Ciliary Extracellular Vesicles: Txt Msg Organelles

Abstract

Cilia are sensory organelles that protrude from cell surfaces to monitor the surrounding environment. In addition to its role as sensory receiver, the cilium also releases extracellular vesicles (EVs). The release of sub-micron sized EVs is a conserved form of intercellular communication used by all three kingdoms of life. These extracellular organelles play important roles in both short and long range signaling between donor and target cells and may coordinate systemic responses within an organism in normal and diseased states. EV shedding from ciliated cells and EV–cilia interactions are evolutionarily conserved phenomena, yet remarkably little is known about the relationship between the cilia and EVs and the fundamental biology of EVs. Studies in the model organisms Chlamydomonas and Caenorhabditis elegans have begun to shed light on ciliary EVs. Chlamydomonas EVs are shed from tips of flagella and are bioactive. Caenorhabditis elegans EVs are shed and released by ciliated sensory neurons in an intraflagellar transport-dependent manner. Caenorhabditis elegans EVs play a role in modulating animal-to-animal communication, and this EV bioactivity is dependent on EV cargo content. Some ciliary pathologies, or ciliopathies, are associated with abnormal EV shedding or with abnormal cilia–EV interactions. Until the 21st century, both cilia and EVs were ignored as vestigial or cellular junk. As research interest in these two organelles continues to gain momentum, we envision a new field of cell biology emerging. Here, we propose that the cilium is a dedicated organelle for EV biogenesis and EV reception. We will also discuss possible mechanisms by which EVs exert bioactivity and explain how what is learned in model organisms regarding EV biogenesis and function may provide insight to human ciliopathies.

Introduction

“Cilia” is a broad term that includes primary or sensory cilia, motile cilia, and flagella. Most cilia share a similar microtubule-based architecture and are constructed by the evolutionarily conserved intraflagellar transport (IFT) machinery (Ishikawa and Marshall 2011; Rosenbaum and Witman 2002). Motile cilia propel fluid surrounding tissues or may move the cell itself (Zhou and Roy 2015). Primary cilia are found on most non-dividing cells in the human body, play central roles in signal transduction and coordination of cellular behaviors, and thus are important for development and health (Zimmerman and Yoder 2015). Disruption of ciliary formation or function causes a wide spectrum of syndromes termed ciliopathies, which show overlapping characteristic symptoms such as cystic kidney disease, neurological disorders, retinal degeneration, polydactyly, situs inversus, and obesity (Pazour and Rosenbaum 2002).

“Ciliary extracellular vesicles (EVs)” is a comprehensive term we use to refer to ciliary-derived or ciliary-associated EVs. EV is also a broad term referring to any extracellular membrane bound vesicle that includes exosomes derived from multivesicular bodies and ectosomes formed via outward budding of the plasma membrane (Colombo et al. 2014; El Andaloussi et al. 2013; Lo Cicero et al. 2015). Exosomes and ectosomes are tiny: less than 100 nm for exosomes and between 100 nm and 1 μm for ectosomes. Cells release EVs containing specific protein, lipid, and nucleic acid-based cargo (Lo Cicero et al. 2015). These cell-specific markers may reveal the source of the EVs, and may be used as biomarkers for normal and diseased states. EV composition and abundance change under different physiological and pathological conditions, therefore understanding the basic process of EV biogenesis, EV cargo selection, and EV-target cell interaction holds exciting significance for diagnostics and therapeutics.

Very recently, cilia have been observed as a site for EV release and for EV attachment (Wood and Rosenbaum 2015). In this review, we discuss this new emerging field of research and offer our perspective on how understanding the relationship between EVs and cilia might advance understanding and treatment of human ciliopathies.

