INTRODUCTION
The cilium is a hair-like cell surface organelle. Although present on most types of vertebrate cells, with a few exceptions such as nodal cilia in left-right asymmetry of the body plan and cilia on olfactory sensory neurons in odorant perception (
Reese, 1965;
Nonaka et al., 1998), functions of the cilium in most cells had rarely been studied as late as a decade ago. This picture has been completely changed in the past decade, during which we witnessed an explosion of cilia studies. We now know the basics of the core machinery for cilia biogenesis, the role of cilia in the Hedgehog pathway and the connection between cilia and a wide spectrum of human disease. Yet the regulation of cilia formation in the context of animal development, the roles of cilia in other major signaling pathways and the underlying mechanisms of cilia-related diseases are far from lucid. What is clear is that we can expect cilia research continues to be an active area in the next decade and new findings will provide crucial insights for understanding both basic biology and human diseases.
CILIARY STRUCTURE AND FUNCTION
The cilium is an appendage-like organelle extending from the surface of a cell. The ciliary membrane is contiguous with the plasma membrane, and covers a microtubule-based structure known as the axoneme (Fig. 1). Anchoring the cilium in place is a modified centrosome, the basal body, at the proximal end of the axoneme (Fig. 1). The ciliary axoneme is composed by a series of nine outer doublets arranged in a concentric circular pattern, and depending on the cell type, these doublets may surround an additional inner doublet in the middle of the circle (“9 + 2”), or the inner doublet may alternatively be absent (“9 + 0”) (Fig. 1). In addition, inner and outer dynein arms connected to the outer doublets are responsible for ciliary motility. Historically the 9 + 2 doublet configuration has been associated with a motile function for cilia, whereas the 9 + 0 configuration has been associated with sensory cilia. However, the motile primary cilium in the mouse node (
Nonaka et al., 1998), which is critical for left-right asymmetry, has a 9 + 0 configuration, demonstrating that this distinction may be blurred.
At the distal end of the basal body is the transition zone, which is structurally composed of a network of fibers that connects the basal body to the plasma membrane and is thought to form a permeable barrier (
Deane et al., 2001). This transition zone has been proposed to serve as a type of pore, through which protein and membrane components can be actively shuttled between cilium and cell body (
Deane et al., 2001). In this context, the transition zone may be the location for regulation of the trafficking of ciliary components.
In multicellular organisms, cilia are found on most cells, including the nephron of the kidney, epithelial cells of the trachea, and the external granule layer of the cerebellum, which gives rise to the cerebellar vermis (
Sanderson and Sleigh, 1981;
Wheatley, 1995;
Spassky et al., 2008). A modified cilium comprises the connecting cilium and outer segments of photoreceptor cells in the retina (
Wright et al., 2010). Depending on the ciliated tissue type, cilia may serve a variety of biological functions, such as maintaining movement of cerebrospinal fluid in the cerebellum, and sensing noxious substances and sweeping foreign material out of the trachea (
Ibañez-Tallon et al., 2004;
Shah et al., 2009).
CILIA BIOGENESIS
Since no protein synthesis machinery exists within the cilium and the axoneme of cilium assembles at the tip (
Marshall and Rosenbaum, 2001) instead of the base, protein components must be synthesized within the cell body, and then transported into the cilium by a process known as intraflagellar transport (IFT) (
Pedersen and Rosenbaum, 2008). By IFT, protein complexes have been observed to travel in both anterograde (toward the tip of the cilium) and retrograde (back toward the basal body) directions, powered by anterograde kinesin and retrograde dynein motors (
Kozminski et al., 1993).
IFT particles are composed of complex A and complex B subunits, which play different roles in the bidirectional movement of cargos along the axoneme of cilia: while complex A is associated more with retrograde transport (
Pazour et al., 1998;
Piperno et al., 1998), complex B is involved in anterograde transport (
Kozminski et al., 1993;
Cole et al., 1998). Consistently, mutants of these two complexes show overlapping but distinct phenotypes in cilia morphology: while mutants of complex B genes display severely shortened or total absence of cilia, complex A mutants show stunted cilia with a bulge at the tip filled with complex B components (
Perkins et al., 1986;
Collet et al., 1998;
Piperno et al., 1998;
Qin et al., 2001;
Schafer et al., 2003;
Iomini et al., 2009).
