Article Images (0)
×![]()
featured_play_listSave article to folder
X
Create new folder
Full Text
PDF
Table of Contents
Abstract
Polycystin Channel Complexes
Abstract
INTRODUCTION
Significance Of Polycystins In Human Development And Disease
HOMOMERIC POLYCYSTIN CHANNELS
Homomeric Assembly And Topology
Voltage Sensor Domain.
TOP Domain.
Pore Domain.
C-Terminal Domain.
Proposed Cell And Physiological Functions
PKD2 Channels.
PKD2L1 Channels.
PKD2L2 Channels.
HETEROMERIC POLYCYSTIN CHANNELS
Heteromeric Assembly And Topology
Ectodomain.
Pore Domain And Proposed Gating Mechanisms.
Proposed Cell And Physiological Functions
PKD1-PKD2 Channels.
PKD1L1-PKD2 Channels.
PKDREJ-PKD2 And PKDREJ-PKD2L1 Channels.
PKD1L1-PKD2L1 Channels.
PKD1L3-PKD2L1 Channels.
SUMMARY
Table 1
SUMMARY POINTS
FUTURE ISSUES
ACKNOWLEDGMENTS
Footnotes
Abstract
Polycystin subunits can form hetero- and homotetrameric ion channels in the membranes of various compartments of the cell. Homotetrameric polycystin channels are voltage- and calcium-modulated, whereas heterotetrameric versions are proposed to be ligand- or autoproteolytically regulated. Their importance is underscored by variants associated with autosomal dominant polycystic kidney disease and by vital roles in fertilization and embryonic development. The diversity in polycystin assembly and subcellular distribution allows for a multitude of sensory functions by this class of channels. In this review, we highlight their recent structural and functional characterization, which has provided a molecular blueprint to investigate the conformational changes required for channel opening in response to unique stimuli. We consider each hetero- and homotetrameric polycystin channel type individually, discussing how they contribute to organelle and cell biology, as well as their impact on the physiology of various tissues.
Polycystin Channel Complexes
Annu Rev Physiol. Author manuscript; available in PMC 2023 Mar 21. Published in final edited form as: Annu Rev Physiol. 2023 Feb 10; 85: 425–448. doi: 10.1146/annurev-physiol-031522-084334 PMCID: PMC10029091 NIHMSID: NIHMS1882989 PMID: 36763973 Orhi Esarte Palomero , * Megan Larmore , * and Paul G. DeCaen Author information Copyright and License information Disclaimer Department of Pharmacology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois, USA * These authors contributed equally to this article ude.nretsewhtron@neaced.luap Copyright notice The publisher's final edited version of this article is available free at Annu Rev Physiol
Abstract
Polycystin subunits can form hetero- and homotetrameric ion channels in the membranes of various compartments of the cell. Homotetrameric polycystin channels are voltage- and calcium-modulated, whereas heterotetrameric versions are proposed to be ligand- or autoproteolytically regulated. Their importance is underscored by variants associated with autosomal dominant polycystic kidney disease and by vital roles in fertilization and embryonic development. The diversity in polycystin assembly and subcellular distribution allows for a multitude of sensory functions by this class of channels. In this review, we highlight their recent structural and functional characterization, which has provided a molecular blueprint to investigate the conformational changes required for channel opening in response to unique stimuli. We consider each hetero- and homotetrameric polycystin channel type individually, discussing how they contribute to organelle and cell biology, as well as their impact on the physiology of various tissues. Keywords: polycystin, autosomal dominant polycystic kidney disease, ADPKD, TRP channel, gating mechanism, primary cilia, ciliopathies
Keywords: polycystin, autosomal dominant polycystic kidney disease, ADPKD, TRP channel, gating mechanism, primary cilia, ciliopathies
INTRODUCTION
The founding members of the polycystins, PKD1 and PKD2, were first identified due to mutations in genes causing autosomal dominant polycystic kidney disease (ADPKD) ( 2 – 5 ). Since then, the polycystin family has expanded to eight proteins ( 6 – 11 ). Based on sequence and structural homology, polycystin members can be grouped into two types of channel subunits ( Figure 1a , ,b). b ). PKD1-related subunits (PKD1, PKD1L1, PKD1L2, PKD1L3, and PKDREJ) have 11 transmembrane segments (S1–S11) that combine features of ion channels and adhesion G protein–coupled receptors (aGPCRs). PKD2-related subunits (PKD2, PKD2L1, and PKD2L2) have six transmembrane segments (S1–S6) and are members of the transient receptor potential (TRP) family of ion channels ( 1 , 12 ). PKD2-related subunits can form homotetrameric ion channels or oligomerize with PKD1-related members to form heterotetrameric channels. Polycystin channels assembled in these forms are uniquely regulated by structural motifs of the constituent subunits—features that have only recently been determined by cryo-electron microscopy (cryo-EM) and electrophysiology techniques ( 13 – 19 ). Polycystin orthologs can be found from yeast to humans ( 20 ); however, we primarily discuss the mammalian proteins, summarizing recent discoveries regarding their unique monomeric architectures, oligomeric assembly, and ion channel function. While there are several reports of polycystin subunits forming heteromeric complexes with divergent channel subunits (e.g., TRPV4) ( 21 , 22 ), we focus exclusively on the features of complexes formed within the polycystin class. The composition, structure, and putative physiological regulation of polycystin complexes are organized under subsections devoted to each channel type. Open in a separate window Figure 1 Topologies of polycystin channel subunits. The eight members are divided into PKD2-related ( a ) and PKD1-related ( b ) channel subunits. The Uniprot IDs of the human polycystin subunits are P98161 (PKD1), Q8TDX9 (PKD1L1), Q7Z442 (PKD1L2), Q7Z443 (PKD1L3), Q9NTG1 (PKDREJ), Q13563 (PKD2), Q9P0L9 (PKD2L1), and Q9NZM6 (PKD2L2). Abbreviations: CTD, C-terminal domain; GPS, G protein–coupled receptor proteolytic site; LDL-A, low density lipoprotein A; PD, pore domain; PLAT, polycystin-1 lipoxygenase and alpha toxin (domain); REJ, receptor for egg jelly domain; TOP, tetragonal opening for polycystins (domain); VSD, voltage sensor domain; SUEL, sea urchin egg lectin; WSC, cell Wall integrity and Stress Component protein. Significance of Polycystins in Human Development and Disease The importance of polycystin dysfunction in human disease is underscored by association with Autosomal Dominant Polycystic Kidney Disease (ADPKD)—a monogenetic and progressive form of kidney disease that impacts millions of people worldwide ( 23 , 24 ). Variants in PKD1 and PKD2 account for ~95% of patients with ADPKD, who typically require kidney transplants in midlife. The human genetic basis of ADPKD and related murine PKD phenotypes caused by loss of PKD1 and PKD2 expression are well characterized ( 25 , 26 ). Although it is considered a dominant monogenic disease, ADPKD is recessive at the cellular level, where cysts develop from cells after acquiring a second somatic mutation to deactivate the remaining normal allele ( 27 ). Assembled polycystin channels frequently localize in discrete cellular compartments, exemplified by their sensory function in the primary cilia—an immotile antenna-like organelle that resides on the apical membrane. Despite their minute dimensions (~5 μm long, <0.5 μm in diameter), primary cilia have profound implications for human development and disease ( 28 ). They are found in all tissues, and mutations in genes encoding ciliary proteins are associated with inherited developmental disorders known as ciliopathies ( 29 ). Several lines of evidence support classification of ADPKD as a ciliopathy—normally PKD1 and PKD2 subunits traffic to the primary cilia of the nephron, where their exclusion promotes kidney cystogenesis in ADPKD mouse models. Human variants also cause defective channel function in the ciliary membrane ( 30 , 31 ). Beyond ADPKD and the kidney collecting duct, evidence for polycystin function in primary cilia comes from dysregulated PKD1L1 ciliary signaling in the embryonic node ( 119 ), which is associated with laterality defects and complex congenital heart disease ( 32 ). In this review, we discuss phenotypes from polycystin gene knockout mice, which have provided clues regarding their independent roles in various organ systems, noting that our understanding of polycystin function in human physiology is still limited and needs further elucidation. We emphasize recent findings that present new opportunities to understand this integral class of membrane proteins.
Significance of Polycystins in Human Development and Disease
The importance of polycystin dysfunction in human disease is underscored by association with Autosomal Dominant Polycystic Kidney Disease (ADPKD)—a monogenetic and progressive form of kidney disease that impacts millions of people worldwide ( 23 , 24 ). Variants in PKD1 and PKD2 account for ~95% of patients with ADPKD, who typically require kidney transplants in midlife. The human genetic basis of ADPKD and related murine PKD phenotypes caused by loss of PKD1 and PKD2 expression are well characterized ( 25 , 26 ). Although it is considered a dominant monogenic disease, ADPKD is recessive at the cellular level, where cysts develop from cells after acquiring a second somatic mutation to deactivate the remaining normal allele ( 27 ). Assembled polycystin channels frequently localize in discrete cellular compartments, exemplified by their sensory function in the primary cilia—an immotile antenna-like organelle that resides on the apical membrane. Despite their minute dimensions (~5 μm long, <0.5 μm in diameter), primary cilia have profound implications for human development and disease ( 28 ). They are found in all tissues, and mutations in genes encoding ciliary proteins are associated with inherited developmental disorders known as ciliopathies ( 29 ). Several lines of evidence support classification of ADPKD as a ciliopathy—normally PKD1 and PKD2 subunits traffic to the primary cilia of the nephron, where their exclusion promotes kidney cystogenesis in ADPKD mouse models. Human variants also cause defective channel function in the ciliary membrane ( 30 , 31 ). Beyond ADPKD and the kidney collecting duct, evidence for polycystin function in primary cilia comes from dysregulated PKD1L1 ciliary signaling in the embryonic node ( 119 ), which is associated with laterality defects and complex congenital heart disease ( 32 ). In this review, we discuss phenotypes from polycystin gene knockout mice, which have provided clues regarding their independent roles in various organ systems, noting that our understanding of polycystin function in human physiology is still limited and needs further elucidation. We emphasize recent findings that present new opportunities to understand this integral class of membrane proteins.
HOMOMERIC POLYCYSTIN CHANNELS
Homotetrameric PKD2-related polycystin subunits domain swap to form functional ion channels. The opening and closing (gating) of these channels is controlled by membrane potential (voltage) and modulated by membrane composition and intracellular Ca 2+ . In the following subsections, we discuss the structural features responsible for their molecular regulation. Homomeric Assembly and Topology Homotetrameric polycystin channels are the least structurally complex and most-characterized version of their oligomeric assemblies. The PKD2-related subunits have six transmembrane segments (S1–S6) and are members of the TRP family of ion channels, which are part of the larger voltage-gated ion channel (VGIC) superfamily ( 33 ) ( Figure 1a ). TRP channels form cation-permeable ion channels that are widely known for their role as transducers of sensory modalities ( 34 ). PKD2-related subunits form the polycystin TRP subclass, which are also called TRPPs (see sidebar titled Polycystin Nomenclature ). PKD2-related subunits have five structural domains: an intracellular N-terminal domain, a voltage sensor domain (VSD; S1–S4), an extracellular tetragonal opening for polycystins domain (TOP; also called the polycystin domain), a pore domain (PD; S5–S6), and an intracellular C-terminal domain (CTD) ( Figure 1a ). Multiple structures of PKD2 and PKD2L1 have been resolved by cryo-EM, demonstrating that these subunits assemble as homotetrameric channels, where four subunits combine to form an ion conductive unit ( 13 – 15 , 35 ) ( Figure 2 ). Each of the channel subunits form interfaces via their PD and TOP domains, implicating their role in oligomeric assembly. Open in a separate window Figure 2 Assembly and proposed gating mechanism of homomeric polycystin channels. ( a ) Transmembrane view of a single PKD2 subunit with identified structural domains and location of ADPKD-causing missense variants ( 13 , 50 ). ( b ) Transmembrane view of overlaid PKD2 ( gray ; PDB 5T4D) and PKD2L1 ( tan ; PDB 5Z1W) channels, with expanded views of the VSD and PD ( 13 , 15 ). ( c ) Extracellular and transmembrane view of domain-swapped PKD2 channels compared to nondomain-swapped Kv11 Eag1 channel. Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; Kv, member of voltage-gated potassium channel family; PD, pore domain; PDB, protein data bank; TOP, tetragonal opening for polycystins domain; VSD, voltage sensor domain; VSDL, voltage sensor-like domain. SIDEBAR 1: POLYCYSTIN NOMENCLATURE The revised and current International Union of Basic and Clinical Pharmacology/British Pharmacological Society (IUPHAR/BPS) nomenclature creates ambiguity regarding the genetic identity of the polycystin family members of transient receptor potential ion channels (TRPPs), especially when cross-referencing manuscripts that describe subunits using the former system ( 1 ). Traditionally, the products of polycystin genes (e.g., PKD1 ) are referred to as polycystin proteins (e.g., polycystin-1; see Table 1 ). However, this nomenclature is neither practical nor available for all of the genes within this family. For simplicity and to prevent confusion, we refer to the polycystin proteins by their gene name rather than differentiating gene and protein with separate names. PKD2 and PKD2L1 adopt domain-swapped architectures ( Figure 2c ). Domain swapping occurs when the VSD from one subunit packs against the PD helices of the neighboring subunit. Details of domain swapping provide mechanistic insights into how the channel opens in response to voltage. In the structures of PKD2-related channels, the VSD in one subunit is connected to the PD of a neighboring subunit by a 12-Å-long S4–S5 linker helix, which runs parallel to the plane of the membrane’s cytoplasmic surface and forms contacts with S5–S6 helices of the PD, analogous to the bundle strap of a floral bouquet. Here, the state of the VSD (activated or deactivated) is linked to the state of the PD (open or closed) through structural interactions of the S4–S5 linker helix with the pore-forming S6 ( 36 ) ( Figure 2c ). In contrast, the VSD and PD from nondomain-swapped channels (e.g., Eag11 KV channels) pack as contiguous units ( 37 , 38 ). Nondomain-swapped channels do not have a helical S4–S5 linker, and their mechanism of gate control is largely undetermined. . Voltage sensor domain. Additional evidence of the voltage-dependent gating of PKD2-related channels comes from the structures of their VSDs. PKD2 and PKD2L1 contain positively charged residues called gating charges in their S4 ( 39 ) ( Figure 2b , bottom ). Typically, VGICs have at least four positive gating charges, either lysine (K1–K4) or arginine (R1–R4), which are found at every third residue. However, in PKD2-related channels, only the lower two positions (K3 and K4) are conserved ( 40 ). One helical turn of the S4 in PKD2 and PKD2L1 adopts a 3 10 -helix that aligns the K3 and K4 with their side chains pointing into the interior of the VSD ( 13 ) ( Figure 2b ). Here, the gating charges are proposed to be stabilized by state-dependent cation-π (K3-Y366 deactivated state) and charge-counter charge interactions (K3-D390 deactivated state, K4-D390 activated state) formed by side chains of the residues (Y366, K390, D390) extending from the S2 and S3 helices. An ADPKD-associated mutation, D511V (equivalent to PKD2L1 D390V), abolishes PKD2 function in reconstitution assays by presumably disrupting the deactivated state interaction with the internal gating charge ( 41 ). Because these are positively charged residues within a transmembrane environment, the S4 moves to the activated state in response to membrane depolarization and returns to the S4 deactivated state upon repolarization. The measured transfer of PKD2L1 gating charges is equal to 4.4 elementary charges (q s ) per channel, or the equivalent of transferring one lysine per VSD ( 39 ). The S4 is predicted to move at least 6 Å outwardly, or the equivalent of 2 helical turns, during the activation process. The outward movement is coupled to the lateral displacement of the S4–S5 linker and opening of the lower pore gate ( 39 ). At the interface between the VSD and PD, cholesterol and phosphatidylinositol-like densities have been resolved for PKD2 ( 42 – 44 ). Cholesterol and phosphatidylinositol are known modulators of ion channels, and their distribution is unique among organelle membranes ( 45 , 46 ). Thus, further interrogations of lipid-channel interactions in primary cilia and endoplasmic reticulum (ER) membranes will likely elucidate critical mechanisms of channel gating and trafficking within the cell. . TOP domain. The TOP domain sits on the extracellular side of the VSD, extending from the S1 and S2 helices to form extracellular contacts with the S3–S4 loop ( Figures 1a and and2a). 2a ). Although this protein fold is unique to polycystins, it is structurally similar to the extraluminal domain of the lysosomal mucolipin TRP channels ( 47 , 48 ). The TOP domain of PKD2 contains several conserved asparagine-linked glycosylation sites (N299, N305, N328, N362, and N375) that are also conserved in PKD2L1 and may facilitate their trafficking to the primary cilia or other cell membranes ( 49 ) ( Figure 1a ). Importantly, the TOP domain of PKD2 is a hotspot of ADPKD missense variants ( 50 ) ( Figure 2a ). The TOP domains of each PKD2 and PKD2L1 subunit are composed of five beta strands (β1–β5) with two intervening alpha helices. The β1–β5 stretch forms two prominent hairpin turns called fingers 1 and 2, which interface between each channel subunit. Several ADPKD-causing variants (K322Q/W, R325Q/P, and C331S) destabilize finger 1 hydrogen bond, disulfide, and cation-π interactions that disrupt channel function by shifting the voltage dependence of PKD2 opening by more than +80 mV ( 51 ). Mutations at these conserved positions in PKD2L1 cause similar shifts in voltage dependence, suggesting that the function of the TOP domain is shared by the two channels ( 51 ). As the TOP domain rests on the extracellular side of the VSD, one apparent function of this domain is to directly regulate the state of the VSD, which in turn controls the ion conductive state of the PD. This form of allosteric regulation of polycystin channels may be analogous to how peptide toxins from spiders and scorpions block cardiac and neuronal sodium and potassium VGIC function by binding and immobilizing their VSDs from the extracellular side of the membrane ( 52 , 53 ). . Pore domain. The PD forms the central ion permeation pathway and houses the selectivity filter responsible for cation preference ( Figure 2b , top ). The selectivity filters of PKD2 and PKD2L1 are scaffolded by two short alpha helices (PH1, PH2) similar to Na v s and some TRP channels ( 54 ). The negative dipole created by the orientation of the amino acids in the eight pore helices (two helices per subunit) contributes to the electronegativity of the PD. Both PKD2 and PKD2L1 channels conduct multiple cations but differ in their selectivity for Ca 2+ . PKD2L1 is more selective for Ca 2+ than PKD2, which is conferred by a second ring of aspartate residues that is not present in PKD2 ( 55 ). Based on the cryo-EM structure and functional analysis, PKD2-related channels are proposed to have multiple gates—hydrophobic constriction sites that move in and out of the ion conduction pathway in a state-dependent manner ( 13 , 15 ). The lower gate is found on the S6 and is proposed to be mobilized by uncoiling its secondary structure from a π-helix to an α-helix ( 56 ). This gate closely resembles those found in VGICs and involves a lateral displacement or splaying of the S6 through a series of hydrogen bonds formed with residues on the S4–S5 linker. The proposed upper gate is found at sites within the selectivity filter: L521 and D523 in PKD2L1 and L641 and D643 in PKD2 ( 13 , 15 ). However, the proposition of an upper gate is controversial within the TRP channel literature. It is unclear how motions from the VSD would be transferred to the upper gate or whether this site is most akin to the inactivation gate found in VGICs. Collapse of the selectivity filter has been widely described in Na v and K v channels and is functionally linked to inactivation modes. Thus, it is possible that polycystin inactivation, which has been reported for PKD2L1, might be conferred by structural changes in the upper gate ( 57 ). . C-terminal domain. The CTDs of all three PKD2-related channel subunits have an EF-hand motif, followed by a coiled-coil domain and multiple intermittent serine phosphorylation sites. Phosphorylation of these sites enhances PKD2 current density in heterologous expression conditions, indicating this domain’s role in channel membrane localization ( 58 – 61 ). The contiguous cytoplasmic CTDs are not structurally determined in all reported PKD2 and PKD2L1 structures, but the EF-hand and coiled-coil structures have been solved as fragments ( 62 – 65 ). The EF-hand from PKD2 and PKD2L1 coordinates one Ca 2+ ion, with reported affinities approximating 19–22 μM ( K d ) ( 31 , 66 ). However, abolishing Ca 2+ affinity of the EF-hand does not alter Ca 2+ -dependent kinetics of the PKD2 channel current ( 31 ). Furthermore, deletions or neutralizations of this site do not contribute to cystic kidney phenotypes in mice. These findings suggest that additional Ca 2+ -binding sites within PKD2 or Ca 2+ -dependent modifiers are responsible for regulating the channel’s activity. The putative function of the coiled-coil motif in PKD2-related channels is also controversial. The isolated coiled-coil motifs of PKD2 and PKD2L1 channels oligomerize as trimers when expressed as fragments, but this stoichiometry opposes the homotetrameric assembly of whole channel structures ( 67 ). Although removal of the coiled-coil motif has little impact on PKD2L1 trafficking and function in heterologous expression systems, nonsense variants that cleave this motif in PKD2 are associated with ADPKD and are required for its heteromeric assembly with PKD1 and PKD1L1 ( 55 , 64 , 68 – 70 ). Taken together, these findings suggest that the CTDs of PKD2-related channels may function differently depending on their assembly (homomeric or heteromeric) with other polycystin subunits. Proposed Cell and Physiological Functions Homomeric polycystin channels are expressed in a variety of tissues. The biophysical features and evidence supporting each channel’s proposed physiological function are described in the following subsections. . PKD2 channels. PKD2 is the founding member of the polycystin family of TRP channels. Results from RNA sequencing and northern blotting suggest that PKD2 has low tissue specificity and is expressed at similar levels in adults and during early development ( 3 , 71 ). Approximately 15% of ADPKD patients have germline variants of PKD2 and thus the preponderance of knowledge about PKD2 function comes from studies carried out in kidney cells. Functional and immunohistochemical evidence for this channel is reported in the ER and primary cilia of the thick ascending limbs of the loop of Henle, the distal convoluted tubule, and the collecting ducts ( 72 ). In these locations, PKD2 amplifies the cytosolic Ca 2+ release through direct association with the inositol trisphosphate receptor ( 41 , 73 , 74 ). In addition, reduced PKD2 expression attenuates mitochondrial Ca 2+ buffering and increases the fragmentation of the mitochondrial network ( 75 ). Previous reports characterizing reconstituted PKD2 proteins isolated from the ER have produced divergent results regarding ion selectivity and conductance. However, voltage-clamp measurements of endogenous and heterologous PKD2 channels from the primary cilia are more consistent (γK = 139 pS, γNa = 89 pS, γCa = 4 pS), where ion permeation favors monovalent cations (P Ca /P Na = 0.1) ( 19 , 76 ). The Ca 2+ concentration gradient is 10,000 times greater on the outside of the cilium. Although PKD2 has a relatively low Ca 2+ selectivity, its activity would have a major influence on internal Ca 2+ conditions and downstream Ca 2+ -dependent signaling. PKD2 voltage dependence is outwardly rectifying, where single channel opening probability increases at positive potentials. Resting Ca 2+ in primary cilia (390–580 nM) is higher than in the cytosol (40–90 nM) ( 77 ). Elevated intraciliary Ca 2+ modulates PKD2 voltage dependence through processes called Ca 2+ -dependent modulation and subsequent Ca 2+ -dependent desensitization—two biophysical regulatory mechanisms that control its function at physiological membrane potentials ( 31 , 78 ). Extrarenal manifestations of ADPKD implicate PKD2 function in the cardiovascular system ( 79 ). In cardiomyocytes, PKD2 modulates Ca 2+ release from stimulated ryanodine receptors through direct association ( 80 ). Reports of PKD2 function in the vasculature are paradoxical. In arterial smooth muscle cells, PKD2 is proposed to mediate systemic blood pressure and contribute to the myogenic response in cerebral arteries though vasoconstriction ( 81 , 82 ). However, in the vascular endothelium, PKD2 is also proposed to mediate vasodilation through nitric oxide synthase activation and reduce blood pressure in mesenteric arteries ( 83 ). Thus, PKD2 function in smooth muscle cells and endothelial cells might be specific to the location within the vasculature. . PKD2L1 channels. Human PKD2L1 is highly expressed in the spleen, testis, retina, heart, and brain ( 9 , 84 ). The PKD2L1 peptide is 56% identical and 76% similar to PKD2, with the greatest divergence in the cytoplasmic CTD ( 9 ). Unlike PKD2, PKD2L1 has been excluded as a candidate gene for all forms of polycystic kidney disease, with little to no expression in the kidney. Thus far, no variants in PKD2L1 have been associated with human disease. Nonetheless, the phenotypes of murine PKD2L1 knockout models implicate this channel’s function in neurons of the central nervous system, cardiomyocytes, and early development. Ablation of PKD2L1 causes hippocampal and thalamocortical hyperexcitability, resulting in susceptibility to drug-induced seizures ( 85 ). PKD2L1 traffics to the primary cilia of pyramidal neurons and is reported to colocalize with the β2-adrenergic receptor. However, PKD2L1 function in hippocampal neurons has not been directly measured, and the mechanism of ciliary PKD2L1 regulation of neuronal excitability is unknown. PKD2L1 channel activity has been directly measured in subependymal cerebrospinal fluid contacting neurons (CSF-cNs), which have an apical protrusion ending with a primary cilium that extends into the central canal ( 86 , 87 ). CSF-cNs receive GABAergic synaptic input, where PKD2L1 channels are hypothesized to receive chemical or mechanical input from the spinal fluid ( 88 ). In teleosts, CSF-cNs fire action potentials in response to alkaline pH changes in the spinal canal, which causes depression of motor activity ( 89 ). However, the physiological function of CSF-cNs in mammals and the role of PKD2L1 in these cells is undetermined. Unlike PKD2, ion currents conducted from homomeric PKD2L1 channels can be measured from the plasma membrane when heterologously expressed. The PKD2L1 conductance is nonselective (γK + = 189 pS, γNa + = 156 pS, γCa 2+ = 53 pS) and outwardly rectifying ( 13 ). However, PKD2L1 channels are more selective for Ca 2+ (P Ca /P Na ≈ 15) than are PKD2 channels, which is conferred by a second aspartate residue (D525) within the selectivity filter ( 55 ). Unlike most members of the TRP channel family, PKD2L1 has many features of voltage-dependent channels, including large tail currents that are activated on membrane repolarization, can be fitted to a Boltzmann equation (V 1/2 = 47 mV), observed Cole-Moore gating properties, and voltage-dependent inactivation (V 1/2 = 15 mV) ( 90 – 92 ). Homomeric PKD2L1 channels are activated by both acidic and basic external pH when heterologously expressed in oocytes and mammalian cells ( 93 , 94 ). This feature is often described as the acid off response, where shifting external pH from low to neutral pH elicits a large transient current ( 91 , 95 ). Like PKD2, PKD2L1 channels are sensitized and desensitized by internal Ca 2+ ( 55 ). These functional characteristics suggest the existence of Ca 2+ -binding sites and protonatable moieties on the protein, which modulate its intrinsic voltage sensitivity and pore inactivation. However, these sites have thus far escaped structural identification in homomeric PKD2L1 channels. . PKD2L2 channels. PKD2L2 channels are the least-characterized polycystin channels in terms of both conductive properties and physiological function. Human PKD2L2 RNA and protein are most highly expressed in the testis and brain ( 10 ). Like PKD2L1 , there are no reports of human disease associated with variants in this gene. Mouse PKD2L2 RNA and protein expression is reported in developing spermatocytes and spermatids, but this channel’s impact on male reproductive fitness is undetermined ( 96 ). Based on the two reports, homomeric PKD2L2 forms functional channels when overexpressed in the plasma membrane of human embryonic kidney (HEK) cells. The apparent PKD2L2 currents are not voltage sensitive and have a lower single channel conductance (γNa ≈ 25 pS) when compared to other PKD2-related channels ( 97 , 98 ). There are no reports of endogenous currents supported by PKD2L2 channel subunits. Thus, PKD2L2’s physiological function as a homomeric or heteromeric channel subunit in vivo remains unestablished.
