Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats

YTAAP-13495; No of Pages 9 Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Toxicology and Applied Pharmacology journal homepage: www.elsevier.com/locate/ytaap

Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats Wenhu Huang ⁎, Walter Collette III, Michelle Twamley, Shirley A. Aguirre, Aida Sacaan Drug Safety Research and Development, Pfizer Global Research and Development, San Diego, CA 92121, USA

a r t i c l e

i n f o

Article history: Received 24 July 2015 Revised 13 October 2015 Accepted 14 October 2015 Available online xxxx Keywords: Electroretinography Pan-cyclin dependent kinase inhibitors Retina Toxicity Rats

a b s t r a c t Retinal ocular toxicity is among the leading causes of drug development attrition in the pharmaceutical industry. Electroretinography (ERG) is a non-invasive functional assay used to assess neuro-retinal physiological integrity by measuring the electrical responses. To directly assess the utility of ERG, a series of studies was conducted following intravitreal and/or iv administration of pan-cyclin-dependent kinase inhibitors: AG-012,986 and AG024,322 in rats. Both compounds have previously shown to induce retinal toxicity. Retinal injury was evaluated by ERG, histopathology and TUNEL staining. Intravitreal injection of AG-012,986 at ≥ 10 μg/eye resulted in decreases (60%) in ERG b-wave and microscopic changes of mild to moderate retinal degeneration, and at 30 μg/eye led to additional ophthalmic findings. Intravenous administration of AG-012,986 daily at ≥5 mg/kg resulted in dose-related decreases (25 to 40%) in b-wave and sporadic to intense positive TUNEL staining. Intravitreal injection of AG-024,322 at 30 μg/eye also resulted in decreases (50 to 60%) in b-wave, mild to marked retinal degeneration and mild vitreous debris. These experiments demonstrate that ERG can be used as a sensitive and reliable functional tool to evaluate retinal toxicity induced by test compounds in rats complementing other classical ocular safety measurements. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Data summarized from Pfizer's internal drug development database revealed that ocular toxicity accounts for approximately 7% of therapeutic candidate attrition, the 4th most common cause of attrition after cardiovascular, liver, and renal toxicity. Among the aforementioned ocular toxicity, 99% were attributed to retinal toxicity. Sensory retina is the most complex of all ocular tissues, composed of multiple lightsensitive neuronal layers (nerve fiber, ganglion, bipolar, photoreceptor layers, plus retinal pigment epithelium lining the inner surface of the eye (Yang and Huang, 2012). Retinal exposure to xenobiotics/chemical compounds has the potential to lead to interruption of visual signal transmission or changes in structure morphology, resulting in retinal toxicity (Huang et al., 2009).

Abbreviations: ERG, electroretinography; pan-CDK inhibitor, pan-cyclin dependent kinase inhibitor; CDK, cyclin-dependent kinase; APAP, N-acetyl-para-aminophenol/acetaminophen; Hz, hertz; H&E, hematoxylin and eosin; TUNEL, terminal deoxynucleotide transferase (TdT)-mediated dUTP nick end labeling; ANOVA, analysis of variance analysis; PRL, photoreceptor layer; OPL, outer plexiform layer; ONL, outer nuclear layer; INL, inner nuclear layer; GCL, ganglion cell layer; RPE, retinal pigment epithelium; NFL, nerve fiber layer; OLM, outer limiting membrane; NSF, no significant findings; SID, once daily; BID, twice daily. ⁎ Corresponding author at: Pfizer Global Research and Development, La Jolla Laboratories, 10777 Science Center Drive, San Diego, CA 92121, USA. E-mail address: wenhu.huang@pfizer.com (W. Huang).

Current practices to assess ocular toxicity consist of in vitro and in vivo approaches. In vitro retinal toxicity screening assays, using retinal cell lines, mixed cells, or organotypic/tissue cultures, are relatively less costly and can be performed in high throughput. However, the predictivity of these assays remains debatable. By contrast, in vivo animal model screening is more translatable and definitive for retinal toxicity evaluation. Therefore in vivo animal testing is considered a standard step in preclinical safety evaluation in drug development, and is well accepted by regulatory agencies. The most common available in vivo ocular safety assessment tools are ophthalmic examinations, including ophthalmoscopy, slit-lamp, fluorescein angiogram, and histopathology, all of which are focused on the evaluation of structural/morphologic changes in the retina. Electroretinography (ERG) on the other hand, is a non-invasive test used to assess the integrity of neuro-retinal physiology by measuring the electrical responses of various retinal cell types. The ERG waveform represents the electrical response that is generated by the entire retina when stimulated by a brief stimulus of light. The light stimulus elicits a biphasic waveform recordable at the cornea. The two components of the ERG which are most commonly measured and reported are the aand b-waves. The a-wave is the first large negative component produced by the photoreceptors (cones and rods) within the retina. The b-wave is a large positive component generated by the bipolar cells within the inner nuclear layer of the retina. Together the two waves comprise the main portion of the ERG waveform and represent a

http://dx.doi.org/10.1016/j.taap.2015.10.008 0041-008X/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

