Expression of the miR-302/367 microRNA cluster is regulated by a conserved long non-coding host-gene

Abstract

MicroRNAs are important regulators of cellular functions. MiR-302/367 is a polycistronic miRNA cluster that can induce and maintain pluripotency. Here we investigate the transcriptional control and the processing of the miR-302 host-gene in mice. Our results indicate that the mmu-miR-302 host-gene is alternatively spliced, polyadenylated and exported from the nucleus. The regulatory sequences extend at least 2 kb upstream of the transcription start site and contain several conserved binding sites for both transcriptional activators and repressors. The gene structure and regulatory elements are highly conserved between mouse and human. So far, regulating miR-302 expression is the only known function of the miR-302 host-gene. Even though we here only provide one example, regulation of microRNA transcription might be a so far little recognized function of long non-coding RNA genes.

Figure 1

Comparison of murine (a) and human (b) miR-302 host-gene structures (LOC110008574 (Gm51018) and LOC109864269 respectively). The genomic coordinates for the mmu-miR-302 host-gene and the hsa-miR-302 host-gene are located at chr3:127,545,089–127,546,901 plus strand (GRCm38/mm10) and chr4:112,646,720–112,648,703 minus strand (GRCh38/hg38) respectively. The murine gene lacks the central exon but has four potential poly adenylation signals. Note the different scales.

Figure 2

Processing of the mmu-miR-302 host transcript. The numbering is based on the longest variant of the transcript. (a–c) Transcription start sites. (d) Alternative splicing, splice donors shown in red. (e, f, g) Different 3′ terminations. The corresponding sequnces were deposited to GenBank with accession numbers: a: KT932380; b: KT932381; c: KT932382; d: KT932383; e: KT932386; f: KT932384; g: KT932385.

Figure 3

Distribution of the mmu-miR-302 host RNA in the cell. (a) Western blot analysis of sub cellular fractions for two biological replicates showing little β-tubuli in the nucleus and high enrichment of histone H3 in the nucleus indicating efficient separation of cytoplasmic and nuclear compartments. Equal amounts of the fractions were loaded. The original full picture of the entire Western blot membranes are shown in the suplemental Figure S2. QRT-PCR quantification of the Gapdh (b), Hprt (c) and Oct4 (d) as cytoplasmic refernce RNA transcript and Malat-1 (e) as nuclear reference RNA confirmed efficient fractionation. MiR-302 host RNA (f) found in both cytoplasmic (> 54%) and nucleus (> 45%) indicating it is significantly exported to the cytoplasm, however its accumulated level in the cytoplasm is lower and not comparable to the reference RNA transcripts. Bar plots show the average of the transformed Ct (2^-Ct) calculated to visualize the relative localization level of each respective transcript in percentage of the cytoplasmic and nuclear fractions of the same sample. Error bars representing ± SD, n = 3 and, p < 0.05 as significance level.

Figure 4

Functional analysis of the 2.1 kb upstream genomic region of the mmu-miR-302 host-gene. (a) Schematic representation of the 3 regions analyzed: part A (+ 45 to -595), part AB (+ 45 to −856), and part ABC (+ 45 to −2120). (b) Comparison of the promoter/enhancer activity of the three regions in different cell types, normalized to the expression level of the promoter-less reporter. The baseline expression level is relatively high, which allows to test for repressor activity, which appears as values below 1. F9 cells (2, 6, 10) show consistently a higher reporter expression than ES cells (1, 5, 9) or teratoma derived stem cells selected for miR-302 host-gene expression (teratoma + G418: 3,7,11). If selection is stopped, teratoma derived cells start to differentiate and lose reporter expression (teratoma -G418: 4, 8, 12). Bars represent mean ± SD, n = 3. Statistical comparison of data: 1 vs. 4 p < 0.01, 2 vs. 4 p < 0.0001, 3 vs. 4 p < 0.01, 5 vs. 8 p < 0.01, 6 vs. 8 p < 0.0001, 7 vs. 8 p < 0.01, 9 vs. 10 p < 0.05, 9 vs. 12 p < 0.0001, 10 vs. 11 p < 0.001, 10 vs. 12 p < 0.0001 and 11 vs. 12 p < 0.0001, 1 vs. 9 p < 0.05, 5 vs. 9 p < 0.01, 2 vs. 10 p < 0.01 and 6 vs. 10 p < 0.001. All other comparisons are not significant (p > 0.05). Note the logarithmic scale.