miR targets in the HCS 3prime-UTR.

The 3′-UTR of human HCS was screened in silico for miR target sequences by using TargetScan (28) and microRNA.org (29). Both softwares predicted a miR-539 target site 1763–1785 bp downstream from the HCS stop codon. The importance of this site in HCS regulation was tested using plasmid CMV-miR539-CMV-Luc-539T (Supplemental Methods).

Luciferase assay.

Luciferase activity was assayed in triplicate, using independent samples; empty vectors and β-galactosidase vectors were used as baseline and transfection controls, respectively (30).

Overexpression of miR-539 in HEK-293 cells.

A miR-539 overexpression plasmid (CMV-miR539-CMV-GFP) was built to determine whether miR-539 decreases the expression of endogenous HCS (Supplemental Methods).

miR-539 promoter activity.

miR-539 promoter activity was quantified using plasmid miR539Pro-GFP in stably transformed HEK-293 cells (Supplemental Methods).

Quantitative real-time PCR (qRT-PCR).

Total RNA was isolated from cells using RNA Spin Mini isolation kits (GE Health). The abundance of miR-539 was quantified by stem-loop qRT-PCR (31). Briefly, primer 5′-GTCGTATCCAGTGCAGGGTCCGAGGTATTCGCACTGGATACGACACACAC-3′ was used for reverse transcription of miR-539 by using the ImProm-II Reverse Transcription System (Promega). miR-539 was quantified by qRT-PCR, using forward primer 5′-CGGCGGGGAGAAATTATCCT-3′ and reverse primer 5′-GTGCAGGGTCCGAGGT-3′. The relative expression of miR-539 was normalized using U6 snRNA as control; U6 snRNA was reverse transcribed using 5′- GTCAGGCAGCGTGCAGGGTCCGAGGTATTCGCACGCTGCCTGACAAAAAT-3′, and PCR amplified using forward primer 5′-CGCAAGGATGACACGCAAATT-3′and reverse primer 5′-GTGCAGGGTCCGAGGT-3′. All reverse transcriptions were conducted using 2 μg of total RNA. For qRT-PCR, the thermocycler was set at 95°C for 15 s and 55°C for 60 s per cycle for a total 40 cycles. Relative changes in mRNA abundances were quantified by using the C_t method; glyceraldehyde-5-phosphate dehydrogenase (GAPDH) mRNA and U6 RNA were used as reference amplicons for data normalization (32, 33).

RNA Spin Mini Isolation kits (GE Health) were used for analysis of luciferase, GFP, and HCS; 1.5 μg of total RNA was reverse transcribed using random hexamer primers (Fermentas). In qRT-PCR, the cDNA was amplified using forward primer 5′-ATGGAAGATAGACTCCACAT-3′ and reverse primer 5′-TGAGACCTGATCCTTAACTTCC-3′ for HCS analysis; forward primer 5′-CGGATTACCAGGGATTTCAGTC-3′ and reverse primer 5′-AAACCGGGAGGTAGATGAGATG-3′ for luciferase analysis; and forward primer 5′-GACCACTACCAGCAGAACAC-3′ and reverse primer 5′- GAACTCCAGCAGGACCATG-3′ for GFP analysis. GAPDH mRNA abundance was used for data normalization; GAPDH cDNA was amplified using forward primer 5′-TCCACTGGCGTCTTCACC-3′ and reverse primer: 5′- GGCAGAGATGATGACCCTTT-3′. All qRT-PCR reactions were performed at 95°C for 15 s and 52°C for 60 s per cycle for a total of 40 cycles.

Northern blots.

Total RNA was isolated from cells using the mirPremier miRNA isolation kit (Sigma) and miR was visualized by using the MicroRNA Northern Blot assay kit (Signosis) according to the manufacturers’ instructions. Biotinylated probes for miR-539 and U6 snRNA (control) were purchased from Signosis.

Western blots.