In addition to IFT genes, which encode the core machinery for ciliogenesis, an increasing list of genes have been implicated in cilia formation, including genes that regulate cilia length and trafficking of specific components to cilia. In
Chlamydomonas, several long flagella mutants have been isolated (
McVittie, 1972;
Jarvik and Rosenbaum, 1980;
Barsel et al., 1988;
Asleson and Lefebvre, 1998). Intriguingly, two of them encode kinases, suggesting the involvement of a signaling cascade in regulating cilia length (
Berman et al., 2003;
Tam et al., 2007). Currently, the underlying mechanism for cilia length control remains elusive. Similarly, the functions of many newly identified genes remain to be fully characterized. Nonetheless, studies on these genes are starting to yield mechanistic insights. An example is BBS genes. After careful analysis, many BBS genes appear to be dispensable for cilia formation at least in some tissues. They are, however, intimately involved in protein trafficking and in mediating ciliary signaling (
Lechtreck et al., 2009;
Jin et al., 2010).
Another aspect of cilia biogenesis is cell fate determination. In vertebrates, most cells harbor a single cilium. However, in some organs, such as the airway of mammals, the skin of frogs and the pronephric duct of zebrafish, multi-ciliated cells intercalate with single-ciliated cells. Lateral inhibition mediated by the Notch pathway seems to play a critical role in specifying the multi-ciliated versus the single-ciliated fate (
Marnellos et al., 2000;
Liu et al., 2007;
Ma and Jiang, 2007;
Tsao et al., 2009).
MODEL ORGANISMS FOR CILIA STUDIES
Model organisms serve as surrogates to allow experiments that are otherwise not possible in an original system. Multiple systems have been used to analyze the formation and function of cilia. Here, we provide a brief introduction of frequently used organisms. It is worth noting that cultured cell lines have also been used successfully in cilia studies.
Chlamydomonas
Cilia/Flagella have been extensively studied in the green algae
Chlamydomonas; in fact, the major machinery important for the formation and function of cilia, IFT particles, were first discovered in
Chlamydomonas (
Kozminski et al., 1993). The simple unicellular and biflagellate structure of
Chlamy-domonas makes it relatively easy to isolate cilia and perform biochemical analysis on ciliary proteins (
Cole et al., 1998;
Piperno et al., 1998;
Rosenbaum and Witman, 2002), a feat that is nearly impossible to achieve in more complex organisms. In addition, the simple structure of
Chlamydomonas allows the visualization of IFT without the aid of GFP, the resolution of which is difficult to reach within other organisms (
Kozminski et al., 1993).
Chlamydomonas is also amendable to genetic analysis, allowing for screens, as well as the generation of a large repertoire of cilia mutants. The release of the nuclear and mitochondrial genome sequence of
Chlamydomonas provides researchers with better tools for genetic manipulation and informatics. Lessons learned from
Chlamydomonas studies on cilia formation and its regulation can then be used to gain insights into the function of cilia in human physiology and diseases.
C. elegans
C. elegans lacks ciliated epithelial tissues but cilia can be found in the dendritic endings of a large number of sensory neurons. Proteins important for normal cilia formation and functions in mammalian species are well conserved in
C. elegans. These genes include IFT genes, such as
IFT88, as well as cilia-related disease genes such as
PKD1 and
PKD2 (
Barr and Sternberg, 1999;
Barr et al., 2001;
Qin et al., 2001). Studies of cilia genes in
C. elegans can also give insight to the important functional role of cilia in biological processes. For example, homologs of
PKD1 and
PKD2 in
C. elegans,
lov-1 and
pkd-2, have been shown to be necessary for normal male mating behavior (
Barr and Sternberg, 1999;
Barr et al., 2001). However, not all mammalian genes important for cilia function are present in the
C. elegans genome.
PKHD1, for example, which encodes Fibrocystin, a protein that is associated with human autosomal recessive polycystic kidney disease, does not have an obvious homolog in
C. elegans. In addition to the array of conserved cilia related genes, the large number of tools available in
C. elegans, such as the completely sequenced genome, the ease of RNAi knockdown technology, the ability to visualize IFT with the aid of GFP, and the feasibility to conduct genome-wide screens, allow effective study of cilia in
C. elegans. However, the research of cilia in
C. elegans is limited by the fact that cilia can only be found in sensory neurons, making it difficult to study tissue-specific roles of cilia.
