Description

A high diversity of cilium types have evolved to carry out a multitude of sensorial and motility functions in animals. It is therefore not surprising that defects in cilium structure and function have been associated with various severe multi-organ disorders, commonly known as ciliopathies (Sreekumar and Norris 2019). Regardless of their type, cilium assembly relies on bidirectional motor-driven transport known as intraflagellar transport (IFT; Prevo et al. 2017). Anterograde IFT driven by kinesin-2 transports different cargos in IFT trains along axonemal microtubules from the ciliary base to the tip of the cilium. Upon reaching the tip, these IFT trains are dismounted, rearranged and prepared for transport in the opposite direction back to the ciliary base by the dynein-2 motor complex. IFT trains are composed of at least 16 IFT-B subunits and 6 IFT-A subunits (Prevo et al. 2017). IFT20 is part of the IFT-B complex and has been shown to be of critical importance for ciliogenesis in human and mouse models (Follit et al. 2006; Jonassen et al. 2008; Katoh et al. 2017; Lim et al. 2020).

In C. elegans, IFT-20 is encoded by the Y110A7A.20 locus on chromosome I. Y110A7A.20 was identified in 2004 as a locus containing a X-box sequence (Fan et al. 2004; Blacque et al. 2005; Efimenko et al. 2005) under the control of DAF-19, a RFX-type transcription factor that regulates ciliary gene expression in ciliated sensory neurons (Swoboda et al. 2000) (the only ciliated cells in C. elegans). Expression of a GFP reporter under the control of the ift-20 promoter revealed that IFT-20 expression is restricted to sensory neurons, similarly to other IFT components (Blacque et al. 2005). In fact, disruption of daf-19 resulted in the almost complete loss of expression of this GFP reporter (Blacque et al. 2005), confirming that DAF-19 drives IFT-20 expression most likely by recognizing the X-box at position -59 of the ift-20 promoter (Blacque et al. 2005; Efimenko et al. 2005). Later, Ou and colleagues constructed a vector for IFT-20::GFP expression, and found that the C. elegans IFT20 homolog is also recruited to cilia where it undergoes IFT with similar kinetics to other IFT subunits (Ou et al. 2007). However, in contrast to its homolog in mammalian cells, IFT-20 was not detectable at the Golgi and does not seem to interact with the C. elegans homolog of GMAP210 (Broekhuis et al. 2013), which anchors IFT20 to the Golgi in mammalian cells (Follit et al. 2008).

In spite of these early analyses of IFT-20 expression and localization, the extent to which IFT-20 contributes to ciliogenesis in C. elegans remains unknown. To dissect the importance of IFT-20, we obtained and mapped a null deletion allele (ok3191) of the ift-20 locus (Figure 1A). This allele carries a deletion of 386 bps that includes part of the 5′ UTR, the X-box motif that controls ift-20 expression, the start codon, and all of the 1st exon (sequencing data was submitted to WormBase). Because the expression of ift-20 was previously shown to be under the control of the DAF-19 transcription factor, and since ok3191 lacks the complete first exon, we expect that this allele results in the complete loss of IFT-20. Thus, we refer to the ift-20 (ok3191) allele as an ift-20 null.

We started our analysis of ift-20 null worms by performing the classic dye-filling assay to test whether cilia are able to form normally in the absence of IFT-20. This assay takes advantage of the fact that the lipophilic, fluorescent DiI dye is specifically incorporated into a subset of ciliated sensory neurons through their cilia (if these are present and in contact with the outside environment; Extended Data Figure 1A). In contrast to the wild-type controls that incorporated the dye, none of the ift-20 null worms had DiI signal in their neurons (Figure 1B; Extended Data Figure 1B). To confirm that this dye-filling phenotype results from ciliary defects rather than abnormal dendritic morphology/extension, we expressed cytoplasmic GFP in sensory neurons (driven by the osm-6/ift-52 promoter) to analyze their morphology. We found no significant differences in the length and morphology of sensory neurons (Extended Data Figure 1C), which, together with our dye-filling results, suggest that IFT-20 loss leads to defects in sensory cilia.