2

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

majority of the function within the retina. When applied at the preclinical stage, ERG may be utilized to evaluate potential ocular toxicity of drug candidates. A class of small molecule compounds with reported ocular toxicities in pre-clinical species is the pan-cyclin dependent kinase (pan-CDK) inhibitors. Cyclin-dependent kinases (CDKs) are a family of serinethreonine kinases which when activated by the cyclin regulatory subunit, control the progression of normal mammalian cells through the cell cycle (Lee and Nurse, 1987). Since inappropriate cell proliferation in cancer was identified to be closely related with overactivity of cell cycle CDKs, de-regulating the specific protein kinases has become widely pursued drug targets in pharmaceutical research. Efforts have been mainly focused on developing potent and safer CDK inhibitors for a variety of cancer indications in the past decade. (Shapiro, 2006; Krystof and Uldrijan, 2010). However, several pan-CDK inhibitors have been reported to cause unexpected retinal and peripheral nerve toxicity in mice and specific photoreceptor layer damage in Cynomolgus monkeys (Illanes et al., 2006; Saturno et al., 2007). One of these pan-CDK inhibitors is AG-012,986 which upon iv administration resulted in retinal degeneration or atrophy in CD-1 mice after 21-days of drug administration (Illanes et al., 2006). Interestingly, compound-related microscopic findings were not detected by routine histology exam in mice. Instead, a more specific staining assay such as the TUNEL assay was further applied to visualize the apoptotic retinal cells (Illanes et al., 2006). Thus sensitive or early in vivo toxicological screening is very important for the development of new drugs, and rats are widely used as one of the standard rodent species and accepted by regulatory agencies as a test species in toxicity studies. To save money and time, developing more robust, sensitive, and noninvasive assays to detect rat retinal toxicity earlier on is therefore beneficial. A retinal functional assay such as ERG is a superlative tool to evaluate ocular toxicity in drug development (Rosolen et al., 2005) and panCDK inhibitors would be ideal bench marker compounds to validate such assays. In the present article, we conducted a comprehensive evaluation of retinal toxicity, using a functional test (ERG), in vivo ophthalmic examinations, and histopathology, in rats treated with pan-CDK inhibitors, AG-012,986 and AG-024,322. The sensitivity and specificity of a fullfield scotopic ERG in detecting retinal toxicity were confirmed and compared to the morphological or structural changes and their relationship to in vivo ophthalmology exams. The effects of intravitreal versus iv administration of test compounds on retinal toxicity in the rat model were also studied. Additionally, the progression of the retinal lesions and possible reversibility were also investigated.

2. Materials and methods Studies were designed to assess the sensitivity and specificity of ERG to detect ocular toxicity in Wistar Han rats administered pan-cyclin dependent kinase inhibitors by intravitreal or iv routes.

2.1. Animals and husbandry In all studies, standard procedures and conditions were applied in a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International. All procedures involving laboratory animals were reviewed and approved by Pfizer's Institutional Animal Care and Use Committee. All studies were conducted in male Wistar Han IGS (CRL:WI [Han]) rats that were supplied by Charles River Laboratories, Inc., (Portage, MI, USA). The rats were 6 to 8 weeks old and weighed 200–250 g at study initiation. Environmental controls for the animal room were set to maintain a temperature range of 20 to 26 °C, relative humidity of 30% to 70%, a routine 12 h light/dark cycle and a minimum of 12 air changes per hour.

2.2. Test articles AG-012,986 (4-((4-amino-5-(2,6-difluorobenzoyl)-1,3-thiazol-2yl)amino)-N-((1R)-2-(dimethylamino)-1-methylethyl)benzamide) and AG-024,322 (N-((5-(3-(4,6-difluoro-1 H-benzo[d]imidazol-2-yl)-1 H-indazol-4-yl)-4-methylpyridin-3-yl)methyl)ethanamine) are chemically distinct, early generation small molecule pan-CDK inhibitors developed at Pfizer. The pan-CDK inhibitors were formulated as a nano-suspension in 2.5% povidone (PVP) K30, 0.2% Tween80 in water at 0.2, 0.6, 2.0 and 6.0 mg/ml or 0.1 M NaCl, 0.1 M acetic acid, pH 4.5 at 0.5, 1.0 and 2.0 mg/ml, for intravitreal (5 μl) or iv administration (dose volume 10 ml/kg), respectively. N-acetyl-paraaminophenol (APAP)/acetaminophen has no reported ocular effects (FDA) and was utilized as a negative control. APAP was supplied by Sigma-Aldrich (Milwaukee, WI) and formulated in 2% polyvinyl alcohol (PVA) as a nano-suspension at 12 mg/ml.

2.3. Ocular toxicity studies Rat study designs are summarized in Table 1. Briefly, AG012,986 was given either by iv or intravitreal injections, while AG024,322 and APAP were administered by intravitreal injection. Groups of male Wistar Han rats were administered AG-012,986 (5 or 10 mg/kg) iv once a day, (10 mg/kg) iv twice a day to achieve better tolerability, or a single intravitreal injection of vehicle, APAP (30 and 60 μg/eye) or AG-012,986 (1, 3, 10 and 30 μg/eye) to both eyes. In preparation for intravitreal dosing, animals were anesthetized using 2.5% isofluorane. Two drops of 2.5% phenylephrine (hydrochloride ophthalmic solution, USP, Akorn Inc., Lake Forest, IL) and 1% tropicamide (Akorn Inc., Lake Forest, IL) were administered, topically, to each eye to facilitate pupil dilation. Immediately prior to the injection, one drop of 0.5% proparacaine hydrochloride (Akorn Inc., Lake Forest, IL) was administered to numb the eyes. The animals were each placed under a surgical microscope and a contact lens was placed on the cornea using a small drop of 0.3% hypromellose (GenTeal, Novartis Pharmaceutial Corp, East Hanover, NJ). Rats were given a single 5 μl intravitreal injection, using a 30 gauge needle attached to a 10-μl Hamilton syringe, approximately 1 mm from the corneal limbus at the lateral canthus of the globe of each eye. The test article was injected slowly and the needle remained in the eye for a few seconds to allow for the dosing solution to equilibrate with the vitreous humor. The needle was then slowly removed so as to minimize the chance of dosing solution/vitreous reflux through the injection site. Topical ophthalmic antibiotic ointment was placed on the ocular surface following the procedure. Ocular signs of toxicity were assessed predose and on days 4, 8, and/or 14 of the study. Ocular irritation was scored in accordance with the Organization for Economic Cooperation and Development 1987 guidelines (OECD, 2012). Ophthalmic examination was conducted with a slit lamp and an indirect ophthalmoscope. Each eye was grossly examined and graded on a scale for the severity of changes using a modified MacDonald–Shadduck Scoring System (Aguirre et al., 2009). Fundus examinations were performed using an indirect ophthalmoscope in eyes with pupils dilated with 2.5% phenylephrine and 1% tropicamide. Intravenous doses were administered daily via the lateral tail vein at a dose volume of 10 ml/kg. The dose volume was based on the most recent individual animal body weight (predose Day 1). The animals were assessed daily for mortality, abnormalities, and signs of pain or distress. Clinical signs of toxicity were assessed daily and body weights were recorded weekly. Ophthalmic examinations were conducted post ERG measurements predose and 4, 8 and/ or 14 days post dose. Body weights were recorded during the pretest period and weekly thereafter (on days 8 and/or 15 postdose).

