Ankyrin repeat and suppressor of cytokine signaling box containing protein-10 is associated with ubiquitin-mediated degradation pathways in trabecular meshwork cells

Primary cell culture

Primary human TM (HTM) cells were isolated and cultured as described previously [28,29]. Briefly, TM tissue was dissected from human donor eyes acquired from Lions Eye Bank (Portland, OR). Use of human cells and tissue was approved by the Oregon Health & Science University Institutional Review Board, and experiments were conducted in accordance with the tenets of the Declaration of Helsinki. HTM cells from four individuals were evaluated (average age=25 years; range=4–49 years). Results shown were consistent among all four cell lines used. HTM cells were cultured in medium-glucose Dulbecco’s Modified Eagle’s Medium (DMEM; Invitrogen, Carlsbad, CA) containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin-gentamicin [30]. Primary HTM cells were used until passage 6.

Dermal fibroblasts were grown from punch skin biopsies from two of the GLC1F family members and an unrelated individual. One family member was diagnosed with POAG, while the other had no clinical signs of glaucoma noted during routine eye examinations. Informed consent was obtained from each individual. Skin biopsies were placed in flasks containing low glucose DMEM supplemented with 10% FBS and 1% penicillin-streptomycin. In addition, normal adult human dermal fibroblasts were obtained from ATCC (Manassas, VA) as an additional unrelated control. The fibroblasts were grown in fibroblast basal medium supplemented with the low serum fibroblast growth kit (ATCC). DNA sequencing confirmed that these fibroblasts were derived from an individual without the T255T synonymous mutation and had no other ASB10 exonic mutations. Dermal fibroblasts were used up to passage 8.

Antibodies

Two polyclonal antibodies against ASB10 were used: a rabbit polyclonal (Sigma Aldrich, St. Louis, MO) and a goat polyclonal (Santa Cruz Biotechnology, Santa Cruz, CA). The rabbit polyclonal recognizes ASB10 ankyrin repeats 2 to 5 (exons 2 and 3), and the goat polyclonal recognizes an internal region of ASB10. The following antibodies were also used: mouse monoclonal ubiquitin and K48-ubiquitin (05–1307) and K63-ubiquitin (05–1308) rabbit monoclonal antibodies (EMD Millipore, Billerica, MA); mouse monoclonal alpha4 subunit of 20S proteasome and mouse monoclonal HSP70 antibodies (ENZO Life Sciences, Farmingdale, NY); mouse monoclonal antibodies HDAC6, LAMP1, and Rab7 (Abcam, Cambridge, MA); and mouse monoclonal antibodies p62 and LC3 (MBL International, Woburn, MA).

Western immunoblotting

Radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 50 mM Tris, pH 7.2, 0.5% sodium deoxycholate, 1% NP-40, and 0.1% sodium dodecyl sulfate) containing protease inhibitor cocktail (Sigma) was added to flasks of confluent HTM cells. Proteins in cell lysates were separated on 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels (BioRad Labs, Hercules, CA) under reducing conditions and transferred to nitrocellulose. Immunoblots were probed with rabbit or goat anti-ASB10 polyclonal antibodies. Secondary antibodies were IRDye 700–conjugated antirabbit and IRDye 800-conjugated antigoat (Rockland Immunochemicals, Gilbertsville, PA). Immunoblots were imaged using the Odyssey infrared imaging system (Li-Cor Biosciences, Lincoln, NE).

Recombinant ankyrin repeat and suppressor of cytokine signaling box containing protein-10 and immunoprecipitation

To generate recombinant ASB10, transfection-ready ASB10 variant 3 (rASB10v3) DNA in the pCMV6-AC vector containing a C-terminal green fluorescent protein (GFP) tag was purchased from Origene Technologies (Rockville, MD). Eight µg of plasmid DNA was transfected into 293 cells using Lipofectamine, and RIPA lysates were harvested 48 h later. Control lysates were prepared from a mock-transfection. Immunoblots were run as above and probed with rabbit or goat anti-ASB10 antibodies and mouse antiturbo GFP antibody (Origene).

