15th Annual
2023 Midwest Eye Research Symposium
Supported by an unrestricted grant from Research to Prevent Blindness to the University of Iowa Department of Ophthalmology and Visual Sciences
Afternoon Poster Presentations
Abstracts
Poster #1
Umbilical amnion for posterior lamellar eyelid reconstruction and anopthalmic socket reconstruction: a retrospective chart review
Sanchez, Peter1; Mansoor, Mahsaw2; Keen, Jamie2; Carter, Keith2; Pham, Chau2; Shriver, Erin2
1University of Iowa Carver College of Medicine; 2University of Iowa Hospitals and Clinics Department of Ophthalmology and Visual Sciences
Purpose: Traditionally, autogenous grafts such as oral mucous membrane, hard palate, and dermis fat grafts have been used for posterior lamellar and anophthalmic socket reconstruction. Due to donor site morbidity, other materials, such as umbilical amnion, have been studied in small case studies for anophthalmic socket reconstruction and cicatricial entropion repair. However, its use in primary eyelid reconstruction has not been described.
Methods: An IRB-approved retrospective review was performed to identify patients undergoing anophthalmic socket and eyelid reconstruction with umbilical amnion at a single, academic tertiary care institution between June 2019 and March 2023. Data collected included: patient age; indication for surgery; length of post-operative follow-up; and outcome measures of eyelid position, surgical site infection, wound dehiscence, reoperation, and cosmetic result. Surgeries were considered successful if the socket retained a prosthesis; the corneal epithelium was stable or improved post-operatively; the cosmetic appearance was favorable; and there was no post-operative eyelid malposition, retraction, epiphora, or reoperation.
Results: Twenty-two patients with a mean age of 53 and an average follow-up of 10 months received umbilical amnion. Thirteen anophthalmic socket reconstructions were performed on 13 sockets in 12 patients. Reoperation was not required in 10 sockets and 5 sockets had proper eyelid position. Complications included: 1 surgical site infection, 1 dehiscence, 8 mispositioned eyelids, and reoperation on 3 sockets. Umbilical amnion was used in 10 sighted patients for 3 primary posterior lamellar eyelid reconstructions after Mohs surgery, 1 medial canthus reconstruction after lesion excision, 3 palpebral conjunctival reconstructions, 1 symblepharon repair, and 2 cicatricial eyelid retractions. Of these patients, 1 patient was lost to follow-up. There were no surgical site infections, 1 wound dehiscence, 2 mispositioned eyelids, and 2 reoperations.
Conclusions: Umbilical amnion can be effective for anophthalmic socket contracture reconstruction and provides successful outcomes in patients undergoing primary posterior lamellar eyelid reconstruction. However, due to the nature of injuries seen in some patients with anophthalmic sockets, healing can be difficult and multiple surgeries may be necessary. Nonetheless, umbilical amnion is well tolerated, eliminates donor site morbidity, and reduces surgical times. Further studies are needed to compare the effectiveness of umbilical amnion to other grafts in primary posterior lamellar reconstruction of the eyelid.
Poster #3
Inhibition of HIF-1 Signaling Reduces Visual Impairment in a Mouse Model of Multiple Sclerosis-like Optic Neuritis
Anders, Jeffrey1; Elwood, Benjamin2,3; Kardon, Randy2,3; Gramlich, Oliver2,3
1Biomedical Sciences Program, University of Iowa; 2Department of Ophthalmology and Visual Sciences, The University of Iowa; 3VA Center for the Prevention and Treatment of Visual Loss, Iowa City, IA
Purpose: Optic neuritis (ON) is often an early sign of multiple sclerosis (MS) and recent studies show a link between HIF-1 pathway activation and inflammation. This study aimed to determine if inhibition of the HIF-1 pathway using the HIF-1α antagonist Acriflavine (ACF) can reduce clinical progression and rescue the ocular phenotype in an experimental autoimmune encephalomyelitis (EAE) ON model.
Methods: EAE-related ON was induced in 40 female C57BL/6J mice by immunization with MOG33-55, Complete Freund’s Adjuvant and pertussis toxin. Twenty EAE mice received daily systemic injections of ACF at 5 mg/kg, 20 EAE mice were placebo injected and another 20 naive mice served as controls. Clinical progression was monitored using a 5-point EAE scoring scheme. Visual acuity was assessed weekly and RGC complex thickness was measured using optical coherence tomography (OCT). Retinal Ganglion Cells (RGC) and optic nerve axons were counted using RBPMS and PPD staining. Grades of demyelination (0-3 grading) and numbers of CD3+ cells in optic nerves were evaluated by H&E/LFB and immunohistochemistry. Differences were analyzed by ANOVA and Tukey’s post-hoc test.