Whole cell proteins were extracted from HEK-293 cells and resolved by gel electrophoresis using NuPage 4–12% Bis-Tris gels (3). Transblots were probed using the following antibodies: goat anti-luciferase (Promega), rabbit anti-human HCS (C-terminus) (8), and mouse anti-GFP (Roche Applied Science). Equal loading was confirmed using goat anti-GAPDH (Santa Cruz). Histones were extracted from HEK-293 cells by using 1 mol/L HCl and resolved by gel electrophoresis (34). Transblots were probed using goat anti-biotin (Abcam). Equal loading was confirmed using goat anti-human histone H3, C-terminus (Santa Cruz). Proteins were visualized using fluorophore-labeled secondary antibodies and an Odyssey Infrared Imaging system (LICOR).

Chromatin immunoprecipitation (ChIP) assays.

ChIP assays were conducted in HCS-GFP-transformed cells as described (7, 16), with the following minor modifications. Anti-GFP was used for immunoprecipitations and the relative abundance of miR-539 in immunoprecipitated DNA and input DNA was quantified by qRT-PCR using primers 5′ GGTTTTATGACTAAAATCTCTCCATGGG -3′ (forward) and 5′ CCATTGTAAACACCGTGTATGCTC -3′ (reverse), and 100 ng of DNA template. The relative enrichments were normalized for nucleosomal occupancy at the miR-539 locus by precipitating chromatin with rabbit antiserum to the C-terminus in human histone H3 (Abcam). Data are expressed in units of percent of input DNA that was precipitated with antibodies.

Statistics.

Variances among groups were homogenous, as judged by Bartlett‘s test (35). When more than 2 groups were compared, significance of differences among groups was tested by 1-way ANOVA. Fisher's protected least significant difference procedure was used for post hoc testing. When 2 groups were compared, significance of differences was tested by paired t test. StatView 5.0.1 (SAS Institute) was used for statistical analyses. Differences were considered significant if P < 0.05. Data are reported as means ± SD.

Regulation of HCS expression by miR-539.

In silico analysis by TargetScan and microRna.org predicted a miR-539 binding site (5′-AACGAUGUAGGUUUCAUUUCUCA-3′) 1763–1785 basepairs downstream from the stop codon in human HCS mRNA. This prediction was confirmed experimentally by using pVgRXR-transformed HEK-293 cells. Ecdysone receptor and retinoid X receptor (RXR) positive clones were selected with zeocin before transfection with Ecd-mir539-CMV-Luc-HCS3UTR and selection with G418. The activity of luciferase was greater in ponasterone A (PonA)-treated cells (48 h) than in PonA-free controls (Fig. 1A). The observed effect was dose-dependent; inhibition of luciferase activity reached a plateau when the concentration of PonA was 5 μmol/L (Fig. 1A). All subsequent experiments were conducted using 5 μmol/L PonA. In control experiments, we confirmed that treatment with PonA increased the abundance of miR-539 in transformed HEK-293 cells by using qRT-PCR (Fig. 1B), and northern blot analysis (Fig. 1C). qRT-PCR data revealed that miR-539 concentrations in PonA-treated cells were 2.6-fold greater than in PonA-free controls. The low luciferase activity (= HCS reporter) observed in PonA-treated HEK-293 cells was paralleled by a low abundance of both luciferase mRNA (Fig. 1D) and protein (Fig. 1E). Importantly, the abundance of endogenous HCS mRNA was 63% lower in PonA-treated (5 μmol/L) HEK-293 cells than in PonA-free controls (Fig. 1F).

Gene repression by miRNA at the mRNA level is initiated by the formation of a miR-mRNA duplex in the 3′-UTR of mRNA (36). The miR-539 target site in human HCS is evolutionary conserved in mammals (Fig. 2A). The binding site perfectly matches to the 5′ end of the miR-539 with the exception of 2 mismatches in rat HCS.

