Identification and Characterization of a Novel Allele of Caenorhabditis elegans bbs-7

Primary cilia play a role in the sensation of and response to the surrounding environment. Caenorhabditis elegans (C. elegans) have primary cilia only on the distal tips of some dendrites. In order to better understand the relationship between receptor localization to cilia, cilia structure and cilia function, we have characterized a mutation originally identified in a forward genetic screen for mutants with defective PKD-2 ciliary localization. Through behavioral assays and examination of the structure of cilia in the cil-5 (my13) mutant animals, we have found that my13 disrupts not only receptor localization, but also some cilia-mediated sensory behaviors and cilia structural integrity. We have identified the my13 lesion and found that it is a missense mutation in bbs-7, an ortholog of human BBS-7, a gene known to affect human cilia and to be involved in Bardet-Biedl syndrome. Finally, we show that bbs-7(my13) also affects the glia cells which support the cilia.


Introduction
Primary cilia play a role in sensation of and response to the surrounding environment. They do so for many different cell types. For example, in humans, cilia on kidney epithelial cells act as flow sensors for urine, cilia on ectodermal and mesenchymal cells in the limb bud are required for proper patterning, and cilia on the rod and cone photoreceptor cells are required for maintenance and function of these cells [1][2][3]. The structure of primary cilia and the process by which they are formed are well-conserved across multiple cell types and organisms. The cilium protrudes from the cell membrane and contains a 9+0 arrangement of One question that can be addressed using C. elegans is how PKD-2 is transported from the golgi body to the ciliary base and then localized on the ciliary membrane. Transport of PKD-2 within vesicles has been proposed for movement to the ciliary base [21]. However, unlike some other receptor proteins, such as OSM-9 and OCR-2, PKD-2 is not localized to the ciliary membrane by intraflagellar transport and the mechanism by which this transport occurs is unclear [21,22]. A forward genetic screen to identify additional factors involved in PKD-2 ciliary localization (the Cil phenotype) [17] isolated cil-5(my13). Cilia of male-specific neurons of C. elegans homozygous for my13 curve slightly inward and accumulate additional PKD-2::GFP at the ciliary base [17].
We present further characterization and cloning of cil-5(my13), a novel allele of bbs-7. Here, we show that bbs-7(my13) mutant animals exhibit specific chemosensory defects and aberrations in cilia structure which do not overlap entirely with phenotypes of previously reported bbs-7 alleles. In addition, although the BBS genes are expressed exclusively in ciliated neurons [23], we observed changes in glial cells in the bbs-7(my13) mutant animals. These results provide additional evidence for the role of bbs-7 in cilia structure and cilia function, and crosstalk within the sensilla between neurons and glial cells.

SNP Mapping and Genome Sequencing
bbs-7(my13) homozygous mutant animals carrying PKD-2::GFP were crossed to the polymorphic Hawaiian strain, CB4856. F1 heterozygotes were identified by the presence of PKD-2::GFP and allowed to self-fertilize. F2 homozygous mutant animals were identified by dye-filling (tailDyf), allowed to self-fertilize, and the F3 progeny were pooled. Genomic DNA isolation was performed using the Gentra Puregene kit (Qiagen) as described in Doitsidou et al. 2010 [24]. Genomic DNA samples were also prepared using CB4856 and JLG3 strains as controls. Library preparation and whole genome sequencing were performed at the University of Wisconsin Biotechnology Center. Analysis of the Hawaiian/Bristol ratios of SNP loci was done using Next Generation EMS Mutation Mapping (http://bar. utoronto.ca/ngm/).

Dye-filling
Dye filling was performed as described [25] with the following modifications. For the assay (dye concentration 25 mg/mL), worms were exposed to the dye for two hours. Dye-filling was assessed by observing the extent of dye-filling in the amphids (nose tip, dendrites and cell bodies) and the phasmids.