Ciliary Proteins are EV Cargo

Mutations in the polycystin-encoding genes cause autosomal dominant polycystic kidney disease (ADPKD) (Ong and Harris 2015) and polycystin ciliary trafficking defects are thought to be an underlying cause of this ciliopathy (Cai et al. 2014). The mammalian polycystins localize to cilia as well as urinary EVs released from renal epithelial cells (Hogan et al. 2009; Pazour et al. 2002; Pisitkun et al. 2004; Yoder et al. 2002). Pisitkun et al. first observed polycystin-1 in urinary exosomes (Pisitkun et al. 2004). Ward and colleagues then showed that polycystin-1 and -2 and the autosomal recessive PKD-gene product fibrocystin colocalized to urinary exosome-like vesicles (Hogan et al. 2009). Polycystin-2 and a peripheral ciliary membrane protein retinitis pigmentosa 2 (RP2), are shed into media in the MDCK canine kidney cell culture (Hurd et al. 2010). The components of exocyst and the ciliary GPCR Smoothened are found in the urinary exosome-like vesicle proteome (Chacon-Heszele et al. 2014). Bioinformatics analysis revealed that some of the known interactors of Smoothened, polycystin-1, and polycystin-2 are present in the urinary exosome-like vesicle proteome (Chacon-Heszele et al. 2014), suggesting that signal transduction modules may be contained in EVs.

The embryonic nodal floor releases membrane bound particles called ‘nodal vesicular parcels’ (Tanaka et al. 2005). These nodal vesicular parcels contain Sonic hedgehog and other signaling molecules that are swept by nodal flow to the left side. Hirokawa and colleagues proposed that nodal vesicular particles interact with immotile cilia for left–right axis determination (Tanaka et al. 2005). This hypothesis, while controversial and unresolved, was the first to propose that ciliary-associated EVs carry morphogens that influence important aspects of animal development.

In the green algae Chlamydomonas reinhardtii, the flagella membrane is continuously shed by releasing flagellar EVs, with a complete turnover of flagellar membrane protein every 6 h (Dentler 2013). A major flagellar membrane protein FMG1, a large glycoprotein (~350 kD), is shed into media in a distinct manner. After crosslinking with lectins or antibodies, FMG1 first aggregates at the flagellar tip, then is transported back in the direction of the cell body and shed into media at the flagellar base in a membrane-bound form (Bloodgood et al. 1986). Chlamydomonas ciliary ectosomes are shed from the flagellar tip and perhaps along the entire flagellar membrane, and also contain polycystin-2 (Cao et al. 2015; Wood et al. 2013).

Ciliary EVs are not simply the breaking off of membrane from cilia. Proteins found in ciliary EVs and cilia are not identical. In Chlamydomonas, Wood et al. (2013) showed that released EVs contain a specific complement of proteins, different from the proteins of the isolated flagellar membranes. Thus, EVs are not just being broken off the flagellar membrane. Caenorhabditis elegans EVs contain the polycystins and CIL-7, a myristoylated EV biogenesis regulator, but not the ciliary kinesin-3 motor KLP-6, components of the IFT machinery, or ciliary beta-tubulin TBB-4 (Maguire et al. 2015; Wang et al. 2014). Chlamydomonas ciliary ectosomes contain polycystin-2, FMG1, the flagellar membrane polypeptide SAG1-C65, and vegetative lytic enzyme VLE but not typical ciliary proteins including IFT81, IFT139, alpha tubulin, and axonemal protein Bug22 (Cao et al. 2015; Wood et al. 2013). Combined, these results suggest that an active EV sorting machinery selects some cargo but not others for packaging in EVs.

Thus, various ciliary proteins are also present on ciliary-derived EVs that are released into the extracellular space. While extensive proteomic analysis of different types of EVs and cilia are available, a systematic look for overlap is lacking. How are cargo targeted to ciliary EVs? What are the functions of the ciliary proteins and other cargo in EVs? Do cilia provide a unique ciliary EV composition to elicit a synchronized response within target cells and tissues? We propose that ciliary EVs act as vessels with a high concentration of cargo for a robust signal transmission and provide a stable environment to enable long distance travel in time and space.

