The lin-4 microRNA targets the LIN-14 transcription factor to inhibit netrin-mediated axon attraction

Abstract

miR-125 microRNAs, such as lin-4 in Caenorhabditis elegans, were among the first microRNAs discovered, are phylogenetically conserved, and have been implicated in regulating developmental timing. Here, we showed that loss-of-function mutations in lin-4 microRNA increased axon attraction mediated by the netrin homolog UNC-6. The absence of lin-4 microRNA suppressed the axon guidance defects of anterior ventral microtubule (AVM) neurons caused by loss-of-function mutations in slt-1, which encodes a repulsive guidance cue. Selective expression of lin-4 microRNA in AVM neurons of lin-4-null animals indicated that the effect of lin-4 on AVM axon guidance was cell-autonomous. Promoter reporter analysis suggested that lin-4 was likely expressed strongly in AVM neurons during the developmental time frame that the axons are guided to their targets. In contrast, the lin-4 reporter was barely detectable in anterior lateral microtubule (ALM) neurons, axon guidance of which is insensitive to netrin. In AVM neurons, the transcription factor LIN-14, a target of lin-4 microRNA, stimulated UNC-6-mediated ventral guidance of the AVM axon. LIN-14 promoted attraction of the AVM axon through the UNC-6 receptor UNC-40 [the worm homolog of vertebrate Deleted in Colorectal Cancer (DCC)] and its cofactor MADD-2, which signals through both the UNC-34 (Ena) and the CED-10 (Rac1) downstream pathways. LIN-14 stimulated UNC-6-mediated axon attraction in part by increasing UNC-40 abundance. Our study indicated that lin-4 microRNA reduced the activity of LIN-14 to terminate UNC-6-mediated axon guidance of AVM neurons.

Fig. 1. Elimination of alg-1 enhances netrin-dependent axon attraction, suppressing AVM axon ventral guidance defects in slt-1 mutants

(A) Schematic drawing highlights signaling pathways that influence AVM axon ventral guidance in development. (B) Schematic diagram of wild-type and mutant AVM axons. (C) alg-1 mutation specifically suppressed AVM axon ventral guidance defects in slt-1 (slit) but not unc-6 (netrin) mutants. alg-1(gk214) and alg-2(ok304) used in this study are null alleles. dcr-1(mg375) is a reduction-of-function allele. Error bars represent standard error of the proportion. Asterisks and brackets represent P < 0.001 by Z-test for two proportions.

Fig. 2. Global survey of microRNA expression in C. elegans development

(A) Quantification of microRNAs involves two steps, stem-loop RT and conventional PCR. The stem-loop RT primer binds to the 3′ portion of the mature microRNA and is reverse transcribed with reverse transcriptase. Subsequently, the RT product is quantified using conventional PCR that includes a microRNA-specific forward primer and an universal reverse primer. The forward primer is tailed at 5′ end to increase Tm. (B) Stem-loop RT-PCR analysis of RNA isolated from populations of staged animals. Equal amount of RNA preparation from staged N2 animals was used in the RT-PCR amplification of microRNA and actin transcripts. E, L1, L2, L3, L4, and YA indicate embryonic, the first larval, the second larval, the third larval, the forth larval, and the young adult stages, respectively. MicroRNAs were clustered based on timing of their half-maximal expression.

Fig. 3. Fluorescent protein promoter reporters for identifying neuronal expression of lin-4 and stem-loop RT-PCR for detecting specifically the mature lin-4 microRNA

(A) Sequence alignment of C. elegans lin-4 microRNA with Drosophila melanogaster miR-125, Xenopus miR-125a and miR-125b, and Mouse miR-125a and miR-125b. Letters in red highlight identical sequences at the 5′ seed region, which is important for target mRNA recognition. (B–O) Expression of lin-4 microRNA in neurons. (B) A late embryo. (C) A late L1-stage animal, whole body. Anterior body in a late L1-stage animal (D–F), a L2-stage animal (G–I), a L3-stage animal (J–L), and a L4-stage animal (M–O). A 1.9-kb lin-4 promoter drives GFP expression in the distinct cells (B, C, D, G, J, and M). Limited lin-4 expression in the late embryonic stage in a few head neurons. Broader lin-4 expression in the late L1 stage in many neurons in the head, the body, and the tail. mec-4::dsred reporter was used to label the AVM neurons (E, H, K, and N). The superimposed images identify the GFP expressing cells as the AVM neurons (F, I, L, and O). Asterisk indicates the ALM neuron. Anterior is to left, dorsal up. Scale bar, 20 μm. (P) Measuring mature lin-4 microRNA level by stem-loop RT-PCR in alg-1 mutants versus wild-type (N2) animals. Equal amount of RNA preparation from either N2 or alg-1 mutants was used in the stem-loop RT-PCR reaction for amplifying lin-4 transcripts. Comparable amounts of actin transcripts amplified from the same RNA preparations from N2 and alg-1 mutants serve as an internal control. Value shown above each lane indicates amount of lin-4 normalized by actin.