Homomeric Assembly and Topology
Homotetrameric polycystin channels are the least structurally complex and most-characterized version of their oligomeric assemblies. The PKD2-related subunits have six transmembrane segments (S1–S6) and are members of the TRP family of ion channels, which are part of the larger voltage-gated ion channel (VGIC) superfamily ( 33 ) ( Figure 1a ). TRP channels form cation-permeable ion channels that are widely known for their role as transducers of sensory modalities ( 34 ). PKD2-related subunits form the polycystin TRP subclass, which are also called TRPPs (see sidebar titled Polycystin Nomenclature ). PKD2-related subunits have five structural domains: an intracellular N-terminal domain, a voltage sensor domain (VSD; S1–S4), an extracellular tetragonal opening for polycystins domain (TOP; also called the polycystin domain), a pore domain (PD; S5–S6), and an intracellular C-terminal domain (CTD) ( Figure 1a ). Multiple structures of PKD2 and PKD2L1 have been resolved by cryo-EM, demonstrating that these subunits assemble as homotetrameric channels, where four subunits combine to form an ion conductive unit ( 13 – 15 , 35 ) ( Figure 2 ). Each of the channel subunits form interfaces via their PD and TOP domains, implicating their role in oligomeric assembly. Open in a separate window Figure 2 Assembly and proposed gating mechanism of homomeric polycystin channels. ( a ) Transmembrane view of a single PKD2 subunit with identified structural domains and location of ADPKD-causing missense variants ( 13 , 50 ). ( b ) Transmembrane view of overlaid PKD2 ( gray ; PDB 5T4D) and PKD2L1 ( tan ; PDB 5Z1W) channels, with expanded views of the VSD and PD ( 13 , 15 ). ( c ) Extracellular and transmembrane view of domain-swapped PKD2 channels compared to nondomain-swapped Kv11 Eag1 channel. Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; Kv, member of voltage-gated potassium channel family; PD, pore domain; PDB, protein data bank; TOP, tetragonal opening for polycystins domain; VSD, voltage sensor domain; VSDL, voltage sensor-like domain. SIDEBAR 1: POLYCYSTIN NOMENCLATURE The revised and current International Union of Basic and Clinical Pharmacology/British Pharmacological Society (IUPHAR/BPS) nomenclature creates ambiguity regarding the genetic identity of the polycystin family members of transient receptor potential ion channels (TRPPs), especially when cross-referencing manuscripts that describe subunits using the former system ( 1 ). Traditionally, the products of polycystin genes (e.g., PKD1 ) are referred to as polycystin proteins (e.g., polycystin-1; see Table 1 ). However, this nomenclature is neither practical nor available for all of the genes within this family. For simplicity and to prevent confusion, we refer to the polycystin proteins by their gene name rather than differentiating gene and protein with separate names. PKD2 and PKD2L1 adopt domain-swapped architectures ( Figure 2c ). Domain swapping occurs when the VSD from one subunit packs against the PD helices of the neighboring subunit. Details of domain swapping provide mechanistic insights into how the channel opens in response to voltage. In the structures of PKD2-related channels, the VSD in one subunit is connected to the PD of a neighboring subunit by a 12-Å-long S4–S5 linker helix, which runs parallel to the plane of the membrane’s cytoplasmic surface and forms contacts with S5–S6 helices of the PD, analogous to the bundle strap of a floral bouquet. Here, the state of the VSD (activated or deactivated) is linked to the state of the PD (open or closed) through structural interactions of the S4–S5 linker helix with the pore-forming S6 ( 36 ) ( Figure 2c ). In contrast, the VSD and PD from nondomain-swapped channels (e.g., Eag11 KV channels) pack as contiguous units ( 37 , 38 ). Nondomain-swapped channels do not have a helical S4–S5 linker, and their mechanism of gate control is largely undetermined. . Voltage sensor domain. Additional evidence of the voltage-dependent gating of PKD2-related channels comes from the structures of their VSDs. PKD2 and PKD2L1 contain positively charged residues called gating charges in their S4 ( 39 ) ( Figure 2b , bottom ). Typically, VGICs have at least four positive gating charges, either lysine (K1–K4) or arginine (R1–R4), which are found at every third residue. However, in PKD2-related channels, only the lower two positions (K3 and K4) are conserved ( 40 ). One helical turn of the S4 in PKD2 and PKD2L1 adopts a 3 10 -helix that aligns the K3 and K4 with their side chains pointing into the interior of the VSD ( 13 ) ( Figure 2b ). Here, the gating charges are proposed to be stabilized by state-dependent cation-π (K3-Y366 deactivated state) and charge-counter charge interactions (K3-D390 deactivated state, K4-D390 activated state) formed by side chains of the residues (Y366, K390, D390) extending from the S2 and S3 helices. An ADPKD-associated mutation, D511V (equivalent to PKD2L1 D390V), abolishes PKD2 function in reconstitution assays by presumably disrupting the deactivated state interaction with the internal gating charge ( 41 ). Because these are positively charged residues within a transmembrane environment, the S4 moves to the activated state in response to membrane depolarization and returns to the S4 deactivated state upon repolarization. The measured transfer of PKD2L1 gating charges is equal to 4.4 elementary charges (q s ) per channel, or the equivalent of transferring one lysine per VSD ( 39 ). The S4 is predicted to move at least 6 Å outwardly, or the equivalent of 2 helical turns, during the activation process. The outward movement is coupled to the lateral displacement of the S4–S5 linker and opening of the lower pore gate ( 39 ). At the interface between the VSD and PD, cholesterol and phosphatidylinositol-like densities have been resolved for PKD2 ( 42 – 44 ). Cholesterol and phosphatidylinositol are known modulators of ion channels, and their distribution is unique among organelle membranes ( 45 , 46 ). Thus, further interrogations of lipid-channel interactions in primary cilia and endoplasmic reticulum (ER) membranes will likely elucidate critical mechanisms of channel gating and trafficking within the cell. . TOP domain. The TOP domain sits on the extracellular side of the VSD, extending from the S1 and S2 helices to form extracellular contacts with the S3–S4 loop ( Figures 1a and and2a). 2a ). Although this protein fold is unique to polycystins, it is structurally similar to the extraluminal domain of the lysosomal mucolipin TRP channels ( 47 , 48 ). The TOP domain of PKD2 contains several conserved asparagine-linked glycosylation sites (N299, N305, N328, N362, and N375) that are also conserved in PKD2L1 and may facilitate their trafficking to the primary cilia or other cell membranes ( 49 ) ( Figure 1a ). Importantly, the TOP domain of PKD2 is a hotspot of ADPKD missense variants ( 50 ) ( Figure 2a ). The TOP domains of each PKD2 and PKD2L1 subunit are composed of five beta strands (β1–β5) with two intervening alpha helices. The β1–β5 stretch forms two prominent hairpin turns called fingers 1 and 2, which interface between each channel subunit. Several ADPKD-causing variants (K322Q/W, R325Q/P, and C331S) destabilize finger 1 hydrogen bond, disulfide, and cation-π interactions that disrupt channel function by shifting the voltage dependence of PKD2 opening by more than +80 mV ( 51 ). Mutations at these conserved positions in PKD2L1 cause similar shifts in voltage dependence, suggesting that the function of the TOP domain is shared by the two channels ( 51 ). As the TOP domain rests on the extracellular side of the VSD, one apparent function of this domain is to directly regulate the state of the VSD, which in turn controls the ion conductive state of the PD. This form of allosteric regulation of polycystin channels may be analogous to how peptide toxins from spiders and scorpions block cardiac and neuronal sodium and potassium VGIC function by binding and immobilizing their VSDs from the extracellular side of the membrane ( 52 , 53 ). . Pore domain. The PD forms the central ion permeation pathway and houses the selectivity filter responsible for cation preference ( Figure 2b , top ). The selectivity filters of PKD2 and PKD2L1 are scaffolded by two short alpha helices (PH1, PH2) similar to Na v s and some TRP channels ( 54 ). The negative dipole created by the orientation of the amino acids in the eight pore helices (two helices per subunit) contributes to the electronegativity of the PD. Both PKD2 and PKD2L1 channels conduct multiple cations but differ in their selectivity for Ca 2+ . PKD2L1 is more selective for Ca 2+ than PKD2, which is conferred by a second ring of aspartate residues that is not present in PKD2 ( 55 ). Based on the cryo-EM structure and functional analysis, PKD2-related channels are proposed to have multiple gates—hydrophobic constriction sites that move in and out of the ion conduction pathway in a state-dependent manner ( 13 , 15 ). The lower gate is found on the S6 and is proposed to be mobilized by uncoiling its secondary structure from a π-helix to an α-helix ( 56 ). This gate closely resembles those found in VGICs and involves a lateral displacement or splaying of the S6 through a series of hydrogen bonds formed with residues on the S4–S5 linker. The proposed upper gate is found at sites within the selectivity filter: L521 and D523 in PKD2L1 and L641 and D643 in PKD2 ( 13 , 15 ). However, the proposition of an upper gate is controversial within the TRP channel literature. It is unclear how motions from the VSD would be transferred to the upper gate or whether this site is most akin to the inactivation gate found in VGICs. Collapse of the selectivity filter has been widely described in Na v and K v channels and is functionally linked to inactivation modes. Thus, it is possible that polycystin inactivation, which has been reported for PKD2L1, might be conferred by structural changes in the upper gate ( 57 ). . C-terminal domain. The CTDs of all three PKD2-related channel subunits have an EF-hand motif, followed by a coiled-coil domain and multiple intermittent serine phosphorylation sites. Phosphorylation of these sites enhances PKD2 current density in heterologous expression conditions, indicating this domain’s role in channel membrane localization ( 58 – 61 ). The contiguous cytoplasmic CTDs are not structurally determined in all reported PKD2 and PKD2L1 structures, but the EF-hand and coiled-coil structures have been solved as fragments ( 62 – 65 ). The EF-hand from PKD2 and PKD2L1 coordinates one Ca 2+ ion, with reported affinities approximating 19–22 μM ( K d ) ( 31 , 66 ). However, abolishing Ca 2+ affinity of the EF-hand does not alter Ca 2+ -dependent kinetics of the PKD2 channel current ( 31 ). Furthermore, deletions or neutralizations of this site do not contribute to cystic kidney phenotypes in mice. These findings suggest that additional Ca 2+ -binding sites within PKD2 or Ca 2+ -dependent modifiers are responsible for regulating the channel’s activity. The putative function of the coiled-coil motif in PKD2-related channels is also controversial. The isolated coiled-coil motifs of PKD2 and PKD2L1 channels oligomerize as trimers when expressed as fragments, but this stoichiometry opposes the homotetrameric assembly of whole channel structures ( 67 ). Although removal of the coiled-coil motif has little impact on PKD2L1 trafficking and function in heterologous expression systems, nonsense variants that cleave this motif in PKD2 are associated with ADPKD and are required for its heteromeric assembly with PKD1 and PKD1L1 ( 55 , 64 , 68 – 70 ). Taken together, these findings suggest that the CTDs of PKD2-related channels may function differently depending on their assembly (homomeric or heteromeric) with other polycystin subunits.
Voltage sensor domain.
. Additional evidence of the voltage-dependent gating of PKD2-related channels comes from the structures of their VSDs. PKD2 and PKD2L1 contain positively charged residues called gating charges in their S4 ( 39 ) ( Figure 2b , bottom ). Typically, VGICs have at least four positive gating charges, either lysine (K1–K4) or arginine (R1–R4), which are found at every third residue. However, in PKD2-related channels, only the lower two positions (K3 and K4) are conserved ( 40 ). One helical turn of the S4 in PKD2 and PKD2L1 adopts a 3 10 -helix that aligns the K3 and K4 with their side chains pointing into the interior of the VSD ( 13 ) ( Figure 2b ). Here, the gating charges are proposed to be stabilized by state-dependent cation-π (K3-Y366 deactivated state) and charge-counter charge interactions (K3-D390 deactivated state, K4-D390 activated state) formed by side chains of the residues (Y366, K390, D390) extending from the S2 and S3 helices. An ADPKD-associated mutation, D511V (equivalent to PKD2L1 D390V), abolishes PKD2 function in reconstitution assays by presumably disrupting the deactivated state interaction with the internal gating charge ( 41 ). Because these are positively charged residues within a transmembrane environment, the S4 moves to the activated state in response to membrane depolarization and returns to the S4 deactivated state upon repolarization. The measured transfer of PKD2L1 gating charges is equal to 4.4 elementary charges (q s ) per channel, or the equivalent of transferring one lysine per VSD ( 39 ). The S4 is predicted to move at least 6 Å outwardly, or the equivalent of 2 helical turns, during the activation process. The outward movement is coupled to the lateral displacement of the S4–S5 linker and opening of the lower pore gate ( 39 ). At the interface between the VSD and PD, cholesterol and phosphatidylinositol-like densities have been resolved for PKD2 ( 42 – 44 ). Cholesterol and phosphatidylinositol are known modulators of ion channels, and their distribution is unique among organelle membranes ( 45 , 46 ). Thus, further interrogations of lipid-channel interactions in primary cilia and endoplasmic reticulum (ER) membranes will likely elucidate critical mechanisms of channel gating and trafficking within the cell.
TOP domain.
. The TOP domain sits on the extracellular side of the VSD, extending from the S1 and S2 helices to form extracellular contacts with the S3–S4 loop ( Figures 1a and and2a). 2a ). Although this protein fold is unique to polycystins, it is structurally similar to the extraluminal domain of the lysosomal mucolipin TRP channels ( 47 , 48 ). The TOP domain of PKD2 contains several conserved asparagine-linked glycosylation sites (N299, N305, N328, N362, and N375) that are also conserved in PKD2L1 and may facilitate their trafficking to the primary cilia or other cell membranes ( 49 ) ( Figure 1a ). Importantly, the TOP domain of PKD2 is a hotspot of ADPKD missense variants ( 50 ) ( Figure 2a ). The TOP domains of each PKD2 and PKD2L1 subunit are composed of five beta strands (β1–β5) with two intervening alpha helices. The β1–β5 stretch forms two prominent hairpin turns called fingers 1 and 2, which interface between each channel subunit. Several ADPKD-causing variants (K322Q/W, R325Q/P, and C331S) destabilize finger 1 hydrogen bond, disulfide, and cation-π interactions that disrupt channel function by shifting the voltage dependence of PKD2 opening by more than +80 mV ( 51 ). Mutations at these conserved positions in PKD2L1 cause similar shifts in voltage dependence, suggesting that the function of the TOP domain is shared by the two channels ( 51 ). As the TOP domain rests on the extracellular side of the VSD, one apparent function of this domain is to directly regulate the state of the VSD, which in turn controls the ion conductive state of the PD. This form of allosteric regulation of polycystin channels may be analogous to how peptide toxins from spiders and scorpions block cardiac and neuronal sodium and potassium VGIC function by binding and immobilizing their VSDs from the extracellular side of the membrane ( 52 , 53 ).
Pore domain.
. The PD forms the central ion permeation pathway and houses the selectivity filter responsible for cation preference ( Figure 2b , top ). The selectivity filters of PKD2 and PKD2L1 are scaffolded by two short alpha helices (PH1, PH2) similar to Na v s and some TRP channels ( 54 ). The negative dipole created by the orientation of the amino acids in the eight pore helices (two helices per subunit) contributes to the electronegativity of the PD. Both PKD2 and PKD2L1 channels conduct multiple cations but differ in their selectivity for Ca 2+ . PKD2L1 is more selective for Ca 2+ than PKD2, which is conferred by a second ring of aspartate residues that is not present in PKD2 ( 55 ). Based on the cryo-EM structure and functional analysis, PKD2-related channels are proposed to have multiple gates—hydrophobic constriction sites that move in and out of the ion conduction pathway in a state-dependent manner ( 13 , 15 ). The lower gate is found on the S6 and is proposed to be mobilized by uncoiling its secondary structure from a π-helix to an α-helix ( 56 ). This gate closely resembles those found in VGICs and involves a lateral displacement or splaying of the S6 through a series of hydrogen bonds formed with residues on the S4–S5 linker. The proposed upper gate is found at sites within the selectivity filter: L521 and D523 in PKD2L1 and L641 and D643 in PKD2 ( 13 , 15 ). However, the proposition of an upper gate is controversial within the TRP channel literature. It is unclear how motions from the VSD would be transferred to the upper gate or whether this site is most akin to the inactivation gate found in VGICs. Collapse of the selectivity filter has been widely described in Na v and K v channels and is functionally linked to inactivation modes. Thus, it is possible that polycystin inactivation, which has been reported for PKD2L1, might be conferred by structural changes in the upper gate ( 57 ).
C-terminal domain.
. The CTDs of all three PKD2-related channel subunits have an EF-hand motif, followed by a coiled-coil domain and multiple intermittent serine phosphorylation sites. Phosphorylation of these sites enhances PKD2 current density in heterologous expression conditions, indicating this domain’s role in channel membrane localization ( 58 – 61 ). The contiguous cytoplasmic CTDs are not structurally determined in all reported PKD2 and PKD2L1 structures, but the EF-hand and coiled-coil structures have been solved as fragments ( 62 – 65 ). The EF-hand from PKD2 and PKD2L1 coordinates one Ca 2+ ion, with reported affinities approximating 19–22 μM ( K d ) ( 31 , 66 ). However, abolishing Ca 2+ affinity of the EF-hand does not alter Ca 2+ -dependent kinetics of the PKD2 channel current ( 31 ). Furthermore, deletions or neutralizations of this site do not contribute to cystic kidney phenotypes in mice. These findings suggest that additional Ca 2+ -binding sites within PKD2 or Ca 2+ -dependent modifiers are responsible for regulating the channel’s activity. The putative function of the coiled-coil motif in PKD2-related channels is also controversial. The isolated coiled-coil motifs of PKD2 and PKD2L1 channels oligomerize as trimers when expressed as fragments, but this stoichiometry opposes the homotetrameric assembly of whole channel structures ( 67 ). Although removal of the coiled-coil motif has little impact on PKD2L1 trafficking and function in heterologous expression systems, nonsense variants that cleave this motif in PKD2 are associated with ADPKD and are required for its heteromeric assembly with PKD1 and PKD1L1 ( 55 , 64 , 68 – 70 ). Taken together, these findings suggest that the CTDs of PKD2-related channels may function differently depending on their assembly (homomeric or heteromeric) with other polycystin subunits.
