Tubby is required for trafficking G protein-coupled receptors to neuronal cilia
© Sun et al; licensee BioMed Central Ltd. 2012
Received: 23 May 2012
Accepted: 7 August 2012
Published: 1 November 2012
Skip to main content
© Sun et al; licensee BioMed Central Ltd. 2012
Received: 23 May 2012
Accepted: 7 August 2012
Published: 1 November 2012
Tubby is the founding member of the tubby-like family of proteins. The naturally occurring tubby mutation in mice causes retinitis pigmentosa, hearing loss and obesity. Tubby has been proposed to function as an accessory factor in ciliary trafficking. We directly examined a role for tubby in ciliary trafficking in vivo.
We used immunofluoresence labeling to examine the subcellular localization of rhodopsin, somatostatin receptor 3 (SSTR3) and melanin concentrating hormone receptor 1 (MCHR1), all of which are G protein-coupled receptors (GPCR), in the retina and brain of wild type (WT) and tubby mutant mice.
In tubby mouse retina, rhodopsin is not fully transported across the connecting cilia to the outer segments with ensuing photoreceptor degeneration. In the tubby mouse brain, SSTR3 and MCHR1 fail to localize at the neuronal primary cilia in regions where these receptors play critical roles in neural signaling. The tubby mutant does not manifest a generalized defect in ciliogenesis or protein trafficking.
Tubby plays a critical role in trafficking select GPCRs to the cilia. This role is reminiscent of tubby-like proteins 1 and 3, which have been proposed to facilitate trafficking of rhodopsin and select GPCRs in photoreceptors and the developing neural tube, respectively. Thus tubby-like proteins may be generally involved in transciliary trafficking of GPCRs.
The tubby-like proteins are defined by a highly conserved carboxyl terminal half of their primary sequence known as the tubby signature domain [1, 2]. This family of proteins includes the prototype tubby, and TULP1, 2 and 3, for tubby-like proteins 1, 2 and 3 [3–5]. Other than members of the tubby family, search of sequence databases reveals no significant homology with known proteins or functional motifs. The tubby gene (Tub) was originally discovered by way of a spontaneously arisen obesity model in mice, and other members of the family were subsequently identified by homology cloning . Mutations in human TULP1 are a cause of retinitis pigmentosa . Loss of TULP1 function in mice replicates this rapid photoreceptor degeneration phenotype [7, 8]. Prior to photoreceptor degeneration in the mouse retina, pronounced ectopic distribution of rhodopsin is apparent indicating a defect in trafficking across the connecting cilia to reach their normal destination, the outer segments . Loss of TULP3 function in mice leads to neural tube patterning defects and embryonic lethality , and the cellular basis can be traced to a failure of Hedgehog signaling due to defective ciliary trafficking . Little is known about Tulp2, but its Chlamydomonas ortholog was identified as one of strongly induced genes during flagellar regeneration  and it was also reported as a candidate gene for human obesity in linkage analysis . The tubby signature domain binds polyphosphorylated phosphatidylinositol , but their N-terminal domain is much more diverse. In the best characterized example, the TULP3 N-terminal domain binds to the IFT-A complex, which is part of the essential cellular machinery for ciliary transport, through a short conserved motif. In cultured cells, TULP3 facilitates membrane receptor trafficking to primary cilia. Thus it serves as bipartite bridges through their phosphoinositide-binding tubby domain and N-terminal IFT-binding motif, coordinating multiple signaling pathways including membrane receptor trafficking .