Results: ACF-treated EAE animals show significantly less motor-sensory impairment (Area under curve EAE:51±3 vs. EAE+ACF: 28±3; p=0.0001). ACF treatment also preserved visual acuity (EAE:0.24±0.06 c/d vs. EAE+ACF: 0.290.006 c/d; p=0.01) and maintained RGC complex thickness when compared to untreated EAE mice (EAE:61±3µm vs. EAE+ACF:64±3µm; p=0.003). Higher RGC density (EAE:248±19 RBPMS cells/field vs. EAE+ACF: 275±16 RBPMS cells/field; p=0.02) and axon numbers (EAE:1310±562 axons/field vs. EAE+ACF: 2096±527; p=0.008) were observed in ACF treated EAE mice when compared to EAE animals in the placebo group. Grade of optic nerve demyelination was lessened (EAE:2.1±0.5 vs. EAE+ACF:0.9±0.9; p=0.047) and less infiltrating CD3+ cells were observed (EAE:98±32 cell/mm vs. EAE+ACF: 31±20 cells/mm; p=0.0001) in ACF treated EAE mice.
Conclusions: In EAE mice, daily ACF injections reduce motor-sensory impairments as well as increase visual system recovery, including preservation of the RGC density and optic nerve axon number. In conclusion, these improvements seem to be attributed to the inhibition of the HIF-1 pathway by ACF. Further experiments are needed to determine ACF’s mechanisms of action in more detail but is a promising approach for MS therapies.
Poster #5
Utilization of Ketogenic Diet Worsens Visual and Motor-Sensory Deficits in An Animal Model of Multiple Sclerosis
Capper, Erin1; Anders, Jeffrey2; Elwood, Benjamin2,3; Kardon, Randy2,3; Gramlich, Oliver2,3
1Carver College of Medicine, University of Iowa; 2Department of Ophthalmology and Visual Sciences, The University of Iowa; 3VA Center for the Prevention and Treatment of Visual Loss, Iowa City, IA
Purpose: Controversy exists regarding the benefits of ketogenic diet (KD) utilization as adjuvant therapy in multiple sclerosis (MS). The purpose of this study was to establish the primary effects of KD on visual function and structure in an experimental autoimmune encephalomyelitis (EAE) mouse model of MS.
Methods: EAE was induced in the 80 female C57BL/6J by immunization with MOG33-55, complete Freund’s Adjuvant, and pertussis toxin while another 16 mice served as a naïve control group. The 80 EAE induced mice were assigned into cohorts (n=20/group) to stay on the standard rodent chow (EAE) or to start KD either 2 weeks before EAE induction (preconditioning), at EAE induction (prophylactic), or at the onset of symptoms (late intervention). Mice were scored daily for motor-sensory impairments using a mobility scale (0=normal to 5=death). Visual acuity was assessed using optokinetic responses (OKR). Visual structure was assessed by measuring the retinal nerve fiber layer (RNFL) thickness. Pattern electroretinography (pERG) and visual evoked potentials (VEP) were recorded at the end of the experiment. Tissue from the retinas, optic nerves, brain, and spinal cord were sectioned for subsequent analysis with microscopy. All data were analyzed using one- and two- way ANOVA followed by post hoc tests.
Results: EAE animals from the preconditioned KD group showed significantly worse motor-sensory impairment relative to EAE controls (Area under curve EAE score: EAE: 58±2, pre KD: 68±3, p<0.001; pro KD: 60±3, p=0.09; late KD: 58±3). Similarly, visual acuity data showed worse OKR tracking in the preconditioned KD group (0.23±0.05 cycles/degree) compared to naïve (0.38±0.03 cycles/degree, p<0.0001) and EAE controls (0.26±0.05 cycles/degree; p=0.024). There was no significant difference between the standard diet EAE controls and the pro KD (0.24±0.06 cycles/degree) and late KD group (0.25±0.05 cycles/degree), but all three EAE groups had significantly lower visual acuity and EAE scores than the naïve group (p<0.0001). Compared to naïve mice, average RNFL thickness decreased significantly in all EAE induced mice (naïve: 69±2µm vs. EAE: 66±4µm; p=0.001, pre KD: 66±4µm; p=0.002, pro KD: 67±4µm; p=0.02, late KD: 66±3µm; p=0.001) whereas differences between these EAE induced groups were not significant.
Conclusions: This study identified that implementing KD negatively influenced motor-sensory and visual function, and that preconditioning the mice with KD before EAE induction resulted in the worst structural and functional outcome. These data suggest that KD should not be recommended for patients with MS.
