A Novel, Long-Acting, Small Molecule PKM2 Activator and Its Potential Broad Application Against Photoreceptor Degeneration

Abstract

Purpose

Activating pyruvate kinase M2 (PKM2) has been shown to be neuroprotective in preclinical models of photoreceptor degeneration. We recently developed novel, small molecule activators for ocular delivery. Here, we sought to characterize the ocular pharmacology, toxicity, and efficacy of MCTI-566, a novel PKM2 activator, to translate this therapeutic strategy to the clinic.

Methods

X-ray protein crystallography and isothermal titration calorimetry assessed the interaction of MCTI-566 with PKM2. PKM2 activation and tissue pharmacokinetics were examined after intravitreal or systemic administration of MCTI-566. Retinal toxicity was evaluated in rats after intravitreal injection. The effect of MCTI-566 on photoreceptor death was assessed using in vitro and in vivo models of outer retinal stress and on the inflammatory response in the rd10 retina using flow cytometry and quantitative real-time polymerase chain reaction.

Results

The PKM2-MCTI-566 co-crystal structure demonstrated a binding pocket distinct from endogenous activators. MCTI-566 increases retinal PK activity 200% following intravitreal or systemic administration. MCTI-566 distributed to the retina after intravitreal or systemic administration, activated the target for ≥90 days and was specific for photoreceptor PKM2. No retinal toxicity was observed after repeated intravitreal administration. MCTI-566 reduced photoreceptor apoptosis in a model of retinal detachment, and delayed photoreceptor degeneration and altered the inflammatory response in the rd10 retina.

Conclusions

MCTI-566 is a small molecule drug candidate for photoreceptor neuroprotection.

Translational Relevance

MCTI-566, a long-acting and well-tolerated ocular PKM2 activator, may be a potential therapeutic to combat currently untreatable retinal degenerations.

Introduction

Retinal diseases, including retinal detachment and inherited retinal diseases (IRDs), lead to irreversible vision loss because of the death of photoreceptor (PR) cells.¹ Of all cells in the body, PRs have the highest metabolic demands,² which renders them especially vulnerable to dysregulation of their metabolic microenvironment.^3–7 Recent work has shown that directly altering the metabolism of PRs or that of the outer retina/retinal pigment epithelium (RPE) complex can be associated with PR degeneration,^8–14 and rewiring PR metabolism during retinal disease can be protective against such degeneration.^4,15–18 Cellular metabolism has also been shown to modulate the immune response,¹⁹ which plays a role in retinal degenerative diseases.^5,20–23 Alterations in cellular metabolism may therefore be directly or indirectly responsible for PR degeneration, and targeting cellular metabolism may be an attractive, gene-agnostic approach to prevent vision loss in retinal degenerative disease.

A key regulatory enzyme of cellular metabolism is pyruvate kinase (PK). This protein is responsible for the final rate-limiting step of glycolysis by converting phosphoenolpyruvate (PEP) and adenosine diphosphate (ADP) to pyruvate and ATP.²⁴ Most tissues, including the retina, express either pyruvate kinase muscle isoform 1 (PKM1) and/or PKM2.^4,25–27 These two enzymes are generated by alternative splicing of the PKM gene and differ primarily in the regulation of their activity. PKM1 is found almost exclusively in the tetrameric form, which is constitutively active, whereas the quaternary structure of PKM2 and, as a result, its catalytic activity is tightly regulated by numerous different metabolites, growth factors, and post-translational modifications.²⁸ PKM2 can exist either in a low catalytic activity monomer/dimer or a high catalytic activity tetramer. Importantly, PKM2 is predominantly expressed in PRs in the retina whereas PKM1 is expressed in the inner retinal neurons.⁴ We and others have demonstrated that increasing retinal PK activity by genetically or pharmacologically activating PR PKM2 is neuroprotective in rodent models of PR stress, validating PKM2 as a novel neuroprotective target.^4,16,29–31

Immune cells also express PKM1 and/or PKM2 and increase the expression of PKM2 upon activation.^25,26,32 Interestingly, it is the upregulation of the low activity, monomer/dimer form of PKM2 that mediates many of the pro-inflammatory effects of immune cell activation and increasing PKM2 activity by promoting its tetramerization alleviates the inflammatory phenotype.^{25,26,32–34} Similarly, targeting microglial PKM2 in neurodegenerative diseases has demonstrated a neuroprotective effect.^35–37 Hence, modulating PKM2 oligomeric state and in return, its activity, in retinal degenerative diseases has the potential to alter the inflammatory response to prolong PR function and survival.

Despite these promising results, current small molecule PKM2 activators, such as ML-265,³⁴ effectively induce PKM2 activation by stabilizing the highly active tetramer but its functional moieties are associated with chemically reactive metabolites and costly development, and its lack of solubility requires formulation with organic solvents (i.e., dimethyl sulfoxide [DMSO]) that have known ocular toxicity.^16,30 These characteristics limit the translatability of current PKM2 activators, particularly through intravitreal (IVT) injections. To overcome these translational hurdles, we recently developed novel small molecules that retain selectivity for and potency with PKM2 while simultaneously increasing solubility and removing structural alerts.³⁰ Here, we evaluate the pharmacology, safety, and efficacy of the lead compound from our previously published series, MCTI-566, to de-risk the technology and accelerate its translation into the clinic for the treatment of retinal degenerative diseases, including IRDs.

