Anti-OCRL antibodies have been previously described [35]. Anti-INPP5B antibody was purchased from Proteintech (Immunogen: 1–309 amino acid of human INPP5B)(Chicago, IL). Antibodies against gamma-tubulin, and acetylated alpha-tubulin were purchased from Sigma (St. Louis, MO). Secondary antibodies AlexaFluor 488 and 546 -conjugated donkey anti-mouse IgG (1∶1000), Cy3-conjugated donkey anti-mouse IgG (1∶500), horseradish peroxidase-conjugated goat anti-rabbit and anti-mouse IgG were obtained from Jackson ImmunoResearch Laboratories, Inc (West Grove, PA). IRDye goat anti-mouse and anti-rabbit (680 and 800) were obtained from Li-cor Bioscience (Lincoln, NB).

FLAG-INPP5B, FLAG-Inpp5b and lentiviral GFP-Inpp5b were generated using Creator (Clontech) based vectors as described [38]. FLAG-INPP5B-delta-CAAX, FLAG-Inpp5b-delta-CAAX and lentiviral GFP-Inpp5b-delta-CAAX mutant were generated using QuikChangeII kit from Stratagene (Santa Clara, CA).

All experiments were conducted in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement on the Use of Animals in Ophthalmic and Vision Research. All procedures in rats were approved by the Institutional Animal Care and Use Committees at Indiana University (Study number #0000003163). Zebrafish studies were approved by the Institutional Animal Care and Use Committees at Indiana University (Study number #3843). Human eyes specimens were obtained with approval of Health Sciences and Behavioral Sciences Institutional Review Boards (IRB-HSBS) of University of Michigan (HUM00034652) and was previously published [35]. Mice mOcrl^−/−:mInpp5b^−/−:hINPP5B^+/+ (ocular samples were generous gifts of Dr. R. Nussbaum, University of California, SF, CA) were previously described [21].

hTERT-RPE1 (American Type Culture Collection) were grown in DMEM/F-12+GlutaMAX (GIBCO) with 10% FCS, penicillin–streptomycin at 37 C in 5% CO₂. Primary human trabecular meshwork (HTM) cells were generous gifts of Dr. Juanita Dortch-Carnes (Morehouse school of Medicine, Atlanta, GA) [39]. Normal human fibroblast (NHF) is a gift of Dr. Dan Spandau (Indiana University, Indianapolis, IN) as previously described [35]. Lowe 1676 and Lowe 3265 patient fibroblasts (Coriell Institute) have markedly decreased OCRL protein levels and were previously characterized [35]. Transfections were performed using Polyethylenimine according to previous published protocol [40].

Immunohistochemistry was performed by obtaining ten micrometer sections from formalin fixed paraffin-embedded tissue, stained with H&E as previously described [35]. Immunofluorescence slides were treated with paraformaldehyde (PFA) for fixation for 10 minutes at room temperature (RT) followed by permeabilization with 0.5% Triton X-100, then blocked with Phosphate buffered saline (PBS)/0.5% Bovine Serum Albumin (BSA)/10% Normal Goat Serum (NGS) for 30 minutes at RT. Cover slips were incubated with the primary antibodies overnight at 4°C, secondary antibodies for 45 minutes at RT, and mounted in ProLong Antifade reagent (Invitrogen). The hTERT-RPE1 cells were synchronized by serum starvation according to previous protocols [41]. To evaluate cilia length, cells were grown to confluence for 48 hours followed by serum deprivation, then fixed with 4% PFA and cilia growth was analyzed by immunostaining with acetylated α-tubulin from z-stacks (0.56 micron step size) taken with LEICA SP6 confocal microscope or Zeiss LSM-700 confocal microscope. Ciliary length was measured with NIH Image J software and statistical analysis was performed with SAS.

Cell lysates were subjected to SDS-PAGE on 10%–12% gel and transferred to nitrocellulose membrane (BioRad). Membranes were blocked in 5% non-fat dried milk in PBS. Primary and secondary antibodies were diluted in concentrations as described above [35]. Odyssey Imaging system (Li-Cor Bioscience) was used to analyze the immunoblots.

INPP5B and control knockdown shRNA lentivirus was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Lentiviral particles were transduced into hTERT-RPE1 cells using standard protocols. Knockdown of INPP5B was verified by immunoblotting as described [42].