Proposed Cell and Physiological Functions
Homomeric polycystin channels are expressed in a variety of tissues. The biophysical features and evidence supporting each channel’s proposed physiological function are described in the following subsections. . PKD2 channels. PKD2 is the founding member of the polycystin family of TRP channels. Results from RNA sequencing and northern blotting suggest that PKD2 has low tissue specificity and is expressed at similar levels in adults and during early development ( 3 , 71 ). Approximately 15% of ADPKD patients have germline variants of PKD2 and thus the preponderance of knowledge about PKD2 function comes from studies carried out in kidney cells. Functional and immunohistochemical evidence for this channel is reported in the ER and primary cilia of the thick ascending limbs of the loop of Henle, the distal convoluted tubule, and the collecting ducts ( 72 ). In these locations, PKD2 amplifies the cytosolic Ca 2+ release through direct association with the inositol trisphosphate receptor ( 41 , 73 , 74 ). In addition, reduced PKD2 expression attenuates mitochondrial Ca 2+ buffering and increases the fragmentation of the mitochondrial network ( 75 ). Previous reports characterizing reconstituted PKD2 proteins isolated from the ER have produced divergent results regarding ion selectivity and conductance. However, voltage-clamp measurements of endogenous and heterologous PKD2 channels from the primary cilia are more consistent (γK = 139 pS, γNa = 89 pS, γCa = 4 pS), where ion permeation favors monovalent cations (P Ca /P Na = 0.1) ( 19 , 76 ). The Ca 2+ concentration gradient is 10,000 times greater on the outside of the cilium. Although PKD2 has a relatively low Ca 2+ selectivity, its activity would have a major influence on internal Ca 2+ conditions and downstream Ca 2+ -dependent signaling. PKD2 voltage dependence is outwardly rectifying, where single channel opening probability increases at positive potentials. Resting Ca 2+ in primary cilia (390–580 nM) is higher than in the cytosol (40–90 nM) ( 77 ). Elevated intraciliary Ca 2+ modulates PKD2 voltage dependence through processes called Ca 2+ -dependent modulation and subsequent Ca 2+ -dependent desensitization—two biophysical regulatory mechanisms that control its function at physiological membrane potentials ( 31 , 78 ). Extrarenal manifestations of ADPKD implicate PKD2 function in the cardiovascular system ( 79 ). In cardiomyocytes, PKD2 modulates Ca 2+ release from stimulated ryanodine receptors through direct association ( 80 ). Reports of PKD2 function in the vasculature are paradoxical. In arterial smooth muscle cells, PKD2 is proposed to mediate systemic blood pressure and contribute to the myogenic response in cerebral arteries though vasoconstriction ( 81 , 82 ). However, in the vascular endothelium, PKD2 is also proposed to mediate vasodilation through nitric oxide synthase activation and reduce blood pressure in mesenteric arteries ( 83 ). Thus, PKD2 function in smooth muscle cells and endothelial cells might be specific to the location within the vasculature. . PKD2L1 channels. Human PKD2L1 is highly expressed in the spleen, testis, retina, heart, and brain ( 9 , 84 ). The PKD2L1 peptide is 56% identical and 76% similar to PKD2, with the greatest divergence in the cytoplasmic CTD ( 9 ). Unlike PKD2, PKD2L1 has been excluded as a candidate gene for all forms of polycystic kidney disease, with little to no expression in the kidney. Thus far, no variants in PKD2L1 have been associated with human disease. Nonetheless, the phenotypes of murine PKD2L1 knockout models implicate this channel’s function in neurons of the central nervous system, cardiomyocytes, and early development. Ablation of PKD2L1 causes hippocampal and thalamocortical hyperexcitability, resulting in susceptibility to drug-induced seizures ( 85 ). PKD2L1 traffics to the primary cilia of pyramidal neurons and is reported to colocalize with the β2-adrenergic receptor. However, PKD2L1 function in hippocampal neurons has not been directly measured, and the mechanism of ciliary PKD2L1 regulation of neuronal excitability is unknown. PKD2L1 channel activity has been directly measured in subependymal cerebrospinal fluid contacting neurons (CSF-cNs), which have an apical protrusion ending with a primary cilium that extends into the central canal ( 86 , 87 ). CSF-cNs receive GABAergic synaptic input, where PKD2L1 channels are hypothesized to receive chemical or mechanical input from the spinal fluid ( 88 ). In teleosts, CSF-cNs fire action potentials in response to alkaline pH changes in the spinal canal, which causes depression of motor activity ( 89 ). However, the physiological function of CSF-cNs in mammals and the role of PKD2L1 in these cells is undetermined. Unlike PKD2, ion currents conducted from homomeric PKD2L1 channels can be measured from the plasma membrane when heterologously expressed. The PKD2L1 conductance is nonselective (γK + = 189 pS, γNa + = 156 pS, γCa 2+ = 53 pS) and outwardly rectifying ( 13 ). However, PKD2L1 channels are more selective for Ca 2+ (P Ca /P Na ≈ 15) than are PKD2 channels, which is conferred by a second aspartate residue (D525) within the selectivity filter ( 55 ). Unlike most members of the TRP channel family, PKD2L1 has many features of voltage-dependent channels, including large tail currents that are activated on membrane repolarization, can be fitted to a Boltzmann equation (V 1/2 = 47 mV), observed Cole-Moore gating properties, and voltage-dependent inactivation (V 1/2 = 15 mV) ( 90 – 92 ). Homomeric PKD2L1 channels are activated by both acidic and basic external pH when heterologously expressed in oocytes and mammalian cells ( 93 , 94 ). This feature is often described as the acid off response, where shifting external pH from low to neutral pH elicits a large transient current ( 91 , 95 ). Like PKD2, PKD2L1 channels are sensitized and desensitized by internal Ca 2+ ( 55 ). These functional characteristics suggest the existence of Ca 2+ -binding sites and protonatable moieties on the protein, which modulate its intrinsic voltage sensitivity and pore inactivation. However, these sites have thus far escaped structural identification in homomeric PKD2L1 channels. . PKD2L2 channels. PKD2L2 channels are the least-characterized polycystin channels in terms of both conductive properties and physiological function. Human PKD2L2 RNA and protein are most highly expressed in the testis and brain ( 10 ). Like PKD2L1 , there are no reports of human disease associated with variants in this gene. Mouse PKD2L2 RNA and protein expression is reported in developing spermatocytes and spermatids, but this channel’s impact on male reproductive fitness is undetermined ( 96 ). Based on the two reports, homomeric PKD2L2 forms functional channels when overexpressed in the plasma membrane of human embryonic kidney (HEK) cells. The apparent PKD2L2 currents are not voltage sensitive and have a lower single channel conductance (γNa ≈ 25 pS) when compared to other PKD2-related channels ( 97 , 98 ). There are no reports of endogenous currents supported by PKD2L2 channel subunits. Thus, PKD2L2’s physiological function as a homomeric or heteromeric channel subunit in vivo remains unestablished.
PKD2 channels.
. PKD2 is the founding member of the polycystin family of TRP channels. Results from RNA sequencing and northern blotting suggest that PKD2 has low tissue specificity and is expressed at similar levels in adults and during early development ( 3 , 71 ). Approximately 15% of ADPKD patients have germline variants of PKD2 and thus the preponderance of knowledge about PKD2 function comes from studies carried out in kidney cells. Functional and immunohistochemical evidence for this channel is reported in the ER and primary cilia of the thick ascending limbs of the loop of Henle, the distal convoluted tubule, and the collecting ducts ( 72 ). In these locations, PKD2 amplifies the cytosolic Ca 2+ release through direct association with the inositol trisphosphate receptor ( 41 , 73 , 74 ). In addition, reduced PKD2 expression attenuates mitochondrial Ca 2+ buffering and increases the fragmentation of the mitochondrial network ( 75 ). Previous reports characterizing reconstituted PKD2 proteins isolated from the ER have produced divergent results regarding ion selectivity and conductance. However, voltage-clamp measurements of endogenous and heterologous PKD2 channels from the primary cilia are more consistent (γK = 139 pS, γNa = 89 pS, γCa = 4 pS), where ion permeation favors monovalent cations (P Ca /P Na = 0.1) ( 19 , 76 ). The Ca 2+ concentration gradient is 10,000 times greater on the outside of the cilium. Although PKD2 has a relatively low Ca 2+ selectivity, its activity would have a major influence on internal Ca 2+ conditions and downstream Ca 2+ -dependent signaling. PKD2 voltage dependence is outwardly rectifying, where single channel opening probability increases at positive potentials. Resting Ca 2+ in primary cilia (390–580 nM) is higher than in the cytosol (40–90 nM) ( 77 ). Elevated intraciliary Ca 2+ modulates PKD2 voltage dependence through processes called Ca 2+ -dependent modulation and subsequent Ca 2+ -dependent desensitization—two biophysical regulatory mechanisms that control its function at physiological membrane potentials ( 31 , 78 ). Extrarenal manifestations of ADPKD implicate PKD2 function in the cardiovascular system ( 79 ). In cardiomyocytes, PKD2 modulates Ca 2+ release from stimulated ryanodine receptors through direct association ( 80 ). Reports of PKD2 function in the vasculature are paradoxical. In arterial smooth muscle cells, PKD2 is proposed to mediate systemic blood pressure and contribute to the myogenic response in cerebral arteries though vasoconstriction ( 81 , 82 ). However, in the vascular endothelium, PKD2 is also proposed to mediate vasodilation through nitric oxide synthase activation and reduce blood pressure in mesenteric arteries ( 83 ). Thus, PKD2 function in smooth muscle cells and endothelial cells might be specific to the location within the vasculature.
PKD2L1 channels.
. Human PKD2L1 is highly expressed in the spleen, testis, retina, heart, and brain ( 9 , 84 ). The PKD2L1 peptide is 56% identical and 76% similar to PKD2, with the greatest divergence in the cytoplasmic CTD ( 9 ). Unlike PKD2, PKD2L1 has been excluded as a candidate gene for all forms of polycystic kidney disease, with little to no expression in the kidney. Thus far, no variants in PKD2L1 have been associated with human disease. Nonetheless, the phenotypes of murine PKD2L1 knockout models implicate this channel’s function in neurons of the central nervous system, cardiomyocytes, and early development. Ablation of PKD2L1 causes hippocampal and thalamocortical hyperexcitability, resulting in susceptibility to drug-induced seizures ( 85 ). PKD2L1 traffics to the primary cilia of pyramidal neurons and is reported to colocalize with the β2-adrenergic receptor. However, PKD2L1 function in hippocampal neurons has not been directly measured, and the mechanism of ciliary PKD2L1 regulation of neuronal excitability is unknown. PKD2L1 channel activity has been directly measured in subependymal cerebrospinal fluid contacting neurons (CSF-cNs), which have an apical protrusion ending with a primary cilium that extends into the central canal ( 86 , 87 ). CSF-cNs receive GABAergic synaptic input, where PKD2L1 channels are hypothesized to receive chemical or mechanical input from the spinal fluid ( 88 ). In teleosts, CSF-cNs fire action potentials in response to alkaline pH changes in the spinal canal, which causes depression of motor activity ( 89 ). However, the physiological function of CSF-cNs in mammals and the role of PKD2L1 in these cells is undetermined. Unlike PKD2, ion currents conducted from homomeric PKD2L1 channels can be measured from the plasma membrane when heterologously expressed. The PKD2L1 conductance is nonselective (γK + = 189 pS, γNa + = 156 pS, γCa 2+ = 53 pS) and outwardly rectifying ( 13 ). However, PKD2L1 channels are more selective for Ca 2+ (P Ca /P Na ≈ 15) than are PKD2 channels, which is conferred by a second aspartate residue (D525) within the selectivity filter ( 55 ). Unlike most members of the TRP channel family, PKD2L1 has many features of voltage-dependent channels, including large tail currents that are activated on membrane repolarization, can be fitted to a Boltzmann equation (V 1/2 = 47 mV), observed Cole-Moore gating properties, and voltage-dependent inactivation (V 1/2 = 15 mV) ( 90 – 92 ). Homomeric PKD2L1 channels are activated by both acidic and basic external pH when heterologously expressed in oocytes and mammalian cells ( 93 , 94 ). This feature is often described as the acid off response, where shifting external pH from low to neutral pH elicits a large transient current ( 91 , 95 ). Like PKD2, PKD2L1 channels are sensitized and desensitized by internal Ca 2+ ( 55 ). These functional characteristics suggest the existence of Ca 2+ -binding sites and protonatable moieties on the protein, which modulate its intrinsic voltage sensitivity and pore inactivation. However, these sites have thus far escaped structural identification in homomeric
PKD2L2 channels.
. PKD2L2 channels are the least-characterized polycystin channels in terms of both conductive properties and physiological function. Human PKD2L2 RNA and protein are most highly expressed in the testis and brain ( 10 ). Like PKD2L1 , there are no reports of human disease associated with variants in this gene. Mouse PKD2L2 RNA and protein expression is reported in developing spermatocytes and spermatids, but this channel’s impact on male reproductive fitness is undetermined ( 96 ). Based on the two reports, homomeric PKD2L2 forms functional channels when overexpressed in the plasma membrane of human embryonic kidney (HEK) cells. The apparent PKD2L2 currents are not voltage sensitive and have a lower single channel conductance (γNa ≈ 25 pS) when compared to other PKD2-related channels ( 97 , 98 ). There are no reports of endogenous currents supported by PKD2L2 channel subunits. Thus, PKD2L2’s physiological function as a homomeric or heteromeric channel subunit in vivo remains unestablished.