Not All Ciliated Cells Shed EVs Equally: C. elegans as a Model for Ciliary EV Biogenesis and Function

Caenorhabditis elegans cilia are situated on distal dendritic endings of sensory neurons (Bae and Barr 2008; Inglis et al. 2007). The self-fertilizing hermaphrodite has a simple nervous system of 302 neurons of which 60 are ciliated. 32 out of 60 of ciliated sensory neurons in the hermaphrodite, including the amphid, phasmid, and inner labial organs are either directly or indirectly exposed to the environment through openings generated by glial cells (Ward et al. 1975). The hermaphrodite releases EVs from six inner labial type 2 (IL2)-ciliated sensory neurons (Fig. 1a), suggesting that EV shedding is an intrinsic property of certain ciliated neurons (Wang et al. 2014).

Fig. 1.

Caenorhabditis elegans EV-releasing neurons (EVNs). Caenorhabditis elegans cilia are located on distal dendritic endings of sensory neurons. There are 60 and 108 sensory ciliated neurons in C. elegans hermaphrodite and male, respectively. However, only six IL2 neurons in hermaphrodites and 27 neurons in males shed and release EVs into environment. a Names and anatomical position of the EVNs in the hermaphrodite and male. The number of each EVN type is in parenthesis. b EVN sensory organs display common ultrastructure features. Each EVN has a sister neuron, their cilia and ciliary bases are isolated in a lumen formed by two glial cells, the sheath cell, and the socket cell. The lumen is continuous with a cuticular pore, from which the EVN cilium protrudes into environment directly while the sister cilium does not. Only the male-specific cephalic sensory organ components are shown. c A model of the cephalic sensory organ based on electron tomography reproduced from (Wang et al. 2014). The cephalic sensillum contains CEM and CEP cilia, CEM-derived EVs, and the lumen formed by sheath and socket cell. The CEM cilium sheds EVs into the lumen that may be released through the cuticular pore to environment. In a klp-6 (an EVN-specific ciliary kinesin) or a cil-7 (a myristoylated coiled-coil protein) mutant, EVs accumulate in the cephalic lumen as diagnosed by transmission electron microscopy and PKD-2::GFP EVs are not released outside (Maguire et al. 2015; Wang et al. 2014). Mutation in either klp-6 or cil-7 disrupts EVN-mediated sensory functions (Color figure online)

The C. elegans male has 385 neurons, sharing 60 ciliated sensory neurons with the hermaphrodite and also possessing 48 male-specific ciliated sensory neurons devoted to sexual behaviors (O’Hagan et al. 2014; Sammut et al. 2015; Sulston et al. 1980). Caenorhabditis elegans males shed and release EVs from 27 ciliated extracellular vesicle releasing neurons (EVNs) including six shared IL2 neurons and 21 male-specific polycystin-expressing EVNs in the head (four CEM neurons) and tail (16 ray B-type RnB neurons and one hook B-type HOB neuron) (Fig. 1a) (Wang et al. 2014). In these male-specific EVNs, the polycystins lov-1 and pkd-2 are required for male sex drive, response to mate contact, and vulva location (Barr et al. 2001; Barr and Sternberg 1999; Barrios et al. 2008). The C. elegans cilium is also a source of bioactive polycystin-containing EVs: the C. elegans polycystins homologs LOV-1 and PKD-2 are shed into environment from male-specific EVNs (Wang et al. 2014). That C. elegans and mammalian polycystins localize to cilia and EVs, and act in a sensory capacity indicates remarkable ancient functions.

In addition to their EV shedding ability, the 27 ciliated EVNs display several unique ultrastructural characteristics. Most notably, the tips of EVN cilia protrude into the environment from sensory organs (sensilla). This is in contrast to the chemosensory amphid and phasmid cilia that are protected within channels or pores that have access to the surrounding environment. An EVN cilium including the distal-most dendritic ending is surrounded by an extracellular lumen formed by two glial cells, a sheath cell, and a socket cell (Fig. 1b). In the cephalic sensillum that houses the cephalic male (CEM) EVN, glial cell bodies are close to the CEM cell body, and send their processes along the CEM dendrite to wrap the neuron with a tight junction at the distal-most dendritic ending (Fig. 1c). The CEM neuron shares the extracellular lumen with CEP neuron present in both males and hermaphrodites, but only cilium of CEM enters the cuticular pore and is directly exposed to the environment (Fig. 1c). The glial-encased portion of the distal dendrite is slightly larger than the rest of dendrite and may correspond to the periciliary compartment of the amphid neurons, which is dedicated to exocytosis and endocytosis (Kaplan et al. 2012).