Zebrafish (wildtype strain: AB Tubingen) were raised and maintained at the Laboratory Animal Resource Center of Indiana University. All animal procedures were performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the NIH. The protocol was approved by the Indiana University Animal Care and Use committee (number 3843). All surgery was performed under paracaine anesthesia. Embryos were fixed overnight at 4°C in 4% paraformaldehyde and 1% sucrose in PBS. Embryos were dechorionated and washed with PBST for 6–8 times 10 min each. After blocking them for 2–4 hr with 10% NGS and 0.5% BSA, immunostaining was performed with 1∶200 anti-acetylated tubulin monoclonal antibody and 1∶500 AlexaFluor 546 donkey anti-mouse conjugate separately at 4°C overnight. KV cilia measurements were performed as described [43]. Cresyl violet staining were performed as described [42].

Antisense MOs were designed and purchased from Gene Tools, Inc. (Philo, OR). Two morpholinos were generated; inpp5b translational blocking MO (inpp5b MO: sequence CTCCTGAAACCCGCCATCAGCGTTC), mismatch (control MO: sequence ATGCGAAATCAAGGTTCGATCATCA), and p53 (p53 MO: sequence GCGCCATTGCTTTGCAAGAATTG) morpholino served as a negative control. Morpholinos stocks were dissolved at 1 mM in water: 2 or 4 nl of injection solution (0.25% phenol red) containing 125–500 mM morpholino was injected into fertilized eggs at the one- to two-cell stage using a pressure injector (Pressure System IIe, Toohey Company, Fairfield, NJ). Synthetic mRNA was prepared from linearized human DNA with Ambion mMessage mMachine high-yield Capped RNA transcription kit, purified with phenol-chloroform, and mRNA was coinjected for rescue experiments as previously described [35].

Melanosome transport assay was performed as described [42]. Briefly, zebrafish 5 dpf larvae were exposed to epinephrine (50 mg/ml, Sigma) at the final concentration of 2 mg/ml in a dark room, and melanosome retractions were observed under the brightfield microscope Leica DFC310 FX. The end of melanosome transport was marked when all melanosomes in the head were perinuclear.

OCRL mutations have been found in Lowe syndrome, which usually presents with bilateral congenital cataracts and glaucoma [4]; however, the pathogenesis of these ocular phenotypes is not known. In the anterior segment of the eye, the trabecular meshwork and Schlemm’s canal endothelial cells regulate aqueous outflow and are often affected in glaucoma patients (Fig. 1A). We hypothesized that OCRL is expressed in human tissues that are responsible for cataract and glaucoma development. To test this hypothesis, we examined human eyes previously enucleated for solitary choroidal melanoma that retained normal anterior and posterior segments. These sections were stained by H&E and immunohistochemistry was carried out with a previously characterized affinity-purified anti-OCRL antibody [35]. Indeed, OCRL was observed in punctate vesicles in the trabecular meshwork and lens epithelial cells; in addition, low level of protein expression was also detected in the Schlemm’s canal endothelial cells (Fig. 1B). In contrast, staining of eye sections using an anti-INPP5B antibody showed nearly undetectable signal in human trabecular meshwork as well the Schlemm’s canal endothelial cells.

The epithelial cells of the lens are a monolayer of epithelium with the basement membrane directed outwardly towards the anterior chamber. Similarly to the trabecular meshwork staining, OCRL is detected in a vesicular pattern but not INPP5B (Fig. 1C). The ciliary body epithelium, which is composed of outer pigmented and inner non-pigmented epithelium, was also examined for inositol phosphatase distribution. The non-pigmented epithelium (NPCE) is responsible for aqueous secretion and is the target for many anti-glaucoma medications. OCRL was detected in human ciliary body epithelium primarily in the outer pigmented epithelial cells; however, INPP5B protein was not detected (Fig. 1D). In the posterior segment of the eye, we found INPP5B immunoreactivity in the outer segment of the photoreceptors and retinal pigment epithelial cells (RPE); the pattern of localization corresponds to the outer photoreceptor segments (Fig. S1A-B in File S1). Similar immunohistochemical results for INPP5B and OCRL were obtained from rat eyes (data not shown).

OCRL and INPP5B are paralogous genes with overlapping enzymatic functions [19]. Although the loss of OCRL in humans results in severe multi-organ phenotypes, inactivation of Ocrl in mice does not phenocopy human disease. This may be due to redundancy of Inpp5b, which is supported by the early embryonic lethality of loss of both Ocrl and Inpp5b. Thus we compared the expression of Ocrl and Inpp5b in murine trabecular meshworks (MTM). The trabecular meshwork in murine eyes stained similarly for both OCRL and INPP5B (Fig. 1E). Using a purified antibody against INPP5B, we show the specificity by immunoblotting using lysates of hTERT-RPE1 knockdown of INPP5B (Fig. 1F). However, a comparison of their expression patterns in human trabecular meshwork (HTM) reveals that the protein level for OCRL was higher in HTM cells while INPP5B was relatively in MTM cells (Fig. 1G). Taken together, OCRL and INPP5B are differentially expressed in the ocular tissues involved in glaucoma and cataract development.