Poster #7
In vivo investigation of mia2/mia3 in protein transport within zebrafish photoreceptors
Donohue, Michael1; Clark, Eric1; Nonarath, Hannah1; Collery, Ross1; Link, Brian1
1Department of Cell Biology, Neurobiology and Anatomy, Medical College of Wisconsin, Milwaukee, WI
Purpose: From a mutational screen, we identified two related factors essential for maintaining age-related photoreceptor health in zebrafish, Mia2 and Mia3. Of significance, mutations in human and dog MIA3 have very recently been shown cause retinopathies. How the proteins Mia2 and Mia3, whose known role is assisting protein transport, affect photoreceptor health is uncharacterized. MIA2/MIA3 have been described to act at ER exit sites to assist large protein transport from the ER to Golgi. In addition to large protein transport, unconventional protein trafficking that circumvents the Golgi occurs in photoreceptors through poorly defined mechanism(s). We hypothesize Mia2/Mia3 serve unique functions within photoreceptors to accommodate trafficking of large proteins, but also functions in the unique unconventional trafficking pathway. To test this hypothesis, Mia2/Mia3 will be investigated through gene manipulation followed by detailed phenotype analysis.
Methods: mia2 and mia3 mutants were created using Crispr/Cas9 genomic deletions. Transgenic Markers 5XATF6RE:d2GFP and gnat2:Lc3-eGFP-mcherry were used to quantify ER stress and autophagy respectively. Electron microscopy was utilized to characterize photoreceptor ultrastructure and OCT was used to detect altered photoreceptor morphology within mia2/ mia3 Mutants.
Results: Zebrafish mia2 mutants showed neuronal defects including photoreceptor degeneration. Consistent with problems with protein transport mia2 mutants showed increased ER Stress. Electron microscopy revealed defects in photoreceptor ultrastructure, presenting with ER vesiculation, autolysosome vacuolation, and stunted outer segments. Additionally, OCT imaging of mia2 mutants revealed a reduction of photoreceptors that decreased further as the fish aged into adulthood. Interestingly and in contrast to mia2, mia3 mutants alone did not display any photoreceptor defects. However, double mia2/mia3 mutants showed exacerbated phenotypes of mia2 mutants.
Conclusions: mia2 mutants display photoreceptor defects while mia3 mutants do not show photoreceptor defects. During the completion of this research a new isoform of mia3; Mia3s was described. The mutation created for mia3 currently only affects the Mia3l isoform. Research is underway to generate mia3 mutants which knock out Mia3s and both Mia3s and the Mia3l concurrently. In addition to detailed phenotype analyses on these newly generated mutants, we will identify binding partners of Mia2/3s/3l through in vivo proximity proteomics.
Poster #9
Labeling and enrichment of retinal proteins using non-canonical amino acids
Inamdar, Shivangi1; Laird, Joseph1; Baker, Sheila1
1University of Iowa Department of Biochemistry and Molecular Biology
Purpose: Understanding the spatial and temporal expression pattern of different proteins in each of the ~100 retinal cell types provides essential information about the dynamic regulation of diverse processes in the retina under normal and diseased conditions. However, analysis of the retinal proteome is limited by the inability to apply a single type of chemistry to label proteins for subsequent purification and analysis. Bio-Orthogonal Non-Canonical Amino acid Tagging (BONCAT) may provide a solution to that problem. The goal of this project is to test the efficiency of applying BONCAT to the mouse retina.
Methods: Azidohomoalanine (AHA) is a non-canonical amino acid with an azide functional group that is used as a methionine substitute. AHA, or PBS as the negative control, was delivered to C57Bl6/J mice (male and female; age 1-6 months) by intraperitoneal injection. Mice were fed either normal chow or a low methionine diet. A range of AHA doses (2-10 mg per 20 g body weight) were delivered for 1-4 days. Subsequently, retinas were harvested and stored at -80C. Proteins from stored retinas were extracted using a lysis buffer containing a broad-spectrum protease and phosphatase inhibitor cocktail (Roche). AHA containing proteins were labeled using a copper-dependent click reaction (Click-&-Go Protein Reaction buffer kit, Click Chemistry Tools) with alkyne-biotin. Proteins were methanol-chloroform precipitated prior to analysis by western blotting, with or without concentration on neutravidin beads. Retina function was tested using ERG.
Results: Western blotting using a streptavidin antibody demonstrated labeling of biotin-clicked retina proteins in AHA injected mice with low non-specific labeling in PBS injected mice. Placing mice on a low methionine diet throughout the experiment increased the amount of labeled protein. An AHA dose of 2 mg per 20 g body weight is sufficient. Overall labeling intensity increased with the number of daily AHA injections. This dose did not cause acute retina toxicity since there was no change in ERG.