High-resolution reconstruction of the adult hermaphrodite anterior sensilla found no EVs in the amphid channel and cephalic lumen and a few EVs in the inner labial lumen, the latter housing the IL2 EVNs (Doroquez et al. 2014). Electron tomography of the anterior sensilla in the male nose reveals 62–259 EVs in cephalic lumen surrounding the CEM cilium (Wang et al. 2014). EVs in the lumen surrounding the CEM cilia may be the source of the GFP-labeled EVs visibly released outside of the worm. In the amphid channel lumen that surrounds 10 cilia, we observed fewer than ten EVs. The position of the EVs found in amphid lumen corresponds to distal regions of amphid channel cilia, whereas EVs are found in the cephalic lumen are situated along the length of the entire CEM cilium, including the periciliary compartment (Wang et al. 2014). We observed one EV budding from or fusing with the ciliary membrane in the CEM ciliary base area (Fig. 2a), which suggests that EVs may be shed from this region. However, we could not resolve the most distal ciliary tip and cannot eliminate the possibility that EVs are also shed from this site, as in EV shedding from Chlamydomonas flagellar tips (Fig. 2a). These results suggest that EVs in the male cephalic and amphid channel lumen are quantitatively and qualitatively different, and are consistent with specific ciliated cell types possessing the ability to shed and release ciliary EVs.

Fig. 2.

Model depicting the cilium as an EV donor (a) and EV acceptor (b). Cilia are highly compartmentalized organelles with an microtubule axonemal skeleton (gray lines) (Blacque and Sanders 2014). The ciliary membrane has different domains that are enriched with certain proteins. The IFT machinery and other ciliary complexes contribute to ciliary compartmentalization. The IFT machinery is required for some aspect of EV biogenesis (Wang et al. 2014). Therefore, we propose that a ciliary compartmentalization might be used for sorting of ciliary EV cargoes. EVs may be released by different ciliary regions including the base (green) or from the tip (purple) of the ciliary membrane. b Cilia are sensory organelles that may interact with EVs (gray bubbles). EVs may originate from neighbors or from cells at long distances. EVs may interact with the ciliary membrane (blue) or may fuse with the ciliary membrane (gray dashes) to promote signal transduction (green arrow). In a disease state, aberrant EVs (red) may trigger a pathological signal (b′) or abnormal cilia (red) may fail to transduce an EV-induced signal (b″) (Color figure online)

The kinesin-3 protein KLP-6 and myristoylated coiled–coil protein CIL-7 are exclusively expressed in the 27 EVNs and regulate ciliary EV biogenesis (Maguire et al. 2015; Wang et al. 2014). These 100–150-nm-sized EVs can be visualized via GFP-labeled EV cargo, which includes PKD-2 and LOV-1 shed and released from male-specific EVNs and CIL-7 from all 27 EVNs. The ability to visualize GFP-tagged EVs combined with the powerful C. elegans molecular genetic toolkit enables the study of EV biogenesis, shedding, release, and signaling in a living animal.

Are Ciliary EVs Exosomes, Ectosomes, or Both?

Determining the identity of an EV is no small task. Exosomes are generated by fusion of multivesicular bodies (MVBs) with the plasma membrane and subsequent release of MVB intraluminal vesicles (Colombo et al. 2014; El Andaloussi et al. 2013). Ectosomes are formed via outward budding of the plasma membrane, although mechanisms controlling formation of ectosomes are not well understood. Exosomes contain certain cargo that may be used as identifiers. However, ectosomes may also carry exosomal markers. Exosomes and ectosomes are shed into bodily fluids and proteomic analysis cannot determine with certainty the nature of the vesicle. To diagnose whether an EV is an ectosomes or exosome, one would ideally visualize EV release in real time. Given their small size, this is a great technical challenge.