Epifluorescence Microscopic Analysis
Animals were mounted as detailed in [17]. Fluorescence analysis was performed on a Nikon i80 using a 60x (NA1.4) oil objective. Images are either of single focal planes or optical sections projected as Z-series that were manipulated using NIS elements software. The imaging of the animals with F16F9.3pro: mCherry was performed on a confocal microscope at the model core of the Mayo Translational PKD Center.

Chemotaxis Assays
60mm plates containing nematode growth medium were divided into 5 sections of equal width (labeled A-E). 2 ml of the test chemical were placed at one end of the plate (section A) and 2 ml of the diluent (ethanol) were placed at the other end of the plate (section E). After rinsing age-synchronized 1 day old adult worms in M9, approximately 50 worms were placed in the center of the plate, and their movements were tracked by counting the number of animals in each section every 10 minutes for an hour. The worm chemotaxis index (WCI) was calculated using the equation: WCI 5 [(A+B) -(D+E)]/(A+B+D+E). In order to measure whether there was any difference in the degree of attraction between the strains, we calculated the percentage of animals in the section closest to the chemical compared to the number of total animals that approached the chemical: this is reported as percent strong attraction. (Strength of avoidance was similarly calculated by observing the number of animals in the section furthest from the chemical.)

Quantification of Vacuoles
Animals of each strain were age-matched by bleaching gravid hermaphrodites to obtain synchronous clutches of embryos. The presence of vacuoles in the sheath cells of CHB28 and JLG23 animals was assessed by looking for round structures that were devoid of F16F9.3 pro: mCherry expression in the sheath cells. Each animal was scored as either having vacuoles or not having vacuoles. In each experiment, between five and fifteen animals per age per strain were assessed. The experiment was repeated four times, and the aggregate data are presented.

bbs-7 (my13) Affects Cilia Structure and Function
Results bbs-7 is required for amphid, phasmid and CEM neuron structure C. elegans homozygous for bbs-7(my13) exhibit defects in male-specific ciliated neurons and some ciliated neurons common to both males and hermaphrodites. As previously reported, PKD-2::GFP localization in the male-specific CEM neurons is aberrant [17]. The CEM cilia of mutant animals curve inward instead of extending straight toward the nose tip indicating possible structural defects in the cilia. There is also increased accumulation of PKD-2::GFP at the ciliary base in the mutant animals compared to wild-type ( fig. 1 A and B). Both males and hermaphrodites contain amphid and phasmid neurons. A subset of wild-type amphid and phasmid neurons are exposed to the environment and the lipophilic fluorescent dye, DiI, labels these neurons. The phasmids of my13 mutant hermaphrodites fail to take up DiI (the tailDyf phenotype) while the amphids fill with DiI ( fig. 1 C-H and [17]). However, the amphid dye-filling phenotype is qualitatively different in bbs-7(my13) mutant animals compared to wild-type. In wild-type animals, the amphid dendrites fill uniformly with dye, but in mutant animals, labeling is not consistent and the dye appears to accumulate in pockets along the length of the dendrites ( fig. 1). Both the phasmid and the amphid dyefilling defects are completely penetrant.

