Materials and Methods

Yeast two-hybrid screening and assays.
The LD of human RanBP2 (residues 62–711) was subcloned in-frame into the GAL4 DNA-binding domain of bait vector, pBUTE (a kanamycin-resistant version of GAL4 bait vector pGBDUC1) [62] and HybriZAP pBD-GAL4 vector (Stratagene, La Jolla, California, United States). The former was used to screen ~18 million clones via mating from murine 9- to 10-d-old embryo and brain cDNA libraries at the Molecular Interaction Facility (MIF), University of Wisconsin, Madison, Wisconsin. The latter was used to screen via transformation of ~5 million clones from bovine retinal cDNA libraries [32]. The screens generated six clones. One and two in-frame clones were independently isolated from the embryonic and adult brain libraries, respectively, and the interactions were validated. The three clones encoded Cox11. Interactions between Cox11, LD, and subdomains thereof were quantified by liquid β-galactosidase (Applied Biosystems, Foster City, California, United States) and growth assays [63]. The maximum specific growth speed (μmax) was determined by calculating μmax = (ln(x t) − ln(x 0))/t, where x t is the OD600 of the culture at t = t, x 0 is the OD600 at t = 0 and t is the time between x 0 and x t. Assays were performed with three independent clones and three samples of each clone, the results were averaged, and the standard deviations calculated.

Site-directed and deletion mutagenesis and plasmid construction.
Deletion mutagenesis was carried out with pairs of primers against domains of interest described in the figure legends. PCR products were subcloned into pGEX-KG [64], HybriZAP pBD-GAL4 vector (Stratagene), and pBUTE [62] vectors. GST-LDZIP alone comprised residues 447–483 of human RanBP2.

GST pull-down and immunoprecipitation assays.
CHAPS-solubilized retinal extracts, expression, and purification of GST-fused constructs were prepared as previously described [65]. GST pull-down assays were carried out with 0.5–2.2 μM of GST-fused proteins [65]. Co-precipitates were resolved on SDS-PAGE and analyzed by Western blot with antibodies described in the Results section. Unfolded and partially denatured Cox11 were generated by incubating recombinant and native Cox11 overnight with 5.6 and 7 M guanidine hydrochloride and urea, respectively. The Cox11 conformers were then diluted ~20-fold in CHAPS-binding buffer containing GST-LD. Immunoprecipitation assays with 5 μg of antibody and Western blots (~ 200–400 ng/ml of antibody) were performed exactly as described previously [13].

Hexokinase I assay.
Hexokinase I activity was determined spectrophotometrically at 25 °C by the method of coupling the glucose-6-phosphate production via glucose-6-phosphate dehydrogenase with the change in the absorbance of NADPH at 340 nm and as described by [66]. The reaction was started by the addition of purified brain hexokinase I (0.24 μg) (gift from J. Wilson) to 1.0 ml of reaction mixture containing 0.05 M Tris-HCl (pH 8.5), 7.4 mM MgCl2, 6.6 mM ATP, 0.65 mM NADP, 11.1 mM monothioglycerol, and 1 unit of glucose 6-phosphate dehydrogenase. In the case of hexokinase activity measured in the presence of Cox11 and LD, the purified HKI was incubated with the recombinant proteins for 15 min at 4 °C before measurement of the activity. The data were fitted directly into the Michaelis-Menten equation using SIGMAPLOT (SPSS Science).

Antibodies.
Rabbit antisera were raised against the recombinant murine Cox11 (residues 40–275) and affinity-purified under non-denaturing conditions (Stereogene) as previously described [13]. Cox11 antibodies were used at 200 ng/ml for Western analysis. The monoclonal antibody Hsp70 was from Affinity Bioreagents (Golden, Colorado, United States) and antibodies against the KBD (JX2) and ZnF domains of RanBP2 have been described [13,16]. Polyclonal and monoclonal antibodies against Hexokinase I were provided by J. Wilson. The mAb414 was purchased from Abcam (Cambridge, Massachusetts, United States).

Immunocytochemistry and microscopy.
Retina dissections, radial cryosections (~6 μm), and immunohistochemistry procedures were carried out as described elsewhere [19]. Brains from 3- to 5-mo-old C57Bl/6 mice were fixed overnight in 2% paraformaldehyde, infused with 30% sucrose, and processed for cryosectioning. Primary brain neurons and glial cells were prepared from the cerebral cortex. This was macerated in the Hank's Balanced Salt solution and triturated. Primary cells were cultured in DMEM (GIBCO, San Diego, California, United States) and collagen-coated 35-mm glass bottom culture dishes (MatTek Corporation, Ashland, Maine) at 5% CO2/37 °C for ~2 d and processed for immunocytochemistry. Primary antibodies were used at concentrations ~2.5–10 μg/ml. Alexa 488- and Alexa 594-conjugated secondary antibodies (2.5 μg/ml) (Molecular Probes) were used for visualization of proteins. Crossover of fluorescent probes, background, and autofluorescence were found to be negligible. Visualization of specimens and localization of proteins were carried out by wide-field epifluorescence microscopy on Nikon (Tokyo, Japan) E600 upright and TE2000U inverted research microscopes equipped with similar Apochromat objectives. Images with the Nikon E600 were acquired with a SPOT-RT digital camera coupled to the microscope and driven by SPOT Imaging v4.0 software (Diagnostic Instruments, Sterling Heights, Michigan, United States). All images were captured at nonsaturating integration levels, 12-bit mono black/white, and then pseudo-colored. Protein colocalization analysis was carried out on the Nikon TE2000U microscope equipped with appropriate excitation and emission filter wheels, cube filters, 100 W mercury light source, Nomarski/DIC, and Plan Apochromat optics (100×, 60×, and 40× oil objectives with NA of 1.4, 40 ×LWD, and 20 ×LWD objectives and encoded motorized Z-Stage (Prior Scientific, Rockland, Maryland, United States). Images with the inverted Nikon TE2000U microscope were obtained using a CCD camera (CoolSNAP HQ; Roper Scientific, Trenton, New Jersey, United States). Images acquired under identical acquisition parameters were analyzed using Metamorph Software v6.2 (Universal Imaging, Glendale, Wisconsin, United States). Whenever applicable, serial optical Z-stacks (20–30 focal planes at 100-nm intervals) were captured and computationally processed by 3-D blind deconvolution methods with the same software.