Because of the compensatory function of Ocrl and Inpp5b in mice, we hypothesized that INPP5B is also a ciliary protein. To address this possibility, we used hTERT-RPE1 cells, which is a well-described ciliated epithelial cell line [44]. Ciliogenesis was induced in hTERT-RPE1 cells by serum-starvation for 24 to 48 hr and the presence of INPP5B in the primary cilia was then examined with acetylated alpha-tubulin immunostaining. This showed that INPP5B localizes distinctly along the primary cilia (Fig. 2A). The immunostaining was absent in cells treated with shRNA INPP5B via a lentiviral vector, which resulted in an approximately 90% decrease in INPP5B expression as compared to controls (Fig. 2B–C). INPP5B was also detected with gamma-tubulin, a basal body marker (Fig. S2 in File S1).

Consequently, a role of INPP5B in cilia development was explored. INPP5B knockdown (KD) resulted in a significant decrease in the percentage of ciliated cells in comparison to control KD cells. Approximately 21% and 50% less ciliated cells were observed for INPP5B KD at 24 hour and 48 hour, respectively (8.2%±2% at 24 hour control KD vs 6.5%±1.5% in INPP5B KD cells, p = 0.04; 16%±2% at 48 hour control KD vs 8%±2% in INPP5B KD cells, p = 2.1E-08) (Fig. 2D). While decreased levels of INPP5B affected ciliation, we also assessed the cilia length in cells serum-starved at 24 and 48 hr. We found that the lengths of the cilia in INPP5B KD cells were significantly reduced as compared to control KD cells. After serum-starvation for 24 hr, hTERT-RPE1 control and INPP5B KD cells showed the cilia length to be 25% less than control cells (3.2±0.4 micron in control KD vs 2.4±0.3 micron in INPP5B KD cells, p = 0.1). About 45% less cells with INPP5B knockdown developed cilia when compared to controls cells after 48 hours of serum-starvation (5.1±0.6 micron in control KD vs 2.8±0.3 micron in INPP5B KD cells, p = 3.12E-09) (Fig. 2E). Therefore, INPP5B is important in development and regulation of primary cilia.

To further determine the functional significance of INPP5B expression in cilia development, morpholino targeting inpp5b expression in zebrafish was employed. The degree of homolog between zebrafish and human INPP5B is approximately 61.0% identity and 76.0% similarity. For this purpose, an antisense oligonucleotide morpholino and a 5-base-pair mismatched control MO were designed and tested. In zebrafish, a single copy of inpp5b is present, which was conveniently targeted using antisense morpholinos. As shown in Figure 3A, inpp5b morpholino-injected embryos specifically decreased Inpp5b protein expression as compared to beta-actin. Expression of inpp5b was verified by immunoblot analysis to demonstrate the specificity of anti-INPP5B antibody. At the 48 hpf stage, inpp5b morphants exhibited dose-dependent microphthalmia, body axis asymmetry, and kinked tail; an approximately 70% of morphants exhibited microphthalmia, which were observed over multiple independent sets of experiments (Fig. 3B–C). To verify that the phenotypes are specific to inpp5b knockdown, p53 MO were injected into zebrafish embryos. This showed that p53 MO alone did not result in decreased eye size, generalized edema, or body axis asymmetry in zebrafish, but these phenotypes appeared in the embryos when co-injected with inpp5b MO and p53 MO (Fig. 3B). These differences in eye size were statistically significant at 48 hpf between p53 morphants and p53 plus inpp5b morphants (236±3.5 micron vs 135±10 micron, p = 4.98E-07) (Fig. 3D). The significant difference in eye size was also found at 72 hpf between p53 morphants and p53 plus inpp5b morphants (293±9 micron vs 180±12 micron, p = 2.1E-10) (Fig. 3D).

Detailed analysis of these ocular phenotypes found the inpp5b morphants developed dystrophic retina and thinner RPE versus control (Fig. 3E). The phenotypes of edema, body asymmetry, and kinked tail in inpp5b morphants are characteristics of functional defects of cilia, and can be partly rescued by co-injection of murine Inpp5b (Inpp5b) WT mRNA with inpp5b MO (Fig. 3F–H). The eye size of morphants at 48 hpf co-injected with Inpp5b WT mRNA was 220±18 micron, compared with the morphants injected with Inpp5b MO (162±15 micron) and control MO (245±7.9 micron) (Fig. 3G).