Methods

Animals

All animals were treated in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. This study was conducted with approval from the Institutional Animal Care and Use Committee at the University of Michigan (PRO00011133). All animals were housed under standard conditions with a 12/12 light/dark cycle. Adult Brown-Norway rats and Dutch-belted rabbits were purchased from Charles River Laboratories (Wilmington, MA, USA). The rd10 (strain no. 004297) and B6;129S-Pkm^tm1.1Mgvh/J (Pkm2^fl/fl, strain no. 024048) strains were obtained from Jackson laboratories.³⁸ To specifically delete the PKM2 isoform from rod PRs, mice carrying the Pkm2^fl/fl alleles were crossed with animals transgenic for a Cre-recombinase under the control of the rhodopsin promoter. This results in selective deletion of the alternative exon encoding for Pkm2 while still allowing for normal Pkm1 splicing and expression in rod PRs.^4,39,40

Cell Culture

The 661W PR-like cell line was generously provided by Dr. Muayyad al-Ubaidi from the Department of Cell Biology at the University of Oklahoma Health Sciences.⁴¹ Cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂ and 95% air cultured and grown with Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 90 units/mL penicillin, 0.09 mg/mL streptomycin, 32 mg/L putrescine, 40 µL/L of β-mercaptoethanol, and 40 µg/L of both hydrocortisone 21-hemisuccinate and progesterone.

Chemicals

All reagents used were of analytical grade and were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA) or Combi Blocks Inc. (San Diego, CA, USA). MCTI-566 was synthesized as previously described (compound 19 in Wubben et al.).³⁰

IVT Injection of MCTI-566

IVT injections were performed as previously described.¹⁶ Briefly, rodents were anesthetized with a mixture of ketamine (90 mg/kg) and xylazine (5 mg/kg) whereas inhaled isoflurane or sevoflurane was used to anesthetize rabbits. Before injection, pupils were dilated with topical phenylephrine (2.5%) and tropicamide (1%), and eyes numbed using topical proparacaine. A sclerotomy just posterior to the limbus was created with a 25-gauge microvitreoretinal blade (Walcott Rx Products, Ocean View, NJ, USA). A 34-gauge blunt cannula was then introduced through the sclerotomy and into the vitreous cavity to deliver 2 µL to rat eyes, 1 µL to mouse eyes, or 50 µL to rabbit eyes using a micropump system. To estimate final vitreous concentrations, we assumed vitreous volumes of 6 µL for mice, 15 µL for rats and 1.15 mL for rabbits.^42–44

PK Activity Enzyme Assay

To estimate the in vivo half-maximum activating concentration (AC₅₀) of MCTI-566 in retinal tissue, a single IVT injection of various concentrations of MCTI-566 or vehicle was performed in rats. Tissue was harvested 4 hours later and individual retinas were homogenized in RIPA Lysis and Extraction Buffer (cat.no. 89900; Thermo Fisher Scientific, Waltham, MA, USA) supplemented with Halt protease inhibitor (cat.no. 87785; Thermo Fisher Scientific). Total PK activity in each sample was estimated using an enzyme-coupled assay as previously described.^16,30 Briefly, this assay measures a change in absorbance at 340 nm after the conversion of NADH to NAD⁺ by the lactate dehydrogenase enzyme in the process of converting pyruvate to lactate. The rate of change in absorbance depends on the total PK enzyme activity (converting PEP to pyruvate). The assay was performed in 96-well format using 200 µL/well assay volume. The assay contained 4, 6, or 8 µL of retinal lysate in a buffer consisting of 50 mM Tris-HCl, pH 7.4, 100 mM KCl, 5 mM MgCl₂, 1 mM ADP, 0.5 mM PEP, and 0.2 mM NADH with 8 U of lactate dehydrogenase. Absorbance at 340 nm was measured using a SPECTROstar Omega plate reader (BMG Labtech Inc., Cary, NC, USA). The rate of NADH depletion was determined using the MARS software suite and data were normalized to vehicle-treated enzyme activity. Total PK activity was then normalized to total protein content of each sample.^4,27 To evaluate PK activity one week after oral delivery in a similar fashion, MCTI-566 was dissolved in 0.5% carboxy methyl cellulose (400–1000 cps; cat.no. M0262; Millipore Sigma, Burlington, MA, USA) with 0.1% v/v Tween 20 (cat.no. P1379; Millipore Sigma,) and delivered by oral gavage.

Pharmacokinetics

Estimation of drug concentration in various tissues and fluids was completed in collaboration with the Pharmacokinetic and Mass Spectrometry Core at the University of Michigan as previously described.¹⁶ Dutch-belted rabbits received 50 µL IVT injection to reach 1 or 100 µM final vitreous concentration of MCTI-566 as described above. Rabbits were euthanized at one hour, one day, two days, seven days, 14 days, and 28 days after injection. Plasma, vitreous, and retina/choroid tissue samples were harvested.

To assess cell and protein specificity in vivo, rd10 mice (aged to postnatal day 70) and Pkm2^fl/fl;Rho-Cre⁺ mice between the ages of two and four months (prior to any significant PR degeneration)^4,11 were used. IVT injections were performed as described above to achieve 1 µM MCTI-566 in the vitreous.⁴⁵ Retina was harvested at one hour, two days, and 14 days after IVT injection. Four retinas were pooled from two mice for pharmacokinetic analysis.