Using established markers for the cilium base (HYLS-1) and the ciliary transition zone (CCEP-290) (Schouteden et al. 2015), we found that ift-20 null worms still positioned their basal body normally at the extremity of the dendrite in phasmid neurons and were able to form at least the CCEP-290 module of the ciliary transition zone (Figure 1C). We then took advantage of the incorporation of the soluble/cytoplasmic GFP into cilia of phasmid sensory neurons to visualize cilia and measure their length in ift-20 null worms. Strikingly, we found that severely shortened/vestigial cilia were still able to form in phasmid neurons of ift-20 null animals (Figure 1D,E,G). This suggests that IFT-20 loss leads to a severe defect in ciliary axoneme assembly but does not completely block ciliogenesis in C. elegans.

In order to directly visualize IFT particles in cilia of ift-20 null animals, we analyzed the distribution of IFT-74::GFP (Yi et al. 2017), another subunit of the IFT-B complex. IFT-74::GFP was readily detectable in dendrites of amphid and phasmid neurons, which allowed us to further confirm the normal development and morphology of these neurons in ift-20 null worms (Extended Data Figure 1D). In controls, IFT-74::GFP was strongly recruited to cilia, labeling the complete axoneme (Figure 1F,H). In contrast, IFT-74::GFP-labeled cilia were severely shortened in the majority of ift-20 null phasmid neurons (67.8%; Figure 1F,H), consistent with our observations using soluble GFP-labeled cilia. Interestingly, however, no IFT-74::GFP-labeled cilia were detectable in the remaining fraction of phasmid neurons (32.2%; Figure 1F), suggesting that IFT-20 is also required for IFT-74 incorporation into cilia. To further confirm this possibility, we quantified the distribution of any detectable IFT-74::GFP signal in phasmid cilia of ift-20 null worms. We found that the signal intensity of IFT-74::GFP recruited to the ciliary base and inside cilia of phasmid sensory neurons was indeed strongly reduced in ift-20 null worms (Figure 1I). These results reveal that, in addition to being important for cilia assembly in C. elegans, IFT-20 plays a role in the ciliary recruitment and entry of IFT subunits.

The fact that short cilia could still form in phasmid sensory neurons of ift-20 null worms was somewhat unexpected given the critical role of IFT20 for the initiation of ciliogenesis in most mammalian models (Follit et al. 2006; Jonassen et al. 2008; Katoh et al. 2017). Interestingly, however, we note that very short flagella with abnormal axonemes (but still motile) can form in sperm of mice with male germline-specific disruption of IFT20 (Zhang et al. 2016).

To assess the degree of functionality of the IFT-20-deficient vestigial cilia, we performed osmotic and mating assays. When performing the osmotic avoidance assay, virtually all ift-20 null worms readily crossed the glycerol rings that wild-type worms avoided (Figure 1J,K). This suggests that IFT-20-deficient cilia fail to sense normally repulsive concentrations of glycerol. In agreement with defective ciliary sensory functions, all ift-20 null male worms failed to mate (Figure 1L), suggesting that IFT-20-deficient cilia in male tails were unable to detect wild-type hermaphrodites for mating. Together, the results of these behavioral assays strongly suggest that IFT-20 loss completely impairs cilia-dependent functions in C. elegans. Although we did not analyze the morphology and IFT of all types of sensory cilia in the ift-20 null, our analysis of cilia-dependent behavior is consistent with the cilium assembly defects that we observed in ift-20 null phasmid sensory neurons.

Altogether, our results show that IFT-20 is required for ciliogenesis and cilia sensory functions in C. elegans. Our data also suggest that IFT-20 is important for the robust recruitment and entry of IFT subunits into vestigial cilia that assemble in ift-20 null animals.