For the immunoprecipitation experiments, HTM cells were transfected with 4 µg rASB10v3 with Amaxa nucleofection and grown for 3 days. Cells were washed with phosphate buffered solution (PBS; 150 mM NaCl, 2.7 mM Na₂HPO₄, pH 7.2), lysed, and immunoprecipitated with the Pierce Crosslink Immunoprecipitation kit (Thermo Fisher Scientific, Rockford, IL) and the mouse antiturbo GFP antibody (Origene). Protein complexes were separated on 4%–12% gels, and immunoblots were probed with antiubiquitin or anti-HSP70. Western immunoblots were imaged on the Odyssey as above.

Generation of ankyrin repeat and suppressor of cytokine signaling box containing protein-10-silencing lentivirus

Silencing lentivirus that targets all ASB10 transcripts (sh1; exon 2) and control short, hairpin ribonucleic acid (shRNA) (shCtrl) that does not target any known transcript were designed and prepared as described previously [7,31]. For this study, we also designed a second shRNA that targets exons 5–6 (sh177). The sequence of the sh177 shRNA is as follows: 5′-CAC CGC TCT ACT AGA TGT CCA TGG CCG AAG CCA TGG ACA TCT AGT AGA GC-3′.

Silencing lentivirus was generated in 293FT cells with ViraPower packaging mix (Invitrogen) following the manufacturer's instructions [7,31]. To analyze the efficacy of messenger ribonucleic acid (mRNA) and protein knockdown in TM cells, shRNA silencing lentivirus (10⁶ pfu) was added with 6 µg/ml polybrene (Sigma, St. Louis, MO) at the time of plating in serum-containing medium. HTM cells were grown for 3 days to allow infection and silencing. RNA was isolated, and quantitative PCR (qRT-PCR) was performed using the DyNAmo HS SYBR Green qPCR kit (Thermo Sci., Lafayette, CO) and a Chromo4-equipped thermocycler (Biorad) [32]. ASB10 mRNA levels were normalized for 18S RNA, which acted as a housekeeping gene. To ensure knockdown of the ASB10 protein, western immunoblots of shRNA-silenced RIPA lysates were performed using the rabbit polyclonal antibody as above. Human anterior segment perfusion culture was set up and performed exactly as described for the sh1 ASB10-silencing lentivirus [7].

Immunofluorescence and microscopy methods

HTM cells were grown on collagen-coated Bioflex membranes (FlexCell Int, Hillsborough, NC), which provides a more compliant surface for TM cells than glass or plastic, for 3 days [30,33]. Cells were then fixed with 4% paraformaldehyde (PFA) and blocked with CAS universal blocking reagent (Invitrogen). Membranes were removed from the dish and immunostained. Secondary antibodies were Alexa Fluor 488-conjugated antimouse or antigoat and Alexa Fluor 594-conjugated antirabbit or antigoat (Molecular Probes, Eugene, OR). Coverslips were mounted in ProLong gold containing 4',6-diamidino-2-phenylindole dihydrochloride nuclear stain (Invitrogen). Images were obtained with a Fluoview laser scanning confocal microscope (Olympus, San Diego, CA) and processed with FIJI software.

For super-resolution structured illumination microscopy (SIM), cells were grown and immunostained as described above. However, following the secondary antibody application and final wash step, antibodies were post-fixed with 4% PFA, and #1.5 coverslips were used. Images were acquired using a Zeiss (Thornwood, NY) ELYRA PS.1 system that reconstructs super-resolution images from a series of images acquired under spatially structured illumination [34,35]. Images were processed for SIM reconstruction using Zen 2010D software (Zeiss), and selected Z-planes were exported as TIFF images.