Conclusions: Systemically delivered AHA is robustly incorporated into mouse retina proteins. Click chemistry can be used to covalently attach biotin tags for subsequent proteomic analysis.
Poster #11
Corneal sensitivity in chronic orbital pain treated with peripheral trigeminal nerve blocks
Lee, Grace1; Nellis, Julie2,3; Pham, Chau2; Kardon, Randy2,3; Shriver, Erin M.2
1University of Iowa Carver College of Medicine; 2Department of Ophthalmology and Visual Sciences, University of Iowa; 3VA Center for the Prevention and Treatment of Visual Loss, Iowa City, IA
Purpose: To determine corneal sensitivity in chronic orbital pain patients, risk for corneal hypoesthesia associated with peripheral trigeminal nerve blocks, and whether corneal sensitivity correlates with the subjective complaints of orbital pain patients.
Methods: Chronic orbital pain patients treated with unilateral peripheral trigeminal nerve blocks and controls were enrolled in the IRB-approved prospective trial. Pain subjects completed Cochet-Bonnet (CB) aesthesiometer evaluations prior to injection, 30 minutes post-injection, and at 2-4 weeks follow-up. Controls underwent CB evaluations once. All subjects completed the Headache Impact Test (HIT-6), the Leeds Assessment of Neuropathic Symptoms and Signs (LANSS), and the Utah Photophobia Symptom Impact Scale (UPSIS-12) for subjective evaluations of their headache, neuropathic pain, and photophobia.
Results: Mean baseline corneal sensitivity thresholds of the injection and fellow eyes of 6 orbital pain patients were 5.75 ± 0.42 cm and 5.83 ± 0.41 cm, respectively, with no significant inter-eye asymmetry (p=0.7). The mean sensitivity thresholds for the 11 controls were 5.95 ± 0.11 cm (OD) and 5.94 ± 0.15 cm (OS) (p=0.8). The baseline thresholds for pain patients were not significantly different from those of controls (p=0.3). The mean sensitivity threshold of the injection eyes was 5.78 ± 0.25 cm at 30 minutes post-injection (p=0.9) and 5.92 ± 0.20 cm at follow-up (p=0.5), similar to baseline. A higher neuropathic pain score was significantly correlated with lower corneal sensitivity after adjusting for age when evaluating all subjects (R = -0.7, p = 0.005), but the correlation was not significant when evaluating the pain group alone (R = -0.8, p = 0.1). Headache and photophobia symptoms did not correlate significantly with corneal sensitivity.
Conclusions: Chronic orbital pain patients do not display corneal hyperesthesia. Peripheral trigeminal nerve blocks do not alter corneal sensitivity. Neuropathic symptoms reflective of trigeminal neurocutaneous hypersensitivity may correlate with corneal hypoesthesia.
Poster #13
Methylenetetrahydrofolate reductase expression in human diabetic retinas
Patel, Serena1; Barner, Mitch1; Shubert, Brynn1; Markand, Shanu1
1A.T. Still University- Kirksville College of Osteopathic Medicine, Kirksville, MO
Purpose: Methylenetetrahydrofolate reductase (MTHFR) is a key enzyme responsible for remethylating folate, a vital cofactor necessary for converting homocysteine to methionine. When the MTHFR gene is mutated, homocysteine, a non-proteinogenic amino acid, begins to accumulate. Increased homocysteine levels have been associated with oxidative stress. These high levels have been linked to several retinal diseases, such as diabetic retinopathy. Previously, it has been found that MTHFR expression is associated with the function of retinal pigment epithelium. The health of the retina is largely dependent on the role of folate. When that pathway is disturbed, the proper function of the retina is at risk. Research quantifying the expression of MTHFR in human eyes with diabetic retinopathy has not been studied. This study aims to assess the retinal expression of MTHFR protein in Lewis rat controls, normal human eyes, and human diabetic retinas.
Methods: Analysis of MTHFR protein expression was performed using immunohistochemical analysis (IHC). Samples included a Lewis rat eye as a control, and human eye samples from National Disease Research Interchange. The human eye samples were from both normal and diabetic retina cross sections. These sections were stained with MTHFR and PDGFR-β primary antibodies. This was followed by labeling with secondary antibodies anti-mouse for PDGFR-β and anti-rabbit for MTHFR. Primary antibody was not added to the rat control sample. DAPI was used as a nuclear stain and mounting medium. Analysis to quantify fluorescence intensity was measured using ImageJ software. Background signal was accounted for by using the equation for corrected total cell fluorescence.
