Role of PRY-1/Axin in heterochronic miRNA-mediated seam cell development

Abstract

Background: Caenorhabditis elegans seam cells serve as a good model to understand how genes and signaling pathways interact to control asymmetric cell fates. The stage-specific pattern of seam cell division is coordinated by a genetic network that includes WNT asymmetry pathway components WRM-1, LIT-1, and POP-1, as well as heterochronic microRNAs (miRNAs) and their downstream targets. Mutations in pry-1, a negative regulator of WNT signaling that belongs to the Axin family, were shown to cause seam cell defects; however, the mechanism of PRY-1 action and its interactions with miRNAs remain unclear.

Results: We found that pry-1 mutants in C. elegans exhibit seam cell, cuticle, and alae defects. To examine this further, a miRNA transcriptome analysis was carried out, which showed that let-7 (miR-48, miR-84, miR-241) and lin-4 (lin-4, miR-237) family members were upregulated in the absence of pry-1 function. Similar phenotypes and patterns of miRNA overexpression were also observed in C. briggsae pry-1 mutants, a species that is closely related to C. elegans. RNA interference-mediated silencing of wrm-1 and lit-1 in the C. elegans pry-1 mutants rescued the seam cell defect, whereas pop-1 silencing enhanced the phenotype, suggesting that all three proteins are likely important for PRY-1 function in seam cells. We also found that these miRNAs were overexpressed in pop-1 hypomorphic animals, suggesting that PRY-1 may be required for POP-1-mediated miRNA suppression. Analysis of the let-7 and lin-4-family heterochronic targets, lin-28 and hbl-1, showed that both genes were significantly downregulated in pry-1 mutants, and furthermore, lin-28 silencing reduced the number of seam cells in mutant animals.

Conclusions: Our results show that PRY-1 plays a conserved role to maintain normal expression of heterochronic miRNAs in nematodes. Furthermore, we demonstrated that PRY-1 acts upstream of the WNT asymmetry pathway components WRM-1, LIT-1, and POP-1, and miRNA target genes in seam cell development.

Keywords: Axin; C. briggsae; C. elegans; Heterochronic development; Pry-1; Seam cell; WNT asymmetry pathway; miRNA.

Fig. 1

Seam cell asymmetric divisions during postembryonic development in C. elegans. During postembryonic development, C. elegans larvae undergo a series of molts, each of which is associated with carefully timed seam cell divisions. In the V1–4, and V6, cells undergo stem cell-like division where the anterior daughter fuses with the hyp7 syncytium. The posterior daughter then self-renews as another ‘stem cell-like’ seam cell. The exception to the asymmetric divisions is in early L2 stage where they undergo one symmetrical division prior to dividing again in an asymmetric manner

Fig. 2

pry-1 mutants display seam cell defects in C. elegans and C. briggsae. Mutations in pry-1 lead to somatic defects. (a) Cuticle is weaker as demonstrated by the cuticle break assay. (b, c) pry-1 mutants having defective alae formation. (b) Control N2 animals have distinct rows of alae. pry-1(mu38) mutants have defect with frequent breaks (see arrow). (c) Quantification of alae defect in control N2 and pry-1(mu38) animals. (d-f) Heterochronic phenotypes in C. briggsae AF16 and Cbr-pry-1 mutants. (d) Alae defects are visible in Cbr-pry-1(sy5353). (e, f) Seam cells in control AF16 (e) and Cbr-pry-1(sy5353) (f) are visualized by adherens-junction-associated marker Cel-dlg-1p::GFP. Cbr-pry-1(sy5353) animals show defects in cell fusion (scale bar 0.1 mm). Boxed areas, marked by dotted lines, have been enlarged in the second row. Scale bars in b, d-f are 0.01 mm. (g) Both the pry-1(mu38) and pry-1(gk3682) animals show increased seam cell numbers compared to control N2 (scale bar 0.1 mm). (h) Each control animal has exactly 16 seam cells, whereas an average of 21 and 19 cells are found in pry-1(mu38) and pry-1(gk3682) mutants, respectively. (i) pry-1(mu38) animals show increased seam cell number by the end of the L2 stage. In panels a, h, and i, each data point represents the mean of at least two replicates (each batch with 30 or more worms) and error bar represents the STD. Student’s t-test was used to determine the statistical significance: *p < 0.05