Drosophila
Similar to
C. elegans,
Drosophila does not have ciliated epithelial tissues, and modified cilia are found in the sensory neurons of specialized sensory organs such as bristles. The modified cilia are important for mechano-sensation, audition and olfaction. Some cilia genes are also conserved in
Drosophila, such as IFT and BBS genes, which are expressed by the cilia in the sensory neurons (
Avidor-Reiss, 2010). The lack of cilia on most
Drosophila cells has important functional implications. Specifically, the role of cilia in the Hedgehog (Hh) signal transduction pathway is different between mouse and
Drosophila, even though some major principles and players of the pathway are fairly conserved. In
Drosophila, in the absence of the Hh ligand, the receptor Patched (Ptc) prevents the protein Smoothened (Smo) from trafficking to the cell surface. Upon binding to the Hh ligand, Ptc is degraded, and Smo is able to traffick to the cell surface and activate downstream effectors (
Denef et al., 2000). In contrast, in mouse and other mammals, in the absence of the Hedgehog ligand, Ptc prevents Smo to localize to the cilium, and upon binding of Hh to Ptc, Smo gains the ability to translocate to the cilium, and activates downstream signals (
Huangfu et al., 2003;
Corbit et al., 2005;
Haycraft et al., 2005;
Rohatgi et al., 2007). As a result, mouse IFT mutants exhibit altered hedgehog signaling (
Huangfu et al., 2003;
Liu et al., 2005;
May et al., 2005;
Huangfu and Anderson, 2006), while
Drosophila IFT mutants do not have Hh phenotypes (
Avidor-Reiss, 2010).
Xenopus
Xenopus laevis is a classic model system for developmental biology studies. The large size and
ex utero development of
Xenopus embryos make them readily accessible to micro-dissection and transplantation. In addition, overexpression, morpholino knockdown and transgenesis can be used to manipulate gene functions. Convenient for cilia studies, abundant multi-ciliated cells can be found on the skin of developing
Xenopus embryos. Cilia in these cells are oriented, move in concert and drive liquid flow that can be easily visualized. These features have been successfully utilized to study the specification of multi-ciliated cells, the biogenesis of cilia, the relationship among planar cell polarity, flow direction and cilia orientation; all of which are conserved phenomena in mammals (
König and Hausen, 1993;
Deblandre et al., 1999;
Park et al., 2006;
Hayes et al., 2007;
Mitchell et al., 2007;
Park et al., 2008;
Vladar and Axelrod, 2008;
Mitchell et al., 2009). On the other hand, the tetraploid genome of
Xenopus laevis poses a challenge to effective genetic manipulations in this system. Interestingly,
Xenopus tropicalis has a diploid genome and shorter generation time. In recent years, it is being actively developed as a genetic model system (reviewed in (
Hirsch et al., 2002;
Carruthers and Stemple, 2006)).
Zebrafish
Several unique features of zebrafish make it an excellent model to study the cilium and its role in development and disease. Zebrafish is small in size and each pair of adult fish can produce hundreds of offspring at weekly intervals, making them accessible for large-scale genetic and chemical screens. The optical transparency of zebrafish embryos allows for monitoring of phenotypes in live embryos. This feature is further enhanced by the creation of
casper, a fish line that is transparent even as adults (
White et al., 2008). In addition, morpholino antisense oligos can be used very effectively and at a relatively high-throughput scale in zebrafish to disrupt gene function (
Ekker, 2000;
Nasevicius and Ekker, 2000) and recently a zinc-finger endonuclease technique has been used successfully for targeted mutagenesis (
Doyon et al., 2008;
Meng et al., 2008;
Foley et al., 2009).