Originally designated rd5, the spontaneously arisen tubby mouse mutant manifests retinal degeneration, hearing loss and obesity, a tripartite phenotype that resembles mouse models of Bardet-Biedl syndrome (BBS) [17, 18]. The tubby mutation is a G-to-T transversion that abolishes the donor splice site in the penultimate exon (exon 11), resulting in an aberrant transcript . This leads to the substitution of the tubby C-terminal 44 amino acids with 24 different residues encoded by the intron. The spontaneous mutation in the tubby mouse (Tub tub/tub ) appears to cause a loss of function, as targeted disruption of the tub gene gives a similar phenotype . The original study on tubby mice found moderate and progressive hearing loss, and a moderate retinal degeneration . A modifier gene Mtap1a modulates the severity of hearing loss and retinal degeneration in the tubby mutant mice [21, 22]. One interesting feature of the tubby mutant that is shared with the Tulp1 knockout mouse is the extracellular accumulation of rhodopsin-laden vesicles in the interphotoreceptor space surrounding the photoreceptor inner segments, which peaks at around 17 to 21 days of age when rhodopsin is rapidly synthesized to build up the outer segments . The vesicles are relatively uniform in size averaging 0.1 to 0.2 μm in diameter and bounded by a single membrane. This distinct phenotype is also seen in transgenic mice carrying a C-terminal rhodopsin mutation known to affect specifically the trafficking of rhodopsin to the outer segments . It was, therefore, hypothesized that the extracellular vesicle accumulation might be a hallmark of defect in the directional transport of nascent rhodopsin to the outer segments, thus implying a role for tubby and TULP1 in rhodopsin trafficking in photoreceptors . In further support of this hypothesis, mice doubly mutant for tubby and Tulp1 have a much more severe retinal phenotype than either mutant alone, manifesting a complete failure of rhodopsin trafficking and outer segment formation, and rapid cell death. These data would appear to suggest that tubby may function synergistically with TULP1 in a pathway that facilitates rhodopsin trafficking to the outer segments . Differing from TULP1, which is photoreceptor-specific, tubby has a wider range of expression but appears enriched in neuronal tissues.
Based on a structure-directed approach, it has been proposed that tubby-like proteins are a unique family of bipartite transcription factors [14, 24]. The molecular architecture of tubby-like proteins is seen as well suited for a function in transcriptional modulation. There is the nuclear localization signal at the N terminus of tubby, an ability of the N-terminal domain to activate transcription when fused to a DNA binding motif and the ability of the conserved C-terminal tubby domain to bind DNA and phosphatidylinositol 4, 5-bisphosphate (PIP2). That the tubby domain binds specifically to PIP2 has been well established  but the transcriptional target genes of tubby have remained unknown. In another series of studies, tubby was proposed to be a MerTK ligand that mediates phagocytosis of the photoreceptor outer segments by retinal pigment epithelia . These findings represented advances in the molecular dissection of tubby function, but how they relate to the in vivo role of tubby and the tubby mutant phenotype has been less clear. In this study, we examined the subcellular distribution of a number GPCRs and show that tubby is essential for GPCR trafficking in the neuronal and sensory cilia.
All animal care and procedures were approved by Animal Care and Use Committees at the Dean A. McGee Eye Institute and the National Eye Institute. Mice were maintained in an animal facility under a 12-h light/12-h dark lighting cycle. Genotyping for tubby mutation was based on a published protocol . WT and tubby mutant mice at 1 month of age were used for analysis of the brain tissue, and mice at 1 month and at 12 days were used to analyze the retinal tissues.
A His-tagged fusion protein encompassing the N-terminal 200 amino acid residues of mouse tubby protein was expressed in E. coli, purified and used to generate a polyclonal antibody in rabbit. The antibody was affinity-purified.
Mice were euthanized and their retinas and brain were dissected out. Tissues were homogenized in RIPA buffer, boiled in Laemmli buffer and separated on 10% SDS-PAGE gels. Proteins were blotted to polyvinylidene difluoride (PVDF) membrane by electrotransfer. After blocking with 5% non-fat milk, the membranes were incubated with primary antibodies overnight at room temperature. After washing, membranes were incubated with peroxidase-conjugated secondary antibodies. SuperSignal® West Pico Chemiluminescent Substrate (Thermo Fisher Scientific, Rockford, IL, USA) was used for detection. For normalization, protein samples were separated on standard SDS-PAGE and probed with an anti-actin antibody.