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

3

Table 1 Summary of study designs for rats administered AG-012,986, AG-024,322, and APAP. Compound

Vehicle A Vehicle B Vehicle C AG-012,986

Route of administration

Intravitreal iv Intravitreal Intravitreal

iv

AG-024,322

Intravitreal

APAP

Intravitreal

Dose

0 μg/eye 0 μg/kg 0 μg/eye 1 μg/eye 3 μg/eye 10 μg/eye 30 μg/eye 5 mg/kga 10 mg/kga 20 mg/kgb 1 μg/eye 30 μg/eye 30 μg/eye 60 μg/eye

N

4 3 3 4 4 6 4 4 4 4 3 3 3 3

No. of eyes

8 6 6 8 8 6 8 8 8 8 6 6 6 6

Test time (days) Ophthalmic examinations

ERG assessmentsc

Necropsy

−1, 2, 8, 15 −1, 8 −1, 8 −1, 2, 8, 15 −1, 2, 8, 15 −1, 4 −1, 14 −1, 14 −1, 14 −1, 15 −1, 8, 14 −1, 8, 14 −1, 8 −1, 8

−1, 2–5, 8–12, 15 −1, 4, 8 −1, 4, 8 −1, 2–5, 8–12, 15 −1, 2–5, 8–12, 15 −1, 4, 8–9 −1, 8–9, 14 −1, 4, 8, 14 −1, 4, 8, 14 −1, 4, 8, 15, 22 −1, 4, 8, 14 −1, 4, 8, 14 −1, 4, 8 −1, 4, 8

15 8 8 15 15 7 15 14 14 5 (Early sacrifice) 15 15 8 8

N = number of animals in the group; Day “−1”: baseline measurements. Vehicle A: 2.5% povidone (PVP) K30, 0.2% Tween80 in water; used when formulating AG-012,986 and AG-024,322 for intravitreal exposure. Vehicle B: 0.1 M NaCl/0.1 M acetic acid pH 4.5; used when formulating AG-012,986 and AG-024,322 for systemic exposure. Vehicle C: 2% PVA; used when formulating APA. a Compound administered once a day (SID). b Compound administered twice a day (BID). c The interval means ERG assessment was staggered during these days.

2.4. Electroretinography Retinal function was evaluated by recording of dark-adapted ERG using the Espion E2 System (Diagnosys LLC, Littleton, MA). Rats were dark adapted for a minimum of 10 h or overnight before ERG recording, and all procedures were performed under dim red light. Rats were anesthetized with a ketamine/xylazine cocktail (100 mg/kg ketamine and 10 mg/kg xylazine, intraperitoneally) and their pupils dilated as described above. Two drops of 0.5% proparacaine hydrochloride were administered to numb the eyes prior to electroretinography. For the ERG recordings, gold electrodes were placed on the surface of each eye with 0.3% hypromellose as a coupling agent. Aluminum reference and ground electrodes were placed subcutaneously in the neck-back region and flank regions, respectively. Scotopic luminance responses were performed on both eyes of each animal. Scotopic ERG was recorded as an average of 3 ERG traces per luminance intensity, with a range of 0.001 to 10 cd s/m2, for each tested eye. ERG signals were high-pass filtered at 0.3 Hz and low-pass filtered at 500 Hz. Individual traces were stored on a local drive. Immediately post ERG, the animals were assessed for pupil dilation and ocular local irritancy. The a-wave amplitude was measured from the baseline to the trough of the a-wave, and implicit time was measured between stimulus and a-wave trough. The b-wave amplitude was measured from the trough of the a-wave to the peak of the b-wave, and b-wave implicit time was measured from stimulus to b-wave peak. The following parameters were analyzed statistically: ERG a-wave amplitudes and implicit times, and ERG b-wave amplitudes and implicit times. The percentage of ERG changes presented here was calculated on the 5th luminance step (0.04 cd s/m2) of the 10 step series, or each corresponding luminance intensity compared to vehicle control.

2.5. Histopathology At termination, eyes with optic nerve were collected and placed in Davidson's fixative. Following fixation, tissues were embedded in paraffin blocks, sectioned (5 μm) and placed onto glass slides. For each eye, a horizontal section just above the optic disk and at least five step sections taken at 100 μm intervals were examined starting above the optic disk toward the injection site. In addition, one to three 100 μm step sections through the optic nerve were examined. All sections were stained with H&E for histopathology evaluation. Microscopic evaluation was performed on all control and treated animals.

The grading scheme for the retina histopathological changes used in all studies ranged from minimal to severe and was based on the percent of the tissue affected. A grade of minimal or 1 corresponded to ≤10%, mild or 2 corresponded to N 10% to ≤25%, moderate or 3 corresponded with N25% to ≤50%, and severe or 4 corresponded to N50% of the tissue affected. 2.6. Retina in situ cell apoptotic assessment Cellular apoptosis assays were used to evaluate potential toxicity, cell damage and tissue necrosis. Apoptosis assays were performed on retinal sections using the DeadEnd™ Colorimetric Kit terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labeling (TUNEL) assay (Promega US, Madison, WI). 2.7. Statistical analysis Treatment effects were analyzed by one-way ANOVA for ERG changes at different days of treatment with Tukey's honestly significant difference (HSD) as post hoc test or Student t-tests where applicable using GraphPad Prism® 5 (GraphPad Software Inc., La Jolla, CA, USA). Results were considered statistically significant if p values were b 0.05. Data is presented as mean ± SD unless otherwise stated. 3. Results 3.1. Body weight analysis, clinical signs, and indirect ophthalmic examination of rats administered pan-CDK inhibitors, and APAP by intravitreal injection No significant treatment-related body weight changes or untoward health issues were observed throughout these studies. While clinical signs were recorded daily, evidence of ocular effect was only observed at the time of ophthalmic examination in the 30 μg/eye AG-012,986treated group and consisted of mild exophthalmoses and pupil dilation. Indirect ophthalmic examinations revealed mild to moderate retinopathy including vasculature abnormality and hemorrhage in four of eight 30 μg/eye AG-012,986-treated eyes on Day 8. APAP (30 and 60 μg/eye), AG-024,322 (1 and 30 μg/eye), and AG-012,986 (1, 3, and 10 μg/eye) had no significant ocular findings during the testing period (Table 2). Incidental lens opacity was noticed across all dose groups, which was considered trauma-related to the dosing procedure.