In Chlamydomonas, EV budding from the flagellar tip can be directly observed by differential interference microscopy (Wood et al. 2013). As Chlamydomonas flagella are devoid of MVBs, these ciliary EVs are clearly ectosomes (Wood and Rosenbaum, 2015). In C. elegans, ciliary EV identification as exosome or ectosome is not so obvious. GFP-labeled EVs are shed and released from ciliated EVNs, but the site of release (at the ciliary base, along the length of the cilium, or at the ciliary tip) cannot be resolved using standard fluorescence-based microscopy. Super-resolution microscopy might address this question. Using transmission electron microscopy and electron tomography, we observed one EV connected to the ciliary base membrane with a long stalk, indicating that this EV was either budding off of or fusing with the membrane (Wang et al. 2014). We did not observe MVB structures spanning the CEM neuron from the distal dendrite to the ciliary tip. Moreover the MVB components STAM-1 (ESCRT-0), MVB-12 (ESCRT-1), and ALIX-1 (required for endosomal intraluminal vesicle formation) are not required for PKD-2::GFP EV shedding or release (Wang et al. 2014). Combined, these results suggest but do not prove that C. elegans EVs are ectosomes.

EVs and Mammalian Cilia

EVs are associated also with mammalian cilia. The stem-cell marker prominin-1 labels primary cilia and EVs associated with cilia tips of mouse neuroepithelial cells (Dubreuil et al. 2007). Intriguingly, prominin-1 labeled EVs are observed surrounding short cilia. Authors propose that ciliary membrane budding may be a mechanism to control ciliary length, which varies depending on the neuroepithelial stage of development and cell cycle. EVs bind to primary cilia of mouse embryonic fibroblasts (Pampliega et al. 2013). Moreover, EVs surround cholangiocyte primary cilia in the autosomal recessive PKD mouse model, whereas a single EV was occasionally observed attached to a wild-type cilium (Woollard et al. 2007). In three cases, it is not known whether cilia were shedding outward-bound EVs or receiving inward-bound EVs.

Polycystin-containing exosome-like vesicles isolated from mammalian urine interact with primary cilia of kidney and biliary epithelial cells (Hogan et al. 2009). These EVs are termed exosome-like vesicles because they express exosome markers and display exosome morphology. Transmission electron microscopy and immunogold labeling showed polycystin-1 on intraluminal vesicles in rat cholangioctyes MVBs. While these data suggest an MVB origin, Ward and colleagues conclude that budding from apical membranes or cilia cannot be eliminated (Hogan et al. 2009). While unambiguous data supporting a universal origin of ciliary EVs are lacking, we propose that ciliated cells may use multiple mechanisms to shed EVs (Fig. 2a). Cilia may shed EVs at the ciliary base via budding as ectosomes or via fusion of MVBs and release of intraluminal vesicles as exosomes. Cilia may also shed EVs as ectosomes at the ciliary tip, as demonstrated in Chlamydomonas by Rosenbaum and colleagues (Wood et al. 2013), or along the length of the ciliary membrane. These ciliary EVs may have different functions depending on their origins and cargo composition.

Regulation of Ciliary EV Biogenesis, Shedding, and Release

Intraflagellar transport (IFT) is an evolutionarily conserved process required for the ciliary formation and maintenance in organisms as diverse as algae, worms, mice, and men (Rosenbaum and Witman 2002). The IFT train is composed of IFT-A and IFT-B multiprotein complexes and is transported by anterograde kinesin-2 from ciliary base to tip and retrograde dynein motor from the tip to base. The IFT machinery may also have a role beyond ciliary construction and maintenance (Baldari and Rosenbaum 2010). Some clues from C. elegans and Chlamydomonas suggest that the IFT machinery may be required for EV biogenesis, EV shedding, EV release, and/or cilia–EV interactions.