The cilium is easily recognizable in zebrafish and is especially enriched in the Kupffer’s vesicle, the kidney duct and the neural tube (
Kramer-Zucker et al., 2005). Many zebrafish homologs of human disease genes, when mutated, lead to phenotypes that can be directly compared to human symptoms. For example, we demonstrated previously that homologs of
vHNF1, a gene associated with human familial GCKD (glomerulocystic kidney disease) (
Bingham et al., 2001), and
PKD2, a gene associated with autosomal dominant PKD (ADPKD) (
Mochizuki et al., 1996), can cause kidney cyst when mutated in zebrafish (
Sun et al., 2004). Conversely,
arl13b/sco was initially identified as a cystic kidney gene in zebrafish and later associated with Joubert Syndrome in human (
Cantagrel et al., 2008). Finally, a collection of cilia mutants, including multiple mutants of IFT genes, was already generated in previous studies (
Sun et al., 2004;
Tsujikawa and Malicki, 2004;
Kramer-Zucker et al., 2005). Taken together, these attributes make zebrafish an excellent system to study the cilium and related kidney diseases.
Mouse
Cilia studies in mouse led to the appreciation and understanding of the diverse roles cilia play in mammalian development and disease. Cilia are present on almost all mammalian cells, including kidney epithelia, the embryonic node and neurons. With the assistance of the powerful tools in mouse genetics, a large number of cilia mutants are available. These mutants have been used effectively to reveal the role of cilia in Hedgehog signaling (
Huangfu et al., 2003;
Corbit et al., 2005;
Haycraft et al., 2005;
Rohatgi et al., 2007) and the establishment of the left-right asymmetry of the body plan (
Nonaka et al., 1998,
2002). Furthermore, tissue specific disruption of cilia biogenesis allows researchers to bypass the limitation posed by early lethality of cilia mutants to investigate ciliary functions in different organs and at later developmental stages. For example, using the Cre-lox system to disrupt cilia biogenesis in different tissues, researchers showed that cilia are involved in hair follicle development, bone formation, branching morphogenesis of the mammary gland and in regulating satiety responses (
Davenport et al., 2007;
Haycraft et al., 2007;
Lehman et al., 2009;
McDermott et al., 2010).
CILIA PATHWAYS AND SIGNAL TRANSDUCTION
Hedgehog
In vertebrate, one signaling pathway that has been shown to have a close relationship with the cilium is the Hedgehog pathway (
Huangfu et al., 2003;
Liu et al., 2005;
May et al., 2005;
Huangfu and Anderson, 2006). Hh signaling governs a variety of cellular processes in development, for example, in mammals, the proliferation of the granule cell precursors that eventually give rise to the cerebellar vermis (
Dahmane and Ruiz i Altaba, 1999;
Wallace, 1999;
Wechsler-Reya and Scott, 1999;
Spassky et al., 2008). A null mutation of mouse
Ift172,
wimple, results in significantly altered
Patched1 expression in the neural tube (
Huangfu et al., 2003). Target genes of Hh signaling are regulated by the Hh-responsive Gli family of transcription factors, and overexpressed Gli1, Gli2 and Gli3 have been shown to localize to the distal tip of cilia (
Haycraft et al., 2005;
Eggenschwiler and Anderson, 2007). In zebrafish, a loss-of-function allele of
gli1 and a dominant repressor form of
gli2 cause a body axis curvature defect (
Karlstrom et al., 2003). Although the dependence of Hh signaling on the cilium is thought to be conserved in vertebrates including zebrafish and mammals, there are subtle differences between different organisms. For example,
gli1 is expressed in zebrafish in the absence of Hh signaling, whereas in mouse Hh signaling is required for
gli1 expression (
Bai et al., 2002;
Karlstrom et al., 2003). Consistently, a maternal-zygotic
ift88 mutant zebrafish that lacks all cilia has a reduced induction of
gli1 expression, but expresses low levels of
gli1 ectopically (
Huang and Schier, 2009). This difference between zebrafish and mammals in the behavior of Hh pathway-responsive genes, such as
gli1, may account for the slight difference in phenotypes in zebrafish mutants, compared to the mammalian phenotypes.
Interestingly, available evidence suggests that complex A and B might have different effects on Hh signaling. In complex B mutants, Hh signaling is mainly inhibited in the neural tube (
Huangfu et al., 2003;
Haycraft et al., 2005;
Liu et al., 2005;
May et al., 2005). However, by contrast, in the only two reported mutational studies on IFT complex A components in mice, one on
Ift139 and one on
Ift122, both mutants show over-activation of the Hh pathway in the neural tube (
Tran et al., 2008;
Cortellino et al., 2009). Together, these data suggest that A and B complexes of IFT have distinct functions in overall cilia biogenesis and signaling.