To analyze the area of ASB10-stained structures, the freehand selection tool in FIJI software was used to manually demarcate the circumference of each structure in a field. This was repeated for 25 separate confocal images from HTM cells derived from four individuals. The area of each structure was then calculated using the measure function. To analyze the number of ASB10-stained structures, immunofluorescence fields were imaged on a Leica DM500 fluorescence microscope with an attached digital camera and IM50 software. From the acquired images, the number of ASB10-stained large structures, that is, structures >5 µm in diameter, and the number of 4',6-diamidino-2-phenylindole dihydrochloride–stained nuclei were counted per field. At least 24 fields were counted for each treatment. The number of large structures per field was then divided by the number of nuclei, and an average number of large structures per nuclei per field±standard error of the mean (SEM) was calculated. To determine significance, analysis of variance (ANOVA) was used to compare ASB10-silenced or treated cells to control cells. p<0.05 was determined to be significant. In addition, Poisson probability distributions, which are used to describe the distribution of events with random occurrences, were calculated.

Imaris Bitplane software was used to provide a quantitative measurement of the colocalization of ASB10 with all the UPS and AL biomarkers used. Pearson’s correlation coefficient is a quantitative measurement that estimates the degree of overlap between fluorescence signals obtained in two channels [36]. Differences in the intensities of the signals are ignored, but similarity between shapes is considered. The Pearson’s coefficient values range from 1.0, an indication of complete colocalization of two structures, 0, which indicates no significant correlation, to −1, which indicates complete separation of two signals [36,37]. A Pearson’s correlation coefficient was calculated for each antibody pairing from raw compressed confocal z-stacks acquired using sequential scanning. Background correction values were identical for all images. A small box was drawn around the region of interest—either a large ASB-stained structure or small vesicles—using the square tool. Pearson’s correlation coefficients were then calculated by the software. The Pearson coefficients were calculated from multiple images. The Pearson coefficients were averaged, and a standard error of the mean was calculated. The degree of colocalization from the Pearson’s coefficient values was categorized as very strong (0.85–1.0), strong (0.49–0.84), moderate (0.1–0.48), weak (−0.26 to 0.09), and very weak (−1 to −0.27) based on a previously published description [37].

Autophagy inhibition and activation assays

Pharmacological manipulators of autophagy were added to the HTM cells in culture. The HTM cells were plated and grown for 36 h before 5 µM MG132 (Enzo Life Sciences), 10 µM wortmannin, or 1 µM bafilomycin A1 (both Sigma) was added for an additional 18 h. MG132 inhibits proteasome degradation and activates autophagy, while wortmannin inhibits phosphatidylinositol 3-kinase (PI3K) and thus autophagosome formation and autophagy [38-40]. Neither agent is entirely specific to the UPS and AL pathways. Control cells were treated with an equal volume of vehicle (dimethyl sulfoxide). At the end of treatment, the TM cells were immunostained for confocal microscopy, and the number of ASB10-stained structures was counted as described above. For the outflow experiments, MG132 or vehicle control was applied at time point 0, and outflow was monitored for a further 43 h. Data for each eye were normalized to the average flow rate before treatment, and then data from individual eyes were averaged and the SEM was calculated. Data were analyzed with ANOVA where p<0.05 was considered significant.

Characterization of ankyrin repeat and suppressor of cytokine signaling box containing protein-10 antibodies