Chlamydomonas sheds ciliary ectosomes that contain a lytic enzyme that digests the mother cell wall and is required for post-mitotic hatching of daughters (Kubo et al. 2009; Wood et al. 2013). Flagella-less ift88-null mutants cannot hatch and addition of ciliary ectosomes isolated from wild-type cells induced hatching. These results indicate that an intact flagellum is required for the production of functional EVs. The role of the IFT machinery in Chlamydomonas ciliary shedding is yet to be explored.

In C. elegans EVNs, IFT-A and IFT-B components, IFT kinesin-2 and dynein motors, and the EVN-specific ciliary kinesin-3 KLP-6 are required for the release of PKD-2::GFP-labeled EVs (Wang et al. 2014). In these mutant backgrounds, the quantity of environmentally released EVs is significantly reduced (Wang et al. 2014) and PKD-2::GFP accumulates in the ciliary base region, resulting in a similar ciliary localization defective or Cil phenotype (Bae et al. 2006, 2008; Peden and Barr 2005; Qin et al. 2005). Whether PKD-2::GFP accumulates in the distal dendrite or cephalic lumen cannot be resolved by light microscopy. Transmission electron microscopy and electron tomography showed that klp-6 mutant males accumulate a large number of luminal EVs and possess a lumen doubled in volume compared to wild type (Fig. 1c) (Wang et al. 2014). The excessive accumulation of luminal EVs indicates that klp-6 is either a negative regulator of EV biogenesis and shedding, or a positive regulator of EV environmental release. In EVN cilia, klp-6 also modulates IFT (Morsci and Barr 2011). The similar Cil phenotype of klp-6 and IFT mutants suggest a similar mechanism of action. Future ultrastructural analysis of IFT mutants will reveal what role the IFT machinery plays in EV biology.

How might IFT contribute to EV biogenesis, shedding, release, and/or signaling? In Chlamydomonas, polystyrene microspheres adhere to and are moved bidirectionally along the external flagellar surface (Bloodgood 1995, 1988). IFT drives flagellar gliding motility and the extraflagellar transport of the major flagellar surface glycoprotein protein FMG1-B (Shih et al. 2013). When an anti-FMG1-B antibody is attached to beads, beads and IFT trains move in similar speeds. In Ctenophores, or comb jellies, individual cells are transported up the external ciliary surface, independent of ciliary beating, to build the statolith, a gravity-sensing organ (Noda and Tamm 2014). In a similar scenario, IFT and KLP-6 inside the cilium may propel EVs along the outside ciliary surface. In this model, EVs and cilia express unidentified surface proteins that couple the EV to the cilium, enabling internal ciliary motors to propel EVs along the ciliary surface (Maguire et al. 2015). Alternatively, ciliary membrane proteins move laterally within the ciliary membrane bi-layer in an IFT-dependent manner (Huang 2007; Qin et al. 2005). Hence, cargo carried on the ciliary membranes by IFT may potentially become components of released bioactive ectosomes.

Targeting Cargo to Ciliary EVs

Understanding how cargo is directed to EVs has important therapeutic implications (Gyorgy et al. 2015). Cargo sorting to C. elegans ciliary EVs appears to be selective—not all ciliary proteins get packaged into EVs (Wang et al. 2014). Additionally, EV cargo displays cell type specificity. The myristoylated coiled–coil protein CIL-7 localizes to ciliary EVs released by all 27 EVNs (Maguire et al. 2015), while the polycystins PKD-2 and LOV-1 are released in EVs by the male-specific EVNs and not the shared IL2 EVNs (Wang et al. 2014). Conversely, we identified EV cargo released from the IL2 but not male-specific EVNs (Wang et al. 2015). We hypothesize that EVs have at least two types of cargo, one for structure (for example, CIL-7) and the other for function (for example, the polycystins).