To estimate MCTI-566 concentration in plasma following systemic administration, adult SAS Sprague Dawley rats received either intravenous (5 mg/kg) or oral (10 mg/kg) MCTI-566 dissolved in 30% cyclodextrin (pH 4.5-5) to a concentration of 2 mg/mL. Blood samples were collected using heparinized calibrated pipettes at various time points: 0.083, 0.167, 0.25, 0.5, 1, 2, 4, 8, and 24 hours. Blood samples were spun in a centrifuge at 15,000 rpm for 10 minutes, and plasma was collected from the upper layer.

Tissue and fluid samples were then processed by the core for analysis using liquid chromatography-tandem mass spectrometry (LC–MS/MS). Tissue and vitreous samples were suspended in 20% acetonitrile, homogenized, and 20 µL was aliquoted into each well of a 96-well plate. 100 µL of acetonitrile containing 5 ng/mL clofazimine (internal standard) was then added to each sample to precipitate proteins for 10 minutes while vortex-mixing. Samples were spun in a centrifuge at 4000 RPM for 10 minutes at 4°C. Supernatant was collected and 2 µL injected into an LC-MS/MS instrument for drug concentration determination using a standard curve of MCTI-566 in acetonitrile. For plasma samples, 20 µL was aliquoted into a 96-well plate, 20 µL of acetonitrile was added, and the mixture was vortex-mixed (500 RPM) for one minute and processed as above. The monitored reaction transitions were 380.2 and 335.0 Da for MCTI-566 and 473.0 and 431.1 Da for the internal standard.

PKM2 Crystallization and Structure Determination

Full length recombinant PKM2 protein (residues 1–531) with a non-cleavable N-terminal His₆-tag (MGSSHHHHHHSSGLVPRGS) was expressed and purified as previously described.³⁰ Before crystallization, purified PKM2 was incubated with 5 M excess of MCTI-566 (stock concentration of 50 mM in DMSO) overnight at room temperature. Crystals grew at 20°C from drops containing 5 µL of protein complex solution and 5 µL of well solution containing 0.1 M MES pH 6.5 and 25% polyethylene glycol 6000. Crystals were harvested and cryoprotected in well solution containing 25% ethylene glycol.

Diffraction data were collected at the Advanced Photon Source on the LS-CAT 21-ID-F beamline equipped with a MAR300 detector. The data were processed with HKL2000⁴⁶ and the structure solved via molecular replacement⁴⁷ using the tetramer of the activator-bound form of PKM2 lacking the cofactors and the MCTI-313 compound (PDB ID: 8G2E) as the search model.³⁰ The structure solved in space group P1 with two tetramers in the asymmetric unit. Coordinates and restraints for MCTI-566 were created using Grade.⁴⁸ The protein complex was iteratively refined and fit to electron density maps using Buster⁴⁸ and Coot,⁴⁹ respectively. The N-terminal His₆-tag along with the first 10 residues of the protein were disordered in the structure. In addition, the residues in the following loop regions were also disordered in each chain: A:11–12; B:11–13, 123–130, 186–190; D:11–13, 122–133, 189–190, 518–520; E: 11–13, 123–131, 186–190; G:11–13, 122–132, 189–190, 519–520. The data collection and refinement statistics for the structure are shown in Supplementary Table S1.

Isothermal Titration Calorimetry (ITC)

ITC experiments were performed in the Nano ITC Low Volume (TA Instruments, Newcastle, DE, USA). Before performing ITC, PKM2 was dialyzed against 10 mM HEPES pH 7.5, 150 mM KCl, 5 mM TCEP, 5 mM MgCl₂, 5% glycerol and DMSO added to a final concentration of 1.2%. The protein was then concentrated to 0.2 mM. For each ITC run, 400 µL of 0.2 mM PKM2 was added to the cell and 50 µL of 0.3 mM MCTI-566 in dialysis buffer containing a final concentration of 1.2% DMSO was taken into the syringe. The experiments were performed at 25°C with 2 µL of compound added to the protein solution every 200 seconds with a stirring speed of 250 RPM for a total of 25 injections. Dialysis buffer containing 1.2 % DMSO was injected into the protein for use as a blank. The K_d, ∆H, ∆S and n-value of the reactions were calculated using the NanoAnalyze Data Analysis software version 2.1.13 (TA Instruments, Newcastle, DE, USA).

Cell Viability and Caspase Activity Assays

For all viability and caspase activity assays, 2,500 661W cells were seeded onto a white-walled 96-well plate 24 hours before treatment, as previously described.⁵⁰ A luminescent assay kit was used to assess cell viability (RealTime-Glo MT Cell Viability Assay; cat.no. G9711; Promega, Madison, WI, USA). Cells were treated with either DMSO or varying concentrations of MCTI-566, and cell viability was estimated 48 hours later following the manufacturer's protocol. To test the anti-apoptotic activity of MCTI-566, 661W cells were pre-treated with MCTI-566 or DMSO for two hours before the addition of 500 ng/mL FasL (Recombinant Mouse Fas Ligand/TNFSF6 Protein; cat.no. 6128-SA-025; R&D Systems Inc., Minneapolis, MN, USA) and 250 ng/mL anti-hemagglutinin antibody (Hemagglutinin/HA Peptide Antibody; cat.no. MAB060; R&D Systems Inc.). Caspase 3/7 and 8 activity was measured eight hours after FasL addition using the Caspase-Glo 3/7 and 8 Assay Systems (cat. nos. G8090 and G8200; Promega, Madison, WI). Luminescence was measured using the SPECTROstar Omega plate reader (BMG Labtech, Inc, Cary, NC, USA). Data were analyzed using the MARS software suite.