HETEROMERIC POLYCYSTIN CHANNELS
PKD1-related polycystins have unique structural regulatory domains that produce channels with distinct properties when they oligomerize with PKD2-related subunits. In the following subsections, we describe the structural features indicating the gating mechanisms that govern state-dependent ion permeation. Heteromeric Assembly and Topology The majority of mutations that cause ADPKD impact PKD1 subunits ( Figure 3a ). While PKD1-related proteins do not form ion channels on their own, they can form heterotetrameric complexes with PKD2-related subunits in various tissues with a 1:3 stoichiometry, respectively. Recent cryo-EM structural determination of the PKD1-PKD2 and PKD1L3-PKD2L1 complexes by Yigong Shi’s lab have provided a structural basis for understanding their molecular regulation ( 16 , 17 ) ( Figure 3b ). PKD1-related subunits (PKD1, PKD1L1, PKD1L2, PKD1L3, and PKDREJ) are large proteins (208–516 kDa) with 11 helical transmembrane segments (S1–S11). The PKD1 transmembrane core can be separated into two entities: the N-terminal transmembrane domain (NTMD) containing S1–S5 and the C-terminal transmembrane domain (CTMD) containing S6–S11 ( Figures 1b and and3a). 3a ). More than half of each PKD1-related protein resides in the extracellular ectodomain, which is theorized to serve as a mechanical or ligand regulatory domain. Thus far, all structures of heteromeric polycystins have been determined using truncated forms of the PKD1-related subunits—where PKD1 is missing its ectodomain and cytoplasmic C terminus while PKD1L3 is missing its ectodomain along with its entire NTMD ( 16 , 17 ). The NTMD has homology to aGPCRs, but there is considerable divergence when aligning the PKD1 NTMD with transmembrane segments of the reported structures of GPR96 (group 1), GPR110 (group V) and GPR 133 (group VI). ( 99 ). Activation of aGPCRs is proposed to involve the stalk region that proceeds the first transmembrane helix that acts like a tethered agonist by interacting with an eternal binding cavity located on the surface transmembrane region. This mechanism might also be conserved by heteromeric polycystin channel complexes assembled with PKD1-related proteins. Within the NTMD, the polycystin-1 lipoxygenase and alpha toxin domain, implicated in lipid binding and PKD1 membrane trafficking, extends from the intracellular S1–S2 loop ( 100 ). The polycystin-1 lipoxygenase and alpha toxin domain of PKD1 contains several ADPKD missense variants, indicating its importance in channel regulation, but the mechanistic function of this motif is undefined ( Figure 3a ). The CTMD (S5–S11) of PKD1-related proteins packs with similar symmetry to PKD2-related subunits within the heteromeric structure ( 13 – 19 ). The TOP domains of PKD1-related subunits are highly homologous and adopt a similar secondary structure to the TOP domains found in PKD2-related proteins. Here, the TOP domains from PKD1 and PKD1L3 units extend from the S6–S7 loop, forming homotypic contacts with PKD2 and PKD2L1, respectively. The heteromeric TOP domain interactions are not symmetric, leaving a gap at the interface between fingers 1 and 2 and the PKD2-related subunits. As reported for PKD2, ADPKD-causing missense variants aggregate in the PKD1 TOP domain ( 50 ). Open in a separate window Figure 3 Assembly and proposed gating mechanisms of heteromeric polycystin channels. ( a ) Transmembrane view of a single PKD1 subunit with identified structural domains and location of ADPKD-causing missense variants ( 16 ). ( b ) Extracellular and transmembrane views of structural overlaid PKD1-PKD2 (PDB 6A70) and PKD1L3-PKD2L1 (PDB 7D7E) heteromeric channels, with an expanded view of the PKD1-S11 transmembrane segment’s positively charged residues lining the ion conducting pathway ( 16 ). ( c ) Transmembrane view of the PKD1L3-PKD2L1 channel with an expanded view of the pore Ca 2+ -binding site captured in the apo (PDB 7D7E) and Ca 2+ -occupied (PDB 7D7F) states ( 143 ). ( d ) Transmembrane view of the third PKD2L1 subunit (PKD2L1 DIII VSD) of the Ca 2+ -occupied PKD1L3-PKD2L1 channel complex. Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; CTMD, C-terminal transmembrane domain; NTMD, N-terminal transmembrane domain; PDB, protein data bank; PLAT, polycystin-1 lipoxygenase and alpha toxin (domain); TOP, tetragonal opening for polycystins (domain); VSD, voltage sensor domain. Ectodomain. The ectodomains of PKD1-related proteins contain one or multiple motifs proposed to form extracellular protein or cell matrix interactions ( 101 , 102 ) ( Figure 1b ). Among these subunits, PKD1 has the most elaborate ectodomain, with 15 PKD repeat motifs, two complete leucine-rich repeat motifs flanked by cysteine-rich sequences, and a C-type lectin domain ( 2 ). The ectodomains of four of the five PKD1-related proteins contain a ~700 amino acid receptor for egg jelly (REJ) domain—a motif homologous to the REJ protein found in the membrane of the sea urchin sperm acrosome. The REJ domain is required for cleavage at the ~40-residue GPCR proteolytic site (GPS), which represents an integral part of a larger ~320-residue domain called the GPCR autoproteolysis-inducing domain. The domain’s fragment of aGPCRs (CL1 and BAI3) has been reported ( 103 ). Analogously to latrophilin aGPCRs, the N-terminal ectodomain of PKD1-related channels is autoproteolytically cleaved at the GPS, which is proximal to the first transmembrane helix. Cleavage at the GPS has been implicated in liberating nascent PKD1 protein from the ER, but the N-terminal fragment remains noncovalently tethered ( 104 , 105 ). Although mutations within the GPS of PKD1 are associated with ADPKD, the full functional consequences of PKD1 autoproteolysis are poorly understood ( 106 , 107 ). Pore domain and proposed gating mechanisms. As discussed previously, the PDs of homomeric PKD2-related channels form symmetric and canonical TRP channel selectivity filters. A key difference between the PD (S10, S11) of heteromeric (PKD1-PKD2 and PKD1L3-PKD2L1) and homomeric (PKD2, PKD2L1) polycystin channels is the loss of symmetry of the ion-conductive pore caused by the contribution of the S10 and S11 segments of PKD1-related subunits ( Figure 3b ). The S11 segment of PKD1-related proteins in these structures is broken in the middle, with the extracellular half occupying the space normally held by helix PH1 in homomeric PKD2 and PKD2L1 structures. Because the PKD1 pore loop lacks the aspartate residues found in PKD2, the cation selectivity for these heteromeric channels is expected to be distinct from homomeric PKD2-related channels. Along the pore lining S11 of PKD1, three positively charged residues are present (R4100, R4107, and H4111) and plug the ion-conducting pathway ( Figure 3b ). It appears that lateral displacement of PKD1 S11, possibly coupled to conformational changes triggered by distant parts of the complex, may gate the PKD1-PKD2 heteromeric channel. In this scenario, the displaced positively charged residues may facilitate ion conduction by bringing water into the pore vestibule, where partially dehydrated ions moving through the selectivity filter can be rehydrated upon entry. Clearly, structural determination of the PKD1-PKD2 complex in the open state—and perhaps in association with an activating ligand—is needed to elucidate its gating mechanism. The structure of the PKD1L3-PKD2L1 complex was determined in the apo (closed) and Ca 2+ -bound (open) states at 3.1-Å and 3.4-Å resolution, respectively ( 17 ) ( Figure 3b , ,c c , ,d). d ). Comparison of the open and closed states identified two unconventional Ca 2+ -binding sites that gate the channel complex. The first Ca 2+ -binding site is found at an asymmetric position within the selectivity filter. Here, the apo PKD1L3-PKD2L1 complex is blocked by K2069 from PKD1L3, which apparently acts as a plug in the absence of Ca 2+ . However, under high Ca 2+ conditions, the ion occupies the selectivity filter and is coordinated by the D523 side chain of PKD2L1 and main chain carbonyls of both PKD2L1 and PKD1L3, while K2069 of PKD1L3 is displaced ( Figure 3c ). The second is in the extracellular cleft of the VSD within the third PKD2L1 subunit (VSD-III), but the remaining VSDs are unoccupied, suggesting that this site within the heteromeric context is unique ( Figure 3d ). Indeed, mutagenesis and functional work carried out in Xenopus oocytes indicates that Ca 2+ occupancy of VSD-III is responsible for Ca 2+ -dependent activation ( 17 ). Here, the VSD-III S4–S5 linker acts directly on the PD S11 of PKD1L3, indicating a mechanism of allosteric regulation where Ca 2+ binding to the VSD-III can control the opening of the lower gate. Taken together, these structures identify key conformational changes that control PKD1L3-PKD2L1 gating by Ca 2+ . However, this channel’s putative structural regulation by external by pH is unclear, as discussed in the subsection titled PKD1L3-PKD2L1 Channels. Proposed Cell and Physiological Functions Heteromeric polycystin channels assemble in divergent and sometimes overlapping tissues, with unique distributions within the cell. The biophysical features and evidence supporting each channel’s proposed physiological function are described in the following subsections. . PKD1-PKD2 channels. The PKD1-PKD2 channel complex is the most extensively studied type of heteromeric polycystin channel. The preponderance of research methodologies is aimed at understanding its function within renal tubule cells, where cystogenesis occurs in ADPKD. Several lines of genetic evidence underscore the importance of the heteromeric PKD1-PKD2 complex in ADPKD progression. First, the majority of germline ADPKD-causing variants impact the PKD1 gene, which results in more rapidly progressing forms of the disease. Second, the genotype of cyst cells is often transheterozygous, having acquired somatic mutations impacting the other polycystin allele ( PKD1 or PKD2 ) ( 108 ). Despite the intensity of this area of research, many of the basic features of this clinically relevant channel complex remain poorly understood. The physiological localization of the PKD1-PKD2 channel is debated, but most evidence suggests that this channel functions in either the primary cilia or plasma membrane ( Figure 4a ). Contrary to initial reports ( 109 ), heterologous coexpression of PKD1 and PKD2 does not produce constitutively active channels in the plasma membrane ( 19 , 110 ). Interdependence between PKD1 and PKD2 in plasma membrane and ciliary trafficking is widely reported, suggesting that heteromeric PKD1-PKD2 assembly at the primary cilia organelle is a disease-relevant channel population ( 111 , 112 ). Yet, other reports have demonstrated that PKD2 traffics to the cilia without expression of PKD1 ( 19 , 113 ). Although it is undisputed that PKD1 and PKD2 biochemically associate to form structures with channel-like features, endogenous regulators of PKD1-PKD2 gating remain undetermined. However, two recent studies have yielded important clues about its putative function as a channel. Open in a separate window Figure 4 Subcellular localization of polycystin complexes in various human tissues. ( a ) Heteromeric PKD1-PKD2 and homomeric PKD2 channels within the primary cilia, endoplasmic reticulum, and apical plasma membranes of the collecting duct epithelium of the kidney nephron. ( b ) Heteromeric PKD1L1-PKD2 location within the (nonmotile) primary cilia of crown cells within the embryonic node. ( c ) PKDREJ-PKD2 heteromeric channels localize to the acrosomal crest of sperm. Coexpression of PKD1 with the PKD2 gain-of-function mutant F604P produces distinct ion selectivity ( 114 ), with greater Ca 2+ permeability compared to the PKD2 mutant alone. The F604P mutation, which is not associated with ADPKD, causes S6 to bend, which in turn opens the ion conducting pathway of the PKD1-PKD2 F604P complex. The resulting current has no rectification or voltage dependence. In a separate study, the normally low plasma membrane localization of PKD1 was amplified by replacing its native N-terminal signal peptide with a strong Ig κ-chain secretion sequence (sPKD1) ( 110 ). In agreement with previous results, heterologously expressed PKD1-PKD2 channels had no constitutive activity in the plasma membrane, but ion currents could be measured from cells coexpressing sPKD1 and the PKD2 F604P gain-of-function mutant. Consistent with previous reports, the ≈89-pS conductance measured from the primary cilia of collecting duct cells was dependent on the expression of PKD2 and not PKD1 ( 19 , 110 ). However, the channel opening probability at negative membrane potentials was enhanced when PKD1 was ablated, suggesting that PKD1 might be a negative regulator of the polycystin channel conductance. Furthermore, the C-type lectin domain from the PKD1 ectodomain could be used as a soluble activator of PKD1-PKD2 F604P channels ( 110 ). From this work, two plausible features of the native PKD1-PKD2 channel emerge. First, the heteromeric PKD1-PKD2 channel is probably more selective for Ca 2+ than homomeric PKD2 channels, and both polycystin channel types can function in the cilia membrane. Second, PKD1-PKD2 channels are constitutively closed until opened by an unknown regulator, such as the cleaved lectin motif or stalk region of the PKD1 ectodomain ( Figure 4a ). The native and dysregulated function of PKD1 presents a large gap in our understanding of ADPKD, as variants in this gene most frequently affect (≈80%) the ADPKD patient population. While it is likely that ciliary signaling involving several pathways can initiate kidney cystogenesis, future work should be directed at determining whether other membrane populations are physiologically relevant. . PKD1L1-PKD2 channels. The PKD1L1-PKD2 complex is proposed to form a Ca 2+ channel whose activation is required in establishing right-left organ asymmetry in the interior of the human body ( Figure 4b ). Right-left asymmetry is established by Ca 2+ -dependent asymmetric gene expression within cells on one side of the embryonic node—a concave structure located at the midline of the postgastrulated embryo ( 115 ). Two types of cilia on the surface of different node cells are proposed to have essential functions in establishing this asymmetry ( 116 , 117 ). First, pit cells containing solitary motile cilia generate laminar flow directed toward the left side of the node. Then, a second cell type called crown cells, which express primary cilia (immotile) and are located on the edge of the node, receive an undetermined signal or stimulus produced by the fluid flow. Because Ca 2+ transients in crown cells are dependent on the expression of PKD2 and PKD1L1, and both channel subunits localize to the immotile cilia, PKD1L1-PKD2 heteromeric channels are proposed to function as sensors of the nodal flow ( 70 , 118 , 119 ). There are two debated mechanisms by which movement of fluid across the node is translated to gene expression. The mechanosensory hypothesis proposes that pit cell fluid flow bends the membrane of crown cell primary cilia, which activates PKD1L1-PKD2 channels ( 120 , 121 ). Alternatively, the morphogen hypothesis posits that crown cell cilia Ca 2+ transients are initiated by an extracellular gradient of morphogenic molecules such as the Nodal secretory protein. Both models have been challenged on theoretical grounds, arguing that the magnitude of fluid force generated by the motile cilia of the pit cells is insufficient to redistribute small molecule morphogens and/or bend the immotile cilia of the crown cells ( 122 , 123 ). However, larger nodal vesicular parcels, which encapsulate Hedgehog and retinoic acid, are indeed asymmetrically transported by the flow across the node, but their activation of polycystin signaling is undetermined ( 124 ). Nonetheless, genetic data suggest that nodal cilia are essential for breaking symmetry in the mouse embryo. This conclusion is reinforced by the observation that all known variants producing situs inversus in humans also implicate a ciliary mechanism for breaking symmetry ( 125 ). Mutations in PKD1L1 and PKD2 are associated with laterality defects in mice and humans, consistent with the proposed PKD1L1-PKD2 channel function in development. In humans, several deletions and missense bi-allelic mutations in PKD1L1 are associated with laterality defects ( 32 ). Among them, the C1691S variant alters the essential disulfide bond of the GPS domain and results in the incorrect development of internal organ asymmetry and arrangement ( 32 ). In PKD1L1 homozygous knockout mice, approximately one-third showed situs inversus and reduced postnatal viability associated with cardiac and vascular abnormalities ( 126 ). The D411G PKD1L1 point mutation in mice that fails to activate the Nodal signaling cascade results in laterality defects and lethality 15.5 days postconception ( 70 ). Homozygous PKD2 −/− mice are not viable, and one-third of embryos have laterality defects of the heart, stomach, and lungs ( 127 ), whereas less than 10% of PKD2 +/− mice exhibit laterality defects. Interestingly, PKD2L1 −/− mice are viable but approximately half develope situs inversus gut defects, which opens the possibility to this polycystin subunit being involved in organ asymmetry determination ( 77 ). It should be noted that conductance related to PKD1L1-PKD2 or PKD2L1 has not been directly measured in the embryonic node crown cell cilia; thus, the genetic origin of the putative conductance and its gating features are undetermined. Future work on the role of nodal vesicular parcels (or other stimuli) in activation of ciliary conductances will certainly settle this fundamental and fascinating question in biology, as well as elucidate the molecular mechanism that regulates polycystin gating in the node. . PKDREJ-PKD2 and PKDREJ-PKD2L1 channels. Heteromeric polycystin channels are proposed to have a role in the fertilization step of sexual reproduction. Sperm from echinoderms, mice, and humans must undergo capacitation to successfully fuse to the ovum ( Figure 4c ). Capacitation is carried out in two steps. First is the triggering of the acrosome reaction—a destabilization of the acrosome compartment within the sperm head which is required for penetration of the ovum membrane. Second, sperm enter a hyperactivated state where enhanced Ca 2+ permeability and cAMP production in the principal piece enhance their mobility. In human sperm, disordered acrosome reaction and hyperactivation are features associated with impaired sperm-ovum fusion in vitro and with reduced fertility in vivo ( 128 ). The sea urchin orthologs suREJ and suPKD2 are proposed to form a heteromeric channel that conducts Ca 2+ in response to glycoprotein(s) found on the egg jelly coat. In human reproduction, zona pellucida glycoprotein 3 sperm-binding protein (ZP3) produced by the ovum is proposed to interact with a specific receptor on the sperm that initiates sperm capacitation. Human PKDREJ RNA and protein expression is restricted to the spermatogenic lineage and is retained in mature sperm ( 11 , 97 ). Immunoprecipitation experiments indicate that PKDREJ associates with PKD2 and PKD2L1 (but not with PKD2L2) when overexpressed in HEK cells ( 97 ). As discussed previously, both PKD2 and PKD2L1 are expressed in testis, suggesting that their heterologous interaction with PKDREJ might be physiologically relevant. In mice, PKDREJ localizes to the acrosomal crescent ( 129 ). It is important to note that none of the putative heteromeric suREJ-suPKD2, PKDREJ-PKD2, and PKDREJ-PKD2L1 channels have been functionally characterized, either in heterologous expression systems or as endogenous ion channel currents. Genetic studies in mice suggest there are divergent roles in sperm capacitation and requirements of the egg glycoprotein-polycystin interaction compared to echinoderms ( 130 ). Sperm from male mice with deleted PKDREJ alleles are able to capacitate in vitro and are fertile in unrestricted mating trials. These results indicate interactions between the PKDREJ and ZP3 glycoprotein are not required for the acrosome reaction or fertility in mice. However, the acrosome reaction of PKDREJ -null sperm develops more slowly and takes three times longer to reach the ovum when compared to wild-type mice. In contrast, hyperactivated flagellar motility develops on a normal time course ( 130 ). These observations indicate that capacitation, acrosome competence, and altered motility are differentially regulated in murine sperm and PKDREJ contributes to the reproductive fitness of male sperm during sexual reproduction. . PKD1L1-PKD2L1 channels. Heteromeric PKD1L1-PKD2L1 channels regulate ciliary Ca 2+ levels and are reported to regulate downstream Hedgehog-dependent transcription of glioma-associated oncogene homolog 1 ( 77 ). While canonical Hedgehog signaling is implicated in embryonic neural tube formation and the growth of malignant tumors ( 131 , 132 ), the functionality of PKD1L1-PKD2L1 channels in either cancer or the development of organizing centers like the embryonic node is undetermined. The activity of endogenous PKD1L1-PKD2L1 channels can be directly measured by voltage clamping the primary cilia membrane of embryonic fibroblasts and retinal pigmented epithelial cells; thus the channel properties are well characterized ( 18 ). PKD1L1-PKD2L1 are enriched (29 channels per μm 2 ) in the primary cilia membrane, which rivals the voltage-gated channel densities observed in neurons. Open probability of PKD1L1-PKD2L1 is enhanced under high membrane pressure (>60 mm Hg), but these channels lack the sensitivity observed in canonical mechanosensitive channels ( 133 , 134 ). Therefore, the proposed impact of fluid flow or ciliary bending as a physiological stimulus to gate this type of polycystin channel is unclear. Heteromeric PKD1L1-PKD2L1 channels can also be observed in the plasma membrane when heterologously overexpressed. Like all polycystin heteromers, PKD1L1 and PKD2L1 subunits assemble as channels with a 1:3 stoichiometry. Contributions of the PKD1L1 subunit to the functional pore can be captured by comparing the ion-conductive properties of the homomeric and heteromeric channels ( 13 , 18 , 55 ). PKD1L1-PKD2L1 channels have smaller single channel conductance (γNa = 96 pS) than homomeric PKD2L1 channels (γNa = 156 pS). PKD1L1-PKD2L1 channels have moderate ion selectivity that favors Ca 2+ (P Ca /P Na ≈ 6) over monovalent cations, which is low in comparison to PKD2L1 channels (P Ca /P Na ≈ 15). The pore of heteromeric and homomeric channels may have unique structural features, which would explain these functional differences, such as reduced electronegativity, different ion selectivity filter interactions, and a more restricted ion conducting pathway. However, a high-resolution structure of the PKD1L1-PKD2L1 complex to elucidate the physical interactions responsible for the differences in ion permeability has yet to be determined. It is postulated that these channels may contribute to elevated resting Ca 2+ or enable a much higher dynamic range of Ca 2+ inside the primary cilia compartment. Because low levels of PKD1L1 protein and RNA are ubiquitous and only partially overlap with PKD2L1 expression, the physiological role of the PKD1L1-PKD2L1 channel in cilia Ca 2+ regulation remains outstanding ( 8 , 9 ). . PKD1L3-PKD2L1 channels. The PKD1L3-PKD2L1 heteromer was initially proposed to function as a pH receptor in type III gustatory cells, but its role in sour taste reception is controversial ( 84 , 135 – 137 ). Type III cells are one of the three cell types within the taste buds of the foliate and circumvallate papillae of the tongue. Acid-evoked Ca 2+ responses and optogenetic stimulation of PKD2L1-expressing type III cells confirm the role of this cell type in adverse taste reception ( 138 , 139 ); however, PKD1L3 and PKD2L1 are only coexpressed in ≈20% of type III cells ( 136 , 140 ). Initially, mice lacking PKD2L1 expression were shown to be completely devoid of type III cell acid responses in electrophysiology recordings to sour stimuli in vivo ( 84 ). In the circumvallate type III cells, the PKD1L3-PKD2L1 complex was proposed to form the molecular sensor responsible for acid sensing ( 139 ). However, subsequent studies ablating the PKD1L3 gene in mice demonstrate normal taste responsiveness in behavioral and electrophysiological tests when compared with wild-type controls ( 135 ), thus refuting the PKD1L3-PKD2L1 channel as the molecular receptor for acidic taste. Recent results suggest that the Otop1 proton channel, which has no structural similarity to polycystins, appears to function as the sour taste receptor in type III cells ( 141 , 142 ). Besides taste buds, PKD1L3 is highly expressed in the liver and testis, but its function as an ion channel subunit (or otherwise) is undetermined. When PKD1L3 and PKD2L1 subunits are heterologously coexpressed in oocytes and HEK cells, currents produced by putative PKD1L3-PKD2L1 channels are activated by external pH changes (acidic and basic) and Ca 2+ ( 139 , 143 ). Although the Ca 2+ -binding sites and their proposed regulation of the channel have been structurally defined, its structural regulation by external pH is undetermined. Regulation of PKD1L3-PKD2L1 channels by external Ca 2+ is bimodal, first sensitizing and subsequently inactivating the current ( 143 ). PKD1L3-PKD2L1 produces a nonselective conductance with preference for Ca 2+ (P Ca /P Na ≈ 11), and the pH or Ca 2+ -activated current has no voltage dependence ( 137 , 144 ). It is important to note that Ca 2+ and pH regulatory features reported for heteromeric PKD1L3-PKD2L1 channels overlap with those described earlier for homomeric PKD2L1 channels. Thus, it is unclear which channel is being measured in these expression systems when both genes are being coexpressed. Furthermore, it was recently reported that a PKD1L3 splice variant (PKD1L3–1a, L708del, S709del) expressed in the type III taste cells does not undergo GPS cleavage and remains sequestered intracellularly ( 145 ). While PKD1L3 and PKD2L1 undoubtedly can form complexes, their functional features in the context of physiological relevance, either at the plasma membrane or as a molecular chaperone, need further investigation.