Results: IHC analysis revealed robust expression of MTHFR staining throughout the retina with higher expression in the photoreceptor outer layer in both rat and human retinas. The Lewis rat control image revealed a value of 58 a.u. In comparison, the normal human eye sample had 266 a.u, whereas the human diabetic retina showed 86 a.u. After quantifying the intensity of fluorescence of MTHFR throughout the retina, the data revealed a decreased intensity in the diabetic retina.
Conclusions: Diabetic retinopathy is the leading cause of blindness in working-aged adults. The importance of identifying pathways that contribute to the progression of the disease is vital. Knowing that MTHFR plays a role in the functional health of the retina allows for a multifactorial role in understanding diabetic retinopathy. This study revealed that MTHFR protein expression is decreased in human eyes with diabetic retinopathy compared to MTHFR protein expression in normal human eyes.
Poster #15
Potential Treatment of Elevated Homocysteine on Retinal Pigmented Epithelial Cells
Schubert, Brynn1; Markand, Shanu1
1A.T. Still University- Kirksville College of Osteopathic Medicine, Kirksville, MO
Purpose: Homocysteine is a sulfur containing non-proteinogenic amino acid involved in methionine metabolism. Elevated homocysteine levels is known as hyperhomocysteinemia (Hhcy) and is implicated in several ocular disorders such as glaucoma, diabetic retinopathy, and age related macular degeneration (AMD). Retinal pigmented cells (RPE) are the primary cellular target for pathologic changes in AMD. Hhcy is known to induce reactive oxygen species production, which can cause direct injury to proteins, lipids, and nucleic acids- leading to cell death. Melatonin is a hormone important for circadian rhythms, but is also an antioxidant. Limited data in the literature exists regarding the role of melatonin as a potential treatment for Hhcy. The purpose of the current study was to evaluate the therapeutic potential of melatonin in Hhcy using the RPE cell culture model system.
Methods: ARPE-19 cells were grown in DMEM F-12 media supplemented with 10% FBS and 1% Penicillin/Streptomycin. MTT assay, Apoptosis/Necrosis Assay, and oxidative stress assays were used to determine manor of cell death of ARPE-19 cells after treatment with homocysteine (1, 5, 10, 50, 100, 500, 1000 µM) and melatonin (5 µM). Changes in cellular morphology were determined by confocal scanning microscope images taken pre-treatment, 24, 48, and 72 hours post treatment. Hydrogen Peroxide acted as the positive control for cellular death. Data was analyzed using a one-way ANOVA. Statistics were analyzed using the GraphPad Prism 9 system.
Results: As concentrations of homocysteine increased, instances of ARPE-19 cell viability decreased with significant values (p<0.001) at 50, 500, and 1000 µM homocysteine. Cells treated with melatonin or co-treated with homocysteine and melatonin resulted in increased cell viability. Cells treated with homocysteine resulted in cellular death via apoptosis, however, cells co-treated with melatonin showed a decrease in cellular death via apoptosis. Cells treated with melatonin showed a decrease in green fluorescence- corresponding to a decrease in oxidative stress, when compared to the positive control. Abnormal cellular morphology was observed in cells treated with homocysteine at concentrations over 50 µM, and conditions worsened at longer treatment duration.
Conclusions: By treating ARPE-19 cells with homocysteine, we found that homocysteine leads to cell death via apoptosis at concentrations above 50 µM. We also found that APRE-19 cells co-treated with homocysteine and melatonin resulted in an increase in cell proliferation, and that melatonin shows protective effects against oxidative stress. This demonstrates that melatonin has potential protective effects against elevated homocysteine.
Poster #17
Increased T-regulatory cell activity protects retinal ganglion cells in glaucoma
Xian, Shuyu1; Zeng, Huilan1,2; Kuehn, Markus H.1,2
1Department of Ophthalmology and Visual Sciences, University of Iowa; 2VA Center for the Prevention and Treatment of Visual Loss, Iowa City, IA
Purpose: Our previous studies have shown that experimental glaucoma in mice activates T cell-driven autoimmune response that can further contribute to vision loss. Regulatory T cells (Tregs) are a specialized subset of T cells that suppress immune responses, thereby maintaining homeostasis and self-tolerance. Here, we sought to explore whether induction of increased T-regulatory cell (T-reg) activity ameliorates vision loss in mice with glaucoma.
Methods: Lymphocyte activation gene 3 (Lag3), also known as CD223, negatively regulates T cell activation, proliferation, and cytokine production. We induced ocular hypertension (OHT) in Lag3fl/fl.Foxp3CreERT2-GFP mice to establish a chronic glaucoma model. Intraperitoneal injection of tamoxifen was used to conditionally remove the Lag3 gene Treg cells of adult mice to increase their abundance and activity. Mice were divided into 5 groups: naïve (n=10), tamoxifen without OHT (n=7), Untreated (OHT without tamoxifen, n=8), Prevention (tamoxifen before OHT, n=9) and Treatment (tamoxifen after OHT, n=10). Visual function was assessed monthly by optokinetic response testing. 12 weeks after OHT induction, optic nerve damage was determined by p-phenylenediamine (PPD) staining.