Plasmid CMV-miR539-CMV-Luc-539T mediates constitutive expression of both miR-539 and luciferase fused to the putative 22-bp miR-539 binding site from human HCS (Fig. 2B). When HEK-293 cells were transfected with CMV-miR539-CMV-Luc-539T, the abundance of miR-539 was 13-fold greater than in cells transfected with the control plasmid CMV-multiple cloning site (MCS)1-CMV-Luc-539T (Fig. 2C). Luciferase mRNA and activity were 39% and 76% lower, respectively, in cells transfected with CMV-miR539-CMV-Luc-539T compared with CMV-MCS1-CMV-Luc-539T (Fig. 2D,E).

miR-539-mediated downregulation of HCS impairs histone biotinylation.

HCS is known to mediate biotinylation of histones (11, 12). Here, we demonstrated that miR-539 may decrease the expression of HCS to an extent that affects histone biotinylation. HEK-293 cells were stably transformed CMV-miR539-CMV-GFP to overexpress miR-539; controls were transformed with the empty vector CMV-MCS1-CMV-GFP (Fig. 3A). When cells were transformed with CMV-miR539-CMV-GFP, the expression of miR-593 was 2.1-fold greater than in controls (Fig. 3B); the high expression of miR-539 coincided with a mRNA abundance of endogenous HCS that was 43% lower in transformed cells than in controls (Fig. 3C). The low expression of HCS in miR-539 overexpression cells caused substantially lower levels of biotinylated histone H4, as judged by western blot analysis and anti-biotin antibody. Similar loading was confirmed by anti-histone H3 (Fig. 3D).

Repression of HCS expression is sensitive to biotin.

The expression of HCS depends on biotin in Jurkat human lymphoblastoma cells (8). Here, we provide evidence that the biotin-dependent expression of HCS in human cells is mediated, at least in part, by miR-539. First, we screened 4 human cell lines cultured in regular commercial media for the relative abundance of miR-539 and observed large differences (Fig. 4A): IMR-90 primary fibroblasts > HEK-293 embryonic kidney > Caco-2 colon carcinoma > Jurkat cells. In subsequent experiments, we determined whether miR-539 abundance depends on the concentration of biotin in culture media. When HEK-293 cells, Caco-2 cells, Jurkat cells, and IMR-90 fibroblasts were cultured in biotin-defined media (Fig. 4BndashE), the abundance of miR-539 was greater in physiological medium compared with other media in HEK-293 and Jurkat cells (Fig. 4C,D) but not in IMR-90 and Caco-2 cells (Fig. 4B,E); we confirmed this observation in multiple independent repeats and consistently obtained the same trends and significance levels (not shown). We were puzzled by the lack of a linear dose-response curve of miR-539 in biotin-defined cells and conducted additional tests to confirm that miR-539 expression is truly greater in cells cultured in physiological media than in deficient and pharmacological media. In these experiments, HEK-293 cells were stably transformed with plasmid miR539Pro-GFP and cultured in biotin-defined media. GFP mRNA abundance was significantly greater in cells cultured in physiological medium than in cells cultured in deficient or pharmacological media (Fig. 4F), consistent with the patterns in HEK-293 wild-type cells (Fig. 4C). Next, we quantified HCS mRNA levels in biotin-defined cells. In HEK-293 cells, high miR-539 levels coincided with low HCS mRNA levels; i.e., HCS mRNA was lower in physiological medium than in other media (Fig. 5A). A similar trend was observed for IMR-90 fibroblasts but those trends were not statistically significant (P = 0.20) (Fig. 5B). In Jurkat (Fig. 5C) and Caco-2 cells (Fig. 5D), HCS mRNA levels increased with increasing biotin supply. These observations suggest that the abundance of miR-539 depends on the availability of biotin in all the tissues tested here, but that in some tissues alternative mechanisms of HCS regulation override possible effects of miR-539 (see Discussion).

The expression of miR-539 depends on HCS.

The abundance of HCS mRNA was 33-fold greater in HEK-293 cells transformed with HCS-GFP compared with controls (Fig. 6A), whereas the abundance of miR-539 mRNA was 439% greater in HCS-GFP-transformed cells compared with controls (Fig. 6B). The enrichment of HCS at the miR-539 locus was not significantly altered by overexpression of HCS, as judged by ChIP assay (data not shown), suggesting that biotin-dependent chromatin remodeling processes play no role in this feedback loop.