CIL-7 and the kinesin-3 KLP-6 are both expressed in all 27 EVNs, localize to cilia, regulate EV biogenesis, and are required for polycystin-mediated male mating behaviors (Maguire et al. 2015). However, CIL-7 but not KLP-6 is EV cargo (Maguire et al. 2015). The myristoylation site in CIL-7 is necessary for CIL-7 function and localization to EVs but not cilia (Maguire et al. 2015). N-myristoylation is used by proteins for membrane anchoring and for ciliary localization of proteins in Trypanosome flagella, C. elegans sensory neurons, mammalian photoreceptors, and retinal pigment epithelial cells (Evans et al. 2010; Maric et al. 2010; Ramulu and Nathans 2001; Wright et al. 2011). Myristoylation targets proteins to EVs in Jurkat T-cells (Shen et al. 2011). In the cpk mouse model of PKD, the cpk mutation lies in the Cystin gene, which encodes a myristoylated cilia- and EV-associated protein (Hogan et al. 2009; Tao et al. 2009). The Cystin myristoylation signal is necessary for ciliary targeting in inner medullary collecting duct cells (Tao et al. 2009), and perhaps EVs. We conclude that, in C. elegans, myristoylation provides a cis-acting motif for EV targeting in vivo. As ciliary EVs and their cargo are identified and characterized, perhaps “EV zip codes” or “EV localization signals” similar to nuclear localization signals will be discovered.

Ciliary EV Signaling and Bioactivity

EVs mediate a broad range of intercellular communication by carrying bioactive proteins, lipids, and nucleic acids (El Andaloussi et al. 2013). The function of ciliary EVs is largely unknown, and the following speculations are therefore based on limited data. The cilium may shed and release EVs as a rapid way to modulate membrane composition, adjust protein levels, and downregulate signaling molecules. This might be important for the ciliary sensory function, as signaling molecules are directly related to sensitivity and sensory adaptation. In accordance with this theory, the major flagellar glycoprotein FMG1 in Chlamydomonas is constitutively shed, while the SAG1 protein is shed only during signaling events but not in resting gametes (Bloodgood et al. 1986; Cao et al. 2015).

In C. elegans, the ciliary kinesin-3 KLP-6 and myristoylated coiled-coil protein CIL-7 are specifically expressed in the 27 EVNs, regulate EV release, and are required male mating behavior. klp-6 and cil-7 mutant males accumulate excessive amounts of EVs in the glial-surrounded lumen of male-specific sensory organs (Fig. 1c) (Maguire et al. 2015; Wang et al. 2014). In invertebrates and specialized sensory organs of higher animals, sensory neurons are ciliated and either exposed to the environment or associated with extracellular matrix, which may be important for sensory transduction (Andres et al. 2014; Cook et al. 2008; McGlashan et al. 2006). In this context, EVs may contribute to the physiological functioning of C. elegans male-specific sensory organs.

In addition to cilia-dependent functions, ciliary EVs are bioactive and can act non-autonomously. Chlamydomonas flagella release EVs containing an enzyme to digest the mother cell wall and free daughter cells (Wood et al. 2013). In C. elegans, isolated EVs from wild type cause a change in male locomotion and trigger male tail-chasing behavior, whereas EVs isolated from a klp-6 mutant and lacking PKD-2::GFP do not elicit male tail-chasing behavior (Wang et al. 2014). These results show that EV cargo content is essential for EV bioactivity. This is a radically new function for EVs, which are generally thought to influence intercellular communication within an animal, and a first demonstration that EVs are a way to communicate between individuals and influence the behavior of other animals.

To define properties of a ciliated EV-releasing cell, we performed RNAseq on the 27 GFP-labeled EVNs isolated from adult C. elegans (Wang et al. 2015). We identified 335 significantly overrepresented genes, of which 61 were validated by GFP reporters. The EVN transcriptional profile uncovered new pathways controlling EV biogenesis and polycystin signaling and also identified EV cargo, which included an ASIC channel and an antimicrobial peptide. Human urinary exosomes contain antimicrobial peptides and defensins that can lyse Escherichia coli (Hiemstra et al. 2014). Hence, EVs may play a role in innate immunity, protecting C. elegans or the mammalian kidney from pathogens.