Fig. 3

pry-1 is necessary for normal miRNA expression in both C. elegans and C. briggsae. (a) Volcano plot of differentially expressed miRNA genes in pry-1(mu38). Red and green dots mark significantly altered upregulated and downregulated transcripts, respectively, with a p-adj value of < 0.05. Orange dots mark transcripts that are not significantly altered (p-adj > 0.05). (b) Data showing log₂ fold expression of miRNAs from RNA-Seq analysis. (c, d) qRT-PCR analysis of miRNAs in Cel-pry-1(mu38) and Cbr-pry-1(sy5353) animals at the L1 stage showed upregulation of all but miR-246. Each data point represents the mean of two replicates and error bar represents the SEM, **p < 0.01. (e, f) miR-246(n4636) mutants exhibit alae defect in the form of gaps (scale bar 0.01 mm) (e) but show no change in the number of seam cells (f). Each data point represents the mean of at least two replicates (each batch with at least 20 worms) and error bar represents the STD

Fig. 4

DE miRNAs and their predicted target genes in pry-1(mu38) are mostly linked to heterochronic development. (a) Enrichment analysis of potential targets of DE miRNAs using GO-terms associated with biological processes. A significant number of genes (29-fold enrichment) are linked to heterochronic development (colour coded in orange). (b) Venn diagram showing the overlap between previously reported protein-coding DE genes (2,660 genes, [25]) and predicted targets of DE miRNAs identified by Targetscan (435 genes, this study) in pry-1(mu38) animals. Further analysis revealed that 435 potential targets are shared between DE miRNAs (36 by lin-4, 115 by miR-48, 115 by miR-84, 36 by miR-237, 115 by miR-241, and 102 by miR-246). A total 111 genes are common between the two data sets, a number that is statistically significant based on the hypergeometric test. (c) Tissue-enrichment analysis of the common set of genes (111) revealed third highest fold enrichment in hypodermal syncytium cells (colour coded in orange)

Fig. 5

WNT asymmetry pathway components WRM-1, LIT-1, and POP-1 are important for PRY-1-mediated expression of miRNAs and seam cell development. (a) Schematic diagram showing the involvement of PRY-1 and other WNT asymmetry pathway components in the specification of seam cell fates. The levels of nuclear POP-1 are high in the anterior cell but low in the posterior cell. (b) RNAi knockdowns of bar-1, pop-1, wrm-1, sys-1 and lit-1 in control N2 and pry-1(mu38) animals. Each data point represents the mean of at least two batches (each batch with at least 30 worms) and error bar represents the STD. Student’s t-test was used to determine the statistical significance: *p < 0.05 (compared to L4440 control). (c) Representative images of control N2 and pry-1(mu38) animals following pop-1 RNAi. The numbers of seam cells are increased in both cases. The phenotype is particularly enhanced in pry-1 mutants (scale bar 0.1 mm). (d) Representative images of two adjacent POP-1::GFP fluorescing cell pairs from V1-V5 seam cell lineages following control and pry-1 RNAi treatments (A-anterior, P- posterior, each dotted line marks a cell pair, scale bar 20 μm). POP-1 asymmetry is disrupted after pry-1 RNAi, resulting in fewer cell pairs having asymmetric localisation of POP-1::GFP in their nuclei. (e, f) qRT-PCR analysis of miRNAs in pop-1(hu9) and pop-1(RNAi) worms at the L1 stage. Similar to pry-1(mu38), all miRNAs, except miR-246, are overexpressed in both pop-1(hu9) (e) and pop-1(RNAi) (f) animals. Each data point represents the mean of two replicates and error bar represents the SEM, *p < 0.05, **p < 0.01

Fig. 6

Analysis of heterochronic genes in pry-1-mediated seam cell development. (a) qRT-PCR experiments at the L1 stage showed that hbl-1 and lin-28 are significantly downregulated in pry-1(mu38) animals. Each data point represents the mean of two replicates and error bar represents the SEM, **p < 0.01. (b) RNAi KD of lin-28 in pry-1(mu38) mutants rescued the seam cell defect. Each data point represents the mean of three batches (each batch with at least 30 worms) and error bar represents the STD. Student’s t-test was used to determine the statistical significance: *p < 0.05 (compared to L4440 control)

Fig. 7

A model summarizing genetic interactions between PRY-1, WNT asymmetric pathway components (WRM-1, LIT-1, and POP-1), heterochronic miRNAs, and the targets of heterochronic miRNAs during L2 stage seam cell development. Our data supports the Ren and Zhang model [18] and places the WNT asymmetric pathway upstream of miRNAs