Results: Tamoxifen injection increased the proportion of CD4+ Treg cells detected in peripheral lymph nodes (naïve vs tamoxifen injected, 14.6%vs18.8%, p=0.0287) and the spleen (naïve vs tamoxifen injected, 10.8%vs15.5%, p=0.0043). 12 weeks after induction of OHT, mice in the Treatment group showed higher visual function when compared to those in the Untreated group (0.366 c/d vs. 0.322 c/d, p=0.027). However, mice in the Prevention group did not perform significantly better than those in the Untreated group (0.336 c/d, p=0.89). Histologically, significantly fewer damaged axons were observed in the optic nerves of the Treatment group when compared to the Untreated group (9.1±6.2 vs. 20.4±14.2, p=0.036). Animals in the Prevention group displayed slightly less damage than untreated mice, but statistical significance was not reached (13.4±5.4, p=0.3532).
Conclusions: These results demonstrate that increasing Treg cell activity can reduce experimental glaucoma damage in mice. Our findings suggest that immunomodulatory therapy aimed to minimize the impact of T-cell mediated damage in glaucoma may be beneficial for some patients.
Poster #19
A Dose Response Study for Subretinal Gene Therapy for Treating X-Linked Juvenile Retinoschisis in a Retinoschisin 1 Knockout Mouse Model
Hassan, Salma1,2,3; Hsu, Ying1,2; Thompson, Jacob1,2; Drack, Arlene V.1,2,3
1University of Iowa Department of Ophthalmology and Visual Sciences, 2University of Iowa Institute for Vision Research; 3University of Iowa Cell and Developmental Biology Program
Purpose: X-linked Retinoschisis (XLRS) causes macular degeneration due to loss-of-function of the retinoschisin 1 (RS1) gene, causing reduced visual acuity. A human clinical trial for treating XLRS showed that high-dose intravitreal gene therapy led to ocular inflammation without phenotype improvement. Subretinal gene delivery's superiority over intravitreal approach for treating RS1 remains unclear. Therefore, we utilized a novel adeno-associated viral vector (AAV) for subretinal RS1 gene delivery in a pre-clinical Rs1-KO mouse model. This AAV2tYF vector contains three tyrosine to phenylalanine (Y-to-F) mutations which have been shown to improve its transduction efficiency. This vector is FDA-approved for clinical applications. To determine the optimal dose range for treating XLRS via subretinal delivery, we conducted a dose-response study using this vector to deliver the RS1 gene in mice. Long-term effects of retinal rescue after rAAV2tYF-CB-hRS1 subretinal gene delivery were investigated.
Methods: Rs1-KO (Rs1-/y and Rs1-/- ) mice were subretinally injected with 2 µL of rAAV2tYF-CB-hRS1 vector with different doses at postnatal days 24-31. Rs1-KO mice were injected with different doses, receiving either the highest dose of 6E9 viral genomes (vg)/eye (n=5), 6E8 vg/eye (n=5), 6E7 vg/eye (n=5), or injection of the diluent alone as a sham injection (Alcon BSS + 0.014% (v/v) Tween buffer), (n=6) and 22 untreated eyes. Analysis of retinal function by electroretinography (ERG) and structural analysis by measuring cyst severity and outer nuclear layer thickness using optical coherence tomography (OCT) were performed on all mice 1 to 3 months post (MP) injection. Functional vision of treated mice was evaluated using a visually guided swim assay. Rod- and cone-dependent visual pathways as measured by ERG and severity of retinal cysts of vector-treated mice were compared to those of diluent-treated mice (sham control group) and completely untreated mice (treatment naïve group).
Results: At 1 MP treatment, treated Rs1-KO eyes with the 6E8 vg/eye dose had a strong response to the 3.0 cd·sec/m2 bright flash, with higher b-wave values and showed a significant improvement in the amplitudes of the 5-Hz flicker ERG compared to both sham-treated and untreated eyes. This improvement was also observed at 2 and 3 MP treatment. These findings show that subretinal gene delivery was able to reestablish the electrical function in the cone photoreceptor cells in ERG. High and low-dose groups did not improve retinal function in rods or cones pathways, despite reduced cyst severity scores. At all doses, cyst severity scores were significantly reduced compared to the untreated eyes but not to the sham-injected eyes. These observations indicate that the ability to rescue retinal function by gene augmentation in XLRS is highly dose-dependent and that the observed improvements in ERG are not entirely accounted for by the amelioration of cysts in this disease.