bbs-7(my13) modulates behavioral responses to sensory cues
Defects in the mating behavior of bbs-7(my13) males, which is regulated in part by the CEM neurons, have been reported [16]. Because of the dye-filling defects observed in the bbs-7(my13) animals, we asked if behaviors attributed to the function of other ciliated neurons were affected. We tested the responses of the bbs-7(my13) mutant worms to diacetyl, benzaldehyde and isoamyl alcohol because responses to these chemicals are mediated by different subsets of ciliated neurons.
Response to diacetyl is mediated in part by the AWA neurons [26]. Attraction to isoamyl alcohol and low concentrations of benzaldehyde is mediated by the AWC neurons while avoidance of higher concentrations of benzladehyde requires the AWB neurons [26]. bbs-7(my13) animals responded similarly to wild-type animals to both high (100%) and low (1%) concentrations of benzaldehyde and diacetyl ( fig. 2 and data not shown). However, the mutant animals' response to isoamyl alcohol was concentration dependent: while wild-type animals were attracted to both 100% and 1% isoamyl alcohol, bbs-7(my13) animals avoided or showed little attraction to 100% isoamyl alcohol ( fig. 2). Although bbs-7(my13) animals showed a similar attraction to 1% isoamyl alcohol compared to wild-type animals as measured by the Worm Chemotaxis Index, we observed that the mutant animals did not approach the point source of the chemical as closely as the wild-type animals. For 1% isoamyl alcohol, specifically, there was a significant difference in attraction strength for each time point tested. The chemotaxis data suggest that some but not all ciliated neuron functions are affected in bbs-7(my13) animals. The different responses observed for the wild-type and mutant strains could be a function of specific defects in ciliated neurons and receptor localization or due to reduced mobility of the bbs-7(my13) mutant animals. To examine this second possibility, we tested both the roaming ability of individual animals on food and the response of the animals in a chemotaxis assay where the plate contained only diluent. There is no significant difference in the area covered by the bbs-7(my13) mutant animals as compared to wild-type animals when the individual animals are on food (data not shown). However, in the absence of odorant or food on the plate, bbs-7(my13) mutant animals do not travel as far as wild-type animals (table 1). When exposed to a strongly attractive (1% diacetyl) or strongly repulsive (100% benzaldehyde) odorant, bbs-7(my13) mutant animals display an equal or greater ability to move toward or away from the odorant compared to wild-type animals (table 1). Therefore, the differences seen in response to the chemicals are unlikely to be due to a difference in locomotion ability.

bbs-7(my13) affects glial cell morphology
The dendrites of the twelve neurons in each of the amphids are associated with two glial cells, the sheath cell and the socket cell [27,28]. The presence of matrixfilled vesicles in cilia mutants has previously been reported [25]. Similarly, we observed that the sheath cells of bbs-7(my13) mutant animals contain vacuoles in the L4 and adult stages ( fig. 3). By contrast, L4 wild-type animals have no visible vacuoles within their sheath cells, but, as the animals age, small vacuoles begin to build-up within the sheath cells. The vacuoles in the adult mutant animals appear larger and more numerous than those in the wild-type animals. The percentage of bbs-7(my13) animals with vacuoles in the amphid sheath cells approaches 100% at both L4 and adult stages ( fig. 3). Thus, it appears that the mutant animals possibly have a defect in the sheath cells.