For immunofluorescence, eyes were enucleated, placed in fixative and their anterior segments and lens were removed. Brains were placed directly in a fixative containing 2% paraformaldehyde in phosphate buffered saline (PBS) for a total duration of 1.5 to 2 hours. Tissues were embedded in 3% agarose and sectioned at 75-μm thickness using a vibratome. Sections were collected into PBS buffer and remained free floating for the duration of the immunostaining process. For staining neuronal primary cilia in the brain, tissue sections were subjected to heat antigen retrieval at 60°C in PBS overnight. Prior to incubating with primary antibodies, sections were exposed to 50 mM NaCNBH3 to quench background fluorescence and blocked in 5% goat serum/PBS. All antibody incubations were carried out for 16 to 24 hours at ambient temperature. Cell nuclei were counter stained blue by 4′,6-diamidino-2-phenylindole (DAPI). The sections were viewed and photographed on a laser scanning confocal microscope (model TCS SP2; Leica Microsystems, Wetzlar, Germany). Multiple consecutive focal planes (Z-stack), spaced at 0.5-μm intervals, were captured.
Primary antibodies used were anti-GRK1 (MA1-720, ABR), anti-RP1 (a gift of Dr. Eric Pierce), anti-ACIII (sc-588; Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-SSTR3 (ss-830, Biotrend Chemicals, Destin, FL, USA), anti-MCHR1 (sc- 5534; Santa Cruz Biotechnology), anti-Htr6 (NBP1-46557, Novus Biologicals, Littleton, CO, USA), anti-mOR28 (NB110-75089, Novus Biologicals) and anti-rootletin . Secondary antibodies included Alexa Fluor 488-, 546- and 647- conjugated antibodies (Invitrogen, Grand Island, NY, USA), and Cy3-conjugated donkey anti-goat IgG and Cy3-conjugated donkey anti-rabbit IgG (Jackson ImmunoResearch Laboratories, Inc, West Grove, PA, USA).
Total RNA was isolated from age and genetic background-matched WT and tubby mouse brains (at one month of age) using the TRIZOL reagents (Life Technologies, Carlsbad, CA, USA). RNA concentration was measured with NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE, USA) at a wavelength of 260 nm. cDNA synthesis was primed with oligo (dT)20 using Invitrogen SuperScript First-Strand Synthesis System (Life Technologies, Carlsbad, CA, USA). qPCR was carried out on an ABI 7900HT system (Life Technologies). Predesigned TaqMan Gene Expression Assays were purchased from Life Technologies with the following assay IDs: Sstr3, Mm00436695_s1; Mchr1, Mm00653044_m1; Eif2s3y, Mm00468995_g1; Hprt1, Mm01318747_g1; Rps26, Mm02601831_g1; Tuba1a, Mm00846967_g1. PCR reactions were carried out in triplicate in two independent reactions for each gene assayed, and the mean value was used to calculate fold-change of Sstr3 and Mchr1 in tubby vs. WT tissues, after normalizing against the geometric mean of the four housekeeping genes used as internal controls (Hprt1, Eif2s3y, Rsp26, Tuba1a) . The comparative CT method (the 2T -ΔΔC method) was used for analyzing qPCR data .
The photoreceptor outer segment is a modified cilium, where the visual pigment rhodopsin, a G protein-coupled receptor and other components of the phototransduction cascade are concentrated. The proximal end of the outer segment is linked to the cell body (inner segment) via a connecting cilium which is structurally homologous to the transition zone of motile or primary cilia  (Figure 1B). We assessed outer segment protein trafficking in the tubby mutant retina. Previously, we have suggested that tubby may function similarly to TULP1 in facilitating rod and cone opsin trafficking [7, 9], based on the phenotype of the double tubby/Tulp1 mutant mice. In the present study, we assessed the rod and cone visual pigments localization in photoreceptors. As shown in Figure 1B, rhodopsin and cone opsins, the visual pigments in cone photoreceptors, in WT retina are normally localized exclusively in the outer segments. In the mutant, however, a substantial fraction of rhodopsin is mislocalized in the inner segments and cell bodies. Similarly, the cone opsins are also partially distributed in the inner segments, the perinuclear region and the synaptic terminals. Because of its abundance in the outer segments, the visual pigment is also a structural constituent of the outer segments. Decreased transport of the visual pigment likely explains the outer segments being shorter in the mutant. Retention of rhodopsin and most outer segment proteins in the cell body could be a secondary phenomenon to the advanced stage of photoreceptor degeneration. To gain further evidence that visual pigment mislocalization was a primary defect, we expanded our studies to include mutant retinas at postnatal Day 12 and found that opsin is similarly mislocalized (data not shown). In contrast to the mislocalization of the rhodopsin and cone opsins, which are 7-pass transmembrane GPCRs, the localization of non-GPCR membrane proteins, such as peripherin/RDS and rhodopsin kinase (GRK), appear unaffected (Figure 1C). Other membrane associated proteins, such as PDE, and cytoskeleton associated protein, such as RP1, are also unaffected (Figure 1C). These data suggest that transciliary protein trafficking is not generally defective in the tubby mutant. Thus the tubby mutation appears to retard rhodopsin traffic through the connecting cilium. The retinal phenotype of tubby mutant is similar to that of the Tulp1 mutant mice although somewhat milder.