Toxicity

MCTI-566 was dissolved in balanced salt solution (BSS) and weekly IVT injections of either BSS or MCTI-566 to achieve final vitreous concentrations of 10 µM or 100 µM were performed in adult Brown-Norway rats over a period of four weeks. Animals were anesthetized as described above and spectral domain optical coherence tomography (SD-OCT) was performed using the Envisu-R SD-OCT imager (Leica Microsystems Inc., Buffalo Grove, IL, USA) to measure retinal thickness as described previously.⁵¹ A 3 mm horizontal B-scan (1000 A-scans, 100 frames) and a 3 mm × 3 mm rectangular volume (36 B-scans, 1000 A-scans, three frames per B-scan) were collected. Measurements were taken at baseline just before the initial injection and one week after the fourth and final injection. Data were analyzed using Bioptigen Diver software to determine total retinal thickness, outer retinal thickness and combined inner/outer segment thickness using 16 separate points.^4,51

Retinal function was determined using a Diagnosys Espion E2 Electrophysiology System (Diagnosys, Lowell, MA, USA).⁴ Animals were dark-adapted overnight before being anesthetized and eyes dilated as described above. The electroretinogram (ERG) response was measured under both scotopic and photopic conditions.

Stability and Metabolite Identification

The in vitro stability of MCTI-566 in hepatocytes of different species was performed by the contract research organization, Pharmaron (Beijing, China). Briefly, the positive control verapamil and the MCTI-566 compound were prepared in DMSO to a concentration of 10 mM and then diluted to 100 µM combining with a mixture of 50% acetonitrile and 50% water. Cryopreserved hepatocytes from rat (cat. no. M0005; Sprague-Dawley; BioIVT, Westbury, NY, USA), rabbit (New Zealand White, cat. no. M00405; BioIVT), minipig (Gottingen, cat. no. M00615; BioIVT) and human (cat. no. X008000; BioIVT) were used at a concentration of 5 × 10⁵ cells/mL. A portion of the cells was boiled for five minutes to act as the negative control.

Verapamil or MCTI-566, at final concentrations of 1 µM, were then incubated with the hepatocyte cultures in duplicate. Samples were taken at multiple time points (0, 15, 30, 60, 90, and 120 minutes) and immediately quenched with cold acetonitrile and the internal standard (100 nM alprazolam, 100 nM tolbutamide, 100 nM ketoprofen, 200 nM labetalol, 200 nM caffeine, and 200 nM diclofenac). The supernatant was collected and sent for LC-MS/MS analysis.

To identify MCTI-566 metabolites, rat, rabbit, minipig and human hepatocytes were cultured at a density of 1 × 10⁶ cells/ml. These cells were then incubated with MCTI-566 at a concentration of 10 µM and incubated at 37°C. After 240 minutes, the samples were quenched with acetonitrile and spun in a centrifuge at 16,000g for 15 minutes. The supernatants were then collected and dried with an evaporator under nitrogen at room temperature. The concentration of MCTI-566 and its metabolites in each sample were then quantified using LC-MS/MS, which consists of the Dionex UltiMate 3000 UHPLC and Thermo Scientific Q Exactive system (Thermo Fisher Scientific).

Western Blot

Mouse retinas were harvested and lysed in RIPA Lysis and Extraction Buffer (cat. no. 89900; Thermo Fisher Scientific) containing protease and phosphatase inhibitors (cat. nos. 87786 and 78402; Thermo Fisher Scientific). Samples were cleared by centrifugation and protein concentration was estimated using the Pierce BCA protein assay kit (cat. no. 23225; Thermo Fisher Scientific). Protein 25 µg of protein was diluted in Laemmli buffer (cat. no. 1610747; Bio-Rad, Hercules, CA, USA) supplemented with β-mercaptoethanol (cat. no. M6250; Millipore-Sigma, Burlington, MA, USA) and separated using 4%–20% Mini-PROTEAN TGX Precast Gels (cat. no. 4561093; Bio-Rad). Separated protein was transferred to polyvinylidene difluoride membranes using Trans-Blot Turbo Transfer System (25 V for 30 minutes) (cat. no. 1704150; Bio-Rad), blocked with 5% nonfat dry milk in phosphate-buffered saline solution and 0.1% Tween 20 for one hour, incubated overnight with primary antibody (PKM1, cat. no. 7067 [Cell Signaling Technology (CST), Danvers, MA, USA]; PKM2, cat. no. 4053 [CST]; ACTB, cat. no. A5316 [Millipore-Sigma]), washed, and incubated with horse radish peroxidase–conjugated secondary antibody (anti-rabbit, cat. no. P0447; Dako, Glostrup, Denmark; anti-mouse, cat. no. P0448; Dako). Westerns were developed using SuperSignal West Dura Extended Duration substrate (cat. no. 34075; Thermo Fisher) and digital images were captured using the Azure c500 (Azure Biosystems, Dublin, CA, USA). Densitometry measurements were performed using AzureSpot 2.0 software.

Experimental Model of Retinal Detachment

Experimental retinal detachment was induced in adult C57B6/J as previously described.^4,52–55 In brief, mice were anesthetized, eyes dilated and sclerotomy performed as described above. To create the experimental retinal detachment, a 35 g beveled canula was inserted through the sclerotomy and then through a peripheral retinotomy into the subretinal space. Two microliters of sodium hyaluronate (10 mg/mL) (Healon OVD; Abbott Medical Optics, Santa Ana, CA, USA) was injected to detach the retina from the RPE in all experiments. Animals received either IVT MCTI-566 or BSS just before induction of retinal detachment.