Generation of a mouse line with a disrupted RanBP2 locus and histological analysis of trapped RanBP2 mice.
A feeder-independent mouse ES cell line derived from the 129P2/OlaHsd strain and harboring a disrupted RanBP2 locus by insertion mutagenesis with the gene trap vector, pGT0pfs [67,68] (generated by W. Skarnes), was obtained from the Mutant Mouse Regional Resource Center (University of California Davis, Davis, California, United States). The gene trap vector contains a promoterless neo-lacZ fusion gene with a splice acceptor site at the 5′ end followed by an internal ribosome entry site (IRES) and the human placental alkaline phosphatase (PLAP) reporter cassette. The RanBP2 129Ola ES line was injected into C57BL/6J blastocysts and four male chimeras were generated. These were bred into 129P2/OlaHsd (Harlan, Indianapolis, Indiana, United States) and C57BL/6J (Jackson Laboratory, Bar Harbor, Maine, United States) females. The two resulting and independent F1s lines, with 129P2/OlaHsd inbred and 129P2/OlaHsd/C57BL/6J mixed genetic backgrounds, were tested for germline transmission by PCR and Southern blot analysis of tail genomic DNA. β-gal and PLAP activities in whole mount embryos and retinal sections were detected, respectively, by incubation with 5-bromo-4-chloro-3-indolyl β-D-galactopyranoside and AP staining buffer (0.1 mg/ml 5-bromo-4-chloro-3-indolyl phosphate, 1 mg/ml nitroblue tetrazolium in 100 mM Tris-HCl (pH 9.5), 100 mM NaCl, 5 mM MgCl2) as described elsewhere [67,68].

Western blot and quantitation analysis of protein expression and ATP levels.
Homogenates and detergent solubilized extracts of tissues and immunoblots were prepared as described elsewhere [69]. Quantitation of immunoblot bands calibrated against internal marker bands, such as mHsp70 and Nup153, was carried out with Gel Pro Analyzer (MediaCybernetics, San Diego, California, United States). For determination of ATP levels, freshly dissected and flash-frozen tissues were homogenized in 2.5% trichloroacetic acid (TCA), neutralized, and diluted to a final concentration of 0.1% with Tris-Acetate (pH 7.75). ATP measurements were carried out with the ENLITEN ATP Assay System Bioluminescence Detection Kit for ATP (Promega, Madison, Wisconsin, United States) as per the manufacturer's instruction in SpectraMax-M5 (Molecular Devices, Sunnyvale, California, United States). Protein concentration was measured by the BCA method (BioRad, Hercules, California, United States) using BSA as a standard.

Metabolic assays.
Two groups of male mice were kept under separate diet since gestation or birth. One group was fed with a normal chow diet (PicoLab Rodent Diet 20 −5053; LabDiet, Richmond, Indiana, United States), while the other was placed on a higher energy and fat diet (PicoLab Mouse Diet 20 −5058; LabDiet). Glucose tolerance test was performed on overnight-fasted mice injected intraperitoneally with 2 g of glucose per kg of body weight. Venous blood glucose values were determined at immediately before (0 min), and 15, 30, 60, and 120 min after the injection with an automatic glucose monitor instrument (One Touch Ultra, Lifescan). Insulin tolerance test was performed the same way as that described for the glucose tolerance test with the exception that human insulin (0.75 U/kg of body mass) was injected intraperitoneally in non-fasted mice. Pyruvate challenge test was performed the same way as that described for the glucose tolerance test with the exception that pyruvate dissolved in saline (2 g/kg of body mass) was injected intraperitoneally in overnight-fasted mice.

Electroretinography.
Five ~6-mo-old male RanBP2+/+ and RanBP2+/− mice in 129ola background were kept in dim cyclic light (10 lux) for 2 wk before electroretinographic responses were recorded. The mice were anesthetized intraperitoneally with ketamine (80 mg/kg) and xylazine (4 mg/kg). Body temperature was maintained with a heating pad at 37.5 ± 1 °C throughout the recording. The pupils were fully dilated with 0.5% tropicamide and 2.5% phenylephrine HCl, and both eyes were recorded simultaneously. Corneal gold-wire loops were used as the active electrodes under 0.5% proparacaine hydrochloride topical anesthesia. Gold-wire electrodes placed on the sclera near the limbus served as reference electrodes. A ground wire was attached to the ear. Scotopic responses were recorded from threshold to 9 log units of intensity above threshold, amplified, and filtered (5,000 gain, 0.1–1,000 Hz). For photopic recordings background light at 34 cd/m2 was used to suppress rod function.