Conclusions: In conclusion, we conducted a dose-response study to determine the optimal dosing for treating RS1 using subretinal gene therapy. Based on the results we concluded that subretinal gene therapy with a dose of 6E8 vg/eye improves retinal function and structure by sustaining photoreceptor cell viability in treated eyes, restoring the electrical function of cone photoreceptors, and reducing cyst severity scores. However, Cyst severity score results indicate that even a low amount of RS1 protein production in the eye promotes the resolution of schisis cysts. Even though the presence of a low amount of RS1 protein appears to ameliorate schisis in the retina, the rescue of retinal electrical function requires a higher gene dose. The findings in this study provide a vector dose and an approach that facilitates the clinical translation of this therapy and can be extended to human trials.
Poster #21
Automated Optic Disc Finder and Segmentation Using Deep Learning for Blood Flow Studies in the Eye
Takahashi, Noriyoshi1; Wang, Jui-Kai1,2; Ahmad, Noor-Us-Sabah1; Nellis, Julie K.1,2; Garvin, Mona K.2,3; Linton, Edward F.1; Kardon, Randy H.1,2
1The University of Iowa Department of Ophthalmology and Visual Sciences; 2VA Center for the Prevention and Treatment of Visual Loss; 3The University of Iowa, Department of Electrical and Computer Engineering
Purpose: In-vivo imaging with laser speckle flowgraphy (LSFG) enables optical measurement of an index of blood flow in the retina. LSFG can help observe blood flow changes in various diseases, including optic nerve problems such as glaucoma, ischemic optic neuropathy, and others. The purpose of this study is to develop an automated method for the identification of the optic disc in the LSFG images. Automated identification of the optic disc in LSFG images is particularly challenging because of limited contour information in the rendered blood flow maps. In this study, we adopted a state-of-the-art U-Net approach (nnU-Net) to automatically segment the optic disc region based on the paired LSFG blood flow image and infrared light intensity map.
Methods: One hundred subjects (training/test dataset ratio: 70/30) were used in this study. A trained neuro-ophthalmologist (Expert 1) traced the optic discs in color fundus photographs from the same eye, and then these masks were registered into the LSFG domain. The nnU-Net was trained to identify the optic disc just based on the LSFG blood flow composite and infrared light intensity images. After training, we compared the difference between nnU-Net's output and Expert 1 with the difference between Expert 1 and a second clinician (Expert 2) in the test dataset.
Results: The mean (and standard deviation) Dice coefficient was 0.92 ± 0.04 between the nnU-Net and Expert 1’s tracing and 0.89 ± 0.06 between Expert 2’s and Expert 1’s tracings, respectively. The mean intersection over union (IoU) index was 0.86 ± 0.06 between the nnU-Net and Expert 1’s tracing and 0.81 ± 0.09 between Expert 2’s and Expert 1’s tracings. Both evaluation metrics showed the nnU-Net's predictions were significantly closer to Expect 1's tracing than Export 2's (p-values < 0.001).
Conclusions: This study demonstrated that the nnU-Net’s predictions for optic disc segmentation were significantly closer to Expert 1’s manual tracings compared to Expert 2’s tracings. In summary, having a robust optic disc segmentation in LSFG can reduce tedious manual tracing and will also be a foundation for future developments of automated region-based feature extraction in LSFG and potentially for other imaging modalities.
Poster #23
Antioxidant treatment suppresses retinal defects in Drosophila models of SNRNP200-associated Retinitis Pigmentosa
Mayer, Sara1; Christensen, Quinton2; McCoy-Munger, Hailey2; Drack, Arlene1; Wallrath, Lori2
1University of Iowa, Department of Ophthalmology and Vision Sciences, Iowa City, IA; 2University of Iowa, Department of Biochemistry and Molecular Biology; Iowa City, IA
Purpose: Retinitis pigmentosa (RP) is a collection of genetic retinal degeneration disorders that affects 1:4,000 individuals worldwide. RP initially presents with a loss of rod photoreceptors, followed by a loss of cones. The disease is progressive and can result in complete blindness. RP33 is a non-syndromic form of autosomal dominant RP caused by mutations in the SNRNP200 gene encoding a core pre-mRNA splicing factor. The pathological mechanisms of RP are not well understood; however, there is evidence of oxidative stress in the retina. Our goal is to understand how mutations in SNRNP200 causes retinal defects and identify potential treatments.