To evaluate whether Inpp5b is necessary for cilia development, we examined the Kupffer’s vesicle (KV) and pronephros duct of young zebrafish larvae. The KV is a ciliated, fluid-containing structure in the zebrafish, orthologous to the mouse embryonic node; it regulates left-right body axis and organ development through directional cilia rotation [45], [46]. KV was examined by immunohistochemistry with anti-acetylated alpha-tubulin to stain the cilia at the 6-somite stage (Fig. S3A in File S1). Compared with control MO-injected embryos, an approximate 50% reduction in the number of cilia in the KV of inpp5b morphants (64±2 vs 36±3 cilia, p = 3.41E-3) was detected (Fig. S3B in File S1). In addition, the average lengths of cilia within the KV were found to be shorter in the inpp5b morphants than controls embryos (4.3±0.5 micron vs 6.5±0.5 micron, p = 1.89E-23) (Fig. S3C in File S1). Cilia quantity (56±4) and lengths (6.2±0.5 micron) in the KV were rescued after co-injection with Inpp5b WT mRNA. Immunostaining of pronephric duct cilia with anti-acetylated alpha-tubulin at the 24 hpf stage was also carried out. Confocal images of individual cilium, with distinct base and ends (N >200 cilia), were outlined and the calculated length recorded. As compared to control MO-injected embryos, there was an approximately 50% shortening of cilia length in inpp5b morphants vs control embryos (9.5±1.8 micron vs 20±2.1 micron,p = 5.45E-18), and cilia length could be restored (17±1.9 micron) by co-injection with Inpp5b WT mRNA (Fig. S3D–E in File S1). Inpp5b is therefore concluded to exhibit an important role in cilia development and function during zebrafish embryogenesis. In zebrafish embryos co-injected with Inpp5b MO and human INPP5B WT mRNA, a reversal of ciliary phenotype was observed in KV cilia (Fig. 4A–C) and pronephric duct cilia (Fig. 4D–E). Compared with control MO-injected embryos, cilia number of KV return to 59±4 and average cilia length return to 5.7±0.6 micron. The pronephric cilia return to 17±2 micron.

Interestingly, we observed a decreased pigmentation in the ocrl and inpp5b morphants (Fig. 4F). These findings were consistent with our previous studies [42] with the inpp5e morphants suggesting that the ocrl and inpp5b morphant zebrafish may also have defects in melanosome transport. To measure melanosome transport, we determined the rates of epinephrine-induced melanosome retraction in the zebrafish larvae. When 5 dpf embryos were exposed to epinephrine, melanosomes contracted rapidly and the endpoint was marked by perinuclear accumulation of melanosomes [47]. Pigment accumulation of ocrl morphants (7.4±0.9 min) and inpp5b morphants (7.7±1.0 min) was 3 times longer than control morphants (1.8±0.4 min) (Fig. 4G). Importantly, the delayed retraction could be rescued by co-injecting either OCRL mRNA or Inpp5b mRNA (Fig. 4G).

Inositol phosphatase INPP5E contains a prenylation CAAX motif, which was found to play a role in MORM syndrome (a ciliopathy) [27]. A premature stop codon in INPP5E affects ciliary localization but not its enzymatic activity. For this purpose, a CAAX deletion mutant of Inpp5b was assayed for its ability to be recruited to the cilia (Fig. 5A). In hTERT-RPE1 cells transduced with GFP-tagged lentivirus for either Inpp5b WT or Inpp5b-delta-CAAX, the localization of this protein by immunofluorescence was visualized (Fig. 5B). The expression of Inpp5b or Inpp5b-delta-CAAX was verified by immunoblot analysis (data not shown). In 48 hr serum-starved hTERT-RPE1 cells expressing either Inpp5b or Inpp5b-delta-CAAX, the loss of the CAAX motif significantly abolished the ciliary recruitment of INPP5B. In hTERT-RPE1 cells transduced with either Inpp5b or Inpp5b-delta-CAAX lentivirus, the length of GFP-positive cilia was assessed and a significant difference was noted (3.4±0.4 micron vs 2.8±0.3 micron, p = 1.36E-06) (Fig. 5 C).