my13 is an allele of bbs-7, a component of the BBSome The my13 allele was originally isolated because of the PKD-2::GFP mislocalization phenotype exhibited by homozygous mutant animals [17]. In order to identify the residue affected by the my13 mutation, we undertook a whole genome sequencing and single nucleotide polymorphism mapping approach [24]. We identified a non-synonymous SNP in bbs-7, the ortholog of human BBS7 ( fig. 4A). Because the my13 mutation results in a substitution of an adenine for a guanine at the first nucleotide of exon 6, we reasoned that my13 might either result in aberrant splicing of the osm-12 transcript or substitution of glutamic acid for glycine at position 314 in the protein. Amplification of osm-12 mRNA from both wild-type and homozygous mutant animals resulted in products that were the same size ( fig. 4B); sequencing of these products confirmed that only the identity of a single nucleotide differed between them ( fig. 4C). Because the transcripts are the same length and there are only single peaks on the chromatogram, the my13 mutation likely does not affect splicing but instead changes the amino acid sequence of the translated protein. However, we cannot rule out the possibility that some transcripts are mis-spliced and that these affect the mutant phenotype. There are three previously reported alleles of bbs-7 (also referred to as osm-12) in C. elegans. We examined the phenotypes of two of these. osm-12(n1606) results in a nonsense mutation at nucleotide 2899 and is likely a null allele [12] while osm-12(ok1351) results in a loss of exons 6 and 7 and is also likely null [29]. Both osm-12(n1606) and osm-12(ok1351) mutant animals are completely dye-filling defective; they fail to take up any dye in either the amphid or phasmid neurons ( [12,29] and our own data). This phenotype differs from that observed in the bbs-7(my13) mutant animals, which we report to be only partially dye-filling defective. This means that the my13 allele retains function to permit dye-filling in the amphid neurons suggesting a functional distinction between my13 and the other alleles.
Complementation analysis revealed a complex relationship between the my13, n1606, and ok1351 alleles, depending on the phenotype evaluated. When observing the dye-filling phenotype of the trans-heterozygotes for any pairwise bbs-7 (my13) Affects Cilia Structure and Function combination of the three alleles, the weaker mutant phenotype predominates: all animals are dye-filling defective in the phasmids and exhibit abnormal dye-filling of the amphids (table 2). The PKD-2::GFP localization phenotype exhibits partial complementation when examining the my13/n1606 and the my13/ok1351 transheterozygotes: 38% and 44% of animals, respectively, have increased accumulation at the ciliary base accompanied by inward curvature of the CEM cilia (table 2). This partial complementation may indicate that, unlike n1606 and ok1351, my13 does not result in complete loss of function of bbs-7.
We examined whether expression of full-length wild-type BBS-7 in bbs-7(my13) mutant animals could rescue the phasmid dye-filling defects. Overexpression of BBS-7: GFP in a wild-type background itself results in phasmid dye-filling defects some of the time (table 3). This suggests that the level of BBS-7 expression in the phasmids may be important. While expression of BBS-7::GFP in the bbs-7(my13) mutant background does not completely rescue the phasmid dye-filling defects, the percentage of animals that fill normally with dye approaches wild-type levels (75% versus 79% nonDyf).