Quantitative Real-Time Polymerase Chain Reaction (RT-PCR)

Intraperitoneal (IP) injections of MCTI-566 or vehicle were performed in rd10 mice at post-natal day 14 (p14). Retinas were harvested seven days after injection at p21 and immersed in RNAlater (cat. no. AM7024; Thermo Fisher Scientific) before total RNA purification using the Qiagen RNeasy Mini Kit (cat. no. 74104; Qiagen, Hilden, Germany). Quantitative PCR was performed using the Mouse Inflammatory Cytokines and Receptors RT² Profiler PCR Array (cat. no. /ID: PAMM-011ZE; Qiagen) using a CFX96 real time PCR system (Bio-Rad). The RT² Profiler PCR Array Data Analysis Template from Qiagen was used for all data analyses. Genes were excluded from analysis if the Ct threshold value was less than background. Relative gene expression was determined using the 2^−ΔΔCt method. Ct values were calculated using the geometric mean of the Ct value for the housekeeping genes Actb, Gapdh, and Hsp90ab1.

Flow Cytometry

Retinas were harvested three days after experimental retinal detachment, dissociated, stained with the DeadEnd Fluorometric TUNEL system (cat. no. G3250; Promega), and analyzed by flow cytometry as previously described using FCS express software version 7 (De Novo Software, Ontario, CA, USA).⁴⁵ For determination of immune cell infiltration in the rd10 mouse, retinas were harvested at p21 dissociated and processed for flow cytometry in a similar fashion.⁴⁵ Samples were then incubated with the following labeled antibodies: PE-Cy7-conjugated CD11b antibody (Clone M1/70, cat. no. 552850; 1:100; BD Biosciences, Franklin Lakes, NJ, USA), PerCP-Cy5.5-conjugated CD45 monoclonal antibody (Clone 30-F11, cat. no. 5572351:100; BD Biosciences); FITC-conjugated Ly6G (1:100, Clone 1A8, cat. no. 551460; BD Biosciences), and AF700-conjugated Ly6C antibody (1:100, Clone 1A9, cat. no. 561237; BD Biosciences). After incubating with the above antibodies, samples were fixed and permeabilized (Cytofix/Cytoperm fixation/permeabilization kit, cat. no. 554714; BD Biosciences) and subsequently stained with AF647-conjugated Iba1 antibody (1:50, cat. no. 78060; Cell Signaling Technology). After rinsing three times with cold phosphate-buffered saline solution, the cells were analyzed using the Attune NXT flow cytometer (Thermo Fisher Scientific) operating the Attune Cytometric Software version 6.21 (Thermo Fisher Scientific). Events representing debris and clumps of cells were removed using plots of forward scatter area (FSC-A) and side scatter area (SSC-A) and then in plots of forward scatter height (FSC-A) versus SSC-H to obtain singlets. The 7AAD (cat. no. 559925; BD Biosciences) staining was used to exclude dead cells. CD45⁺ cells were first identified from live, singlet cells. CD11b⁺/Lys6G⁻ cells were then taken forward to quantify the number of mononuclear phagocytes (MNPs) by identifying those cells positive for Iba1 and to identify inflammatory monocytes by gating for the common leukocyte marker, CD45, and Ly6C.⁵⁶ Data were analyzed and visualized in FlowJo software version 10 (Becton, Dickinson and Company, Ashland, OR) and a summary of the gating strategy used for these studies is found in Supplementary Figure S1.

Immunofluorescence

Immunohistochemistry was performed on sections obtained from paraffin-embedded retinas as previously described.⁵⁷ Antigen retrieval was performed in boiling sodium citrate buffer (30 mM citric acid (Fisher Scientific, cat. no. A940-1), 0.5% Tween-20 (Fisher Scientific, cat. no. PI85114), pH 6). Slides were then blocked using 10% normal goat serum in phosphate-buffered saline solution supplemented with Tween-20 before incubation with primary antibody overnight at 4°C in a buffer containing 1% normal goat serum and 1% bovine serum albumin (PKM2 – 1:200 Proteintech, Rosemont, IL, USA, cat. no. 60268-1-Ig; Iba1 – 1:250 Wako Chemical, Richmond, VA, USA, cat. no. 019-19741). Slides were then washed before applying secondary antibody for 1 hour at room temperature (anti-rabbit – 1:1000, Thermo Fisher Scientific, cat. no. A11037; anti-mouse – 1:1000, Thermo Fisher Scientific, cat. no. A10680). Slides were then washed again and counterstained with DAPI using prolong gold mounting media (Thermo Fisher Scientific, cat. no. P36935). Slides were imaged using a Leica DM6000 wide-field fluorescent microscope.

Statistical Analysis

Data are reported as mean ± standard error of the mean unless otherwise stated. The significance of the difference between means was determined using a two-tailed student's t-test or one-way analysis of variance (ANOVA) for in vitro and in vivo efficacy data in Prism 9.0 (GraphPad Software, San Diego, CA). Results with a P-value ≤ 0.05 were considered significant. The initial velocity data obtained from the steady state PKM2 assay was analyzed using nonlinear regression curve fitting with the agonist versus response model in Prism 9.0. Nonlinear regression analysis in Prism 9.0 was used to fit the retinal pharmacokinetic data to either a one- or two-phase decay model based on Akaike's Information Criterion. Rat plasma pharmacokinetic parameters were estimated using non-compartmental analyses with Phoenix WinNonlin software (Certara, Radnor, PA).