First, we characterized two commercially available ASB10 antibodies. The epitope of the rabbit polyclonal antibody is from ANK2 to ANK5, while the goat polyclonal recognizes an internal region of ASB10. Both antibodies recognize the predicted molecular size of endogenous ASB10 (48 kDa) in immunoblots of RIPA lysates of HTM cells (Figure 1B). However, the goat polyclonal also recognizes other bands, which could represent alternative splice forms, degradation products, or non-specific bands. Both antibodies recognize recombinant ASB10 variant 3 (rASB10v3) from transfected 293 cells. Since rASB10v3 has a C-terminal GFP tag, the predicted molecular weight was larger, approximately 72 kDa. Again, the goat antibody had less apparent specific cross-reactivity than the rabbit polyclonal antibody. Next, we immunostained primary HTM cells (Figure 1C) and adult human dermal fibroblasts (Figure 1D) with anti-ASB10. Confocal microscopy showed that ASB10 was found in intracellular structures of varying sizes in both cell types. Some structures were perinuclear, but many were not. The rabbit (red) and goat (green) polyclonal antibodies colocalized at the same structures in HTM cells (Figure 1E). A negative control of the TM cells with no primary antibody is shown for comparison (Figure 1F). For the majority of the experiments presented, the rabbit antibody was used, and the goat polyclonal was used only when antibody combinations did not allow us to use the rabbit anti-ASB10.

The area of the ASB10-stained structures was measured (Figure 2A). There were two size ranges: a population of smaller vesicles, with areas from approximately 1 to 14 µm² (diameters, 2–4 µm), and a population of larger structures with approximate areas from 20 to 60 µm² (diameters, 5–9 µm). The average area was 23.76±1.53 µm² (range, 1–128 µm²). The small vesicles were mainly spherical, while the larger structures were more irregularly shaped and included spheres, ovals, and triangular-shaped structures, sometimes with ruffled edges.

The number of large structures (>5 µm diameter) in the normal and ASB10-silenced HTM cells was then counted. Two shRNA silencing lentiviruses were used: One targets exon 2 (sh1) of ASB10, while the other targets exon 5 (sh177). Efficacy of the sh1 ASB10-silencing lentivirus was shown previously [7]. Here, we show that sh177 significantly decreased ASB10 mRNA levels by 70% compared to the control-infected cells (Figure 2B). The ASB10 protein was also decreased in western immunoblots. The densitometry of the bands showed an approximate 35% reduction compared to the control-infected HTM cells. Moreover, this ASB10-silencing lentivirus significantly reduced outflow rate in human anterior segment perfusion culture by a similar amount and in a similar time frame as the sh1 silencing lentivirus (Figure 2C) [7]. This outflow rate reduction when using lentivirus to silence full-length transcripts is consistent with our hypothesis that loss of full-length ASB10 may contribute to IOP increases in patients with glaucoma [7]. ASB10 silencing by two constructs targeting two regions of the gene strengthens the contention that the outflow effect is not due to mis-targeting or non-specific downstream effects of the shRNA.

The sh1 and sh177 ASB10-silencing lentiviruses were then incubated with HTM cells. There was no significant difference between the number of cell nuclei per field in sh1 or sh177 ASB10-silenced cells and the control-infected cells (data not shown). However, there was a significant decrease in the number of large structures present in ASB10-silenced cells for sh1 and sh177 (p<0.035; Figure 2D). The average number of large structures (>5 µm diameter) per nuclei was 0.336, 0.216, and 0.215 for the control-infected, sh1-silenced, and sh177-silenced cells, respectively. The number of ASB10-stained structures was also analyzed with the Poisson probability distribution function (Figure 2E). The number of large structures per field was plotted against the proportion of fields showing the number of structures. The distribution for both silenced groups was narrower and shifted to a lower number of structures per field compared to the control groups, which exhibited a wider distribution curve with a higher probability of large structures per field. Together, these data show that the number of large ASB10-stained structures decreases concomitantly as the ASB10 protein levels decrease.