In addition to carrying protein cargo, EVs may contain small RNAs. One intriguing possibility is that ciliary EVs may transfer small RNAs and cilia may also receive EVs containing small RNAs to influence long-term signaling processes. This would be a means of communication between tissues to coordinate developmental timing (Benkovics and Timmermans 2014), between animals of the same species as a form of quorum sensing (Sarkies and Miska 2013), or between pathogen and host to co-opt the host’s immune response (Buck et al. 2014).

Do Abnormal Cilia–EV Interactions Contribute to Ciliopathies?

Ciliary EVs have specific cargo, including surface-associated proteins, channels, and receptors (Bloodgood 1995; Cao et al. 2015; Wang et al. 2014, 2015; Wood et al. 2013). These EV surface proteins may be important for EV-target cell interactions (Fig. 2b). Fractionated Chlamydomonas EVs bind only to flagella and rarely to the cell body, suggesting that flagellar EVs bind to a specific region of the single-celled algae (Cao et al. 2015). We speculate that the polycystins (or another ciliary receptor) may mediate an interaction between cilia and EVs, allowing a homotypic interaction between polycystin-positive cilia and polycystin-positive EVs (Fig. 2b). Indeed, rapid interaction between exosome-like vesicles and primary cilia has been reported. A defect in ciliary EVs (Fig. 2b′) or a defect in the cilium (Fig. 2b″) would result in abnormal EV–cilia interactions and a potentially pathogenic cellular response.

There is experimental evidence for this hypothesis. The fibrocystin protein, which is encoded by causal gene of ARPKD, pkhd1, is required for the rapid interaction between mammalian primary cilia and EVs (Hogan et al. 2009; Masyuk et al. 2010). In the ARPKD patient or the mouse model (scenario in Fig. 2b″), EVs are found attached to cilia, but not all cilia (Hogan et al. 2009), consistent with selective EV targeting and attachment.

Somlo and colleagues made a surprising discovery that simultaneous inactivation of the polycystins and cilia resulted in a decrease in PKD severity compared with polycystin-only knockout (Ma et al. 2013). Authors propose that the polycystins act as inhibitory signals that modulate an unidentified pathway and that requires intact cilia to function. A tantalizing but untested possibility is that polycystin inactivation results in both abnormal ciliary EVs and abnormal ciliary signaling, resulting in severe pathology. Consistent with this idea, EVs isolated from ADPKD patients display different proteins than unaffected individuals (Hogan et al. 2015). By removing the cilium in a Pkd1 or Pkd2 mutant, abnormal EV–cilia interactions and signaling are abrogated, thereby suppressing cystogenesis. Future studies on the cilia-dependent cyst-activating pathway should reveal whether or not EVs play a role in ADPKD or other ciliopathies.

One critical and perhaps overlooked aspect of the cilium is its dynamic nature: the protein composition changes constantly, the membrane is renewed, cilia length is changing and even the entire cilium can be shed and regenerated! The finding that the cilium shed EVs adds another layer of complexity to this essential organelle. A deeper understanding of the dynamic regulation of cilia may provide insight for intervention or ultimately a cure for ciliopathies. EVs are promising diagnostic tools and have great therapeutic potential to deliver cell-specific treatments (Lo Cicero et al. 2015). Urinary EVs may eventually be used in diagnosing and monitoring ADPKD (Hogan et al. 2015). Many exciting new areas in this emerging new field of ciliary EVs await exploration. The best is yet to come.

Abbreviations

ADPKD

Autosomal dominant polycystic kidney disease

CEM

Cephalic male neuron

EV

Extracellular vesicle

EVN

Extracellular vesicle releasing neuron

IFT

Intraflagellar transport

IL2

Inner labial type 2 neuron

MVB

Multivesicular body