Results

MCTI-566 Has Properties Translatable to Human Therapy and Binds to PKM2

To improve the clinical translatability of pharmacologically activating PKM2 for retinal degenerative disease, we developed MCTI-566, a novel, small molecule PKM2 activator. MCTI-566 has a pyridazinoindolone core and replaces the methyl sulfoxide of ML-265, which complicates drug development, with a dimethylaminomethyl moiety and the aniline of ML-265, which is associated with reactive metabolites,⁵⁸ with 6-methylpyridine (Fig. 1A). We previously demonstrated that this novel chemical entity retains potency and selectivity for the PKM2 isoform while increasing solubility nearly 7-fold as compared to ML-265. The compound is easily formulated as an in situ hydrochloride salt in BSS at neutral pH with an osmolality of 320 mOsm/kg.³⁰

Figure 1.

Structure and binding analysis of MCTI-566. (A) Chemical structure of ML-265 and MCTI-566. ML-265 has a thienopyrrolopyridazinone core with methyl sulfoxide and aniline functional groups while MCTI-566 has a pyridazinoindolone core with dimethylaminomethyl and 6-methylpyridine groups. (B) Tetrameric PKM2 in complex with MCTI-566. Each monomer of the tetramer is colored differently and shown in cartoon mode with the interface between each monomer denoted by the dashed line. MCTI-566 is depicted as orange spheres bound across the A-A′ interface. The endogenous activator, FBP, is depicted as black spheres and phosphate as blue spheres. An oxalate molecule (gray spheres) is bound in the active site of two protomers along with a magnesium ion (green sphere). (C) Molecular level interactions between MCTI-566 and PKM2. The binding site consists of equivalent residues from each protomer that makes up the A-A′ dimer interface. Residues from the different protomers are colored in light blue and aqua, respectively, in accordance with the protomer colors from panel B. Both symmetry-related conformations of MCTI-566 (orange stick mode) are shown. Hydrogen bonds are depicted as dashed lines for one conformation for clarity. Water molecules are depicted as red spheres. (D) ITC analysis of MCTI-566 binding to PKM2. PKM2 (0.2 mM) was titrated with 0.3 mM MCTI-566. The heat associated with each small molecule-protein injection was corrected by subtracting the heat of dilution. The buffer used in all experiments was 10 mM HEPES 7.5, 150 mM KCl, 1 mM TCEP, 5 mM MgCl₂, 5% glycerol.

To explore the binding interaction between MCTI-566 and PKM2 and to determine if the changes in the functional groups addended to the pyridazinoindolone core impact target-ligand interactions, purified recombinant human PKM2 was crystallized in complex with MCTI-566. This co-crystal structure was solved in the P1 space group to 2.42 Å resolution (Supplementary Table S1). The resulting structure demonstrated two molecules of MCTI-566 bound per tetramer (Fig. 1B). In addition to MCTI-566, the protein co-purified with an oxalate (analog of the PKM2 substrate, PEP) in the active site as well as known endogenous activator, fructose 1,6-biphosphate (FBP). The protomers lacking FBP contain a single phosphate in the binding site. Of note, MCTI-566 bound to PKM2 at a site unique from that of the active site or endogenous activator binding site.

Similar to previously solved co-crystal structures of PKM2 with ML-265 or other related activators,^30,34 MCTI-566 bound across the A-A′ interface with equivalent residues from each protomer making up the binding pocket (Fig. 1C). Because of the twofold symmetry of the pocket, the co-crystal structure showed MCTI-566 bound in two orientations (Supplementary Figure S2). Despite the different heterocyclic core and functional groups, the molecular level interactions between MCTI-566 and PKM2 were similar to that previously described by us and others for ML-265 and related PKM2 activators further demonstrating the importance of the phenylalanine residues within the binding pocket and their π-π interactions with the core heterocycle.^30,34

ITC further characterized the binding of MCTI-566 to purified recombinant human PKM2 (Fig. 1D). The titration of MCTI-566 into PKM2 was fit to a single site binding model and showed the interaction of MCTI-566 to PKM2 to be enthalpy driven (Table) consistent with the hydrogen bonds and electrostatic interactions noted in the co-crystal structure (Fig. 1C). The dissociation constant (K_d) for MCTI-566 was determined to be in the low micromolar range (1.377 ± 0.114 µM) with a binding stoichiometry of 0.336 ± 0.006 (Table), which is in accordance with two molecules of MCTI-566 binding to the PKM2 tetramer (i.e., binding stoichiometry of 0.5 MCTI-566 per monomer) as noted in the co-crystal structure (Fig. 1B).

Table.

ITC Parameters for the Interaction of MCTI-566 With PKM2 at pH 7.5 and 25°C

Compound	Kd (µM)	ΔH (kJ/mol)	ΔS (J/mol·K)	Stoichiometry to Monomer
MCTI-566	1.377 ± 0.114	−33.000 ± 1.919	1.527 ± 7.033	0.336 ± 0.006

MCTI-566 Displays Durable Target Engagement and Is Specific for PKM2 In Vivo

Previous structure-activity relationship studies demonstrated MCTI-566’s nanomolar AC₅₀ and greater than 200% activation with human recombinant PKM2 as well as in 661W cells.³⁰ To characterize its potency and ability to enhance total PK activity in the retina in vivo, single IVT injections of increasing amounts of MCTI-566 were performed in rat eyes with the retina harvested four hours later for PK activity determination. Intravitreal injection of MCTI-566 enhanced total PK activity 177% ± 8% with an AC₅₀ of 887 ± 201 nM in the rat retina in vivo (Fig. 2A).