In this study we show that in the absence of tubby, a number of 7-pass transmembrane proteins that function as GPCRs are not properly transported to the primary or sensory cilia. In the brain, we have demonstrated that tubby is required for the transciliary trafficking of two GPCRs, MCHR1 and SSTR3. In multiple brain regions where these GPCRs are found to concentrate in the cilia in WT mice, we show that the same GPCRs are diminished or extinguished from the neuronal primary cilia of tubby mice. By co-labeling with α-acetylated tubulin for ciliary axoneme and with the ciliary membrane protein ACIII, we show that ciliogenesis or maintenance appear unaffected in the tubby mutant. Therefore, the primary defect in the tubby mutant is a block in transport of select GPCRs to the neuronal primary cilia. Tubby is not essential for all GPCR trafficking in neuronal cilia, however. In our study, we found that the odorant receptor mOR28, which is a GPCR that interacts with odorant molecules, remains correctly localized to the distal cilia of olfactory epithelial cells. The entire repertoire of GPCRs that requires tubby protein for trafficking is unlikely to be limited to just MCHR1 and SSTR3. As more cilia restricted GPCRs in the brain are being discovered and tested in the tubby mutant, a more complete list of tubby-dependent GPCR will emerge, which in turn, may help define precisely how tubby functions in a ciliary trafficking pathway.
In the tubby mutant retina, photoreceptor cells accumulate rhodopsin and cone opsins ectopically in the cell body, accompanied by shortened outer segments. Mislocalization of rhodopsin can occur as a secondary phenomenon to retinal degeneration as discussed previously . On the other hand, most retinal degeneration mouse models, in which the affected gene does not function in rhodopsin or ciliary trafficking, do not show overt opsin mislocalization when the photoreceptor layers are largely still intact. For example, in the rhodopsin T17M transgenic mice  and in the P23H transgenic mice and rats [38–40], opsin correctly localize to the outer segments. In contrast, in the rhodopsin P347S mutant, which affects a known C-terminal trafficking signal [41, 42], opsin is prominently mislocalized early on . In the tubby mutant, opsin mislocalization is likely to be a primary defect since we observed rhodopsin mislocalization at an early time point (postnatal Days 12 and 30) while the photoreceptors are still relatively intact and when trafficking of most other membrane proteins remains normal. Furthermore, tubby mice accumulate extracellular vesicles at early postnatal ages which had been shown to be evidence of aberrant rhodopsin trafficking . We believe, therefore, that rhodopsin mislocalization is a primary defect in the tubby mutant photoreceptor, much like that observed in the related Tulp1 mutant mouse in previous studies . The defect in the photoreceptors is limited to GPCRs, rather than a generalized failure of targeting proteins to the outer segments.
In the sensory hair cells of the cochleas, loss of tubby function apparently has a negative impact as well as indicated by the hearing loss in the tubby mice. We hypothesize that the primary defect in the cochlea hair cells might also originate from a defective protein trafficking along the kinocilia. It remains unclear, however, what GPCRs might be candidates for tubby-dependent transport machinery in cochlear hair cells.
Both the tubby mouse phenotype and the cellular defect described in this study appear remarkably similar to those found in the mouse models of BBS. Both models develop obesity, retinitis pigmentosa and hearing loss. Although human BBS patients often show polydactyly, neither mouse models do so, reflecting again the similarity of the two disease models. Human BBS also manifests mental retardation. In BBS2 and BBS4 mouse mutants, both SSTR3 and MCHR1 receptors were found to be absent from neuronal primary cilia . MCHR1 mediated signaling in the hypothalamus regulates food intake and energy homeostasis , and the disruption of its normal localization in the cilia could underlie at least in part the obesity phenotype of the tubby mice. SSTR3 signaling in the cilia of hippocampal neurons appears to couple to ACIII. Disruption of somatostatin signaling in the hippocampus, as revealed in the study of Sstr3 knockout mice, leads to lower cAMP levels in the hippocampus and a defect in novel object learning . This phenotype is similar to that of ACIII knockout mice . Many BBS proteins have been found to localize to the base of cilia. In the case of tubby like proteins, we have previously shown that TULP1 is diffusely distributed throughout the photoreceptor cell body . We were unable to pinpoint tubby protein in the proximity of cilia either in photoreceptors or in CNS neurons (data not shown). A previous work by Ikeda et al. also found a diffuse pattern of tubby staining in the retina. Thus tubby and TULP1 seem to differ from TULP3, which localizes to the base or ciliary shaft in cultured cells [11, 15]. It is possible that tubby will behave differently in cultured neurons vs. in vivo tissues, or that tubby localizes to the cilia under specific conditions. Further experiments will be needed to clarify these points. A human disease attributable to a tubby mutation has not been identified, but BBS is a plausible candidate.
How tubby-like proteins perform their functions remain incompletely understood. Available data together indicate a role for tubby proteins at the cilia. In the case of tubby, TULP1 and TULP3, ciliary trafficking of GPCRs is likely to be a major aspect of their ciliary functions, whereas TULP2 has been implicated in ciliogenesis . A recent study provides important insights into how tubby proteins may execute this process [1, 15]. In that study, it was found that TULP3 interacts with IFT-A particles to traffic membrane receptors to cilia, and modulation by phosphoinositide binding is also required for this process. Furthermore, the IFT-binding sequence in the divergent N-terminal domain is also present in tubby and TULP2, suggesting that other tubby family members may also interact with IFT-A. Thus, a functional interaction with IFT particles during ciliary trafficking may be generally applicable to mechanisms of action by tubby-like proteins. It remains to be determined how the ciliary trafficking role of tubby-like proteins might reconcile with the proposed transcription modulator model . In that model, tubby associates with plasma membranes where PIP2 contents are high. Upon G protein signaling and hydrolysis of PIP2 by phospholipase C, tubby dissociates from plasma membrane and translocates to the nucleus where it binds DNA and modulates gene transcription. It is possible that tubby engages in a similar process, getting on and off plasma membranes depending on the phosphoinositide content along the route of ciliary trafficking. In this regard, it is interesting to note that PIP2 phosphatases, INPP5E and INPP5F, are localized in the cilia and their functional deficits are associated with ciliopathy [46–49], thus creating a potential gradient of PIP2 physiologically at the cilia-plasma membrane junction. As to whether and how tubby-like proteins modulate gene transcription or serve as a ligand for phagocytosis will await further studies.
Tubby protein, and the tubby-like family of proteins in general, are involved in transciliary trafficking of select GPCRs. The GPCR trafficking defects we identified could largely explain the phenotype of the tubby mutant mice.
G protein-coupled receptors
Melanin concentrating hormone receptor 1
Phosphatidylinositol 4, 5-bisphosphate
Somatostatin receptor 3
We thank Robert Fariss for advice with confocal imaging; Rivka Rachel for help with mouse olfactory epithelia preparation; Matthew Brooks, Yan Li and Hyun-Jin Yang for help with qPCR; Nicolas Berbari for discussing choice of antibodies; and Helen May-Simera for commenting on the manuscript. This work was supported by the National Eye Institute intramural program (TL) and by NIH grants COBRE-P20-RR017703 (JM), P30-EY12190 (JM) and R01EY018724 (JM).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.