Wnt
The cilium has also been implicated in Wnt signaling. In zebrafish, disruption of the basal body proteins Bbs1, Bbs4 and Bbs6 results in defects in convergent extension, which is regulated by noncanonical Wnt/planar cell polarity (PCP) signaling (
Gerdes et al., 2007). Suppression of
BBS4 in HEK293T cells also results in the stabilization of β-catenin, a major component of the canonical Wnt pathway (
Gerdes et al., 2007).
Bbs genes have been shown to interact genetically with the PCP gene
Vangl2 in both mice and zebrafish, and Vangl2 localizes to the ciliary axoneme as well as the basal body (
Ross et al., 2005). The
kif3a mutant, which has ciliogenesis defects, exhibits upregulation of canonical Wnt in mouse embryos, and reveals a role for cilia in restraining the canonical Wnt pathway (
Corbit et al., 2008). Inversin, which acts as a molecular switch between canonical and noncanonical Wnt pathways, localizes to cilia in MDCK cells (
Otto et al., 2003;
Simons et al., 2005). In addition, in multiple ciliary mutants, the canonical Wnt pathway was sensitized during early development (
Corbit et al., 2005;
Gerdes et al., 2007). More directly, in
Ift88 and
Ift20 mutant mice, non-canonical Wnt phenotypes were observed in the cochlear and kidney ducts respectively (
Jonassen et al., 2008;
Jones et al., 2008).
Nonetheless other data indicates that the cilium is not overtly required for normal Wnt signaling. Maternal-zygotic
ift88 (MZ
ovl) zebrafish mutants, which lack all cilia, undergo normal convergent extension, as indicated by
krox20 and
myoD expression in the midbrain-hindbrain and somites, respectively (
Huang and Schier, 2009). Moreover, these mutants show normal spatial expression of the canonical Wnt target genes
axin2,
sp5, and
sp5l (
Huang and Schier, 2009). In mouse, no obvious Wnt-like phenotypes were reported in several IFT mutants (
Ocbina et al., 2009). In addition, mutations in canonical Wnt pathway genes
wnt3 or
lef1 tcf1 double mutants are embryonic lethal very early during development, whereas mice carrying null mutations in IFT genes do not display as severe phenotypes (
Eggenschwiler and Anderson, 2007). Currently, the role of cilia in Wnt pathways remains controversial. It is plausible that the role of cilia in Wnt pathways is tissue and stage specific, or that cilia play a minor or redundant role in Wnt pathways.
CILIOPATHIES
Consistent with its almost ubiquitous distribution and functional importance, the cilium has been linked to an ever-expanding list of symptoms in human diseases, ranging from kidney cyst, retinal degeneration and abnormal situs to obesity and diabetes. “Ciliopathy” was coined to describe this diverse spectrum of diseases with a common involvement of the cilium (for a review, see (
Hildebrandt et al., 2009)). The connection between polycystic kidney disease (PKD) and the cilium is among one of the best studied. It is thought that cilia on renal epithelial cells detect environmental signals to promote cell differentiation and prevent cell proliferation. Structural or functional defects of cilia can therefore lead to uncontrolled cell proliferation and eventual kidney cyst formation. The central role of the cilium in PKD is supported by the fact that multiple proteins involved in PKD have been found on the cilium and that disruption of cilia formation or function almost inevitably leads to kidney cyst formation (
Pazour et al., 2000;
Pazour et al., 2002;
Yoder et al., 2002a,
b;
Sun et al., 2004).
Mutations in IFT components generally result in ciliogenesis defects and may cause a wide range of developmental abnormalities, clinical manifestations and ciliopathies such as PKD. For example, an early animal model of PKD was the
orpk mouse, which carried a mutation in
ift88/polaris (
Schrick et al., 1995).
ift57hi3417,
ift81hi409, and
ift172hi2211 mutants in zebrafish develop kidney cysts and body axis curvature (
Sun et al., 2004). Defects in the connecting cilium of retinal pigment epithelial cells may result in progressive degeneration of the photoreceptor epithelium, giving rise to clinical presentations, such as retinitis pigmentosa or Leber congenital amaurosis (
Murga-Zamalloa et al., 2009;
Shintani et al., 2009).