Figure 2.

MCTI-566 increases PK activity, displays prolonged retention in ocular tissues, and is specific for PKM2 in PRs. (A) PK activation in the rat retina following IVT injections of increasing concentrations of MCTI-566. Data are normalized to rat retina PK activity after IVT injection of vehicle (DMSO). N = 3 eyes. (B, C) Concentration of MCTI-566 over time in either vitreous (B) or retina/RPE-choroid (C) following a single IVT injection into rabbit eyes. N = 2 eyes per concentration per time point. Inset shows PK activation in rabbit retina 28 days after single IVT injection of MCTI-566 (100 µM final vitreous concentration) or vehicle (DMSO). Data are reported as mean ± range for inset only. (D) Percent PK activation (over vehicle) in rat retina following a single IVT injection of MCTI-566. N = 3 eyes per treatment per time point. Data are presented as the mean ± standard error of the mean. (E) Concentration of MCTI-566 in retina following single IVT injection (1 µM final vitreous concentration) in either WT or post-natal day 70 rd10 mice. (F) Concentration of MCTI-566 in retina after single IVT injection (1 µM final vitreous concentration) in either WT or Pkm2^fl/fl;Rho-Cre⁺ mice (rod PR-specific PKM2 knockout). Four retinas were pooled from two mice for N = 1 for each time point.

With the understanding that MCTI-566 increases PK activity in the retina in vivo, we next investigated the distribution and duration of MCTI-566 in ocular tissues following IVT administration. Rabbits were utilized initially as they have relatively large eyes that share many anatomical features with humans and have been classically used for intraocular pharmacokinetic studies.⁴² Single IVT injections of two different doses of MCTI-566 were performed to provide final vitreous concentrations of 1 or 100 µM, and the vitreous, retina/choroid, and plasma harvested over time to assess the concentration of drug present via LC-MS/MS. The concentration of MCTI-566 was near or below the level of quantitation (2.5 ng/mL) for all plasma samples collected (Supplementary Table S2). Assessing the in vitro stability of MCTI-566 in hepatocytes from numerous species, including human, dog, minipig, rabbit, rat, and mouse demonstrated the drug has medium to high clearance compared to the high clearance of the drug, verapamil, with MCTI-566 mainly metabolized via oxidation, de-amination, and de-methylation (Supplementary Tables S3, S4). So, any MCTI-566 that enters the systemic circulation after IVT injection is likely quickly metabolized and cleared.

With regard to intraocular pharmacokinetics, MCTI-566 cleared from the rabbit vitreous after a one phase decay model (Fig. 2B). At a final vitreous concentration of 100 µM after a single IVT injection, the half-life (t_1/2) of MCTI-566 in the vitreous was 4.2 days. At 1 µM final vitreous concentration, the half-life of MCTI-566 in the vitreous could not be determined with confidence because of the majority of the compound clearing within one day of IVT injection and the remaining concentrations detected approaching the limit of quantitation. At both final vitreous concentrations, MCTI-566 was identified in the retina/choroid and appeared to follow a two-phase decay model secondary to its biphasic decline. Interestingly, the drug was present at detectable levels and engaged the target in the retina 28 days after a single IVT injection (Fig. 2C). As this potentially advantageous prolonged residence time in the target tissue was unforeseen, the analysis was limited by a lack of later time points to better define the model and the predicted half-life.

To ensure the prolonged tissue residence of MCTI-566 equates to increased PK activity, we performed single IVT injections into rat eyes to give a final vitreous concentration of 1 or 10 µM MCTI-566 and collected tissue out to 90 days for PK activity determination. Similar to the rabbit retina, the PK activity versus time curve was best fit by a two-phase decay model and demonstrated the drug continued to enhance PK activity in the retina up to 90 days after a single IVT injection. This two-phase decay model allowed for an estimation of the half-life of MCTI-566 based on its target engagement. To this end, the first or fast decay phase had a half-life of 2.8 and 2.5 days for the 1 µM and 10 µM groups, respectively, while the second or slow phase had a half-life of 45.6 and 57.8 days, respectively (Fig. 2D). Hence, MCTI-566 resides in the retina and engages the target for a prolonged period after a single IVT injection.

To further assess the cell and target specificity of the prolonged residence and target engagement of MCTI-566 in the retina, we used a single IVT injection in mice because of the well-characterized, reproducible models of PR degeneration in this species and their genetic tractability that allows for generation of cell-specific conditional knockout models. Our previous work demonstrated that MCTI-566 is specific for PKM2 over PKM1.³⁰ In the mammalian retina, PKM2 is expressed almost exclusively by PRs while PKM1 is expressed in inner retinal neurons.⁴ Therefore MCTI-566 should specifically target PRs in the neural retina. To address cell-specificity, we performed a single IVT injection into the eyes of rd10 mice at p70 and harvested the retina at multiple timepoints. rd10 mice experience rapid PR degeneration beginning around p14, which is almost entirely complete by p60.⁵⁹ Unlike wild-type (WT) mice, where MCTI-566 was identified in the retina at least to 14 days after a single IVT injection, MCTI-566 was below the limit of quantification (5 ng/g) in retinal tissue isolated from aged rd10 animals (Fig. 2E), suggesting that this compound is specific for PR cells in the outer retina. To confirm that MCTI-566 is specific for PKM2 within PRs, the Pkm2^fl/fl;Rho-Cre⁺ mouse, where PKM2 is specifically deleted from rod PRs and PKM1 is upregulated in its absence,⁴ was used. These animals do not display significant loss of PRs between two and four months of age.^4,60 Similar to the aged rd10 animals, MCTI-566 was below the limit of quantification in retinas isolated from Pkm2^fl/fl;Rho-Cre⁺ animals after a single IVT injection (Fig. 2F). These data suggest the retention of MCTI-566 in the retina depends on the presence of not only PRs but also PKM2 within PRs.