Discussion
BBS-7 is one of a family of genes known to cause Bardet-Biedl Syndrome, a pleiotropic ciliopathy, in humans [30]. BBS7 forms a complex, known as the BBSome, with six other BBS proteins. Components of the BBSome have been shown to play key roles in the function of primary cilia in both humans and C. elegans: the BBSome complex is necessary for vesicle trafficking to the ciliary membrane, proper intraflagellar transport and, specifically, turnaround of the transport mechanism at the tip of the cilium [31][32][33]. BBS proteins also play roles in assembly of the intraflagellar transport complex necessary for movement of molecules within the cilium [12,34], neuropeptide secretion [35], and gene expression regulation [36]. Dysfunction of the BBS proteins leads to Bardet-Biedl syndrome in humans and sensory defects in C. elegans. Here, we characterize and clone my13, a novel allele of bbs-7 and provide additional evidence for the role of bbs-7 in primary cilia structure, sensory neuron function, and communication between neurons and sheath glia.
my13 affects a conserved residue of bbs-7 We show that my13 results in the substitution of an adenine for a guanine at the first nucleotide of exon 6, which would likely lead to glutamic acid instead of glycine at position 314 in the BBS7 protein. This glycine residue is highly conserved among both the Caenorhabditis genus and vertebrates, indicating possible functional significance. The affected residue also falls in a 252-amino acid region of human BBS7 that is highly conserved between human BBS7 and human BBS2 as well as BBS2 in other species including C. elegans [30]. This common region is predicted to encode a six-bladed b-propeller structure that has been suggested to mediate interactions of BBS7 with other proteins within the BBSome and the chaperonin complex [30,37]. The glycine affected by my13 is predicted to be located in the coil or turn between two b-strands [38], and we postulate that the change from a small hydrophobic glycine to a large negatively-charged glutamic acid may have an effect on the conformation of the BBS7 protein.
The partial dye-filling defect of bbs-7(my13) animals combined with the altered cilia structure observed when examining ciliary receptor localization suggest that bbs-7 is important for primary cilia structure. In BBS-7 and BBS-8 null worms, the IFTA and IFTB complexes dissociate for anterograde transport [32], and interactions between BBS1, BBS7 and BBS9 are disrupted [33]. In mice, BBS7 has been shown to be a component of both the BBSome and the BBS chaperonin complex [39]. BBS6, BBS10, and BBS12 form a complex with the CCT chaperonins in vertebrates, and it has been proposed that these proteins mediate the interactions of CCT chaperonins with BBS2 and BBS7 [40]. It will be interesting to determine whether the my13 mutation results in disruption of the interactions between BBSome components or, alternatively, of the BBSome with the CCT chaperonins.   In nature, C. elegans are soil-dwelling and depend on a relatively simple nervous system of 302 neurons to identify, integrate and respond to the complex array of chemicals present in order to find food and avoid toxins. Distinct subsets of neurons mediate responses to aversive and attractive odors: the AWA and AWC neurons detect attractive odors while the ASH, ADL, AWB and ASK neurons detect aversive odors [26]. bbs-7(my13) animals do not respond like wild-type animals to the volatile odorant isoamyl alcohol, but respond normally to benzaldehyde and diacetyl. The different behavioral responses to the volatile odorants may be due to altered overall perception of specific chemicals or an inability to finely map changing concentrations of the chemicals over time. The phasmid neurons have been shown to be chemosensory and integral for the avoidance of the chemical SDS [41]. It is possible that the chemosensory and phasmid dye-filling defects observed in the bbs-7(my13) animals are interrelated.
Neuron-sheath cell interactions are affected in bbs-7(my13) mutant animals bbs-7 is expressed specifically in ciliated neurons in C. elegans [12,32]. Thus, we would expect to see mutant phenotypes involving the ciliated neurons. We also observe accumulation of vacuoles in the supporting sheath glial cells in bbs-7(my13) mutant animals. These glial cells have been associated with the remodeling of neuronal endings [42] and communication between neurons and glial cells is important for function of both [43]. This primary defect in ciliated neurons resulting in a secondary defect in sheath glial cells is not unique: a similar phenotype has been observed in daf-19 mutants [44]. A number of intraflagellar transport (IFT) and other cilia mutants also have accompanying defects in sheath cell structure including che-11, che-10, osm-3, che-12, and mec-8. Mutations in each of these genes are associated with accumulation of vesicles in the amphid sheath cells [25]. More specifically, these matrix-filled vesicles are indicative of defects in vesicle release. While we have not identified the my13 vacuoles as vesicles, the phenotype that we observe would be consistent with the vesicle accumulation previously reported for other mutants. Proper amphid glia formation has been shown to be required for cilia formation and function of some Table 3. BBS-7::GFP restores phasmid dye-filling in bbs-7(my13) mutant animals.

Strain
Ability to dye-fill (%) him-5(e1490) 100% nonDyf (n562) bbs-7(my13) 100% tailDyf (n564) BBS-7::GFP 79% nonDyf and 21% tailDyf (n562) bbs-7(my13) with BBS-7::GFP 75% nonDyf and 25% tailDyf (n556) Animals homozygous for the my13 allele do not take up fluorescent dye in their phasmids and exhibit abnormal dye-filling in the amphids (tailDyf) while wildtype (him-5(e1490) animals take up dye in both the amphids and phasmids. Some wild-type and bbs-7(my13) animals carrying the BBS-7::GFP extrachromosomal array are tailDyf. doi:10.1371/journal.pone.0113737.t003 of the amphid neurons [28]. Thus, the mutant phenotypes observed in bbs-7(my13) animals are likely to be the result of both cilia dysfunction itself and perturbation of cilia-sheath cell interactions. The identification of a novel allele of bbs-7 provides additional evidence of BBS7 function in cilia structure, sensory function, and interaction between ciliated neurons and glial cells. Future characterization of the nature of the protein encoded by bbs-7(my13) may be useful in providing details about the interactions between BBS7 and other components of the BBSome as well as between BBS7 and the chaperonin complex.