Substrate requirements for let-7 function in the developing zebrafish embryo

Abstract

MicroRNAs (miRNAs) are involved in the regulation of gene expression at the post-transcriptional level by base pairing to the 3'-UTR (untranslated region) of mRNAs. The let-7 miRNA was first discovered in Caenorhabditis elegans and is evolutionarily conserved. We used zebrafish embryos as a vertebrate in vivo system to study substrate requirements for function of let-7. Injection of a double-stranded let-7 miRNA into the zygotes of zebrafish and frogs causes specific phenotypic defects. Only the antisense strand of the let-7 duplex has biological activity. In addition, co-injected mRNA of gfp fused to the 3'-UTR of a zebrafish lin-41 ortholog (a presumed target of let-7) is silenced by let-7. Point mutant studies revealed that the two let-7 target sites in the lin-41 3'-UTR are both essential and sufficient for silencing. let-7 and mir221 together, but not either of them alone, can silence a construct with one of the let-7 target sites replaced by a target site for mir221, showing that two different miRNAs can provide the required cooperative effect. let-7 target sites can be moved around: they are also functional when positioned in the coding sequence or even in the 5'-UTR of gfp. We took advantage of reporter and phenotypic assays to analyze the activity of all possible point mutant derivatives of let-7 and found that only the 5' region is critical for function of let-7.

Figure 1

Misexpression of let-7 in vertebrate embryos. (A) Lateral views of Danio rerio and X.tropicalis embryos injected with let-7 and a mutant version containing five mismatches (underlined). (B) Phenotype induced by injection of a pre-let-7 hairpin. The northern blot shows processing of pre-let-7 in zebrafish embryos. Pictures were taken at 28 h post-fertilization (hpf) and 2 days post-fertilization (dpf) for D.rerio and X.tropicalis, respectively. NIC, not injected control.

Figure 2

Phenotypic effects of injection of 3 let-7 heteroduplexes. (A) Sequences of heteroduplexes. The upper strand is the antisense strand of the let-7 duplex. (B) Phenotypes of 28 h zebrafish embryos injected with heteroduplexes. (C) Bar diagram showing relative length of the yolk sac extension of fish injected with let-7 heteroduplexes.

Figure 3

Effects of let-7 injection on gfp::lin-41 fusions and zebrafish development. (A) Structure of let-7 bound to its target sites in the zebrafish lin-41 3′-UTR. The position of the point mutation in the target and let-7 is indicated in red. At these positions the C in the target is changed to a T and the G in let-7 is changed to an A. (B) Lateral views of 28 h zebrafish embryos injected with gfp with and without the lin-41 3′-UTR. The upper panel shows phenotypes and the lower panel GFP fluorescence. (C) Western and northern analysis of GFP in 28 h zebrafish embryos co-injected with gfp:lin-41 3′-UTR. (D) Phenotypes and GFP expression in 28 h embryos injected with mutant (gfp::lin-41 3′-UTR-2, let-7mm5) and wild-type (gfp::lin-41 3′-UTR-1, let-7) let-7 and gfp::lin-41 3′-UTR fusions.

Figure 4

let-7 target site 1 and 2 are sufficient for silencing. (A) Overview of constructs and miRNAs analyzed in this study. + indicates strong silencing, − indicates no silencing (B) GFP fluorescence in 24 h zebrafish embryos injected with let-7 or let-7mm5 and gfp::let-7tar 5′-UTR or gfp::let-7tar cds.

Figure 5

Mutational analysis of the let-7 miRNA. (A) Phenotypes were scored by measuring the length of the yolk sac extension relative to the diameter of the yolk (bar diagram). For each data point, five individuals were measured. Approximate GFP levels, as assessed from western blots (B), are indicated in the table. − indicates that GFP protein level is similar to a control where only the gfp::lin-41 3′-UTR mRNA was injected; + indicates a significant decline of GFP protein level compared with the control; 0 indicates a moderate decline of GFP protein level compared with the control. At each position, the three types of possible mutations are shown in the sequences beneath the bars. The first series of mutants (purple bars and sequence a) has an equal number of hydrogen bonds compared with the wild-type let-7 duplex. Both other series (dark red and gray bars and sequences b and c) have a different number of hydrogen bonds, depending on the nature of the mutation, i.e. U/A to C/G or vice versa. The let-7 sequence is depicted in red. (B) Western blot analysis of GFP protein level for 3 × 22 mutant derivatives of let-7. Protein was extracted from 5 embryos. For each sample, 5 mg of protein was loaded. As an injection control, the level of gfp::lin-41 3′-UTR mRNA was determined by northern blotting. For each sample, 3 μg of RNA, obtained from 10 individuals, was loaded. Numbering is from the 5′ end of let-7 and corresponds to that in (A).