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

4

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Table 2 Summary of study readouts — ophthalmology findings, ERG, histopathology, and TUNEL. Compound

Route of administration

Dose/day

Ophthalmology finding

ERG

Histological assessment (grade)

TUNEL (grade)

Vehicle A Vehicle B Vehicle C AG- 012,986

Intravitreal iv Intravitreal Intravitreal

0 μg/eye 0 μg/kg 0 μg/eye 1 μg/eye 3 μg/eye 10 μg/eye

NSF NSF NSF NSF NSF NSF

NSF NSF NSF 20–40% b-wave decrease 40% b-wave decrease 60% b-wave decrease

30 μg/eye

Mild exophthalmoses and pupil dilation; mild to moderate retinopathy in 4/8 eyes; vitreal hemorrhage NSF

80% b-wave and 70% a-wave decrease

NSF (0) NSF (0) NSF (0) NSF (0) NSF (0) Mild to moderate retinal degeneration (2-3) Mild to moderate retinal degeneration, mild to moderate vitreal hemorrhage (2-3) NSF (0)

NSF (0) NSF (0) – – – Partial positive staining (2) Intense positive staining (3–4)

iv

5 mg/kga 10 mg/kg

a

20 mg/kgb

NSF

AG- 024,322

Intravitreal

1 μg/eye 30 μg/eye

Retina appeared hyper-reflective; no apparent abnormalities on retina NSF NSF

APAP

Intravitreal

30 μg/eye 60 μg/eye

NSF NSF

Initial 25% b-wave decrease (recovered by day 14) 40% b-wave decrease

NSF (0)

30% b-wave decrease (day 3)

–

NSF 50–60% b-wave decrease NSF NSF

NSF (0) Mild to marked retinal degeneration, mild vitreal body cell debris (1–4) NSF (0) NSF (0)

Sporadic positive staining (1) Intense positive staining (3–4) – – – NSF (0) NSF (0)

NSF: no significant findings. Vehicle A: 2.5% povidone (PVP) K30, 0.2% Tween80 in water; used when formulating AG-012,986 and AG-024,322 for intravitreal exposure. Vehicle B: 0.1 M NaCl/0.1 M acetic acid pH 4.5; used when formulating AG-012,986 and AG-024,322 for systemic exposure. Vehicle C: 2% PVA; used when formulating APA. Histology/TUNEL grading: minimal or 1 corresponded to ≤10%, mild or 2 corresponded to N10% to ≤25%, moderate or 3 corresponded with N25% to ≤50%, and severe or 4 corresponded to N50% of the tissue affected. a Compound administered once a day (SID). b Compound administered twice a day (BID).

3.2. Clinical signs and indirect ophthalmic examination of rats administered AG-012,986 by intravenous injection AG-012,986 administered iv (5 and 10 mg/kg SID; 10 mg/kg BID or 20 mg/kg/day) in Wistar Han rats to evaluate ocular toxicity at a maximum tested dose. Rats did not exhibit any specific abnormalities upon ophthalmic examination, except for hyper reflective retina observed in the 10 mg/kg BID (20 mg/kg/day) treatment group. The animals at 10 mg/kg BID (20 mg/kg/day) were found dead or euthanized early (Day 5) due to excessive systemic toxicities, including decreased activity, debilitated and moribund conditions associated with body weight loss. 3.3. ERG assessments in rats following administration of pan-CDK inhibitors by intravitreal injection Following a single intravitreal administration of AG-012,986 (30 μg/ eye) in dark adapted Wistar Han rats, there were significant ERG bwave amplitude changes against a series of light intensity from 0.001 to 10 cd s/m2. The mean b-wave amplitudes were decreased (80%) on Day 9 when compared to Day − 1 (baseline) (Fig. 1A). These findings are also shown in a representative plot of single-trace recordings of scotopic (dark-adapted) luminance responses (Fig. 1C), where integrity of the wave forms deteriorated after 8 days post-treatment with AG012,986 to a series of white light flash stimuli of increasing intensity from 0.001 to 10 cd s/m2. In addition to the effect on b-wave amplitude, treatment with 30 μg/eye AG-012,986 also resulted in a 70% decrease in a-wave amplitude on Day 9 (Table 2). There was dose-dependent decrease in b-wave amplitude with increasing doses (1, 3, 10 and 30 μg/eye) of AG-012,986, as shown in Table 2. The dose-dependent effect was also reflected in a plot of the light intensity vs. vehicle control in Fig. 1B, with corresponding bwave reduction of 20–40, 40, 60, and 80% at 1, 3, 10 and 30 μg/eye treatment, respectively. Whereas treatment of acetaminophen/APAP (up to

60 μg/eye) had no appreciable effects on b-wave or other ERG parameters when compared to vehicle control. In order to test any possible reversibility of retinal injury in this model, we also performed a single dose two-week recovery study. Following single intravitreal dose of 1 μg/eye AG-012,986, there was 20% decrease in the b-wave amplitude on Day 4 measurement, and the reduction appeared not reversible during the testing period (Fig. 1D). Similar to the effect of AG-012,986, rats administered AG024,322 displayed a dose dependent decrease in b-wave amplitude with 30 μg/eye resulting in 50 to 60% decrease in b-wave amplitude. Whereas 1 μg/eye treatment had no remarkable changes compared to vehicle control (Fig. 1E). All ERG parameters, including both a-wave and b-wave amplitudes and implicit times, were not altered following administration of APAP up to a maximum feasible dose (60 μg/eye based on ophthalmic formulation at 12 mg/ml) and/or vehicle (Table 2). 3.4. ERG assessments in rats administered AG-012,986 by intravenous injection Intravenous administration of 10 mg/kg AG-012,986 SID resulted in 40% decrease in b-wave amplitude on Day 4 (Table 2; Fig. 1F) and this effect persisted throughout the duration of the study (14 days). Administration of AG-021,986 at 5 mg/kg IV (SID) resulted in a 25% b-wave reduction on Day 4; however the decrease in amplitude was not observed (possibly reversed) by Day 14 (Table 2). There were no other ERG parameter changes associated with iv treatment of AG-012,986. In an attempt to increase the dose to explore the effects of retinal structure change, AG-021,986 was given to rats iv 10 mg/kg BID (or 20 mg/kg/day). Following repeat BID dosing of AG-021,986, 30% bwave reduction was noticed on day 3 measurements. However, the rats were not tolerated with severe systemic toxicity, including decreased activity, body weight loss and moribund as describe above, and the study had to be early terminated on day 5.