Joubert syndrome (JS) is a rare recessive ciliopathy with an estimated prevalence in the United States of 1 in 100,000 (
Parisi et al., 2007). JS is caused by a malformation of the cerebellum, classically resulting in hypotonia and mental retardation, and frequently oculomotor apraxia, ataxia and abnormal breathing patterns (
Parisi et al., 2007). In addition to the classical manifestations, some JS patients present with other clinical features, including pre- or post-axial polydactyly, hepatic fibrosis, occipital encephalocele, congenital retinal degeneration and familial juvenile nephronophthisis, a cystic kidney disease that is the most common cause of chronic renal failure in children (
Satran et al., 1999;
Parisi et al., 2007). As a set, these features are known as Joubert syndrome and related disorders (JSRD). The primary clinical diagnostic criterion of classical JS is the appearance of a “molar tooth sign” on magnetic resonance images of the brain, indicating cerebellar vermis hypoplasia (
Maria et al., 1999). While diagnosis of JS with observation of the molar tooth sign is possible by as early as the third trimester of gestation, no specific treatments exist for JS, and patients with nephronophthisis or renal dysplasia typically undergo renal dialysis or transplantation for disease management (
Fluss et al., 2006;
Parisi et al., 2007).
To date, recessive mutations in nine autosomal genes and one X-linked gene have been found to be associated with Joubert syndrome:
NPHP1,
NPHP6/CEP290,
NPHP8/RPGRIP1L,
CC2D2A,
AHI1,
MKS3/TMEM67,
TMEM216,
INPP5E,
OFD1 and
ARL13B (Table 1) (
Dixon-Salazar et al., 2004;
Ferland et al., 2004;
Parisi et al., 2004;
Sayer et al., 2006;
Valente et al., 2006;
Arts et al., 2007;
Baala et al., 2007;
Delous et al., 2007;
Cantagrel et al., 2008;
Gorden et al., 2008;
Bielas et al., 2009;
Coene et al., 2009;
Edvardson et al., 2010). Of these ten, three are also associated with nephronophthisis:
NPHP1 (encodes Nephrocystin),
NPHP6/CEP290 (encodes Nephrocystin-6), and
NPHP8 (encodes RPGR-interacting protein-1-like protein), revealing considerable clinical and genetic overlap between JSRD and isolated nephronophthisis (
Hildebrandt et al., 1997;
Sayer et al., 2006;
Wolf et al., 2007). However, mutations in the ten JSRD-associated genes are found in less than 50% of JS patients in some clinical cohorts, indicating that other JSRD-related genes remain undiscovered (
Doherty, 2009).
Analyses of the ten genes associated with JSRD implicate a role for cilia in this multisystemic disorder. First of all, nine of the ten known JSRD-associated proteins have been shown to localize to the cilium or the basal body (Table 1). In addition, mutations in eight JS-associated genes have been found to result in ciliogenesis defects (Table 1). Furthermore, since ciliary defects are closely associated with NPHP, the common involvement of at least three genes in both JSRD and NPHP provides additional support for the role of cilia in JSRD. Collectively, these findings suggest that JS is a ciliopathy whose extracranial manifestations may be traced back to the diverse functional role of cilia in vertebrates.
CONCLUDING REMARKS
Overlooked for a long time, the antenna-like cilium is now being linked to a growing number of human diseases (
Hildebrandt et al., 2009). Protruding from the cell surface into the environment, the ubiquitous cilium is ideally situated to function as a sensor for vertebrate cells. In agreement with its sensory role, multiple receptors, including the Hedgehog receptor Patched, PDGFRα and G protein coupled receptors, have been found on the cilium (
Händel et al., 1999;
Schneider et al., 2005;
Rohatgi et al., 2007;
Berbari et al., 2008). Although role of cilia in the Hh pathway is now well established, the precise function of cilia in other signaling pathways remains to be elucidated. In depth analysis of cilia function in the context of tissue development and homeostasis will be critical for understanding and treating ciliopathies.
Higher Education Press and Springer-Verlag Berlin Heidelberg 2010