Ankyrin repeat and suppressor of cytokine signaling box containing protein-10 and the ubiquitin-proteasomal pathway

The SOCS box of ASB10 contains consensus sequence motifs for interaction with Cullin5 and elongin BC, which recruit proteins of the E3 ubiquitin ligase complex [14,15]. In other ASB proteins, the SOCS box recruits the ubiquitin complex, which then ubiquitinates the substrate bound to the ANK repeats suggesting a similar role for ASB10 [14-17]. Polyubiquitin chains are built via linkage at specific lysine (K) residues [41]. Proteins covalently tagged with polyubiquitin chains via K48 are destined for UPS degradation (Figure 3A(i)) [41]. Conversely, K63-linked ubiquitin is not typically associated with proteasomal degradation, but proteins tagged with K63 are degraded by the AL pathway (Figure 3A(ii)) [41-43]. Recently, linkage-specific antibodies have become available that can be used to distinguish polyubiquitin chains [44]. These linkage-specific antibodies are therefore powerful tools for interrogating protein modifications and providing clues to the function of endogenous proteins [45].

To examine ubiquitin colocalization with ASB10, we immunostained HTM cells with ubiquitin, ubiquitin K-linkage specific antibodies, and the alpha 4 subunit of the 20S proteasome (Figure 3). The ASB10-stained small vesicles and large structures showed some colocalization of ubiquitin (Figure 3B) and the 20S proteasome (Figure 3C). Using the K-linked ubiquitin antibodies, we detected colocalization with ASB10-stained structures, mostly with the larger structures rather than the smaller vesicles (Figure 3D,E). These immunostaining patterns suggest that a portion of ASB10 may be associated with the UPS and AL degradation pathways. However, in all cases, numerous vesicles contained ubiquitin, the K-linked antibodies, or 20S proteasome alone, which is not surprising since we would not expect full colocalization of these proteins with ASB10.

Next, we investigated whether ASB10 was ubiquitinated. Due to the low expression levels of endogenous ASB10, recombinant ASB10 was made by transfecting HTM cells with a GFP-tagged ASB10 construct. Recombinant ASB10 was immunoprecipitated using a GFP antibody, and immunoblots were probed with the ubiquitin antibody (Figure 3F). Ubiquitin was present in the cell lysates (Ly) and flow-through (Fl) on western immunoblots of the pull-down assays, but was not present in the bound rASB10 lane (IP). This suggests that ASB10 itself is not ubiquitinated.

Ankyrin repeat and suppressor of cytokine signaling box containing protein-10 and autophagy-lysosomal pathway biomarkers

To further explore ASB10’s association with proteolytic degradation pathways, we colocalized ASB10 with various biomarkers of the AL pathway in the HTM cells. This pathway is characterized by the formation of characteristic cellular structures during the step-wise degradation process, each of which is associated with specific biomarkers (Figure 4A). Histone deacetylase 6 (HDAC6) and HSP70 are biomarkers of aggresomes, a preautophagic structure [20,46,47]. In addition, HDAC6 binds the C-terminus of ubiquitin via its zinc finger-binder of ubiquitin zinc finger domain [48,49]. In HTM cells, HDAC6 and HSP70 colocalized with ASB10 (Figure 4B,C). HDAC6 also immunostained some TM cell nuclei in a punctate pattern, consistent with HDAC6’s role as a histone deacetylase (data not shown).

Next, we investigated the expression pattern of the multifunctional scaffold protein, p62. This protein binds to K63-linked ubiquitin chains and is recruited into the double-membraned autophagosome [25,42]. p62 colocalized with large ASB10-stained structures, but small p62-stained vesicles in the cytosol were not colocalized (Figure 4D). Microtubule-associated protein 1 light chain 3 (LC3), a biomarker of autophagosomes, binds to p62 and is found as a cytosolic form (LC3-I) and an autophagosome-associating form (LC3-II) [50]. The antibody used in this study recognizes both forms of LC3. Clusters of LC3-stained vesicles were highly concentrated in the center of the large ASB10-stained structures (Figure 4E). LC3 was also found in single vesicles in a punctate pattern throughout the TM cell cytoplasm, which was not associated with ASB10.