Methods: To determine how mutations in SNRNP200 cause retinal defects, we developed Drosophila models possessing mutations identified in humans with RP. Drosophila has an orthologue of SNRNP200 that we refer to as dSnrnp200. Human and Drosophila SNRNP200 exhibit 74% amino acid identity and 89% similarity. CRISPR mutagenesis was used to introduce mutations into Drosophila dSnrnp200. These mutations include a de novo mutation called Bigfoot because it is a private mutation. The Bigfoot mutation causes amino acid substitution T731I. In addition, a mutation that causes S1087L was generated. RNAi was also used to knock-down dSnrnp200 in differentiated cells of larval eye imaginal discs, which give rise to the adult eye. Retinal defects were detected in in dSNRNP200 mutants and upon depletion. To determine if oxidative stress contributed to these defects, we fed flies either N-acetyl cysteine (NAC) or water as a control. NAC is an FDA approved antioxidant.
Results: Flies with the Bigfoot mutation are homozygous lethal. By contrast, flies with the mutation that causes S1087L are homozygous viable. Both dSnrnp200 mutations increase apoptosis in larval eye imaginal discs, relative to controls. The mutant larvae develop into adults that have abnormal electroretinograms, but no other apparent defects. The photoreceptors in these mutants show an abnormal pattern of organization which was partially corrected upon treatment with NAC. Depletion of dSnrnp200 increases apoptosis in larval eye imaginal discs and produced a rough eye phenotype in adults. NAC treatment decreased apoptosis in the larval eye discs and showed a dosage-dependent suppression of the rough eye phenotype.
Conclusions: Taken together, these data demonstrate that the Drosophila models of SNRNP200-associated RP recapitulate aspects of the human disease. Currently, NAC is being used in a clinical trial for individuals with RP; however, it is not known if individuals with SNRNP200 mutations benefit. Our findings strongly suggest that individuals with SNRNP200-associated RP will benefit from NAC treatment and that oxidative stress plays a key role in the pathogenesis.
Poster #25
Role of SIN3A dysregulation in neural crest differentiation and Witteveen-Kolk syndrome
Al-Kaylani, Hend1; Cheng, Lin2,3; Dumitrescu, Alina V.2; Kuehn, Markus H.2,3
1University of Iowa Carver College of Medicine; 2University of Iowa Department of Ophthalmology and Visual Sciences; 3VA Center for the Prevention and Treatment of Visual Loss, Iowa City, IA
Purpose: Childhood-onset glaucoma is commonly a manifestation of multi-organ syndromes, such as Witteveen-Kolk syndrome (WITKOS). WITKOS was defined in 2016 and is due to a loss of function mutation in SIN3A, which encodes a histone deacetylase scaffold and is critical to neural development. While ocular dysgenesis is not a common feature of WITKOS, a parent and two children presented to UIHC with early-onset glaucoma and a confirmed SIN3A mutation. We aimed to quantify SIN3A gene expression at the neural crest stage of anterior segment development. We also aimed to assess the gene’s potential role in the establishment of neural crest identity.
Methods: Urine epithelial cells were collected non-invasively from all three family members and primary cultures were established; SIN3A mutations were confirmed by Sanger sequencing. Neural crest-like status was induced in human induced pluripotent stem cells (hiPSC) using a specialized growth factor medium. SIN3A gene expression was assessed using real-time quantitative polymerase chain reaction (RT-PCR) in hiPSCs and the induced neural crest cells. Flow cytometry against CD49d and p75/NGFR was used to evaluate neural crest identity.
Results: Primary cell cultures were established from cells collected from the patients’ urine samples. SIN3A mutations were confirmed in DNA extracted from patient cells. Exposure of hiPSCs to an induction medium resulted in a relative increase in expression of markers of neural development.
Conclusions: This project evaluates SIN3A expression in induced neural crest cells to characterize anterior segment development in WITKOS. First, we found that collecting cells from patient urine is a viable approach to establishing primary cell lines and therefore allows a non-invasive method to collect human samples. Second, hiPSCs can be differentiated into neural crest using specialized media. These techniques can be used to investigate molecular pathways disrupted by genetic or epigenetic mutations, as well as to model early ocular dysgenesis. Future goals include defining the changes in Sin3a protein levels after translational inhibition and utilizing CRISPR-Cas9 to create loss of function mutations for functional studies. These experiments will characterize the timeline of anterior segment dysgenesis and/or dysregulation in Witteveen-Kolk syndrome and other syndromic causes of childhood glaucoma. In turn, these findings can inform potential gene therapy treatments as well as clinical timepoints to regulate intraocular pressure in pediatric glaucoma.
Individuals with disabilities are encouraged to attend all University of Iowa sponsored events. If you are a person with a disability who requires a reasonable accommodation in order to participate in this program, please contact Dr. Kuehn in advance at 319-335-9565.
Last Updated: