lable at ScienceDirect

Neuropharmacology 109 (2016) 148e158

Contents lists avai

Neuropharmacology

journal homepage: www.elsevier .com/locate/neuropharm

Evaluating the neurotherapeutic potential of a water-solubleprogesterone analog after traumatic brain injury in rats

Bushra Wali a, *, Iqbal Sayeed a, David B. Guthrie b, Michael G. Natchus b, Nefize Turan c,Dennis C. Liotta b, Donald G. Stein a

a Department of Emergency Medicine, Brain Research Laboratory, Emory University, Atlanta, GA 30322, USAb Emory Institute for Drug Development/Department of Chemistry, Emory University, Atlanta, GA 30322, USAc Department of Neurosurgery, Emory University School of Medicine, Atlanta, GA 30322, USA

a r t i c l e i n f o

Article history:Received 4 March 2016Received in revised form6 May 2016Accepted 24 May 2016Available online 4 June 2016

Keywords:Traumatic brain injuryNeurosteroidProgesteroneProgesterone analogsEdema

Abbreviations: ANOVA, analysis of variance; BWcontrolled cortical impact; CNS, central nervous systeprogesterone; GABA, gamma-aminobutyric acid; GCSglial fibrillary acidic protein; HBC, 2-hydroxypropydensity lipoprotein; i.m., instramuscular; i.v., intraveeter; LDL, low-density lipoprotein; MFC, medial frontamaze; nAChR, nicotinic acetylcholine receptor; NMDAphosphate buffered saline; PK, Pharmacokinetics; p.oreceptor; PROG, progesterone; PXR, pregnane X receSD, Sprague Dawley; TBI, traumatic brain injury.* Corresponding author. Department of Emergen

Laboratory, 1365B Clifton Rd NE, Suite 5100, Emory UUSA.

E-mail address: bwali@emory.edu (B. Wali).

http://dx.doi.org/10.1016/j.neuropharm.2016.05.0170028-3908/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

The poor aqueous solubility of progesterone (PROG) limits its potential use as a therapeutic agent. Wedesigned and tested EIDD-1723, a novel water-soluble analog of PROG with >100-fold higher solubilitythan that of native PROG, as candidate for development as a field-ready treatment for traumatic braininjury (TBI). The pharmacokinetic effects of EIDD-1723 on morphological and functional outcomes in ratswith bilateral cortical impact injury were evaluated. Following TBI, 10-mg/kg doses of EIDD-1723 orPROG were given intramuscularly (i.m.) at 1, 6 and 24 h post-injury, then daily for the next 6 days, withtapering of the last 2 treatments. Rats were tested pre-injury to establish baseline performance on gripstrength and sensory neglect, and then retested at 4, 9 and 21 days post-TBI. Spatial learning wasevaluated from days 11e17 post-TBI. At 22 days post-injury, rats were perfused and brains extracted andprocessed for lesion size. For the edema assay the animals were killed and brains removed at 24 h post-injury. EIDD-1723 significantly reduced cerebral edema and improved recovery from motor, sensory andspatial learning deficits as well as, or better than, native PROG. Pharmacokinetic investigation after asingle i.m. injection in rats revealed that EIDD-1723 was rapidly converted to the active metabolite EIDD-036, demonstrating first-order elimination kinetics and ability to cross the blood-brain barrier. Our re-sults suggest that EIDD-1723 represents a substantial advantage over current PROG formulations becauseit overcomes storage, formulation and delivery limitations of PROG and can thereby reduce the timebetween injury and treatment.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Numerous drugs have shown promise as neuroprotective agents

C, brain water content; CCI,m; ent-PROG, enantiomer of, Glasgow Coma Score; GFAP,l-b-cyclodextrin; HDL, high-nous; JVC, jugular vein cath-l cortex; MWM, Morris water, N-methyl-d-aspartate; PBS,., per oral; PR, progesteroneptor; RT, room temperature;

cy Medicine, Brain Researchniversity, Atlanta, GA 30322,

in preclinical animal models of traumatic brain injury (TBI), butnone have worked in clinical trials. Despite substantial pre-clinicalevidence (Schumacher et al., 2015) supporting the neuroprotectiveeffects of progesterone (PROG), two recently completed phase IIIclinical trials reported that the neurosteroid did not provide a sig-nificant benefit to patients recovering from moderate to severe TBI(Skolnick et al., 2014;Wright et al., 2014). One factor in the failure ofthe trials may have been treatment delays, which ranged from 4 to9 h post-injury and could have compromised the neuroprotectivebenefits of the test drug. A key tenet of brain injury treatment isthat the sooner the intervention takes place, the better the func-tional outcome will be (Stein, 2011), but there must also be goodpenetration of the agent and its metabolites into the damagedtissue to permit neuroprotective activity. Thus another factor in thenegative trial results could have been the poor solubility of PROG,and the intrinsic biological activity of the carrier chosen for thedelivery of PROG (Howard et al., 2015; Stein, 2015). Current clinical

B. Wali et al. / Neuropharmacology 109 (2016) 148e158 149

protocols for PROG administration require that patients be trans-ported to a hospital setting before treatment can begin, losingvaluable time. Time to treatment is an important considerationbecause many agents which show positive effects in the laboratory,where they can be given minutes to several hours after injury, donot show efficacy when treatment is delayed to later in the injurycascade.

PROG thus has certain limitations as a therapeutic intervention,notably its insolubility in aqueous-based formulations, and its shorthalf-life, which restrict its capability for rapid delivery in emer-gency conditions. What is needed is a stable, water-soluble andeasy-to-administer form of PROG that works immediately to rescuedamaged tissue in preparation for a longer course of therapy thatcan be delivered in the hospital.

As noted, the disappointing outcomes of TBI randomizedcontrolled trials may have been impacted by the choice of vehiclewith which PROG was administered intravenously (i.v.). In bothphase III trials (Skolnick et al., 2014; Wright et al., 2014) a lipidcarrier was used, but these agents can have important effects onphysiology and are not completely “neutral” as required for anappropriate control vehicle. Lipid emulsions have been shown toreduce inflammatory reactions in endothelial cells, an effect whichhas implications for modulating vascular repair in the damagedbrain (Harvey et al., 2015). Some lipid carrier effects are beneficial,some are not. Significant elevation in total HDL and LDL cholesterolfollowing 96 h of continuous 10% Intralipid infusion has been re-ported (Wasan et al., 1994). A meta-analysis of case reports foundthat Intralipid administration was effective in reversing multipledrug toxicities (Muller et al., 2015). Studies have shown that i.v.lipid emulsions can be effective in the acute stages of drug intoxi-cation, improving Glasgow Outcome Scale (GOS) scores andreducing blood glucose levels up to 6 h after administration(Taftachi et al., 2012). Intralipid administration has also been shownto increase insulin resistance in heart tissue in an experimentalmodel of type II diabetes in rats, indicating that the agent caninterfere with glucose metabolism under certain conditions thatmight also affect the extent of brain injuries (Lou et al., 2015). Thus,as a vehicle in a clinical trial, Intralipid or other lipid compoundsmay confound and/or mask the effects of PROG given to patients inthe acute stage of the injury cascade. In contrast, virtually all pre-clinical experiments preceding the clinical trials used more solu-ble 2-hydroxypropyl-b-cyclodextrin (HBC) as a vehicle rather thana lipid formulation, another possible reason the pre-clinical reportsfound reliable neuroprotective benefits while the clinical trials didnot. Further, three small, single-center clinical studies using low-dose intramuscular (i.m.) PROG have shown positive outcomes(Xiao et al., 2008; Mofid et al., 2016; Raheja et al., 2016). HBC hasbeen used clinically (Preskorn, 2005; Mermelstein et al., 2013) as avehicle for i.m. dosing of poorly water-soluble drugs. However, anaqueous-based formulation will have the added advantage that itcan be carried as a powder, mixed in a bottle or syringe with awater-based solution, and injected for immediate delivery to braininjury victims before they arrive in the hospital.

We have previously reported synthesis of a novel set of water-soluble analogs of PROG (MacNevin et al., 2009; Guthrie et al.,2012). These analogs were shown to possess intrinsic neuro-protective activity and are well suited for field administration as ani.m. injection, given by i.v. drip, or even potentially by intranasaladministration. Several of the analogs we tested demonstratedneuroprotective effects and reduced brain edema in an in vitro,glutamate-induced, excitotoxic, primary neuronal cell lesion model(MacNevin et al., 2009; Guthrie et al., 2012). Based on these studies,we selected EIDD-1723 for evaluation of early-stage morphologicaland functional efficacy after bilateral fronto-cortical contusion in-juries in rats. In this set of proof-of-concept experiments, our

results show that EIDD-1723 offers substantial treatment advan-tages over the PROG-lipid formulations used in the phase III TBItrials.

2. Materials and methods

2.1. Synthesis of PROG analog EIDD-1723

EIDD-1723 is a substituted C-20 oxime analog of PROG that issynthesized in three synthetic steps from readily available 5-pregnen-3b-ol-20-one (pregnenolone) (Fig. 1). Condensation ofpregnenolone with O-((methylthio)methyl)hydroxylamine, formedin situ from the action of hydrazine on the appropriate O-substituted phthalimide, provides C-20 oxime ether 1 in 74% yield.Oppenauer oxidation of the steroidal A ring gives 2 in 73% yield.Treatment of the thioether moiety with N-iodosuccinimide in thepresence of phosphoric acid installs the phosphonooxymethylgroup; subsequent deprotonation with Na2CO3 and purification byreverse phase chromatography provides EIDD-1723 as a disodiumsalt in 47% yield. Experimental details are available in theSupporting Information.

2.2. Pharmacokinetics study

A pharmacokinetics (PK) experiment was carried out in ratsusing per oral (p.o.), i.m. or i.v. administration to investigate themetabolic conversion of EIDD-1723 to its active metabolite EIDD-036. This study was outsourced to Agilux, Inc. (Worcester, MA,USA) under Protocol #CE-0006-DA-RI. Male Sprague-Dawley (SD)rats fittedwith indwelling jugular vein cannulas (JVC) were allowedto acclimate to the test facility for 2 days prior to the start of thestudy and then randomly assigned to 5 groups (n ¼ 4/group) asdescribed in Table 1. The animals were not fasted, but the p.o. groupwas dosed >2 h after “lights on” to ensure that stomachs wereempty when dosed. All oral doses were administered via gavagetube and all i.v. doses were administered via tail vein catheter.Catheters were flushed with ~0.5 ml of saline immediatelyfollowing dose and prior to catheter removal. All animals wereobserved at dosing and each scheduled collection. Serial sampleswere collected from the JVC according to the schedule in Table 1. Atthe 10-h post-dosing, animals were euthanized and brain sampleswere collected from all the rats. Brains were rinsed with saline,patted dry, weighed, and placed on dry ice prior to storage at�80 �C until transferred to analytical chemistry for analysis. Bloodsamples were collected into K2EDTA tubes and stored on wet iceuntil processed to plasma by centrifugation (3500 rpm at 5 �C)within 1 h of collection. Plasma samples were transferred intomatrix tubes and stored at nominal �80 �C until transferred toanalytical chemistry for analysis.

2.3. Tissue distribution of EIDD-1723 in rat brain

A study of the distribution of the compound in brain tissue wasperformed to determine the time-concentration profile of EIDD-1723 in the CNS after an i.m. injection. This study was conductedby Agilux, Inc. under Protocol #CE-0006-DA-RI. Male SD rats(n ¼ 24) were allowed to acclimate to the test facility for 2 daysprior to the start of the study. The animals were provided food andwater ad libitum. All animals were dosed i.m. with 10 mg/kg ofEIDD-1723 and observed at dosing and at each scheduled collec-tion. No abnormalities were recorded. Four animals were eutha-nized at each of the 0.5, 1, 2, 4, 6 and 8 h time points, and brainsamples were collected. Tissue samples were rinsed with saline,patted dry, weighed, and placed on dry ice prior to storage at�80 �C until transferred to analytical chemistry for analysis. Plasma

Fig. 1. Schematic of EIDD-1723 synthesis.

Table 1Experimental design of pharmacokinetics study.

Group Number of rats Test article Dose level (mg/kg) Conc. (mg/ml) Dose volume (ml/Kg) Vehicle Route Sampling time (h)

1 4 EIDD-1723 30 6 5 PROPYLENEGLYCOL/WATER(40/60) PO 0.25, 0.5, 1,2,3,4,6,8,102 4 EIDD-1723 3 1.5 2 PG/W (40/60) IM 0.25, 0.5, 1,2,3,4,6,8,103 4 EIDD-1723 10 5 2 PG/W (40/60) IM 0.25, 0.5, 1,2,3,4,6,8,104 4 EIDD-1723 30 15 2 PG/W (40/60) IM 0.25, 0.5, 1,2,3,4,6,8,105 4 EIDD-1723 2 1 2 PG/W (40/60) IV 0.083, 0.25, 0.5, 1, 2, 4, 6, 8

B. Wali et al. / Neuropharmacology 109 (2016) 148e158150

and tissue samples were analyzed on a Sciex MS API-6500 instru-ment (SCIEX, Framingham, MA, USA) in positive ionization modefitted with an ACE-3 C18 column using a gradient (2.5 h) mobilephase with mobile phase A ¼ H2O:ACN:FA (95:5:0.1) and mobilephase B ¼ ACN:MeOH:FA (50:50:0.1). Chrysin was used as an in-ternal standard.

2.4. Efficacy study

2.4.1. SubjectsMale SD rats weighing 200e350 g at the time of injury were

used for the efficacy study, which was conducted in our facility atEmory University. The rats were maintained on a reverse 12:12light-dark cycle at 22� ± 1 �C with appropriate humidity levels.After one week of quarantine the rats were handled at least 3 timesprior to surgery. This study was conducted in a facility approved bythe American Association for the Accreditation of Laboratory Ani-mal Care (AAALAC) in accordance with all National Institutes ofHealth guidelines. All experimental animal procedures wereapproved by the Emory University Institutional Animal Care andUse Committee (Protocol # 2002865).

2.4.2. Contusion injuryRats were anesthetized using isoflurane gas (5% induction, 1.5%

maintenance, 700mmHgN2O, 500mmHgO2) and thenmounted ina Kopf stereotaxic device (David Kopf Instruments, Tujunga, CA,USA). The scalp incision area was shaved and sterilized with Beta-dine® antiseptic and 70% isopropanol. Physiological parameterswere monitored with pulse oximetry (SurgiVet™ model V3304):

heart rate was maintained above ~300 beats per minute and SpO2kept above 90%. Core body temperature (~37 �C) was maintainedwith a homeothermic heating blanket system (Harvard Apparatus,Holliston, MA, USA). Under aseptic conditions, a midline incisionwas made into the skin and fascia covering the skull. Cotton swabswere used to staunch and clean any fascial bleeding. Bregma waslocated and a trephine drill was used to perform a 5-mm diametermid-sagittal bilateral craniotomy 3 mm anterior to bregma.Controlled cortical impact (CCI) injury to the medial frontal cortex(MFC) was induced with a magnetic cortical pinpoint contusionimpactor (4-mmdiameter; PC1300; Hatteras Instruments, Cary, NC,USA) to a depth of 2.5 mm at a pressure of 1.7 psi, impact time of100 ms, and velocity of 2.26 m/s. Sutures were used to close theincision after bleeding stopped. Animals were then placed intoheated recovery boxes and allowed to recover from the anesthesiabefore being returned to their home cages. The sham groupreceived no impact and the incisions were sutured closed aftercomparable time under anesthesia.

2.4.3. TreatmentAnimals were randomly assigned to one of four treatment

groups. For the edema assay: sham (n ¼ 3), vehicle (22.5% HBC(n ¼ 4)), PROG (10 mg/kg; n ¼ 4) and EIDD-1723 (10 mg/kg; n ¼ 4).HBC was used because it is water-soluble and because it haslimited, if any, toxicity even when given over long periods of time(Gould and Scott, 2005). Treatment was administered i.m. at 1 and6 h post-TBI. Animals were euthanized at 24 h post-injury and theirbrains removed and sampled for edema assay. Percent change inbrain tissue water content was measured.

B. Wali et al. / Neuropharmacology 109 (2016) 148e158 151

For the behavioral study, rats were randomly assigned (n ¼ 8/group) to sham-vehicle, lesion-vehicle, lesion-PROG and lesion-EIDD-1723 groups and the identity of the groups was coded toavoid experimenter bias. Rats received i.m. injections of PROG(10 mg/kg), EIDD-1723 (10 mg/kg), or HBC alternately into the rightor left hind limb, at 1, 6, and 24 h and 2, 3, 4, 5, 6, and 7 days post-injury. The last 2 injections were tapered (reduced doses) to reducePROG withdrawal syndrome (Cutler et al., 2005). After completionof the behavioral tests the rats were perfused and their brainsextracted at 22 days post-injury for histological assays.

2.4.4. Edema assayEdemawas measured at 24 h after surgery. Briefly, the rats were

decapitated under deep anesthesia and the brain extracted anddissected into anterior and posterior sections. The anterior sectioncontained the entire lesion area. Each section was placed in a pre-labeled and pre-weighed tube that was immediately capped. Eachtube was reweighed, and then opened and placed in a 60 �C ovenwith 15 mmHg vacuum pressure for 48 h. Samples were reweighedafter drying. All weighing was done on the same balance. Thepercent water content was calculated by [(wet wt - dry wt)/wet wt]*100. The percent difference in water content between the anteriorperi-contusional and the posterior distal section was calculated foreach sample by: [(anterior H2O% � posterior H2O%)/(posterior H2O%)] � 100.

2.4.5. Behavioral/functional assays2.4.5.1. Grip strength. The rats were tested pre-injury to establish abaseline performance on grip strength, then retested at 4, 9, and 21days post-TBI. To check forelimb grip strength, we used an elec-tronic digital force gauge grip-strength meter (Columbus In-struments, Columbus, OH, USA) that measures the peak forceexerted by an animal while gripping the sensor bar. A digitalreading (in Newtons) of three successive trials was obtained foreach rat, and the best (highest) score was used for data analysis.

2.4.5.2. Sensory neglect. The animals were tested pre-injury toestablish a baseline performance for sensory neglect, then retestedat 4, 9 and 21 days post-TBI. An adhesive removal test was used toevaluate somatosensory function. Adhesive (half-inch round) labelswere placed on the ventral surface of the forepaw. The latency toremove the adhesive label for each rat was recorded manually. If arat was unable to remove the sticker after 120 s passed, the stickerwas removed by the experimenter and the animal was given amaximum latency of 2 min. Two trials per animal were given andthe best latency was used for final analysis.

2.4.5.3. Morris water maze (MWM) spatial learning performance.All the rats were tested for cognitive and spatial navigational per-formance through MWM starting at day 11 and ending at day 17 (7consecutive days) post-TBI. An overhead camera and computer-assisted tracking system (San Diego Instruments Inc., San Diego,CA, USA) recorded the rat’s position in the maze, swim distance,heading, and time taken to find the platform (latency). Each ratreceived 2 trials per day. The trials were separated by a 5-min in-terval and used a different quadrant location. The rats were allowedto swim in the pool until they reached the platform or until 90 s hadpassed. If rats were unable to reach the platformwithin the allottedtime, the experimenter guided them to the platform. Rats wereallowed to remain on the platform for 20 s and then removed fromthe pool. After 5 min, subjects were again released into the tankfrom a new position and allowed to swim to the platform. Theplatform remained in the same position for both trials on all days oftesting. Between and after trials, animals were placed in holdingcages in front of a warm air blower to dry.

2.4.6. Tissue preparationFor histological analysis brains were extracted after transcardial

perfusion with 4% formalin. After 24 h of post-fixation in 4% para-formaldehyde solution, brains were immersed in 10%, 20% and 30%sucrose solution, thenmounted and frozen using dry ice. The brainswere cut into 20-mm sections on a cryostat and stored at �80 �C on1% gelatin-coated slides. Slides were stained in 0.1% cresyl violetsolution (0.1 g cresyl violet acetate and 0.3 ml of glacial acetic aciddissolved in 100 ml distilled water) for 10 min at 45 �C, and thenrinsed in distilledwater. Then slides werewashed in 95% alcohol for5e10 min and 100% alcohol (2 � 5 min each), and then in xylene(3 � 5 min each). Slides were mounted with xylene-based cytoseal.

2.4.7. Necrotic cavity measurementsSix sections from each rat brain (4.7, 3.7, 2.7, 1.7, 0.7 anterior and

0.3 mm posterior to bregma), were chosen for lesion cavity mea-surements. Stained slides were scanned using a Silverfast PathscanEnabler IV scanner. Scanned images were analyzed with Image-Jsoftware, which permits precise tracing of an image and calcula-tion of its surface area. The percentage of remaining tissue from asingle section was calculated by tracing the perimeter of theremaining brain tissue, determining its surface area, dividing thisby an estimate of the total surface area of the section (taken bytracing both the remaining tissue and the estimated perimeter ofthe necrotic cavity), and multiplying by 100. The percentage ofremaining tissue for each rat was determined by averaging thepercentage of remaining tissue across the six selected sections.

2.4.8. ImmunohistochemistryHistologic cryosections (20 mm in thickness) were selected for

immunohistochemical analysis and immunolabeled for glial fibril-lary acidic protein (GFAP) and microglia-specific protein Iba1. Sec-tions were air-dried for 1 h and washed in 1� phosphate bufferedsaline (PBS) 5 times, then incubated inwith protein blocking serumat room temperature (RT) for 1 h. Sections were then incubatedwith primary antibody GFAP (1:200, Millipore, Billarica, MA, USA)and Iba1 (1:1000, Wako, Richmond, VA, USA) for 1 h at RT and thenwashedwith 1X PBS 5 times followed by incubationwith secondarygoat anti mouse (1:100, Alexa Fluor 488 IgG (H þ L)), A11001,Invitrogen, Carlsbad, CA, USA) and goat anti rabbit (1:100, lifetechnologies IgG (H þ L), F2765, for 30 min at RT. Slides werewashed with 1X PBS 5 times and covered with mounting medium(Vectashield). The slides were examined under a fluorescence mi-croscope and pictures were taken from the peri-contusion areausing a digital camera mounted on a microscope (Olympus BX41,Olympus America, Inc., Center Valley, PA, USA).

2.4.9. StatisticsAll results were expressed as mean ± standard error of the mean

and calculations were obtained using SPSS 11.0 software. Allbehavioral data were analyzed by repeated-measures analysis ofvariance (ANOVA) followed by a Tukey’s post-hoc test for individualcomparisons among the groups. Other results were analyzed withone-way ANOVA followed by LSD post-tests for multiple compar-isons. The criterion for statistical significance was set at p < 0.05.

3. Results

3.1. PK study of EIDD-1723 in rats

EIDD-1723 was dosed i.m. at 3, 10 and 30 mg/kg and theresulting concentration versus time plots (Fig. 2A) show that theparent compound is rapidly cleaved after an i.m. injection andconverted to the active metabolite EIDD-036, which was observedin plasma in a dose-dependent manner and demonstrated first-

Fig. 2. Time vs. concentration graphs of EIDD-036 in rats after a single dose of prodrugEIDD-1723. Plasma concentrations show dose-dependent pharmacokinetics (A) afteroral dosing. The concentrations in brain tissue (B) over time after a 10-mg/kg doseclosely matched those found in plasma, suggesting rapid blood-brain barrierpenetration.

B. Wali et al. / Neuropharmacology 109 (2016) 148e158152

order elimination kinetics. The parent compound EIDD-1723 wasnot detected in plasma at any time point. A fourth group of 4 ratswas dosed i.v. at 2 mg/kg to provide a reference for bioavailabilitycalculations, and a fifth group was dosed p.o. at 30 mg/kg. Resultsshowed that the compound is not orally available.

Brain tissue was collected from all animals 10 h after dosing. Thegroup dosed i.m. with EIDD-1723 at 30 mg/kg showed a meanconcentration of EIDD-036 of 29.4 ± 13.1 ng/g. Concentrations in allother groups was below the limit of detection. The time versusconcentration graph in Fig. 2B shows that levels of EIDD-036 werepresent in concentrations that were slightly lower than those inplasma, thus confirming that it readily crosses the blood-brainbarrier.

Fig. 3. Effect of EIDD-1723 and progesterone on edema level 24 h post-TBI. Testcompounds (10 mg/kg) were administered intramuscularly. Animals treated withprogesterone and EIDD-1723 show significantly less brain edema compared to vehicle.Values are expressed as mean ± SE. *CCI þ Veh vs. Sham þ Veh; # CCI þ Veh vs.CCI þ PROG or EIDD-1723.

3.2. Brain water content (edema)

Brain water content (BWC) was measured to evaluate the po-tential efficacy of the EIDD-1723 analog relative to PROG inreducing cerebral edema following injury. In order to use the mosteffective dose in our proof-of-concept study, we first conducted adose-response study of PROG (3, 10, and 30 mg/kg) efficacy (datanot shown in figures) to reduce edema at 24 h post-CCI. We founddecreases in edema of 30.88, 58.17 and 49.80% with 3, 10, and30 mg/kg doses of PROG, respectively. The most effective dose ofPROG, 10 mg/kg, was used to compare the efficacy at the same doseof EIDD-1723.

Significantly (p < 0.05) greater BWC was found in thelesion þ vehicle group around the peri-contusional site comparedto the sham-vehicle group (Fig. 3). All surgical groups showedhigher BWC than the sham group. Compared with the vehicle

group, both PROG and EIDD-1723 treatments reduced BWC tosignificantly (p < 0.05) lower levels. EIDD-1723 showed equivalentefficacy to PROG in attenuating BWC/edema.

3.3. Functional outcome study

3.3.1. Grip strengthI.m. administration of EIDD-1723 improved grip strength and

showed equivalent efficacy to PROG following CCI. Peak forceexerted to grip the bar assembly showed significant group[F(3,28) ¼ 13.26, p < 0.001] effects. Grip strength decreasedsignificantly (p < 0.05) in rats subjected to CCI (8.13 ± 0.64 N,10.06 ± 0.29 N, 11.38 ± 0.21 N, at 4, 9 and 21 d post-injury,respectively) at all time points compared to sham-operated rats(12.05 ± 0.51 N,13.26 ± 0.58 N,15.09 ± 0.72 N, at 4, 9, and 21 d post-surgery, respectively). Post-hoc analyses showed that repeatedtreatments with 10 mg/kg PROG and EIDD-1723 after CCI signifi-cantly (p < 0.05) improved grip strength at 4, 9 and 21 d (Fig. 4).

3.3.2. Sensory neglectI.m. administration of EIDD-1723 improved sensory neglect and

showed equivalent efficacy to PROG following CCI. Latency toremove adhesive tape from the forepaw showed significant group[F(3,28)¼ 15.11, p¼ 0.001] effects. There were significant (p < 0.05)deficits in latency to remove the sticky tape in CCI- þ vehicle-treated rats (108.13 ± 11.87, 66.13 ± 10.99, 41.25 ± 13.75, at 4, 9 and21 d post-CCI, respectively) at all time points compared to shams(10.25 ± 1.50, 12.13 ± 2.31, 6.63 ± 2.53, at 4, 9, and 21 d post-surgery). Repeated i.m. treatments with PROG and EIDD-1723 af-ter CCI significantly (p < 0.05) decreased the latency to remove thesticker at 4 and 9 d but not at 21d post-CCI when all animals hadrecovered on this task (Fig. 5).

3.3.3. Spatial navigation performance in the MWMThis task allows measurement of working (short-term, trial-to-

trial) and reference (longer-term, day-to-day) memory. During the7 days consecutive days of testing the latency to escape the watermaze and find the platform was measured, and the data wererecorded for two trials 5 min apart. The first trial of the day wastaken to be ameasure of reference or long-termmemory (therewasa 24-h delay between the last trial of the day and the first trial of thenext day). After the CCI, we found significant cognitive impairmentsin the vehicle group in MWM learning. Repeated measures one-way ANOVA on duration showed significant group effects on Trial

Fig. 4. EIDD-1723 improves grip strength in rats with controlled cortical impact (CCI)injury. Values are expressed as mean ± SE. The rats subjected to CCI þ vehicle treat-ment showed significantly (p < 0.05) lower grip-strength scores at all time pointscompared to sham þ vehicle rats. Progesterone- (PROG) and EIDD-1723-treated ratsshowed significantly (p < 0.05) improved grip compared to the vehicle group.

Fig. 5. EIDD-1723 improves sensory neglect in rats with controlled cortical impact(CCI) injury. Values are expressed as mean ± SE. The rats subjected to CCI þ vehicletreatment showed significantly (p < 0.05) higher sensory scores at all time pointscompared to sham þ vehicle rats. The progesterone- (PROG) and EIDD-1723-treatedgroups showed significantly (p < 0.05) lower latency compared to the vehicle group.

Fig. 6. Comparative effect of progesterone (PROG) and EIDD-1723 on spatial learningand memory tasks. During Trials 1 (A) and 2 (B) there was a significant effect ofcontrolled cortical impact (CCI) injury and intramuscular treatments with PROG andEIDD-1723. EIDD-1723 treatment decreased the mean latency to find a hidden plat-form in the Morris water maze compared to the vehicle-treated group. Values areexpressed as mean ± SE. *CCI þ Veh vs. Sham þ Veh; # CCI þ Veh vs. CCI þ PROG orEIDD-1723.

B. Wali et al. / Neuropharmacology 109 (2016) 148e158 153

1 (F(3, 28) ¼ 5.92, p < 0.05) and Trial 2 (F(3, 28) ¼ 7.45, p < 0.001).Post-hoc tests revealed a significant difference between

CCI þ vehicle and sham þ vehicle (p < 0.05) and a significant dif-ference between EIDD-1723 and vehicle-treated groups (p < 0.05).Intramuscular treatment with PROG and EIDD-1723 showed pro-gressive reduction in time taken to reach the hidden platform thatbecame significant (p < 0.05) at day 7 compared with vehiclegroups on both Trial 1 and Trial 2 (Fig. 6). EIDD-1723 showedequivalent efficacy to PROG in latency to escape thewatermaze andfind the platform.

Fig. 7. Lesion reconstruction analysis of brain sections after controlled cortical impact(CCI) and treatment with progesterone (PROG), EIDD-1723 or vehicle. A. Representa-tive brain slices from injured animals injected with vehicle, PROG, or EIDD-1723. B.Mean percent of volumetric tissue loss at 3 weeks post-injury. Values are expressed asmean ± SE. *CCI þ Veh vs. CCI þ PROG or EIDD-1723.

3.4. Necrotic cavity

Briefly, the 6 sections from each rat brain were photographedwith an Epson scanner and ImagePro™ software. Group means ofpercent of damaged tissue averaged across 6 anterior-posteriorsections were used to determine overall lesion size. There was asignificant main effect of lesion/treatment on lesion size (F(2,21) ¼ 3.53, p < 0.05). Post-hoc comparisons further revealed asignificant difference between the CCI þ vehicle and PROG-treatedgroups (p < 0.05) and between the EIDD-1723 and vehicle-treatedgroups (p < 0.05) (Fig. 7).

B. Wali et al. / Neuropharmacology 109 (2016) 148e158154

3.5. Gliosis

The immunohistochemical assay showed a significant increasein astrogliosis (GFAP immunoreactivity) and microgliosis (Iba1immunoreactivity) in the brain tissues of the vehicle group. BothPROG and EIDD-1723 treatments markedly decreased the intensityof GFAP (Fig. 8) and Iba1 (Fig. 9) at 3 weeks post-injury compared tovehicle (p < 0.05). EIDD-1723 showed equivalent efficacy to PROGin attenuating gliosis.

4. Discussion

We compared the neuroprotective actions of a novel PROGanalog, EIDD-1723, to those of PROG. In this proof-of-concept study,compared to vehicle-treated animals, EIDD-1723 showed signifi-cantly better edema reduction, gait improvement, cognitive re-covery, and smaller lesion volumes that were closely comparable tothose following PROG treatment.

Fig. 8. Immunostaining of glial fibrillary acidic protein (GFAP, a marker of astrocytes) in tmagnification (10�) and B, E, H and K at higher magnification (40�). Immunostaining picturgroup (A, E, I). Treatment with PROG (C, G, K) and EIDD-1723 (D, H, L) significantly reducquantification of GFAP intensity. Data expressed as mean ± SE. *CCI þ Veh vs. Sham þ Veh

Multiple studies have reported PROG’s edema-attenuating ef-fects in a number of animal models of brain injury (Guo et al., 2006;Maghool et al., 2013). An inverse correlation between serum PROGlevel and degree of edema has been reported (Wright et al., 2001).More recently it was demonstrated that compared to dexametha-sone, a corticosteroid drug commonly used to reduce brain swellingduring neurosurgical procedures, PROG had no side effects and wasequally effective in reducing brain edema and more effective inattenuating acute cellular inflammatory responses (Xu et al., 2014).

In previous studies evaluating edema by measuring BWC, wescreened multiple PROG analogs for efficacy in a rat CCI model(MacNevin et al., 2009; Guthrie et al., 2012). For the current studywe chose the i.m. route of administration because any field-readytherapy for TBI will likely need to be delivered i.m. in an aqueousformulation to allow for rapid release into circulation. Three small,single-center clinical studies have now shown positive outcomeswith low-dose i.m. PROG (Xiao et al., 2008; Mofid et al., 2016;Raheja et al., 2016), but large phase III clinical trials using i.v.

he peri-contusional area at 22 days post-traumatic brain injury. A, D, G and J at lowes showed higher intensity of GFAP in the vehicle group (B, F, J) compared to the controled the expression of astrocytes. The scale bar represents 10 mm. Graph (M) showing; # CCI þ Veh vs. CCI þ PROG or EIDD-1723.

Fig. 9. Immunofluorescence of Iba1 (a marker of gliosis) in the peri-contusional area at 22 days post-traumatic brain injury. A, B, C and D at low magnification (10�) and E, F, G andH at higher magnification (40�). Immunostaining pictures showed higher intensity of Iba1 in the vehicle group (B, F, J) compared to the control group (A, E, I). Treatment withprogesterone (C, G, K) and EIDD-1723 (D, H, L) significantly reduced the expression of Iba1. The scale bar represents 10 mm. Graph (M) showing quantification of Iba1 intensity at 22days post-injury. Data expressed as mean ± SE. *CCI þ Veh vs. Sham þ Veh; # CCI þ Veh vs. CCI þ PROG or EIDD-1723.

B. Wali et al. / Neuropharmacology 109 (2016) 148e158 155

administration at much higher doses of the hormone had negativeresults (Wright et al., 2007, 2014; Skolnick et al., 2014).

We compared the efficacy of the 10-mg dose of EIDD-1723 tothat of PROG at the same dose. EIDD-1723 showed equivalent ef-ficacy to PROG in attenuating BWC/edema. We found significantdifferences in necrotic cavity size in the injury-alone compared tothe PROG- and EIDD-1723-treated groups. The reduction of theinjury-induced necrotic cavity in the present study is likely theresult of the ability of EIDD-1723 to attenuate edema during theacute phase of injury. The current results are in agreement with ourearlier work showing that, when given after bilateral brain injury,PROG analogs have functional and edema-attenuating activity andsupport the potential utility of these compounds for the treatmentof brain injury (MacNevin et al., 2009; Guthrie et al., 2012).

Although measurement of edema is useful in the screening andevaluation of efficacy, it alone cannot provide a prognosis con-cerning the extent of, and recovery from, cognitive, sensory, andmotor deficits. Behavioral outcome measures are essential to this

purpose. PROG has been shown to promote functional recovery indifferent brain injury models including stroke and TBI (Djebailiet al., 2004; Gibson and Murphy, 2004; Grossman et al., 2004;Gibson et al., 2005; Chen et al., 2007; Wali et al., 2011; Tanget al., 2013; Geddes et al., 2014; Robertson and Saraswati, 2014;Wali et al., 2014; Yousuf et al., 2014a,b; Peterson et al., 2015). Ourresults in the present study show that besides reducing the watercontent in the penumbral region at one day after TBI, EIDD-1723significantly improved functional recovery and reduced accumu-lation of astrocytes and microglia, cells associated with post-injuryinflammatory processes that can impair recovery of function.

In TBI patients, sensorimotor and cognitive deficits are wellrecognized (Writer and Schillerstrom, 2009; Sun and Feng, 2014).The sensory neglect test is used in rats to evaluate the sensory-motor deficits caused by brain injury (Bouet et al., 2009) and itcan be used to assess long-lasting impairments and recovery inanimal models of TBI (Teasdale et al., 1998). In the present study wefound that PROG and EIDD-1723 treatment ameliorated sensory

B. Wali et al. / Neuropharmacology 109 (2016) 148e158156

neglect at 4 and 9 days after CCI. The rats’ grip strength, impaired bythe injury, was also significantly improved in the PROG- and EIDD-1723-treated groups.

Cognitive impairments, such as the partial inability to form orstore new memories, and visuo spatial deficits are among the mostcommon characteristics of mild to severe TBI (Dikmen et al., 2009;Sigurdardottir et al., 2009; Bagiella et al., 2010; Gerrard-Morriset al., 2010). The MWM test is widely used to assess these cogni-tive functional deficits. The hidden platform test employed in thepresent study is considered a reliable measure of spatial workingmemory (Janus, 2004). Our results showing that EIDD-1723 treat-ment improved spatial learning performance in the MWM task arecomparable to the PROG effects observed, and in agreement withprevious reports demonstrating PROG-mediated recovery andimproved behavioral performance in rats (Roof et al., 1994). In thepresent study, the brain-injured groups were impaired on theacquisition phase of MWM testing compared to the control group.However, the PROG- and EIDD-1723 treated groups showedimproved performance in finding the hidden platform over theVehicle animals.

After brain injury, including TBI, GFAP expression increases inastrocytes, and activated astrocytes release inflammatory cytokinesand disrupt functional recovery (Hoane et al., 2003a,b). A largebody of literature now demonstrates PROG is a potent anti-inflammatory agent (Stein, 2008; Schumacher et al., 2015) thatcan modulate microglial activity after brain injury (Drew andChavis, 2000). The analysis of reactive gliosis around the lesionshowed that PROG and EIDD-1723 markedly decreased gliosis(GFAP and Iba1 immunostaining). Previous studies have shownPROG-induced reduction in GFAP expression following TBI (Liuet al., 2014). These results suggest that EIDD-1723 administrationfollowing CCI significantly reduces tissue damage at the site ofinjury. The frontal cortex is especially vulnerable to workingmemory deficits and deficits on the adhesive removal test (Hoaneet al., 2003a,b), so the reduction in lesion size is likely respon-sible for the protection from behavioral impairments in the sensoryneglect and spatial navigation tasks.

We also found that after i.m. injection, EIDD-1723 is rapidlycleaved, generates high levels of the active steroid EIDD-036, andpasses through the blood-brain barrier at therapeutic levels, acritical property for a TBI treatment. EIDD-036 levels in brain wereapproximately the same as those in plasma at any given time point,indicating that the material readily penetrates the blood-brainbarrier. After i.m. injection, only EIDD-036 was observed inplasma, while the parent conjugate, EIDD-1723, was neverobserved in levels above the limit of detection. Our previous studieshave shown the intermediate oximes to be water-insoluble(<15 mg/ml) but intrinsically active metabolites of C-20 PROG an-alogs (Guthrie et al., 2012).

PROG is a pleiotropic hormone that uses multiple signalingpathways including the regulation of gene expression after bindingto intracellular progesterone receptors (PR). In addition to theclassical PR, PROG also interacts with other signal transductionmechanisms such as the s1 receptor, for which it is a competitiveinhibitor and through which it may reduce N-methyl-D-aspartate(NMDA) glutamate signaling (Hanner et al., 1996; Bergeron et al.,1999). PROG also signals at the nicotinic acetylcholine receptor(nAChR) (Valera et al., 1992) and affects gamma-aminobutyric acid(GABA) through its 5a-reduced metabolite allopregnanolone andpositive modulation of the GABAA receptor (Belelli et al., 2002;Losel et al., 2003; Mani, 2006; Schumacher et al., 2007). PROG isalso known to activate the pregnane X receptor (PXR) and themembrane surface receptor 25-Dx (Meffre et al., 2005; Guennounet al., 2008). Our previous findings have shown that progestin-mediated pro-survival response following TBI is regulated either

independently of the classical PR or via nongenomic PR-regulatedactions (Vanlandingham et al., 2006). Taken together, these find-ings suggest that synthetic steroids with physicochemical proper-ties similar to those of natural forms can function at the molecularlevel by mechanisms other than classically mediated transcription.The observation that EIDD-1723 and PROG have comparable func-tional effects suggests that they may act through some of the samemechanisms, but it is also possible that PROG analogs are actingthrough a novel receptor mechanism not yet defined. The focus ofthis study was on comparing the novel analog with PROG’s efficacyon the enhancement of functional recovery. The molecular mech-anisms of EIDD-1723-mediated neuroprotection will requirefurther attention.

Our results demonstrate that EIDD-1723 is equally as effective asPROG in attenuating edema and lesion volume and promotingmotor improvement and cognitive recovery. Importantly, EIDD-1723 is highly soluble and stable in water-based formulations,and can therefore be readily administered in the field by emergencypersonnel. Like the i.v. route, i.m. administration bypasses the he-patic metabolism, avoiding the first-pass effect and resulting inhigher bioavailability. These are crucial properties, as the failure ofvery high-dose, i.v. PROG in the phase III clinical trials (Skolnicket al., 2014; Wright et al., 2014), in addition to other dosing(Howard et al., 2015) and trial design factors, may have been due inpart to the long interval between injury and initial treatmentnecessitated by formulation requirements. This new analog ofPROG will have the same therapeutic benefits of PROG, but withoutthe inherent limitations for administration in emergency situa-tions. The ability of first responders to rapidly administer EIDD-1723 in the field makes it an attractive compound to consider forTBI and/or stroke treatment.

Conflict of interest

IS, DCL, DG, MN and DGS along with Emory University retainpatents related to the use of progesterone analogs and its uses buthave no financial gains, royalties or licensing agreements fromresearch on progesterone analogs. BW and NT declare no conflict ofinterest.

Acknowledgments

This work was supported by Georgia Research Alliance funding(Grant # VL14.C13).

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.neuropharm.2016.05.017.

References

Bagiella, E., Novack, T.A., Ansel, B., Diaz-Arrastia, R., Dikmen, S., Hart, T., Temkin, N.,2010. Measuring outcome in traumatic brain injury treatment trials: recom-mendations from the traumatic brain injury clinical trials network. J. HeadTrauma Rehabil. 25, 375e382.

Belelli, D., Casula, A., Ling, A., Lambert, J.J., 2002. The influence of subunit compo-sition on the interaction of neurosteroids with GABA(A) receptors. Neuro-pharmacology 43, 651e661.

Bergeron, R., de Montigny, C., Debonnel, G., 1999. Pregnancy reduces brain sigmareceptor function. Br. J. Pharmacol. 127, 1769e1776.

Bouet, V., Boulouard, M., Toutain, J., Divoux, D., Bernaudin, M., Schumann-Bard, P.,Freret, T., 2009. The adhesive removal test: a sensitive method to assesssensorimotor deficits in mice. Nat. Protoc. 4, 1560e1564.

Chen, G., Shi, J., Ding, Y., Yin, H., Hang, C., 2007. Progesterone prevents traumaticbrain injury-induced intestinal nuclear factor kappa B activation and proin-flammatory cytokines expression in male rats. Mediat. Inflamm. 2007, 93431.

Cutler, S.M., Pettus, E.H., Hoffman, S.W., Stein, D.G., 2005. Tapered progesteronewithdrawal enhances behavioral and molecular recovery after traumatic brain

B. Wali et al. / Neuropharmacology 109 (2016) 148e158 157

injury. Exp. Neurol. 195, 423e429.Dikmen, S.S., Corrigan, J.D., Levin, H.S., Machamer, J., Stiers, W., Weisskopf, M.G.,

2009. Cognitive outcome following traumatic brain injury. J. Head. TraumaRehabil. 24, 430e438.

Djebaili, M., Hoffman, S.W., Stein, D.G., 2004. Allopregnanolone and progesteronedecrease cell death and cognitive deficits after a contusion of the rat pre-frontalcortex. Neuroscience 123, 349e359.

Drew, P.D., Chavis, J.A., 2000. Female sex steroids: effects upon microglial cellactivation. J. Neuroimmunol. 111, 77e85.

Geddes, R.I., Sribnick, E.A., Sayeed, I., Stein, D.G., 2014. Progesterone treatmentshows benefit in a pediatric model of moderate to severe bilateral brain injury.PLoS One 9, e87252.

Gerrard-Morris, A., Taylor, H.G., Yeates, K.O., Walz, N.C., Stancin, T., Minich, N.,Wade, S.L., 2010. Cognitive development after traumatic brain injury in youngchildren. J. Int. Neuropsychol. Soc. 16, 157e168.

Gibson, C.L., Constantin, D., Prior, M.J., Bath, P.M., Murphy, S.P., 2005. Progesteronesuppresses the inflammatory response and nitric oxide synthase-2 expressionfollowing cerebral ischemia. Exp. Neurol. 193, 522e530.

Gibson, C.L., Murphy, S.P., 2004. Progesterone enhances functional recovery aftermiddle cerebral artery occlusion in male mice. J. Cereb. Blood Flow. Metab. 24,805e813.

Gould, S., Scott, R.C., 2005. 2-Hydroxypropyl-beta-cyclodextrin (HP-beta-CD): atoxicology review. Food Chem. Toxicol. 43, 1451e1459.

Grossman, K.J., Goss, C.W., Stein, D.G., 2004. Effects of progesterone on the in-flammatory response to brain injury in the rat. Brain Res. 1008, 29e39.

Guennoun, R., Meffre, D., Labombarda, F., Gonzalez, S.L., Deniselle, M.C., Stein, D.G.,De Nicola, A.F., Schumacher, M., 2008. The membrane-associated progesterone-binding protein 25-Dx: expression, cellular localization and up-regulation afterbrain and spinal cord injuries. Brain Res. Rev. 57, 493e505.

Guo, Q., Sayeed, I., Baronne, L.M., Hoffman, S.W., Guennoun, R., Stein, D.G., 2006.Progesterone administration modulates AQP4 expression and edema aftertraumatic brain injury in male rats. Exp. Neurol. 198, 469e478.

Guthrie, D.B., Stein, D.G., Liotta, D.C., Lockwood, M.A., Sayeed, I., Atif, F.,Arrendale, R.F., Reddy, G.P., Evers, T.J., Marengo, J.R., Howard, R.B., Culver, D.G.,Natchus, M.G., 2012. Water-soluble progesterone analogues are effective,injectable treatments in animal models of traumatic brain injury. ACS Med.Chem. Lett. 3, 362e366.

Hanner, M., Moebius, F.F., Flandorfer, A., Knaus, H.G., Striessnig, J., Kempner, E.,Glossmann, H., 1996. Purification, molecular cloning, and expression of themammalian sigma1-binding site. Proc. Natl. Acad. Sci. U. S. A. 93, 8072e8077.

Harvey, K.A., Xu, Z., Pavlina, T.M., Zaloga, G.P., Siddiqui, R.A., 2015. Modulation ofendothelial cell integrity and inflammatory activation by commercial lipidemulsions. Lipids health Dis. 14, 9.

Hoane, M.R., Akstulewicz, S.L., Toppen, J., 2003a. Treatment with vitamin B3 im-proves functional recovery and reduces GFAP expression following traumaticbrain injury in rats. J. Neurotrauma 20, 1189e1199.

Hoane, M.R., Knotts, A.A., Akstulewicz, S.L., Aquilano, M., Means, L.W., 2003b. Thebehavioral effects of magnesium therapy on recovery of function followingbilateral anterior medial cortex lesions in the rat. Brain Res. Bull. 60, 105e114.

Howard, R.B., Sayeed, I., Stein, D., 2015 Sep 15. Suboptimal dosing parameters aspossible factors in the negative Phase III clinical trials of progesterone in TBI.J. Neurotrauma. http://dx.doi.org/10.1089/neu.2015.4179 [Epub ahead of print].

Janus, C., 2004. Search strategies used by APP transgenic mice during navigation inthe Morris water maze. Learn. Mem. 11, 337e346.

Liu, F., Liao, F., Li, W., Han, Y., Liao, D., 2014. Progesterone alters Nogo-A, GFAP andGAP-43 expression in a rat model of traumatic brain injury. Mol. Med. Rep. 9,1225e1231.

Losel, R.M., Falkenstein, E., Feuring, M., Schultz, A., Tillmann, H.C., Rossol-Haseroth, K., Wehling, M., 2003. Nongenomic steroid action: controversies,questions, and answers. Physiol. Rev. 83, 965e1016.

Lou, P.H., Lucchinetti, E., Zhang, L., Affolter, A., Gandhi, M., Zhakupova, A.,Hersberger, M., Hornemann, T., Clanachan, A.S., Zaugg, M., 2015. Propofol(Diprivan(R)) and Intralipid(R) exacerbate insulin resistance in type-2 diabetichearts by impairing GLUT4 trafficking. Anesth. Analg. 120, 329e340.

MacNevin, C.J., Atif, F., Sayeed, I., Stein, D.G., Liotta, D.C., 2009. Development andscreening of water-soluble analogues of progesterone and allopregnanolone inmodels of brain injury. J. Med. Chem. 52, 6012e6023.

Maghool, F., Khaksari, M., Siahposht Khachki, A., 2013. Differences in brain edemaand intracranial pressure following traumatic brain injury across the estrouscycle: involvement of female sex steroid hormones. Brain Res. 1497, 61e72.

Mani, S.K., 2006. Signaling mechanisms in progesterone-neurotransmitter in-teractions. Neuroscience 138, 773e781.

Meffre, D., Delespierre, B., Gouezou, M., Leclerc, P., Vinson, G.P., Schumacher, M.,Stein, D.G., Guennoun, R., 2005. The membrane-associated progesterone-binding protein 25-Dx is expressed in brain regions involved in water ho-meostasis and is up-regulated after traumatic brain injury. J. Neurochem. 93,1314e1326.

Mermelstein, F., Hamilton, D.A., Wright, C., Lacouture, P.G., Ramaiya, A., Carr, D.B.,2013. Single-dose and multiple-dose pharmacokinetics and dose proportion-ality of intravenous and intramuscular HPbetaCD-diclofenac (Dyloject)compared with other diclofenac formulations. Pharmacotherapy 33, 1012e1021.

Mofid, B., Soltani, Z., Khaksari, M., Shahrokhi, N., Nakhaee, N., Karamouzian, S.,Ahmadinejad, M., Maiel, M., Khazaeli, P., 2016. What are the progesterone-induced changes of the outcome and the serum markers of injury, oxidantactivity and inflammation in diffuse axonal injury patients? Int.

Immunopharmacol. 32, 103e110.Muller, S.H., Diaz, J.H., Kaye, A.D., 2015. Clinical applications of intravenous lipid

emulsion therapy. J. Anesth. 29, 920e926.Peterson, B.L., Won, S., Geddes, R.I., Sayeed, I., Stein, D.G., 2015. Sex-related differ-

ences in effects of progesterone following neonatal hypoxic brain injury. Behav.Brain Res. 286, 152e165.

Preskorn, S.H., 2005. Pharmacokinetics and therapeutics of acute intramuscularziprasidone. Clin. Pharmacokinet. 44, 1117e1133.

Raheja, A., Sinha, S., Samson, N., Bhoi, S., Subramanian, A., Sharma, P., Sharma, B.S.,2016. Serum biomarkers as predictors of long-term outcome in severe trau-matic brain injury: analysis from a randomized placebo-controlled Phase IIclinical trial. J. Neurosurg. 1e11.

Robertson, C.L., Saraswati, M., 2014. Progesterone protects mitochondrial functionin a rat model of pediatric traumatic brain injury. J. Bioenerg. Biomembr. 47,43e51.

Roof, R.L., Duvdevani, R., Braswell, L., Stein, D.G., 1994. Progesterone facilitatescognitive recovery and reduces secondary neuronal loss caused by corticalcontusion injury in male rats. Exp. Neurol. 129, 64e69.

Schumacher, M., Denier, C., Oudinet, J.P., Adams, D., Guennoun, R., 2015. Proges-terone neuroprotection: the background of clinical trial failure. J. Steroid Bio-chem. Mol. Biol. 160, 53e66.

Schumacher, M., Guennoun, R., Ghoumari, A., Massaad, C., Robert, F., El-Etr, M.,Akwa, Y., Rajkowski, K., Baulieu, E.E., 2007. Novel perspectives for progesteronein hormone replacement therapy, with special reference to the nervous system.Endocr. Rev. 28, 387e439.

Sigurdardottir, S., Andelic, N., Roe, C., Schanke, A.K., 2009. Cognitive recovery andpredictors of functional outcome 1 year after traumatic brain injury. J. Int.Neuropsychol. Soc. 15, 740e750.

Skolnick, B.E., Maas, A.I., Narayan, R.K., van der Hoop, R.G., MacAllister, T., Ward, J.D.,Nelson, N.R., Stocchetti, N., the, S.T.I., 2014. A clinical trial of progesterone forsevere Traumatic Brain Injury. N. Engl. J. Med. 371, 2467e2476.

Stein, D.G., 2008. Progesterone exerts neuroprotective effects after brain injury.Brain Res. Rev. 57, 386e397.

Stein, D.G., 2011. Progesterone in the treatment of acute traumatic brain injury: aclinical perspective and update. Neuroscience 191, 101e106.

Stein, D.G., 2015. Embracing failure: what the Phase III progesterone studies canteach about TBI clinical trials. Brain Inj. 29, 1259e1272.

Sun, Z.L., Feng, D.F., 2014. Biomarkers of cognitive dysfunction in traumatic braininjury. J. Neural Transm. (Vienna) 121, 79e90.

Taftachi, F., Sanaei-Zadeh, H., Sepehrian, B., Zamani, N., 2012. Lipid emulsion im-proves Glasgow coma scale and decreases blood glucose level in the setting ofacute non-local anesthetic drug poisoningea randomized controlled trial. Eur.Rev. Med. Pharmacol. Sci. 16 (Suppl. 1), 38e42.

Tang, H., Hua, F., Wang, J., Sayeed, I., Wang, X., Chen, Z., Yousuf, S., Atif, F., Stein, D.G.,2013. Progesterone and vitamin D: improvement after traumatic brain injury inmiddle-aged rats. Horm. Behav. 64, 527e538.

Teasdale, G.M., Pettigrew, L.E., Wilson, J.T., Murray, G., Jennett, B., 1998. Analyzingoutcome of treatment of severe head injury: a review and update on advancingthe use of the Glasgow outcome scale. J. Neurotrauma 15, 587e597.

Valera, S., Ballivet, M., Bertrand, D., 1992. Progesterone modulates a neuronalnicotinic acetylcholine receptor. Proc. Natl. Acad. Sci. U. S. A. 89, 9949e9953.

Vanlandingham, J.W., Cutler, S.M., Virmani, S., Hoffman, S.W., Covey, D.F.,Krishnan, K., Hammes, S.R., Jamnongjit, M., Stein, D.G., 2006. The enantiomer ofprogesterone acts as a molecular neuroprotectant after traumatic brain injury.Neuropharmacology 51, 1078e1085.

Wali, B., Ishrat, T., Won, S., Stein, D.G., Sayeed, I., 2014. Progesterone in experimentalpermanent stroke: a dose-response and therapeutic time-window study. Brain137, 486e502.

Wali, B., Sayeed, I., Stein, D.G., 2011. Improved behavioral outcomes after proges-terone administration in aged male rats with traumatic brain injury. Rest. Nurol.Neurosci. 29, 61e71.

Wasan, K.M., Grossie Jr., V.B., Lopez-Berestein, G., 1994. Effects of intralipid infusionon rat serum lipoproteins. Lab. Anim. 28, 138e142.

Wright, D.W., Bauer, M.E., Hoffman, S.W., Stein, D.G., 2001. Serum progesteronelevels correlate with decreased cerebral edema after traumatic brain injury inmale rats. J. Neurotrauma 18, 901e909.

Wright, D.W., Kellermann, A.L., Hertzberg, V.S., Clark, P.L., Frankel, M.,Goldstein, F.C., Salomone, J.P., Dent, L.L., Harris, O.A., Ander, D.S., Lowery, D.W.,Patel, M.M., Denson, D.D., Gordon, A.B., Wald, M.M., Gupta, S., Hoffman, S.W.,Stein, D.G., 2007. ProTECT: a randomized clinical trial of progesterone for acutetraumatic brain injury. Ann. Emerg. Med. 49, 391e402.

Wright, D.W., Yeatts, S.D., Silbergleit, R., Palesch, Y.Y., Hertzberg, V.S., Frankel, M.,Goldstein, F.C., Caveney, A.F., Howlett-Smith, H., Bengelink, E.M., Manley, G.T.,Merck, L.H., Janis, L.S., Barsan, W.G., Investigators, N., 2014. Very early admin-istration of progesterone for acute traumatic brain injury. N. Engl. J. Med. 371,2457e2466.

Writer, B.W., Schillerstrom, J.E., 2009. Psychopharmacological treatment forcognitive impairment in survivors of traumatic brain injury: a critical review.J. Neuropsychiatry Clin. Neurosci. 21, 362e370.

Xiao, G., Wei, J., Yan, W., Wang, W., Lu, Z., 2008. Improved outcomes from theadministration of progesterone for patients with acute severe traumatic braininjury: a randomized controlled trial. Crit. CZre 12, R61.

Xu, F.F., Sun, S., Ho, A.S., Lee, D., Kiang, K.M., Zhang, X.Q., Wang, A.M., Wu, E.X.,Lui, W.M., Liu, B.Y., Leung, G.K., 2014. Effects of progesterone vs. dexamethasoneon brain oedema and inflammatory responses following experimental brain

B. Wali et al. / Neuropharmacology 109 (2016) 148e158158

resection. Brain Inj. 28, 1594e1601.Yousuf, S., Atif, F., Sayeed, I., Tang, H., Stein, D.G., 2014a. Progesterone in transient

ischemic stroke: a dose-response study. Psychopharmacol. (Berl.) 231,3313e3323.

Yousuf, S., Sayeed, I., Atif, F., Tang, H., Wang, J., Stein, D.G., 2014b. Delayed proges-terone treatment reduces brain infarction and improves functional outcomesafter ischemic stroke: a time-window study in middle-aged rats. J. Cereb. BloodFlow. Metab. 34, 297e306.