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

5

Fig. 1. (A) Plot of light intensity versus b-wave amplitudes in dark adapted Wistar Han rats after a single intravitreal administration of AG-012,986 (30 μg/eye) on Day 1. The mean b-wave amplitudes are significantly decreased on Day 9, ***: p b 0.001, ANOVA, compared to pre-treatment baseline data (Day −1). (B) Plot of Day 9 light intensity versus percentage of vehicle control b-wave amplitude in dark-adapted Wistar Han rats following AG-012,986 or acetaminophen intravitreal administration on Day 1. (C) Representative, single-trace recordings of scotopic (dark-adapted) luminance responses to a series of white light flash stimuli of increasing intensity from 0.001 to 10 cd s/m2, baseline wave forms (left panel). Eight days after the single intravitreal administration of AG-012,986 (30 μg/eye), these scotopic luminance responses decreased dramatically (right panel). (D) Time course plot of light intensity versus percentage of vehicle control b-wave amplitude in dark-adapted Wistar Han rats after intravitreal AG-012,986 administration (1 μg/eye). After 15 days the b-wave decrease does not return to the pre-dose levels. (E) Plot of light intensity versus b-wave amplitudes in dark-adapted Wistar Han rats 14 days after a single intravitreal administration of AG-024,322. The mean b-wave amplitudes are significantly decreased with the 30 μg/eye AG-024,322 administration. **: p b 0.01, ANOVA, compared to vehicle-treated control data. (F) Plot of light intensity versus percentage of baseline b-wave amplitude in dark adapted Wistar Han rats following intraocular or systemic iv AG-012,986 administration. On Day 4, the exposure AG-012,986 of both intraocular and systemic routes created respective decreases in b-wave amplitude compared to their pre-dose measurements (Day −1) or the vehicle control.

3.5. Histopathology evaluation Evaluation of rat retinas stained with hematoxylin and eosin (H&E) revealed normal retinal structure in rats administered vehicle A (2.5% povidone K30, 0.2% Tween80) (Fig. 2A), B (0.1 M NaCl/0.1 M Acetic Acid pH 4.5) or C (2% PVA) by iv or intravitreal injection (Table 2). Retinas from rats administered 60 μg/eye APAP (Fig. 2B), ≤ 3 μg/eye AG-012,986 or 1 μg/eye AG-024,322 (Fig. 2C and Table 2) were also unremarkable. By contrast, 30 μg/eye of treatment with pan-CDK inhibitors resulted in severe retinal degeneration on Day 15 (Fig. 2D and F). The degeneration was characterized by loss of 50 to 90% of the retinal organizational structure and bilateral. Mild to moderate hemorrhage occurred within the vitreal body of rats given 30 μg/eye AG-012,986 (Table 2) and mild amounts of cell debris was observed within the vitreous body of rats administered 30 μg/eye AG-024,322 (Fig. 2F). When 50% of the retina was still intact, the inner plexiform layer and ganglion cell layer were present which suggested that the initial damage occurred in the outer retinal layers. At 10 μg/eye AG-012,986, mild to moderate retinal degeneration was also observed on termination day 7 (Table 2). For intravitreal treatment, rank quantified retinal damage of histology with H&E staining correlated well in a dose-related manner to the ERG changes assessed during in-life for both compounds (Table 2).

With iv administration, ocular findings were unremarkable in rats administered 5, 10 mg/kg AG-012,986 SID for 2 weeks. There was no treatment related retinal changes at 10 mg/kg BID (or 20 mg/kg/day) of AG-021,986 in moribund rats terminated early on Day 5. Other non-ocular tissue/organ toxicities were observed at high dose of AG021,986, including thymus (atrophy), bone marrow (hyperplasia), mesenteric lymph nodes (lymphoid depletion), and colon (necrosis). Upon intravitreal treatment of AG-024,322, there were no other tissue/organ histology findings except for ocular at any dose levels. 3.6. TUNEL staining in retinal sections Since no microscopic changes were observed in the retina of rats following iv administration of AG-012,986 (up to 10 mg/kg) despite a clear decrease in b-wave amplitude as measured by ERG, the potential for cellular injury was further investigated using TUNEL staining and was compared to retina from rats treated with Vehicle B or APAP. There was no positive TUNEL staining detected in the retina of rats given vehicle (Fig. 3A) or in rats administered APAP (Fig. 3B). Background staining in the ONL was noted in rats administered vehicle (Fig. 3A) and in rats administered APAP (Fig. 3B). No background staining was observed in the INL of rats administered either vehicle (Fig. 3A) or APAP (asterisk, Fig. 3B). In contrast, the INL and ONL of rats administered iv daily dosing

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

6

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

Fig. 2. Hematoxylin and eosin (H&E) staining of rat retinas. (A) Retina from a rat administered 5 μl/eye vehicle (2.5% povidone (PVP) K30, 0.2% Tween80 in water) at 2 weeks (400×). (B) Retina from a rat administered intravitreal 60 μg/eye APAP at 1 week. The retina was normal and similar to the vehicle control (A) (400×). (C) Retina from a rat administered 1 μg/eye AG-012,896 in a 2-week study. The retina was normal and similar to the rats given either the vehicle (A) or 60 μg/eye APAP. (D) Retina from a rat administered 30 μg/eye AG012,896 at 2 weeks. Note the severe retinal degeneration with loss of the photoreceptors (PRL), inner nuclear layer (INL), outer nuclear layer (ONL), inner plexiform layer (IPL), outer plexiform layer (OPL) and ganglion cell layer (GCL) (400×). (E) Retina from a rat administered 1 μg/eye AG-0,024,322 at 2 weeks. Note that the retina was normal and similar to the rats administered either the vehicle (A), 60 μg/eye APAP (B) or 1 μg/eye AG-012,896 (C) (200×). (F) Retina a rat administered 30 μg/eye AG- 024,322 at 2 weeks. Note that the severe retinal degeneration was similar to rats administered 30 μg/eye AG-012,896 (D) with loss of the PRL, INL, IPL, OPL, ONL and GCL and presence of cellular debris within the vitreous body (200×).