Fig. 4. Enhancing netrin-mediated axon attraction by a lin-4 microRNA mutation

(A) Specific rescue of AVM axon guidance defects in slt-1 mutants by a loss-of-function lin-4 microRNA mutation. Anterior is to left, dorsal up. Scale bar, 20 μm. (B) Mutation in lin-4 microRNA specifically suppressed AVM axon ventral guidance defects in slt-1 (slit) and eva-1 but not unc-6 (netrin) and unc-40 (DCC) mutants. Error bars represent standard error of the proportion. Asterisks indicate that comparisons between slt-1 and lin-4; slt-1 and between eva-1 and lin-4; eva-1 are significantly different at P < 0.01 by Z-test for two proportions. (C) Cell-autonomous rescue of lin-4 mutant phenotypes in AVM axon guidance by a mec-4::lin-4 transgene. unc-54 and ajm-1 promoters were used to drive lin-4 expression in body wall muscles and hypodermal cells, respectively. Asterisks indicate that comparison is significantly different between lin-4; slt-1 with and without mec-4::lin-4 transgene at P < 0.01 by Z-test for two proportions.

Fig. 5. The transcription factor lin-14 enhances netrin-mediated axon attraction

(A, B) Expression of lin-14 in neurons. (A) A late embryo. Red arrowheads indicate ventral cord motor neurons. (B) A late L1-stage animal. Scale bar, 20 μm. (C) A gain-of-function lin-14 mutant allele (n355gf) recapitulates lin-4 loss-of-function phenotype in enhancing AVM axon ventral guidance. (D) Reduced lin-14 activity suppresses lin-4 loss-of-function phenotype in AVM axon ventral guidance. (E) Specific rescue of AVM axon guidance defects in slt-1 mutants by overexpressing lin-14 in the AVM neurons. Anterior is to left, dorsal up. Scale bar, 20 μm. (F) A cell-type specific promoter-driven lin-14 expression in the AVM neurons can enhance netrin-dependent attraction, suppressing ventral guidance defects in slt-1 mutants. In all figures, error bars represent standard error of the proportion. Asterisks represent P < 0.01 by Z-test for two proportions. (G) Model of developmental switch of responsiveness to netrin by lin-4 microRNA. netrin stimulates axon attraction in early AVM neurons through the transcription factor lin-14. lin-4 microRNA expressed in later AVM neurons stops netrin-mediated axon attraction through inhibition of lin-14 expression.

Fig. 6. Down-regulation of lin-14 3′UTR by endogenous lin-4 microRNA in the AVM neurons

(A) Organization of Pmec-4::GFP::lin-14 3′UTR sensor construct. Seven lin-4 microRNA binding sites were computationally predicted in the lin-14 3′UTR. Sensor constructs were injected at 1ng μl⁻¹. (B) Representative animals express a sensor with either the lin-14 3′UTR or a control (unc-54) 3′UTR. (C) Quantification of sensor expression in AVM neurons. The fluorescent intensity of Pmec-4::GFP::lin-14 3′UTR sensor is normalized by Pmec-4::mCherry::unc-54 3′UTR. When repression of a lin-14 3′UTR sensor by lin-4 occurs, a substantially higher GFP (normalized by mCherry) signal is detected in the AVM neurons in lin-4 mutants. Error bars indicate the s.e.m. Asterisks represent P < 0.001 by Student’s t-Test.

Fig. 7. Dynamic regulation of lin-14 3′UTR activity during AVM axon guidance in the L1 stage

(A) Temporal expression of Plin-14::GFP and Plin-4::GFP reporters in the AVM neurons. An early L1-stage animal (top panel) showing lin-14 strongly expressed in AVM. An early L1-stage animal (middle panel) showing lin-4 weakly expressed in AVM and a late L1-stage animal (lower panel) showing lin-4 strongly expressed in AVM. The Pmec-4::mCherry reporter was used to label the AVM neurons. The merged images identify the GFP expressing cells as the AVM neurons. Arrowhead indicates the AVM neuron and the open arrowhead marks the ALM neuron. Anterior is to left, dorsal up. Scale bar, 20 μm. The expression intensity of the GFP::lin-14 3′UTR sensor in the AVM neurons was measured every hour at the first larval stage, starting at 4 hours after hatching and ending at 12 hours after hatching, in wild-type animals (B) and lin-4 mutants (C). Eight animals each were measured for wild type and lin-4 mutants. Bars represent the average intensity. ** and *** indicate intensity between wild type and lin-4 mutants is significantly different at P < 0.01 and P < 0.001, respectively. P values were calculated using a Student’s t-Test.

Fig. 8. lin-14 functions through unc-40, madd-2, unc-34 and ced-10 to promote netrin-mediated AVM axon ventral guidance

LIN-14 requires UNC-40, MADD-2, UNC-34 and CED-10 to stimulate netrin signaling. AVM was visualized with zdIs5[mec-4::gfp]. Asterisks indicate data significantly different from Ex[lin-14](-) controls (P < 0.001 by Z-test for two proportions).

Fig. 9. Effects of lin-14 on unc-40 protein distribution in the anterior touch neurons

(A) Distribution of UNC-40::GFP and LIN-14::mCherry fusion proteins in the ALM and AVM neurons. (B) Quantification of anterior touch neurons showing UNC-40 protein distribution either restricted to the perinuclear region or expanded to the whole cell, depending on UNC-40 dosage and LIN-14 amount. Using the expression intensity of the Mos[Pmec-4::unc-40::GFP] single copy insertion line as a reference, we estimated 1 copy is the low expression line, 1.2 copies is the medium expression line, and 9.5 copies is the high expression line. (C) Over-expression of unc-40::GFP in the AVM neurons suppressed the AVM axon ventral guidance defects in slt-1 but not unc-6 mutants. In all figures, error bars represent standard error of the proportion. Asterisks and brackets represent P < 0.01 by Z-test for two proportions.