**1. Introduction**

Diabetic retinopathy (DR) is a neurovascular complication of diabetes mellitus and the leading cause of blindness in working age adults [1]. Diabetic retinal microangiopathy significantly contributes to DR pathogenesis and strategies targeting its occurrence and progression are important for preventing vision loss in affected patients [2,3].

Hyperglycemia-induced retinal vascular pathology is a multi-step process characterized by retinal endothelial cell dysfunction and death, increased vascular permeability leading to diabetic macular edema (DME), and abnormal retinal neovascularization, as seen in proliferative diabetic retinopathy (PDR) [2,3]. Prolonged hyperglycemia affects the retinal microvasculature by altering multiple molecular pathways involving redox imbalance and induction of pro-inflammatory responses [3].

Previously, we showed that hyperglycemia-induced oxidative/nitrative stress accelerates retinal endothelial cell senescence and that this is an important early pathogenic event during the development of diabetic retinal microangiopathy [4]. Retinal oxidative/nitrative stress results from increased production of reactive oxygen and nitrogen species (ROS and RNS, respectively) from cellular and mitochondrial oxidases [5,6] as well as from loss of endogenous antioxidant activities [7,8].

Histone deacetylase 6 (HDAC6) is a class IIb histone deacetylase known to exert important biological functions due to its ability to regulate the acetylation state of nuclear and cytoplasmic proteins [9]. Most known targets of HDAC6 are cytoskeletal proteins, transcription factors [9] and endogenous antioxidants such as peroxiredoxin 1 (Prx1) [10,11]. As a consequence of its multiple biological targets, altered HDAC6 expression and activity is linked to inflammation, oxidative stress and the pathogenesis of a number of neurodegenerative [12] and cardiovascular disorders [13] as well as cancer [14].

The potential impact of HDAC6 on diabetic microvascular complications is understudied. Previous studies have shown beneficial e ffects of HDAC6 inhibition in diabetic kidney disease [15] and diabetic heart disease [16]. Moreover, while little is known on HDAC6 contribution to retinal diabetic disorders, studies have implicated pro-oxidative e ffects of HDAC6 in animal models of retinal neurodegenerative diseases [17,18], potentially suggesting a role for this histone deacetylase in retinal pathologies involving oxidative stress, such as DR.

Based on this evidence, we assessed the e ffects of diabetes on HDAC6 retinal expression and activity in human and experimental DR and we investigated the molecular mechanisms involved in this process in a rat model of type 1 diabetes (streptozotocin-induced diabetic rat = STZ-rat) and in cultures of human retinal microvascular endothelial cells exposed to glucidic stress.

#### **2. Materials and Methods**

#### *2.1. Human Postmortem Samples*

De-identified human postmortem retinal samples were obtained from Georgia Eye Bank, Inc. (Alanta, GA, USA). Retinas were obtained from a total of 8 diabetic and 8 non-diabetic donors that were selected based on DR history or lack of reported ocular pathologies (control). Supplementary Table S1 summarizes the demographic and clinical history information available for all donors.

#### *2.2. Animals and Treatment*

All animal experiments strictly adhered to the Statement of the Association for Research in Vision and Ophthalmology (ARVO) for the humane use of laboratory animals for ophthalmological research and to Augusta University approved protocols (#2009-0181). All animals were housed in the vivarium of Augusta University with a 12-h day/night light cycle with light intensity in the room maintained at 130 lux at cage level, and fed ad libitum. Adult male Sprague–Dawley (SD) rats (250−300 g) obtained from Envigo (Indianapolis, IN, USA) were made diabetic by a single intraperitoneal injection of streptozotocin (STZ) (55 mg/kg dissolved in 0.1 mol/L sodium citrate, pH 4.5) (Sigma-Aldrich, St. Louis, MO, USA). Control rats received injections of vehicle alone. Rats with fasting blood glucose ≥250 mg/dL were considered diabetic. A group of STZ-rats was injected intraperitoneally with 10 mg/kg Tubastatin A (TS) (MedChemExpress, Monmouth Junction, NJ, USA) starting two weeks after the STZ injections and continuing every other day for the next six weeks. Control rats received vehicle phosphate-bu ffered saline (PBS) injection. The diabetic rats were sacrificed by an overdose of anesthesia (ketamine 200 mg/kg and xylazine 60 mg/kg) followed by thoracotomy. Blood glucose levels were measured by ReliOn prime blood glucose monitoring system (Bentonville, AR, USA) and glycated hemoglobin A1c (HbA1c) was measured using A1c Now+ System (PTS Diagnostic, Winter Park, FL, USA). A number of other metabolites were also monitored in diabetic and control rat plasma using a biochemistry panel analyzer (Piccolo Xpress analyzer, Princeton, NJ, USA). Rats metabolic profiles in response to the di fferent treatment protocols are reported in Supplementary Table S2.

#### *2.3. Cells and Treatment*

Human retinal endothelial cells (HuREC) were purchased from Cell Systems Corporation (Kirkland, WA, USA) and maintained using complete medium with normal glucose formulation (Cell Systems) at 37 ◦C and 5% CO2 in a humidified atmosphere. All the experiments were carried out using HuREC between passages 3 to 7 and all the tissue culture flasks/plates were pre-coated with attachment factor (Cell Systems). The cells were switched to serum-free medium (Cell Systems) 10 to 12 h before the experiments. To mimic the e ffects of the hyperglycemia (glucidic stress), HuREC were cultured for 48 h in serum-free medium containing 25 mM d-glucose (high glucose, HG). Similarly, control cells were cultured in serum-free and normal glucose medium (5.5 mM d-glucose, NG). Cells grown in normal glucose media with the addition of l-glucose (5.5 mM d-glucose + 19.5 mM l-glucose, LG) served as an osmotic control. Some of the HuREC were treated with 5 μM TS. To determine TS toxicity towards HuREC, a dose-response curve using MTT cell viability assay (Abcam, Cambridge, MA, country) was conducted following the manufacturer's instructions. MTT dye absorbance was read using a microplate reader at 540 nm (Supplementary Figure S1).