of AG-012,986 at 5 or 10 mg/kg dose exhibited a variable pattern of positive TUNEL or apoptotic nuclear staining, which had a bilateral dose dependent increased intensity in the ONL (Fig. 3C and D). TUNEL staining was positive in the INL and ONL (Fig. 3C) in rats administered 5 mg/kg/dose AG-012,896 and also in the INL (Fig. 3D) of rats administered 10 mg/kg AG-012,986. In addition, positive staining was more intense in the ONL (Fig. 3D) in rats administered 10 mg/kg AG012,986 iv. Further, the number of nuclei within the INL and ONL of rats given iv 10 mg/kg AG-012,986 was decreased compared to rats administered vehicle, APAP or 5 mg/dose AG-012,986. The outer limiting membrane (OLM) was more prominent as a result of the loss of nuclei in the ONL (bold open arrow, Fig. 3D). Furthermore, the positive stains of damaged tissues were rank quantified, which positively correlated to the ERG b-wave changes for both intravitreal and iv administration of AG-012,986 (Table 2).

4. Discussion The current study demonstrated the utility of ERG, as a non-invasive test, in assessing neuro-retinal physiological function in rats. We have not only evaluated ERG's potential application in detecting ocular toxicity early on using compounds with known retinal toxicity but also compared ERG endpoints to histology and TUNEL. ERG was a sensitive method to detect retinal degeneration by two types of chemically distinct panCDK inhibitors when administered to rats either by iv or intravitreal methods. APAP as a negative control and various vehicles were not toxic to retina in both ERG and histopathology analysis. TUNEL assay appeared to be more sensitive in detecting retinal injury, especially apoptosis, and correlated well with ERG findings. To our knowledge, this is the first report that characterized ocular toxicity with pan-CDK inhibitors administered via different dosing routes and

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

7

Fig. 3. TUNEL staining of rat retinas. (A) Retina from a rat administered vehicle (0.1 M NaCl, 0.1 M acetic acid, pH 4.5) iv once a day at 2 weeks. Staining was negative in the INL with background staining present in the ONL (arrowheads) (400×). (B) Retina from a rat administered 60 μg/eye APAP at 1 week. Note the negative staining in the INL (asterisk) and background staining in the ONL that was similar to (A, arrowheads) (400×). (C) Retina from rat administered 5 mg/kg/day AG-012,896 iv once a day for 4 days at 2 weeks. Note the intense nuclear staining of a cell in the INL (long arrow) and ONL (short arrow) (400×). (D) Retina from a rat administered 10 mg/kg/day AG-012,896 iv once a day at 2 weeks. Note the loss of nuclei from the INL and ONL, positive staining of cells in the INL (long arrows) and positive staining of cells in the ONL (short arrows) and a more prominent OLM (bold open arrow, 400×).

using a series of ocular toxicity tests, including ophthalmic examination, ERG, and histopathology. CDK inhibitors have been actively pursued as drug targets by pharmaceutical companies, especially in oncology due to their role in cell cycle regulation (Cicenas and Valius, 2011; Casimiro et al., 2014). Among the various CDK family members, the expression of CDK 1, 2, 4, and 5 was detectable in the retina or ocular tissues in rat, monkey and human evaluated by real-time PCR or immunohistochemistry at mRNA and protein levels (Sakamoto et al., 2011; Saturno et al., 2007). Except for CDK 5 which is expressed across different layers of the retina, all other CDKs tested were mainly present in the outer segment. CDKrelated retinal toxicity was first reported with AG-012,986, an early generation pan-CDK inhibitor, in mice (Illanes et al., 2006). Saturno et al. (2007) attempted to investigate mechanistically if CDK2 inhibition is the cause of retinal toxicity in monkeys treated with a potent aminothiazole CDK2 inhibitor, while Cheng et al. (2003) speculated based on the expression profiles of various CDKs and CDK5 mutant analysis, that CDK5 may be the possible culprit in the observed retinal toxicity. Regardless of the exact cause of retinal toxicity, the two non-specific pan-CDK inhibitors (AG-012,986 and AG-024,322) were used in the current study to induce ocular toxicity in rats, a common species used in toxicology studies. The pan-CDK inhibitor AG-012,986 was administrated by both intravitreal and systemic iv routes. For the majority of general toxicity studies, drugs are administered via systemic routes (e.g. iv, oral, etc.), to mimic the intended routes of clinical testing. In the process of drug safety evaluation, occasionally ocular toxicity may develop sub-chronically or chronically following weeks to months of systemic drug administration. The long duration required for developing ocular lesions makes it difficult to validate ocular findings in a routine 14-day exploratory toxicity study, although it is not unusual to encounter compounds that lead to acute ocular toxicity in short term screening study, like the two compounds presented here. To this end, local or direct ocular application, such as intravitreal administration, delivers the test article directly to the target tissue (retina). This has several advantages including minimal test article requirements, quick onset of exposure and

action thus potentially accelerating any pathogenesis, and resulting in faster assessment of toxicity. The present study compared ocular toxicities induced via both routes using conventional ocular evaluation methods and electroretinography. In the current study, intravitreal administration of AG-012,986 and AG-024,322 resulted in marked retinal degeneration, detectable by all methods used including ophthalmic examinations, histopathology, ERG, as well as ocular biomarkers detected in the plasma (Peng et al., 2014). Intravitreal injection is a well-developed local drug delivery technique; in fact it has been employed to deliver a number of commercially available or investigational therapeutics for ophthalmic indications (Jonas et al., 2005). In preclinical investigational toxicity studies, this efficient dosing route has been widely used in different species for ocular toxicity study to better understand clinical ocular findings and for excipient selection to reduce the risk of attrition during compound formulation development (Huang et al., 2009; Aguirre et al., 2012). The results from the current study point to the utility and benefits of intravitreal administration and ERG in assessing retinal toxicity and ocular functions. A dose dependent decrease in b-wave amplitude was observed with escalating intravitreal doses of 1, 3, 10 and 30 μg/eye AG-012,986. At the highest dose, AG-012,986 also caused a significant decrease in a-wave amplitude of the treated eyes. Similarly, AG024,322 (30 μg/eye) significantly decreased b-wave amplitude. These findings correlated well with in-life ophthalmic examination, and were further supported by histopathology characterization. Upon local intravitreal injection of either of the pan-CDK inhibitors, ERG detected the functional retinal changes, which correlated with morphological changes in a dose-dependent manner. To assess and confirm that the ERG changes observed with the panCDK inhibitors were not due to a nonspecific effect related to the route of administration (intravitreal), a non-retinal toxicant, acetaminophen, was chosen and was formulated in 2% polyvinyl alcohol (PVA) for intravitreal dosing. Up to 60 μg/eye of acetaminophen (maximum feasible dose based on 12 mg/ml ophthalmic formulation) was injected intravitreally, no remarkable ERG findings were noted in the

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

8

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx

acetaminophen-treated eyes, and no test article related retinal changes were detected by microscopic examination. These data provided confidence in the safety of intravitreal dosing procedure, and supported our interpretation of ocular toxicity by excluding any procedure-related artifacts in this rat ocular toxicity model. Whereas intravitreal injection is applicable when early assessment of potential ocular toxicity of a drug is suspected or when not enough material is available to conduct traditional toxicity studies, most toxicity studies utilize systemic routes of administration. To this end we set out to evaluate if ERG changes, indicative of potential ocular toxicity, can be observed following systemic administration. Our results showed that iv injection of AG-012,986 (5 or 10 mg/kg SID) also resulted in retinal degeneration in rats, detectable by ERG and TUNEL assay. Similar to the intravitreal injection study, ERG examination revealed b-wave amplitude reduction that was dose dependent in magnitude and which was reversible at 5 mg/kg SID (25% b-wave amplitude reduction) but was non-reversible (40% reduction) in the 10 mg/kg/day group. Iv administration of AG-012,986 (10 mg/kg BID) led to excessive systemic toxicities which necessitated early termination (Day 5), therefore prevented us from fully evaluating the retinal toxicity at these higher exposures. By contrast, the local delivery approach was better tolerated resulting in low systemic exposure due to retina-blood barrier and allowed us to characterize the ocular toxicity of these compounds at high concentrations and up to 14 days, the typical duration of an early toxicity study. This highlights the benefit of intravitreal administration as a non-invasive procedure to assess retinal toxicity. The utility of ERG as a non-invasive test in assessing neuro-retinal physiological function in rats has been demonstrated in this study. The ERG data captured dose-dependent retinal functional changes induced by pan-CDK inhibitors at the doses used regardless of the route of administration. Even at a low dose of 1 μg/eye AG-012,986 intravitreal injection, the b-wave amplitude was consistently decreased (20–40%). This effect when assessed 15 days later (following a single administration) appeared to not reverse. The lack of reversibility in b-wave amplitude is likely due to the prolonged toxicity effect that outlasted the actual exposure time of the test article in the retina. In a rat PK/PD study (Rittenhouse et al., 2014), the t½ values of a compound in a choroidal neovascularization model were 4–6 h and relatively consistent across all ocular tissues (retina, vitreous) and plasma, but in contrast to the relatively short ocular PK t½, the pharmacological effects lasted much longer. In comparison to ERG, histopathology evaluation revealed retinal degeneration in the eyes following administration of higher doses of the pan CDK inhibitors (10 and 30 μg/eye AG-012,986 and 30 μg/eye AG-024,322) but not in eyes treated with lower doses of the inhibitors (1 and 3 μg/eye AG-012,986 and 1 μg/eye AG-024,322), despite the fact that ERG has demonstrated a decrease in b-wave amplitude at these lower doses. This suggests that ERG is a more sensitive method for detection of ocular toxicity when compared to traditional histopathology. Furthermore, histopathology examination also failed to reveal signs of retinal toxicity following iv administration of AG-012,986 despite a decrease in b-wave amplitude detected by ERG. The lack of morphologic changes in the retinas by routine histology examinations is likely due to the limits of its sensitivity, since routine H&E staining usually requires cellular or tissue structure changes such as cell membrane damage, rupture, or swelling. Another limitation of systemic (iv) administration is tolerability. In this study, we attempted to increase the rat plasma exposure of AG-012,986 by administering it twice daily (10 mg/kg BID or 20 mg/kg/day) to potentially observe histologic changes similar to those observed following intravitreal administration; however, the animals were terminated early due to excessive systemic toxicity. Unlike the systemic route, local intravitreal delivery approach leads to better toleration with no systemic toxicity allowing us to fully investigate any potential retinal structural changes. As discussed above, due to limitation of the retina-blood barrier, systemic exposure is not expected to reach toxic high level, making intravitreal

administration an ideal route for assessing retinal toxicity in this animal model particularly for systemically toxic test articles. Utilizing TUNEL (the terminal deoxynucleotidyl transferase dUTP nick end-labeling) as a more sensitive staining to assess apoptosis in retina sections from animals treated with AG-012,986 at 5 or 10 mg/kg iv, we have demonstrated a dose dependent increase in apoptotic nuclear staining in concordance with the ERG results and pointing that ERG and TUNEL are more sensitive methods than histopathology in detecting retinal toxicity. At 10 mg/kg dose, intense nuclear TUNEL staining was most evident in the outer nuclear layer (50% or more of cells positive), with faint to intense nuclear staining in fewer cells occupying the inner nuclear layer. TUNEL is a preferred assay to detect and quantify apoptotic cells with fragmented DNA and has become widely used in situ staining method to assist in detection of apoptotic cells in tissues sections. Studies have shown that even the detection of a few apoptotic cells can indicate a major tissue remodeling (Pallardy et al., 1999). The routine light microscopy can detect lesion, which relies mainly on morphologic changes or secondary necrosis after distinct apoptotic morphology (Saraste and Pulkki, 2000). TUNEL on the other hand is a sensitive method for quantitative and confirmative detection of apoptotic cells, especially at the early stage when the morphologic changes are not apparent at the light microscopic level. Overall, the ERG results correlated well with findings generated by in vivo ophthalmic exams, TUNEL assay, and histopathology. In practice, the immediate readout of ERG parameters can be qualitatively evaluated in real-time, while further advanced analysis of the a-waves, bwaves, and implicit times can be applied thereafter. Moreover, ERG appears to be more sensitive and can detect retinal functional changes at a very early stage of pathogenesis. Indeed, a single intravitreal injection of AG-012,986 produced a dose dependent decrease in b-wave amplitude as early as Day 3 post dosing in a two-week exploratory toxicity study (data not shown). For future work, this method will be fully validated utilizing additional compounds of different mechanism of action to ensure the sensitivity of ERG in detecting retinal toxicity. Additionally, to clarify the correlation between ERG electrophysiological data and the specific type of retinal lesion, potential miRNA biomarkers and retinal substructure location, future experiments are ongoing with the assistance of immunohistochemical and laser capture microscopic technology. Lastly, the applicability of ERG to other toxicological species, like rabbit and dog, is currently being explored by this laboratory. Transparency document The Transparency document associated with this article can be found, in the online version. Acknowledgements The authors would like to thank Michelle Lee and Patrick Lappin for their excellent technical assistance in generating and interpreting the TUNEL data for this manuscript. References Aguirre, S.A., Collette, W., Gukasyan, H.J., Huang, W., 2012. An assessment of the ocular safety of excipient maleic acid following intravitreal injection in rabbits. Toxicol. Pathol. 40, 797–806. Aguirre, S.A., Huang, W., Prasanna, G., Jessen, B., 2009. Corneal neovascularization and ocular irritancy responses in dogs following topical ocular administration of an EP4prostaglandin E2 agonist. Toxicol. Pathol. 37, 911–920. Casimiro, M.C., Velasco-Velazquez, M., Aguirre-Alvarado, C., Pestell, R.G., 2014. Overview of cyclins D1 function in cancer and the CDK inhibitor landscape: past and present. Expert Opin. Investig. Drugs 23, 295–304. Cicenas, J., Valius, M., 2011. The CDK inhibitors in cancer research and therapy. J. Cancer Res. Clin. Oncol. 137, 1409–1418. Cheng, Q., Sasaki, Y., Shoji, M., Sugiyama, Y., Tanaka, H., Nakayama, T., Mizuki, N., Nakamura, F., Takei, K., Goshima, Y., 2003. Cdk5/p35 and rho-kinase mediate ephrin-A5-induced signaling in retinal ganglion cells. Mol. Cell. Neurosci. 24, 632–645.

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008

W. Huang et al. / Toxicology and Applied Pharmacology xxx (2015) xxx–xxx Huang, W., Yang, A.H., Matsumoto, D., Collette, W., Marroquin, L., Ko, M., Aguirre, S., Younis, H.S., 2009. PD0325901, a mitogen-activated protein kinase kinase inhibitor, produces ocular toxicity in a rabbit animal model of retinal vein occlusion. J. Ocul. Pharmacol. Ther. 25, 519–530. Illanes, O., Anderson, S., Niesman, M., Zwick, L., Jessen, B.A., 2006. Retinal and peripheral nerve toxicity induced by the administration of a pan-cyclin dependent kinase (cdk) inhibitor in mice. Toxicol. Pathol. 34, 243–248. FDA. http://www.fda.gov/ohrms/dockets/ac/02/slides/3882s1_02_gibertson.pdf Jonas, J.B., Kreissig, I., Degenring, R., 2005. Intravitreal triamcinolone acetonide for treatment of intraocular proliferative, exudative and neovascular diseases. Prog. Retin. Eye Res. 24, 587–611. Krystof, V., Uldrijan, S., 2010. Cyclin-dependent kinase inhibitors as anticancer drugs. Curr. Drug Targets 11, 291–302. Lee, M.G., Nurse, P., 1987. Complementation used to clone a human homologue of the fission yeast cell cycle control gene cdc2. Nature 327, 31–35. OECD, 2012. Test no. 405: acute eye irritation/corrosion, OECD guidelines for the testing of chemicals, section 4. Publishing, OECD http://dx.doi.org/10.1787/ 20745788. Pallardy, M., Biola, A., Lebrec, H., Breard, J., 1999. Assessment of apoptosis in xenobioticinduced immunotoxicity. Methods 19, 36–47. Peng, Q., Huang, W., John-Baptiste, A., 2014. Circulating microRNAs as biomarkers of retinal toxicity. J. Appl. Toxicol. 34, 695–702.

9

Rittenhouse, K.D., Johnson, T., Vicini, P., Hirakawa, B., Kalabat, D., Yang, A.H., Huang, W., Basile, A.S., 2014. RTP801 gene expression is differentially upregulated in retinopathy and is silenced by PF-04523655, a 19-Mer siRNA directed against RTP801. Inest. Ophthalmol. Vis. Sci. 55, 1232–1240. Rosolen, S.G., Rigaudière, F., Le Gargasson, J.F., Brigell, M.G., 2005. Recommendations for a toxicological screening ERG procedure in laboratory animals. Doc. Ophthalmol. 110, 57–66. Sakamoto, K., Ohki, K., Sito, M., Nakahara, T., Ishii, K., 2011. Small molecule cyclindependent kinase inhibitors protect against neuronal cell death in the ischemicreperfused rat retina. J. Ocul. Pharmacol. Ther. 27, 419–425. Saraste, A., Pulkki, K., 2000. Morphologic and biochemical hallmarks of apoptosis. Cardiovasc. Res. 45, 528–537. Saturno, G., Pesenti, M., Gavazzoli, C., Rossi, A., Giusti, A.M., Gierke, B., Pawlak, M., Venturi, M., 2007. Expression of serine/threonine protein-kinases and related factors in normal monkey and human retinas: the mechanistic understanding of a CDK2 inhibitor induced retinal toxicity. Toxicol. Pathol. 35, 972–983. Shapiro, G.I., 2006. Cyclin-dependent kinase pathways as targets for cancer treatment. J. Clin. Oncol. 24, 1770–1783. Yang, A.H., Huang, W., 2012. Retinal vein occlusion induced by a MEK inhibitor — impact of oxidative stress on the blood-retinal barrier. In: Lushchak, V.I., Gospodaryov, D.V. (Eds.), Oxidative Stress. InTech, pp. 469–494.

Please cite this article as: Huang, W., et al., Application of electroretinography (ERG) in early drug development for assessing retinal toxicity in rats, Toxicol. Appl. Pharmacol. (2015), http://dx.doi.org/10.1016/j.taap.2015.10.008