The ability of Inpp5b-delta-CAAX and INPP5B-delta-CAAX mutant to rescue the loss of INPP5B was investigated. Since those cilia-associated phenotypes in Inpp5b morphants can be partly rescued by co-injection Inpp5b WT mRNA with inpp5b MO, an Inpp5b-delta-CAAX mutant of mRNA was co-injected with 4 ng inpp5b MO (Fig. S3F-G in File S1). In Inpp5b morphants co-injected with Inpp5b WT mRNA, the KV cilia length was 6.5±0.5 micron; however, the KV cilia in morphants co-injected with Inpp5b-delta-CAAX mRNA showed shortened cilia (4.7±0.4 micron, p = 4.90E-07) (Fig. S3G in File S1). Compared with morphants co-injected with INPP5B WT mRNA, the KV cilia in embryos co-injected with INPP5B-delta-CAAX mRNA was 44±5 (p = 3.90E-26, Fig. 5E) and 4.4±0.5 micron in length (p = 9E-06, Fig. 5F). Therefore, the CAAX domain is important for INPP5B in the formation of primary cilia.

The similar function and structures OCRL and INPP5B suggests that the knockdown of both 5-phosphatases may result in greater ciliary defects than the loss of either enzyme alone. To make such comparisons, a scale of phenotypes observed in zebrafish morphants, which include anophthalmia, microphthalmia, hydrocephalus, left-right axis asymmetry, generalized edema, and pigmentation defects, was established. By this scale, the normal fish was indistinguishable from the wild type fish. This scale consists of mild, with one or two of the phenotypes; moderate, with greater than two of the cilia-related phenotypes; or severe, if four or more abnormalities were present (Fig. 6A). In zebrafish treated with either 4 ng of ocrl or inpp5b MO, approximately 20% of zebrafish developed severe disease whereas when both MO are given at 4 ng, nearly 90% of zebrafish were dead within 54 hours (Fig. 6B). The combination of 2 ng ocrl and inpp5b MO resulted in equivalent number of severe phenotype as the 4 ng groups; whereas 0.5 ng of ocrl and inpp5b MO also resulted in equal numbers of mild phenotypes in morphants (Fig. 6B). Therefore, we conclude Ocrl and Inpp5b play synergistic roles in zebrafish; the loss of both genes results in a greater degree of phenotype than either gene alone.

The transgenic murine model of Lowe syndrome was also examined. Mice with Ocrl^−/−:Inpp5b^−/−:INPP5B^+/+ genotype are previously characterized to show renal phenotypes resembling human disease and that introduction of INPP5B rescues the lethality of the double Ocrl/Inpp5b knockout. In these mice, the loss of Ocrl was compensated by Inpp5b expression as determined by immunohistochemistry showing normal structures of the trabecular meshwork, lens epithelial cells, and retinal outer segment (Fig. 6C–E). Expression of Inpp5b was detected in these tissues while Ocrl was not observed. Thus, human INPP5B likely compensates for the loss of OCRL in ocular tissue in this murine model.

To confirm compensatory roles between OCRL and INPP5B, we examined whether INPP5B could restore ciliary length in Lowe patients fibroblasts previously characterized [35], [36], [37]. Normal human fibroblasts, Lowe 1676 and Lowe 3265 fibroblasts were transduced with GFP-Inpp5b or GFP-Inpp5b-delta-CAAX lentivirus, and cilia formation was induced by serum starvation for 48 hr. Acetylated alpha-tubulin immunostaining was performed to determine ciliary length (Fig. 7A). In Lowe 1676 patient fibroblasts, cilia lengths were noted to be longer in Inpp5b transduced cells than those treated with control lentivirus (3.3±0.3 micron vs 2.4±0.2 micron, p = 0.004). Similarly, in Lowe 3265 patient fibroblasts, cilia lengths were also noted to be longer in Inpp5b transduced cells than the control cells (2.2±0.2 micron vs 3±0.3 micron, p = 0.006) (Fig. 7B). However, Inpp5b-delta-CAAX domain did not restore the ciliary length (2.4±0.4 micron in Lowe 1676 and 2.5±0.4 micron in Lowe 3265) (Fig. 7B).

Additionally, zebrafish inpp5b was assessed for its ability to restore the loss of cilia formation in ocrl morphant zebrafish. The KV cilia in morphants co-injected with ocrl MO and Inpp5b WT mRNA were noted to be longer than those injected with ocrl MO (4.3±0.4 micron vs 3.0±0.3 micron, p = 2.19E-04) (Fig. 7C). Co-injection with Inpp5b-delta-CAAX mRNA did not restore the cilia length in the ocrl morphants (3.2±0.2 micron vs 3.0±0.3 micron, p = 0.07) (Fig. 7D). Therefore, we conclude that INPP5B can rescue the role of OCRL in development and regulation of primary cilia.