Heteromeric Assembly and Topology
The majority of mutations that cause ADPKD impact PKD1 subunits ( Figure 3a ). While PKD1-related proteins do not form ion channels on their own, they can form heterotetrameric complexes with PKD2-related subunits in various tissues with a 1:3 stoichiometry, respectively. Recent cryo-EM structural determination of the PKD1-PKD2 and PKD1L3-PKD2L1 complexes by Yigong Shi’s lab have provided a structural basis for understanding their molecular regulation ( 16 , 17 ) ( Figure 3b ). PKD1-related subunits (PKD1, PKD1L1, PKD1L2, PKD1L3, and PKDREJ) are large proteins (208–516 kDa) with 11 helical transmembrane segments (S1–S11). The PKD1 transmembrane core can be separated into two entities: the N-terminal transmembrane domain (NTMD) containing S1–S5 and the C-terminal transmembrane domain (CTMD) containing S6–S11 ( Figures 1b and and3a). 3a ). More than half of each PKD1-related protein resides in the extracellular ectodomain, which is theorized to serve as a mechanical or ligand regulatory domain. Thus far, all structures of heteromeric polycystins have been determined using truncated forms of the PKD1-related subunits—where PKD1 is missing its ectodomain and cytoplasmic C terminus while PKD1L3 is missing its ectodomain along with its entire NTMD ( 16 , 17 ). The NTMD has homology to aGPCRs, but there is considerable divergence when aligning the PKD1 NTMD with transmembrane segments of the reported structures of GPR96 (group 1), GPR110 (group V) and GPR 133 (group VI). ( 99 ). Activation of aGPCRs is proposed to involve the stalk region that proceeds the first transmembrane helix that acts like a tethered agonist by interacting with an eternal binding cavity located on the surface transmembrane region. This mechanism might also be conserved by heteromeric polycystin channel complexes assembled with PKD1-related proteins. Within the NTMD, the polycystin-1 lipoxygenase and alpha toxin domain, implicated in lipid binding and PKD1 membrane trafficking, extends from the intracellular S1–S2 loop ( 100 ). The polycystin-1 lipoxygenase and alpha toxin domain of PKD1 contains several ADPKD missense variants, indicating its importance in channel regulation, but the mechanistic function of this motif is undefined ( Figure 3a ). The CTMD (S5–S11) of PKD1-related proteins packs with similar symmetry to PKD2-related subunits within the heteromeric structure ( 13 – 19 ). The TOP domains of PKD1-related subunits are highly homologous and adopt a similar secondary structure to the TOP domains found in PKD2-related proteins. Here, the TOP domains from PKD1 and PKD1L3 units extend from the S6–S7 loop, forming homotypic contacts with PKD2 and PKD2L1, respectively. The heteromeric TOP domain interactions are not symmetric, leaving a gap at the interface between fingers 1 and 2 and the PKD2-related subunits. As reported for PKD2, ADPKD-causing missense variants aggregate in the PKD1 TOP domain ( 50 ). Open in a separate window Figure 3 Assembly and proposed gating mechanisms of heteromeric polycystin channels. ( a ) Transmembrane view of a single PKD1 subunit with identified structural domains and location of ADPKD-causing missense variants ( 16 ). ( b ) Extracellular and transmembrane views of structural overlaid PKD1-PKD2 (PDB 6A70) and PKD1L3-PKD2L1 (PDB 7D7E) heteromeric channels, with an expanded view of the PKD1-S11 transmembrane segment’s positively charged residues lining the ion conducting pathway ( 16 ). ( c ) Transmembrane view of the PKD1L3-PKD2L1 channel with an expanded view of the pore Ca 2+ -binding site captured in the apo (PDB 7D7E) and Ca 2+ -occupied (PDB 7D7F) states ( 143 ). ( d ) Transmembrane view of the third PKD2L1 subunit (PKD2L1 DIII VSD) of the Ca 2+ -occupied PKD1L3-PKD2L1 channel complex. Abbreviations: ADPKD, autosomal dominant polycystic kidney disease; CTMD, C-terminal transmembrane domain; NTMD, N-terminal transmembrane domain; PDB, protein data bank; PLAT, polycystin-1 lipoxygenase and alpha toxin (domain); TOP, tetragonal opening for polycystins (domain); VSD, voltage sensor domain.
Ectodomain.
The ectodomains of PKD1-related proteins contain one or multiple motifs proposed to form extracellular protein or cell matrix interactions ( 101 , 102 ) ( Figure 1b ). Among these subunits, PKD1 has the most elaborate ectodomain, with 15 PKD repeat motifs, two complete leucine-rich repeat motifs flanked by cysteine-rich sequences, and a C-type lectin domain ( 2 ). The ectodomains of four of the five PKD1-related proteins contain a ~700 amino acid receptor for egg jelly (REJ) domain—a motif homologous to the REJ protein found in the membrane of the sea urchin sperm acrosome. The REJ domain is required for cleavage at the ~40-residue GPCR proteolytic site (GPS), which represents an integral part of a larger ~320-residue domain called the GPCR autoproteolysis-inducing domain. The domain’s fragment of aGPCRs (CL1 and BAI3) has been reported ( 103 ). Analogously to latrophilin aGPCRs, the N-terminal ectodomain of PKD1-related channels is autoproteolytically cleaved at the GPS, which is proximal to the first transmembrane helix. Cleavage at the GPS has been implicated in liberating nascent PKD1 protein from the ER, but the N-terminal fragment remains noncovalently tethered ( 104 , 105 ). Although mutations within the GPS of PKD1 are associated with ADPKD, the full functional consequences of PKD1 autoproteolysis are poorly understood ( 106 , 107 ).
Pore domain and proposed gating mechanisms.
As discussed previously, the PDs of homomeric PKD2-related channels form symmetric and canonical TRP channel selectivity filters. A key difference between the PD (S10, S11) of heteromeric (PKD1-PKD2 and PKD1L3-PKD2L1) and homomeric (PKD2, PKD2L1) polycystin channels is the loss of symmetry of the ion-conductive pore caused by the contribution of the S10 and S11 segments of PKD1-related subunits ( Figure 3b ). The S11 segment of PKD1-related proteins in these structures is broken in the middle, with the extracellular half occupying the space normally held by helix PH1 in homomeric PKD2 and PKD2L1 structures. Because the PKD1 pore loop lacks the aspartate residues found in PKD2, the cation selectivity for these heteromeric channels is expected to be distinct from homomeric PKD2-related channels. Along the pore lining S11 of PKD1, three positively charged residues are present (R4100, R4107, and H4111) and plug the ion-conducting pathway ( Figure 3b ). It appears that lateral displacement of PKD1 S11, possibly coupled to conformational changes triggered by distant parts of the complex, may gate the PKD1-PKD2 heteromeric channel. In this scenario, the displaced positively charged residues may facilitate ion conduction by bringing water into the pore vestibule, where partially dehydrated ions moving through the selectivity filter can be rehydrated upon entry. Clearly, structural determination of the PKD1-PKD2 complex in the open state—and perhaps in association with an activating ligand—is needed to elucidate its gating mechanism. The structure of the PKD1L3-PKD2L1 complex was determined in the apo (closed) and Ca 2+ -bound (open) states at 3.1-Å and 3.4-Å resolution, respectively ( 17 ) ( Figure 3b , ,c c , ,d). d ). Comparison of the open and closed states identified two unconventional Ca 2+ -binding sites that gate the channel complex. The first Ca 2+ -binding site is found at an asymmetric position within the selectivity filter. Here, the apo PKD1L3-PKD2L1 complex is blocked by K2069 from PKD1L3, which apparently acts as a plug in the absence of Ca 2+ . However, under high Ca 2+ conditions, the ion occupies the selectivity filter and is coordinated by the D523 side chain of PKD2L1 and main chain carbonyls of both PKD2L1 and PKD1L3, while K2069 of PKD1L3 is displaced ( Figure 3c ). The second is in the extracellular cleft of the VSD within the third PKD2L1 subunit (VSD-III), but the remaining VSDs are unoccupied, suggesting that this site within the heteromeric context is unique ( Figure 3d ). Indeed, mutagenesis and functional work carried out in Xenopus oocytes indicates that Ca 2+ occupancy of VSD-III is responsible for Ca 2+ -dependent activation ( 17 ). Here, the VSD-III S4–S5 linker acts directly on the PD S11 of PKD1L3, indicating a mechanism of allosteric regulation where Ca 2+ binding to the VSD-III can control the opening of the lower gate. Taken together, these structures identify key conformational changes that control PKD1L3-PKD2L1 gating by Ca 2+ . However, this channel’s putative structural regulation by external by pH is unclear, as discussed in the subsection titled PKD1L3-PKD2L1 Channels.
Proposed Cell and Physiological Functions
Heteromeric polycystin channels assemble in divergent and sometimes overlapping tissues, with unique distributions within the cell. The biophysical features and evidence supporting each channel’s proposed physiological function are described in the following subsections. . PKD1-PKD2 channels. The PKD1-PKD2 channel complex is the most extensively studied type of heteromeric polycystin channel. The preponderance of research methodologies is aimed at understanding its function within renal tubule cells, where cystogenesis occurs in ADPKD. Several lines of genetic evidence underscore the importance of the heteromeric PKD1-PKD2 complex in ADPKD progression. First, the majority of germline ADPKD-causing variants impact the PKD1 gene, which results in more rapidly progressing forms of the disease. Second, the genotype of cyst cells is often transheterozygous, having acquired somatic mutations impacting the other polycystin allele ( PKD1 or PKD2 ) ( 108 ). Despite the intensity of this area of research, many of the basic features of this clinically relevant channel complex remain poorly understood. The physiological localization of the PKD1-PKD2 channel is debated, but most evidence suggests that this channel functions in either the primary cilia or plasma membrane ( Figure 4a ). Contrary to initial reports ( 109 ), heterologous coexpression of PKD1 and PKD2 does not produce constitutively active channels in the plasma membrane ( 19 , 110 ). Interdependence between PKD1 and PKD2 in plasma membrane and ciliary trafficking is widely reported, suggesting that heteromeric PKD1-PKD2 assembly at the primary cilia organelle is a disease-relevant channel population ( 111 , 112 ). Yet, other reports have demonstrated that PKD2 traffics to the cilia without expression of PKD1 ( 19 , 113 ). Although it is undisputed that PKD1 and PKD2 biochemically associate to form structures with channel-like features, endogenous regulators of PKD1-PKD2 gating remain undetermined. However, two recent studies have yielded important clues about its putative function as a channel. Open in a separate window Figure 4 Subcellular localization of polycystin complexes in various human tissues. ( a ) Heteromeric PKD1-PKD2 and homomeric PKD2 channels within the primary cilia, endoplasmic reticulum, and apical plasma membranes of the collecting duct epithelium of the kidney nephron. ( b ) Heteromeric PKD1L1-PKD2 location within the (nonmotile) primary cilia of crown cells within the embryonic node. ( c ) PKDREJ-PKD2 heteromeric channels localize to the acrosomal crest of sperm. Coexpression of PKD1 with the PKD2 gain-of-function mutant F604P produces distinct ion selectivity ( 114 ), with greater Ca 2+ permeability compared to the PKD2 mutant alone. The F604P mutation, which is not associated with ADPKD, causes S6 to bend, which in turn opens the ion conducting pathway of the PKD1-PKD2 F604P complex. The resulting current has no rectification or voltage dependence. In a separate study, the normally low plasma membrane localization of PKD1 was amplified by replacing its native N-terminal signal peptide with a strong Ig κ-chain secretion sequence (sPKD1) ( 110 ). In agreement with previous results, heterologously expressed PKD1-PKD2 channels had no constitutive activity in the plasma membrane, but ion currents could be measured from cells coexpressing sPKD1 and the PKD2 F604P gain-of-function mutant. Consistent with previous reports, the ≈89-pS conductance measured from the primary cilia of collecting duct cells was dependent on the expression of PKD2 and not PKD1 ( 19 , 110 ). However, the channel opening probability at negative membrane potentials was enhanced when PKD1 was ablated, suggesting that PKD1 might be a negative regulator of the polycystin channel conductance. Furthermore, the C-type lectin domain from the PKD1 ectodomain could be used as a soluble activator of PKD1-PKD2 F604P channels ( 110 ). From this work, two plausible features of the native PKD1-PKD2 channel emerge. First, the heteromeric PKD1-PKD2 channel is probably more selective for Ca 2+ than homomeric PKD2 channels, and both polycystin channel types can function in the cilia membrane. Second, PKD1-PKD2 channels are constitutively closed until opened by an unknown regulator, such as the cleaved lectin motif or stalk region of the PKD1 ectodomain ( Figure 4a ). The native and dysregulated function of PKD1 presents a large gap in our understanding of ADPKD, as variants in this gene most frequently affect (≈80%) the ADPKD patient population. While it is likely that ciliary signaling involving several pathways can initiate kidney cystogenesis, future work should be directed at determining whether other membrane populations are physiologically relevant. . PKD1L1-PKD2 channels. The PKD1L1-PKD2 complex is proposed to form a Ca 2+ channel whose activation is required in establishing right-left organ asymmetry in the interior of the human body ( Figure 4b ). Right-left asymmetry is established by Ca 2+ -dependent asymmetric gene expression within cells on one side of the embryonic node—a concave structure located at the midline of the postgastrulated embryo ( 115 ). Two types of cilia on the surface of different node cells are proposed to have essential functions in establishing this asymmetry ( 116 , 117 ). First, pit cells containing solitary motile cilia generate laminar flow directed toward the left side of the node. Then, a second cell type called crown cells, which express primary cilia (immotile) and are located on the edge of the node, receive an undetermined signal or stimulus produced by the fluid flow. Because Ca 2+ transients in crown cells are dependent on the expression of PKD2 and PKD1L1, and both channel subunits localize to the immotile cilia, PKD1L1-PKD2 heteromeric channels are proposed to function as sensors of the nodal flow ( 70 , 118 , 119 ). There are two debated mechanisms by which movement of fluid across the node is translated to gene expression. The mechanosensory hypothesis proposes that pit cell fluid flow bends the membrane of crown cell primary cilia, which activates PKD1L1-PKD2 channels ( 120 , 121 ). Alternatively, the morphogen hypothesis posits that crown cell cilia Ca 2+ transients are initiated by an extracellular gradient of morphogenic molecules such as the Nodal secretory protein. Both models have been challenged on theoretical grounds, arguing that the magnitude of fluid force generated by the motile cilia of the pit cells is insufficient to redistribute small molecule morphogens and/or bend the immotile cilia of the crown cells ( 122 , 123 ). However, larger nodal vesicular parcels, which encapsulate Hedgehog and retinoic acid, are indeed asymmetrically transported by the flow across the node, but their activation of polycystin signaling is undetermined ( 124 ). Nonetheless, genetic data suggest that nodal cilia are essential for breaking symmetry in the mouse embryo. This conclusion is reinforced by the observation that all known variants producing situs inversus in humans also implicate a ciliary mechanism for breaking symmetry ( 125 ). Mutations in PKD1L1 and PKD2 are associated with laterality defects in mice and humans, consistent with the proposed PKD1L1-PKD2 channel function in development. In humans, several deletions and missense bi-allelic mutations in PKD1L1 are associated with laterality defects ( 32 ). Among them, the C1691S variant alters the essential disulfide bond of the GPS domain and results in the incorrect development of internal organ asymmetry and arrangement ( 32 ). In PKD1L1 homozygous knockout mice, approximately one-third showed situs inversus and reduced postnatal viability associated with cardiac and vascular abnormalities ( 126 ). The D411G PKD1L1 point mutation in mice that fails to activate the Nodal signaling cascade results in laterality defects and lethality 15.5 days postconception ( 70 ). Homozygous PKD2 −/− mice are not viable, and one-third of embryos have laterality defects of the heart, stomach, and lungs ( 127 ), whereas less than 10% of PKD2 +/− mice exhibit laterality defects. Interestingly, PKD2L1 −/− mice are viable but approximately half develope situs inversus gut defects, which opens the possibility to this polycystin subunit being involved in organ asymmetry determination ( 77 ). It should be noted that conductance related to PKD1L1-PKD2 or PKD2L1 has not been directly measured in the embryonic node crown cell cilia; thus, the genetic origin of the putative conductance and its gating features are undetermined. Future work on the role of nodal vesicular parcels (or other stimuli) in activation of ciliary conductances will certainly settle this fundamental and fascinating question in biology, as well as elucidate the molecular mechanism that regulates polycystin gating in the node. . PKDREJ-PKD2 and PKDREJ-PKD2L1 channels. Heteromeric polycystin channels are proposed to have a role in the fertilization step of sexual reproduction. Sperm from echinoderms, mice, and humans must undergo capacitation to successfully fuse to the ovum ( Figure 4c ). Capacitation is carried out in two steps. First is the triggering of the acrosome reaction—a destabilization of the acrosome compartment within the sperm head which is required for penetration of the ovum membrane. Second, sperm enter a hyperactivated state where enhanced Ca 2+ permeability and cAMP production in the principal piece enhance their mobility. In human sperm, disordered acrosome reaction and hyperactivation are features associated with impaired sperm-ovum fusion in vitro and with reduced fertility in vivo ( 128 ). The sea urchin orthologs suREJ and suPKD2 are proposed to form a heteromeric channel that conducts Ca 2+ in response to glycoprotein(s) found on the egg jelly coat. In human reproduction, zona pellucida glycoprotein 3 sperm-binding protein (ZP3) produced by the ovum is proposed to interact with a specific receptor on the sperm that initiates sperm capacitation. Human PKDREJ RNA and protein expression is restricted to the spermatogenic lineage and is retained in mature sperm ( 11 , 97 ). Immunoprecipitation experiments indicate that PKDREJ associates with PKD2 and PKD2L1 (but not with PKD2L2) when overexpressed in HEK cells ( 97 ). As discussed previously, both PKD2 and PKD2L1 are expressed in testis, suggesting that their heterologous interaction with PKDREJ might be physiologically relevant. In mice, PKDREJ localizes to the acrosomal crescent ( 129 ). It is important to note that none of the putative heteromeric suREJ-suPKD2, PKDREJ-PKD2, and PKDREJ-PKD2L1 channels have been functionally characterized, either in heterologous expression systems or as endogenous ion channel currents. Genetic studies in mice suggest there are divergent roles in sperm capacitation and requirements of the egg glycoprotein-polycystin interaction compared to echinoderms ( 130 ). Sperm from male mice with deleted PKDREJ alleles are able to capacitate in vitro and are fertile in unrestricted mating trials. These results indicate interactions between the PKDREJ and ZP3 glycoprotein are not required for the acrosome reaction or fertility in mice. However, the acrosome reaction of PKDREJ -null sperm develops more slowly and takes three times longer to reach the ovum when compared to wild-type mice. In contrast, hyperactivated flagellar motility develops on a normal time course ( 130 ). These observations indicate that capacitation, acrosome competence, and altered motility are differentially regulated in murine sperm and PKDREJ contributes to the reproductive fitness of male sperm during sexual reproduction. . PKD1L1-PKD2L1 channels. Heteromeric PKD1L1-PKD2L1 channels regulate ciliary Ca 2+ levels and are reported to regulate downstream Hedgehog-dependent transcription of glioma-associated oncogene homolog 1 ( 77 ). While canonical Hedgehog signaling is implicated in embryonic neural tube formation and the growth of malignant tumors ( 131 , 132 ), the functionality of PKD1L1-PKD2L1 channels in either cancer or the development of organizing centers like the embryonic node is undetermined. The activity of endogenous PKD1L1-PKD2L1 channels can be directly measured by voltage clamping the primary cilia membrane of embryonic fibroblasts and retinal pigmented epithelial cells; thus the channel properties are well characterized ( 18 ). PKD1L1-PKD2L1 are enriched (29 channels per μm 2 ) in the primary cilia membrane, which rivals the voltage-gated channel densities observed in neurons. Open probability of PKD1L1-PKD2L1 is enhanced under high membrane pressure (>60 mm Hg), but these channels lack the sensitivity observed in canonical mechanosensitive channels ( 133 , 134 ). Therefore, the proposed impact of fluid flow or ciliary bending as a physiological stimulus to gate this type of polycystin channel is unclear. Heteromeric PKD1L1-PKD2L1 channels can also be observed in the plasma membrane when heterologously overexpressed. Like all polycystin heteromers, PKD1L1 and PKD2L1 subunits assemble as channels with a 1:3 stoichiometry. Contributions of the PKD1L1 subunit to the functional pore can be captured by comparing the ion-conductive properties of the homomeric and heteromeric channels ( 13 , 18 , 55 ). PKD1L1-PKD2L1 channels have smaller single channel conductance (γNa = 96 pS) than homomeric PKD2L1 channels (γNa = 156 pS). PKD1L1-PKD2L1 channels have moderate ion selectivity that favors Ca 2+ (P Ca /P Na ≈ 6) over monovalent cations, which is low in comparison to PKD2L1 channels (P Ca /P Na ≈ 15). The pore of heteromeric and homomeric channels may have unique structural features, which would explain these functional differences, such as reduced electronegativity, different ion selectivity filter interactions, and a more restricted ion conducting pathway. However, a high-resolution structure of the PKD1L1-PKD2L1 complex to elucidate the physical interactions responsible for the differences in ion permeability has yet to be determined. It is postulated that these channels may contribute to elevated resting Ca 2+ or enable a much higher dynamic range of Ca 2+ inside the primary cilia compartment. Because low levels of PKD1L1 protein and RNA are ubiquitous and only partially overlap with PKD2L1 expression, the physiological role of the PKD1L1-PKD2L1 channel in cilia Ca 2+ regulation remains outstanding ( 8 , 9 ). . PKD1L3-PKD2L1 channels. The PKD1L3-PKD2L1 heteromer was initially proposed to function as a pH receptor in type III gustatory cells, but its role in sour taste reception is controversial ( 84 , 135 – 137 ). Type III cells are one of the three cell types within the taste buds of the foliate and circumvallate papillae of the tongue. Acid-evoked Ca 2+ responses and optogenetic stimulation of PKD2L1-expressing type III cells confirm the role of this cell type in adverse taste reception ( 138 , 139 ); however, PKD1L3 and PKD2L1 are only coexpressed in ≈20% of type III cells ( 136 , 140 ). Initially, mice lacking PKD2L1 expression were shown to be completely devoid of type III cell acid responses in electrophysiology recordings to sour stimuli in vivo ( 84 ). In the circumvallate type III cells, the PKD1L3-PKD2L1 complex was proposed to form the molecular sensor responsible for acid sensing ( 139 ). However, subsequent studies ablating the PKD1L3 gene in mice demonstrate normal taste responsiveness in behavioral and electrophysiological tests when compared with wild-type controls ( 135 ), thus refuting the PKD1L3-PKD2L1 channel as the molecular receptor for acidic taste. Recent results suggest that the Otop1 proton channel, which has no structural similarity to polycystins, appears to function as the sour taste receptor in type III cells ( 141 , 142 ). Besides taste buds, PKD1L3 is highly expressed in the liver and testis, but its function as an ion channel subunit (or otherwise) is undetermined. When PKD1L3 and PKD2L1 subunits are heterologously coexpressed in oocytes and HEK cells, currents produced by putative PKD1L3-PKD2L1 channels are activated by external pH changes (acidic and basic) and Ca 2+ ( 139 , 143 ). Although the Ca 2+ -binding sites and their proposed regulation of the channel have been structurally defined, its structural regulation by external pH is undetermined. Regulation of PKD1L3-PKD2L1 channels by external Ca 2+ is bimodal, first sensitizing and subsequently inactivating the current ( 143 ). PKD1L3-PKD2L1 produces a nonselective conductance with preference for Ca 2+ (P Ca /P Na ≈ 11), and the pH or Ca 2+ -activated current has no voltage dependence ( 137 , 144 ). It is important to note that Ca 2+ and pH regulatory features reported for heteromeric PKD1L3-PKD2L1 channels overlap with those described earlier for homomeric PKD2L1 channels. Thus, it is unclear which channel is being measured in these expression systems when both genes are being coexpressed. Furthermore, it was recently reported that a PKD1L3 splice variant (PKD1L3–1a, L708del, S709del) expressed in the type III taste cells does not undergo GPS cleavage and remains sequestered intracellularly ( 145 ). While PKD1L3 and PKD2L1 undoubtedly can form complexes, their functional features in the context of physiological relevance, either at the plasma membrane or as a molecular chaperone, need further investigation.
PKD1-PKD2 channels.
. The PKD1-PKD2 channel complex is the most extensively studied type of heteromeric polycystin channel. The preponderance of research methodologies is aimed at understanding its function within renal tubule cells, where cystogenesis occurs in ADPKD. Several lines of genetic evidence underscore the importance of the heteromeric PKD1-PKD2 complex in ADPKD progression. First, the majority of germline ADPKD-causing variants impact the PKD1 gene, which results in more rapidly progressing forms of the disease. Second, the genotype of cyst cells is often transheterozygous, having acquired somatic mutations impacting the other polycystin allele ( PKD1 or PKD2 ) ( 108 ). Despite the intensity of this area of research, many of the basic features of this clinically relevant channel complex remain poorly understood. The physiological localization of the PKD1-PKD2 channel is debated, but most evidence suggests that this channel functions in either the primary cilia or plasma membrane ( Figure 4a ). Contrary to initial reports ( 109 ), heterologous coexpression of PKD1 and PKD2 does not produce constitutively active channels in the plasma membrane ( 19 , 110 ). Interdependence between PKD1 and PKD2 in plasma membrane and ciliary trafficking is widely reported, suggesting that heteromeric PKD1-PKD2 assembly at the primary cilia organelle is a disease-relevant channel population ( 111 , 112 ). Yet, other reports have demonstrated that PKD2 traffics to the cilia without expression of PKD1 ( 19 , 113 ). Although it is undisputed that PKD1 and PKD2 biochemically associate to form structures with channel-like features, endogenous regulators of PKD1-PKD2 gating remain undetermined. However, two recent studies have yielded important clues about its putative function as a channel. Open in a separate window Figure 4 Subcellular localization of polycystin complexes in various human tissues. ( a ) Heteromeric PKD1-PKD2 and homomeric PKD2 channels within the primary cilia, endoplasmic reticulum, and apical plasma membranes of the collecting duct epithelium of the kidney nephron. ( b ) Heteromeric PKD1L1-PKD2 location within the (nonmotile) primary cilia of crown cells within the embryonic node. ( c ) PKDREJ-PKD2 heteromeric channels localize to the acrosomal crest of sperm. Coexpression of PKD1 with the PKD2 gain-of-function mutant F604P produces distinct ion selectivity ( 114 ), with greater Ca 2+ permeability compared to the PKD2 mutant alone. The F604P mutation, which is not associated with ADPKD, causes S6 to bend, which in turn opens the ion conducting pathway of the PKD1-PKD2 F604P complex. The resulting current has no rectification or voltage dependence. In a separate study, the normally low plasma membrane localization of PKD1 was amplified by replacing its native N-terminal signal peptide with a strong Ig κ-chain secretion sequence (sPKD1) ( 110 ). In agreement with previous results, heterologously expressed PKD1-PKD2 channels had no constitutive activity in the plasma membrane, but ion currents could be measured from cells coexpressing sPKD1 and the PKD2 F604P gain-of-function mutant. Consistent with previous reports, the ≈89-pS conductance measured from the primary cilia of collecting duct cells was dependent on the expression of PKD2 and not PKD1 ( 19 , 110 ). However, the channel opening probability at negative membrane potentials was enhanced when PKD1 was ablated, suggesting that PKD1 might be a negative regulator of the polycystin channel conductance. Furthermore, the C-type lectin domain from the PKD1 ectodomain could be used as a soluble activator of PKD1-PKD2 F604P channels ( 110 ). From this work, two plausible features of the native PKD1-PKD2 channel emerge. First, the heteromeric PKD1-PKD2 channel is probably more selective for Ca 2+ than homomeric PKD2 channels, and both polycystin channel types can function in the cilia membrane. Second, PKD1-PKD2 channels are constitutively closed until opened by an unknown regulator, such as the cleaved lectin motif or stalk region of the PKD1 ectodomain ( Figure 4a ). The native and dysregulated function of PKD1 presents a large gap in our understanding of ADPKD, as variants in this gene most frequently affect (≈80%) the ADPKD patient population. While it is likely that ciliary signaling involving several pathways can initiate kidney cystogenesis, future work should be directed at determining whether other membrane populations are physiologically relevant.