We then immunostained HTM cells with the small GTPase Rab7. This is a biomarker of late endosomes within the endocytic pathway, but it can also associate with autophagosomes to form single-membraned amphisomes and mediate their fusion with lysosomes [51-53]. Not all Rab7 associated with ASB10, but that which did seemed to be predominantly located on one side of the ASB10-stained structures (Figure 4F). Finally, we examined immunostaining of LAMP1, an archetypal biomarker of lysosomes. LAMP1 staining was punctate in single vesicles in the cytoplasm of HTM cells, but clusters of LAMP1 vesicles were associated with ASB10-stained structures (Figure 4G). LAMP1 showed intense staining in the center of these structures.

Super-resolution confocal microscopy

SIM is a new microscopy technique that provides substantially increased resolution (approximately 100 nm) compared to conventional confocal microscopy (to approximately 250 nm) [35]. This advanced light microscope methodology has several advantages over electron microscopy for identifying autophagosomes, including simpler immunostaining procedures for visualizing a larger number of cells. We therefore used SIM to further investigate ASB10-stained structures and the colocalization with selected UPS and AL biomarkers (Figure 5). Many of the large ASB10-stained structures appeared to be composed of clusters of smaller vesicles and appeared to form an empty sphere (Figure 5A). Smaller individual vesicles that were not clustered were also apparent (not shown). Since the higher resolution lends to more precise colocalization, SIM was also used to confirm the colocalization of ASB10 and a select number of UPS and AL biomarkers. Ubiquitin partially colocalized with ASB10, as if there was fusion of ASB10- and ubiquitin-stained structures (Figure 5B). This was also observed for the small vesicles. K63-linked ubiquitin chains were highly colocalized with ASB10 in large structures (Figure 5C). HDAC6 also colocalized with ASB10, and HDAC6 appeared as if was partially contained within an ASB10-stained sphere (Figure 5D). LC3-stained vesicles were clustered and surrounded by ASB10-stained structures, and there was partial colocalization (Figure 5E). Finally, Rab7 was associated with the ASB10-stained structures, but again the pattern suggested that there may have been fusion of the two structures (Figure 5F). Only a small amount of colocalization of Rab7 with ASB10 was observed. Together, this high-resolution microscopy data suggested that large ASB10-stained structures may form from clustering of smaller vesicles, and colocalization of ASB10 with K63-linked ubiquitin and HDAC6 was confirmed. Although ASB10 seemed to be located in the same structures with LC3 and Rab7, SIM showed that there was little colocalization.

Quantitative analysis of ankyrin repeat and suppressor of cytokine signaling box containing protein-10 colocalization

To further analyze the colocalization of ASB10 with the UPS and AL biomarkers, we used Imaris Bitplane software to calculate the Pearson correlation coefficients. This quantitative measurement estimates the degree of overlap between fluorescence signals obtained in two channels [36]. Higher Pearson’s values represent a higher degree of colocalization of two signals. The degree of colocalization from the Pearson’s values were categorized as very strong (0.85–1.0), strong (0.49–0.84), moderate (0.1–0.48), weak (−0.26 to 0.09), and very weak (−1 to −0.27) as described previously [37]. Representative large structures and small vesicles were selected from compressed z-stacks by drawing a box around them. The software then calculated a Pearson’s coefficient for each combination of antibodies (Figure 6). Quantitation of multiple images showed that small ASB10-stained vesicles had only moderate colocalization with the AL and UPS biomarkers as the Pearson’s coefficients were at 0.42 and below (Figure 6A). However, analysis of the large ASB10-stained structures showed that ASB10 was strongly colocalized with the α4 subunit of the 20S proteasome, K48 and K63-linked ubiquitin, p62, HSP70, and HDAC6 (Pearson’s range, 0.59–0.82; Figure 6B). There was only moderate colocalization of large ASB10-stained structures with Rab7, LC3, and LAMP1 (Pearson’s range, 0.21–0.41). Of the biomarkers tested, this analysis suggests that ASB10 is more closely associated with biomarkers of the UPS pathway and preautophagic structures of the AL pathway than with mature autophagosomes and lysosomal markers.