PKD1L1-PKD2 channels.
. The PKD1L1-PKD2 complex is proposed to form a Ca 2+ channel whose activation is required in establishing right-left organ asymmetry in the interior of the human body ( Figure 4b ). Right-left asymmetry is established by Ca 2+ -dependent asymmetric gene expression within cells on one side of the embryonic node—a concave structure located at the midline of the postgastrulated embryo ( 115 ). Two types of cilia on the surface of different node cells are proposed to have essential functions in establishing this asymmetry ( 116 , 117 ). First, pit cells containing solitary motile cilia generate laminar flow directed toward the left side of the node. Then, a second cell type called crown cells, which express primary cilia (immotile) and are located on the edge of the node, receive an undetermined signal or stimulus produced by the fluid flow. Because Ca 2+ transients in crown cells are dependent on the expression of PKD2 and PKD1L1, and both channel subunits localize to the immotile cilia, PKD1L1-PKD2 heteromeric channels are proposed to function as sensors of the nodal flow ( 70 , 118 , 119 ). There are two debated mechanisms by which movement of fluid across the node is translated to gene expression. The mechanosensory hypothesis proposes that pit cell fluid flow bends the membrane of crown cell primary cilia, which activates PKD1L1-PKD2 channels ( 120 , 121 ). Alternatively, the morphogen hypothesis posits that crown cell cilia Ca 2+ transients are initiated by an extracellular gradient of morphogenic molecules such as the Nodal secretory protein. Both models have been challenged on theoretical grounds, arguing that the magnitude of fluid force generated by the motile cilia of the pit cells is insufficient to redistribute small molecule morphogens and/or bend the immotile cilia of the crown cells ( 122 , 123 ). However, larger nodal vesicular parcels, which encapsulate Hedgehog and retinoic acid, are indeed asymmetrically transported by the flow across the node, but their activation of polycystin signaling is undetermined ( 124 ). Nonetheless, genetic data suggest that nodal cilia are essential for breaking symmetry in the mouse embryo. This conclusion is reinforced by the observation that all known variants producing situs inversus in humans also implicate a ciliary mechanism for breaking symmetry ( 125 ). Mutations in PKD1L1 and PKD2 are associated with laterality defects in mice and humans, consistent with the proposed PKD1L1-PKD2 channel function in development. In humans, several deletions and missense bi-allelic mutations in PKD1L1 are associated with laterality defects ( 32 ). Among them, the C1691S variant alters the essential disulfide bond of the GPS domain and results in the incorrect development of internal organ asymmetry and arrangement ( 32 ). In PKD1L1 homozygous knockout mice, approximately one-third showed situs inversus and reduced postnatal viability associated with cardiac and vascular abnormalities ( 126 ). The D411G PKD1L1 point mutation in mice that fails to activate the Nodal signaling cascade results in laterality defects and lethality 15.5 days postconception ( 70 ). Homozygous PKD2 −/− mice are not viable, and one-third of embryos have laterality defects of the heart, stomach, and lungs ( 127 ), whereas less than 10% of PKD2 +/− mice exhibit laterality defects. Interestingly, PKD2L1 −/− mice are viable but approximately half develope situs inversus gut defects, which opens the possibility to this polycystin subunit being involved in organ asymmetry determination ( 77 ). It should be noted that conductance related to PKD1L1-PKD2 or PKD2L1 has not been directly measured in the embryonic node crown cell cilia; thus, the genetic origin of the putative conductance and its gating features are undetermined. Future work on the role of nodal vesicular parcels (or other stimuli) in activation of ciliary conductances will certainly settle this fundamental and fascinating question in biology, as well as elucidate the molecular mechanism that regulates polycystin gating in the node.
PKDREJ-PKD2 and PKDREJ-PKD2L1 channels.
. Heteromeric polycystin channels are proposed to have a role in the fertilization step of sexual reproduction. Sperm from echinoderms, mice, and humans must undergo capacitation to successfully fuse to the ovum ( Figure 4c ). Capacitation is carried out in two steps. First is the triggering of the acrosome reaction—a destabilization of the acrosome compartment within the sperm head which is required for penetration of the ovum membrane. Second, sperm enter a hyperactivated state where enhanced Ca 2+ permeability and cAMP production in the principal piece enhance their mobility. In human sperm, disordered acrosome reaction and hyperactivation are features associated with impaired sperm-ovum fusion in vitro and with reduced fertility in vivo ( 128 ). The sea urchin orthologs suREJ and suPKD2 are proposed to form a heteromeric channel that conducts Ca 2+ in response to glycoprotein(s) found on the egg jelly coat. In human reproduction, zona pellucida glycoprotein 3 sperm-binding protein (ZP3) produced by the ovum is proposed to interact with a specific receptor on the sperm that initiates sperm capacitation. Human PKDREJ RNA and protein expression is restricted to the spermatogenic lineage and is retained in mature sperm ( 11 , 97 ). Immunoprecipitation experiments indicate that PKDREJ associates with PKD2 and PKD2L1 (but not with PKD2L2) when overexpressed in HEK cells ( 97 ). As discussed previously, both PKD2 and PKD2L1 are expressed in testis, suggesting that their heterologous interaction with PKDREJ might be physiologically relevant. In mice, PKDREJ localizes to the acrosomal crescent ( 129 ). It is important to note that none of the putative heteromeric suREJ-suPKD2, PKDREJ-PKD2, and PKDREJ-PKD2L1 channels have been functionally characterized, either in heterologous expression systems or as endogenous ion channel currents. Genetic studies in mice suggest there are divergent roles in sperm capacitation and requirements of the egg glycoprotein-polycystin interaction compared to echinoderms ( 130 ). Sperm from male mice with deleted PKDREJ alleles are able to capacitate in vitro and are fertile in unrestricted mating trials. These results indicate interactions between the PKDREJ and ZP3 glycoprotein are not required for the acrosome reaction or fertility in mice. However, the acrosome reaction of PKDREJ -null sperm develops more slowly and takes three times longer to reach the ovum when compared to wild-type mice. In contrast, hyperactivated flagellar motility develops on a normal time course ( 130 ). These observations indicate that capacitation, acrosome competence, and altered motility are differentially regulated in murine sperm and PKDREJ contributes to the reproductive fitness of male sperm during sexual reproduction.
PKD1L1-PKD2L1 channels.
. Heteromeric PKD1L1-PKD2L1 channels regulate ciliary Ca 2+ levels and are reported to regulate downstream Hedgehog-dependent transcription of glioma-associated oncogene homolog 1 ( 77 ). While canonical Hedgehog signaling is implicated in embryonic neural tube formation and the growth of malignant tumors ( 131 , 132 ), the functionality of PKD1L1-PKD2L1 channels in either cancer or the development of organizing centers like the embryonic node is undetermined. The activity of endogenous PKD1L1-PKD2L1 channels can be directly measured by voltage clamping the primary cilia membrane of embryonic fibroblasts and retinal pigmented epithelial cells; thus the channel properties are well characterized ( 18 ). PKD1L1-PKD2L1 are enriched (29 channels per μm 2 ) in the primary cilia membrane, which rivals the voltage-gated channel densities observed in neurons. Open probability of PKD1L1-PKD2L1 is enhanced under high membrane pressure (>60 mm Hg), but these channels lack the sensitivity observed in canonical mechanosensitive channels ( 133 , 134 ). Therefore, the proposed impact of fluid flow or ciliary bending as a physiological stimulus to gate this type of polycystin channel is unclear. Heteromeric PKD1L1-PKD2L1 channels can also be observed in the plasma membrane when heterologously overexpressed. Like all polycystin heteromers, PKD1L1 and PKD2L1 subunits assemble as channels with a 1:3 stoichiometry. Contributions of the PKD1L1 subunit to the functional pore can be captured by comparing the ion-conductive properties of the homomeric and heteromeric channels ( 13 , 18 , 55 ). PKD1L1-PKD2L1 channels have smaller single channel conductance (γNa = 96 pS) than homomeric PKD2L1 channels (γNa = 156 pS). PKD1L1-PKD2L1 channels have moderate ion selectivity that favors Ca 2+ (P Ca /P Na ≈ 6) over monovalent cations, which is low in comparison to PKD2L1 channels (P Ca /P Na ≈ 15). The pore of heteromeric and homomeric channels may have unique structural features, which would explain these functional differences, such as reduced electronegativity, different ion selectivity filter interactions, and a more restricted ion conducting pathway. However, a high-resolution structure of the PKD1L1-PKD2L1 complex to elucidate the physical interactions responsible for the differences in ion permeability has yet to be determined. It is postulated that these channels may contribute to elevated resting Ca 2+ or enable a much higher dynamic range of Ca 2+ inside the primary cilia compartment. Because low levels of PKD1L1 protein and RNA are ubiquitous and only partially overlap with PKD2L1 expression, the physiological role of the PKD1L1-PKD2L1 channel in cilia Ca 2+ regulation remains outstanding ( 8 , 9 ).
PKD1L3-PKD2L1 channels.
. The PKD1L3-PKD2L1 heteromer was initially proposed to function as a pH receptor in type III gustatory cells, but its role in sour taste reception is controversial ( 84 , 135 – 137 ). Type III cells are one of the three cell types within the taste buds of the foliate and circumvallate papillae of the tongue. Acid-evoked Ca 2+ responses and optogenetic stimulation of PKD2L1-expressing type III cells confirm the role of this cell type in adverse taste reception ( 138 , 139 ); however, PKD1L3 and PKD2L1 are only coexpressed in ≈20% of type III cells ( 136 , 140 ). Initially, mice lacking PKD2L1 expression were shown to be completely devoid of type III cell acid responses in electrophysiology recordings to sour stimuli in vivo ( 84 ). In the circumvallate type III cells, the PKD1L3-PKD2L1 complex was proposed to form the molecular sensor responsible for acid sensing ( 139 ). However, subsequent studies ablating the PKD1L3 gene in mice demonstrate normal taste responsiveness in behavioral and electrophysiological tests when compared with wild-type controls ( 135 ), thus refuting the PKD1L3-PKD2L1 channel as the molecular receptor for acidic taste. Recent results suggest that the Otop1 proton channel, which has no structural similarity to polycystins, appears to function as the sour taste receptor in type III cells ( 141 , 142 ). Besides taste buds, PKD1L3 is highly expressed in the liver and testis, but its function as an ion channel subunit (or otherwise) is undetermined. When PKD1L3 and PKD2L1 subunits are heterologously coexpressed in oocytes and HEK cells, currents produced by putative PKD1L3-PKD2L1 channels are activated by external pH changes (acidic and basic) and Ca 2+ ( 139 , 143 ). Although the Ca 2+ -binding sites and their proposed regulation of the channel have been structurally defined, its structural regulation by external pH is undetermined. Regulation of PKD1L3-PKD2L1 channels by external Ca 2+ is bimodal, first sensitizing and subsequently inactivating the current ( 143 ). PKD1L3-PKD2L1 produces a nonselective conductance with preference for Ca 2+ (P Ca /P Na ≈ 11), and the pH or Ca 2+ -activated current has no voltage dependence ( 137 , 144 ). It is important to note that Ca 2+ and pH regulatory features reported for heteromeric PKD1L3-PKD2L1 channels overlap with those described earlier for homomeric PKD2L1 channels. Thus, it is unclear which channel is being measured in these expression systems when both genes are being coexpressed. Furthermore, it was recently reported that a PKD1L3 splice variant (PKD1L3–1a, L708del, S709del) expressed in the type III taste cells does not undergo GPS cleavage and remains sequestered intracellularly ( 145 ). While PKD1L3 and PKD2L1 undoubtedly can form complexes, their functional features in the context of physiological relevance, either at the plasma membrane or as a molecular chaperone, need further investigation.
SUMMARY
We have highlighted recent structural insights in combination with functional characterizations that have elucidated the molecular regulation and biological roles of polycystin complexes. Their unique hetero- and homomeric assemblies host a repertoire of gating mechanisms that respond to diverse stimuli within and outside of the cell. Expression of polycystin complexes is tissue-specific, and they are proposed to carry out specific physiological functions in the kidney and other organs. Their subcellular distribution in organelle membranes like the primary cilia make them the ideal molecular sensory apparatus for cells. Yet, as we discussed, our understanding of many of the hetero- and homomeric polycystin complexes remains understudied, and their molecular regulation and physiological function remain putative. In previous years, the impetus to explore their physiological function came from their dysregulation in human disease (e.g., ADPKD)—work that is now coming to fruition. Future research should be directed at determining the physiological relevance of polycystin populations in the acrosome, embryonic node, and brain.
Table 1
Gene Corresponding protein names and abbreviations PKD1 Polycystin-1 (formerly TRPP1) PKD1L1 Polycystin-1-like protein PKD1L2 Not available PKD1L3 Not available PKDREJ Not available PKD2 Polycystin-2, TRPP1 (formerly TRPP2) PKD2L1 Polycystin-2L1, polycystin-L, TRPP2 (formerly TRPP3) PKD2L2 Polycystin-L2, TRPP3 (formerly TRPP5) Open in a separate window Polycystin nomenclature. SUMMARY POINTS 1 Polycystins form Ca2+-conducting channels in organelle membranes, such as the primary cilium, where they are proposed to receive sensory inputs from the periphery. 2 Variants in polycystins can cause human development defects and polycystic kidney disease. 3 Polycystin subunits can assemble as homomeric and heteromeric complexes that have different modalities of channel regulation. 4 Polycystin tissue expression is widely distributed but often unique to specialize cell types, such as sperm and taste cells. 5 Cryo-EM structures of polycystins have provided insight into channel regulation through unique structural domains. FUTURE ISSUES 1 Capturing polycystins in unique structural states will enhance our understanding of conformational changes that govern their conductive properties. 2 Assessing the impact of disease-causing variants on channel localization and function is required to develop therapeutic approaches to polycystic kidney disease. 3 Further functional characterization of many polycystin channel types is outstanding, yet essential for elucidating their physiological roles. 4 Identification of endogenous ligands that regulate polycystin channels remains largely undefined.
SUMMARY POINTS
1 1 Polycystins form Ca2+-conducting channels in organelle membranes, such as the primary cilium, where they are proposed to receive sensory inputs from the periphery. 2 2 Variants in polycystins can cause human development defects and polycystic kidney disease. 3 3 Polycystin subunits can assemble as homomeric and heteromeric complexes that have different modalities of channel regulation. 4 4 Polycystin tissue expression is widely distributed but often unique to specialize cell types, such as sperm and taste cells. 5 5 Cryo-EM structures of polycystins have provided insight into channel regulation through unique structural domains.
FUTURE ISSUES
1 1 Capturing polycystins in unique structural states will enhance our understanding of conformational changes that govern their conductive properties. 2 2 Assessing the impact of disease-causing variants on channel localization and function is required to develop therapeutic approaches to polycystic kidney disease. 3 3 Further functional characterization of many polycystin channel types is outstanding, yet essential for elucidating their physiological roles. 4 4 Identification of endogenous ligands that regulate polycystin channels remains largely undefined.
ACKNOWLEDGMENTS
The authors were supported by the US National Institutes of Health (U2CDK129917, TL1DK132769, R01DK123463-01, R01 DK131118-01) and National Institute of General Medical Sciences (5T32 GM008382). The content is solely the responsibility of the authors and does not necessarily represent the official views of the US National Institutes of Health.
Footnotes
. DISCLOSURE STATEMENT. The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.
Article Images (0)
×![]()
Bioz is the world’s most comprehensive AI search engine for scientific experimentation. The Bioz search engine offers researchers billions of data-driven product, technique, and protocol recommendations. Bioz taps into the latest advances in Artificial Intelligence (AI) – including Natural Language Processing (NLP), Machine Learning (ML) and Generative AI to mine and structure hundreds of millions of pages of complex and unstructured scientific papers. By harnessing the power of AI, Bioz provides researchers with an unprecedented amount of summarized scientific experimentation knowledge right at their fingertips, ultimately helping to speed up drug discovery and the rate of success in finding cures for diseases. Bioz is used by over 15 Million researchers from over 17,000 different universities and companies in 196 countries.