To verify the colocalization data, we performed coimmunoprecipitation experiments. Recombinant ASB10-GFP or GFP control constructs were transfected into HTM cells and immunoprecipitated with the ASB10 or GFP antibody (Figure 6C). Western immunoblotting showed that HSP70 (expected size, 70 kDa) was in the bound lysates of the pull-down assays of the cells transfected with rASB10, but not in the GFP control-transfected cells. Similarly, a band corresponding to the expected size (28 kDa) of the α subunit of the 20S proteasome was found in the bound lysate of cells transfected with rASB10, but not in the control cells. These coimmunoprecipitation data confirm an association of ASB10 with biomarkers of the UPS and AL pathways.

Autophagy inhibitors and activators

ASB10 immunostaining was analyzed in the presence of known autophagy inhibitors (wortmannin and bafilomycin A1) and an inhibitor of proteasomal degradation and activator of autophagy (MG132). First, conversion of endogenous LC3 from LC3-I to LC3-II was assessed to monitor autophagy (Figure 7) [38,39]. LC3-II is associated with autophagosomes. HTM cells were serum-starved and incubated with the pharmacological agents for 18 h. In the control and MG132-treated HTM cells, LC3-II was the most abundant form, whereas wortmannin partially inhibited LC3-I to LC3-II conversion (Figure 7A). Bafilomycin A1 caused an accumulation of LC3-II, as was expected. Next, we immunostained HTM cells with ASB10 (red) or LC3 (green) antibodies to monitor autophagy (Figure 7B–I). In the presence of MG132, which inhibits proteasomal degradation and activates autophagy, there was a significant increase in the number of ASB10-stained structures, especially small cytosolic vesicles, and a concomitant increase in LC3 immunostaining (Figure 7D,E). Other biomarkers (p62, LAMP1, Rab7) also seemed to increase (data not shown). Conversely, with the wortmannin treatment, which inhibits autophagosome formation, there was a slight decrease in the number of ASB10-stained structures, and there was a substantial decrease in the association of large ASB10 structures and LC3 (Figure 7F,G). Treatment with bafilomycin A1, which inhibits degradation of autophagic cargo inside autolysosomes, caused a large increase in punctate cytosolic ASB10 staining. However, there was no significant change in the number of larger ASB10-stained structures (Figure 7H,I). Bafilomycin A1 treatment caused a large increase in the number of LC3-stained vesicles, as expected. The number of ASB10-stained structures was counted for each treatment (Figure 7J). MG132 and bafilomycin A1 treatment increased the number of small vesicles compared to vehicle control, thus significantly increasing the total number of ASB10-stained structures. Wortmannin treatment significantly decreased the number of small vesicles, but there was no significant change in the large structures or in the overall number of ASB10-stained structures. In addition, we tested the effect of MG132 on outflow rate in human perfusion culture. Application of MG132 significantly decreased outflow rate compared to vehicle control within 3 h, and this decrease was maintained until the end of perfusion at 43 h (Figure 7K).

Ankyrin repeat and suppressor of cytokine signaling box containing protein-10 immunolocalization in adult dermal fibroblasts

Finally, we evaluated ASB10 immunostaining in adult dermal fibroblasts derived from skin biopsies from one affected and two non-affected members of the family with GLC1F POAG and an unrelated individual. Three normal cell lines and one POAG cell line, which carried the T255T synonymous mutation, were tested. ASB10 was again found in intracellular structures in both normal dermal fibroblasts (Figure 8A,C) and in fibroblasts from the patient with POAG (Figure 8B,D). Colocalization of large ASB10-stained structures with ubiquitin and HDAC6, and other UPS and AL biomarkers (data not shown), showed no discernible differences in the fibroblasts from the normal individuals and the patient with POAG.