Systemic MCTI-566 Administration Significantly Increases Retinal PK Activity

The prolonged tissue residence time and target engagement of MCTI-566 after IVT injection is unique for a small molecule and attractive for clinical translation.^42,61 However, to safely extend experimentation into models of early-onset PR degeneration, such as the rd10 mouse model, and potentially reduce patient and physician burden in the clinic, we sought to understand the ability of MCTI-566 to cross the blood retinal barrier and enhance PK activity in the retina after systemic delivery. The retinal drug concentration and PK activity after a single IP administration in mice was first assessed. Analysis of the retinal lysate after a single IP injection of various MCTI-566 concentrations showed significant amounts of compound present out to 90 days after IP injection (Fig. 3A) with a concomitant increase in total PK activity in mouse retinal lysate (Fig. 3B), similar to that observed after a single IVT injection (Figs. 2C, 2D). Fitting the PK activity versus time data to a one phase decay model, the half-life of MCTI-566 is approximately 24 days for the higher concentrations (25 and 50 mg/kg) tested (Fig. 3B). A dose-dependent increase in the retinal lysate PK activity was also observed one week after a single oral gavage of MCTI-566 in mice (Fig. 3C). Interestingly, this prolonged activation of PK in the retina after systemic delivery occurred despite the estimated half-life for MCTI-566 being 0.8 hours in the plasma after intravenous administration (Fig. 3D) and 9.9 hours after oral administration of MCTI-566 in rats (Fig. 3E, Supplementary Table S5). These data coincide with the stability estimates of MCTI-566 in rat hepatocytes in vitro (Supplementary Table S3) and further suggest the drug likely has a firm and long-lasting PKM2-binding property under conditions encountered in the retina in vivo.

Figure 3.

Systemic MCTI-566 administration increases PK activity in the rodent retina. (A) MCTI-566 concentration in the mouse retina after single IP injection. Four retinas pooled from two mice for N = 1 per time point. (B) Percent increase in retinal PK activity as compared to vehicle-treated mice after single IP injection. N = 4 eyes per treatment per time point. Inset presents PK activation at 90 days. (C) Mouse retinal PK activity as compared to vehicle-treated mice measured one week after a single oral gavage of solubilized MCTI-566. N = 3–4 eyes per dose. (D) Plasma concentrations of MCTI-566 in rats over 24 hours after 5 mg/kg delivered intravenously (IV). N = 3 animals. (E) Plasma MCTI-566 concentrations in rats over 24 hours after 10 mg/kg delivered orally. N = 3 animals. Statistical differences in C are based on one-way ANOVA where *P <0.05 and **P <0.01. Data are presented as mean ± standard error of the mean.

MCTI-566 in Aqueous Formulation Is Nontoxic to the Retina

We previously demonstrated that the concentration of DMSO necessary for the IVT delivery of ML-265 attenuates retinal function.¹⁶ The low solubility of ML-265 (153.4 µM at pH 7.4) requires the inclusion of such organic solvents. MCTI-566, on the other hand, is nearly 7-fold more soluble than ML-265 (1013.8 µM at pH 7.4), which allows the compound to be dissolved in aqueous solution devoid of organic solvent.³⁰ To assess whether this aqueous formulation of MCTI-566 acutely impacts retinal structure and function, MCTI-566 was dissolved in BSS and intravitreally injected in rats to achieve final vitreous concentrations of 10 or 100 µM MCTI-566. The IVT injections were performed weekly for four weeks. Longitudinal OCT imaging found no evidence that the vehicle or the drug itself negatively impacts retinal structure (Figs. 4A, 4B). Histological analysis of whole eyes also did not show any gross morphological defects after treatment (Fig. 4C). Animals were also evaluated for changes to retinal function using ERG, which again demonstrated no adverse effect on retinal function over time for the vehicle- or MCTI-566-treated groups (Fig. 4D).

Figure 4.

IVT delivery of MCTI-566 is not toxic to the rat retina. IVT injections of vehicle (BSS) or MCTI-566 were performed in Brown-Norway rats weekly for four weeks and compared to baseline. (A) Representative OCT scans of rats treated with either vehicle or MCTI-566 (10 µM or 100 µM) compared to baseline OCT. N = 6–8 eyes per group. (B) Quantitation of OCT data for total retinal thickness, outer nuclear layer thickness and IS/OS thickness at baseline and one week after the last IVT injection of vehicle or MCTI-566 (10 or 100 µM). (C) Representative hematoxylin and eosin-stained retinal sections from rats one week after the last of four weekly IVT injections of vehicle or MCTI-566 (10 µM or 100 µM). N = 3 animals per group. Scale bar is 50 µm. (D) ERG scotopic and photopic a-wave and b-wave amplitudes in WT mice that received weekly IVT injection of either vehicle or MCTI-566 (10 µM or 100 µM) for four weeks. A flash intensity of 32 cd*s/m² and 100 cd*s/m² was used for scotopic and photopic responses, respectively. N = 3–4 animals per group. No statistical differences were observed in B and D as determined by an unpaired two-tailed Student's t-test. Data are presented as mean ± standard error of the mean. GCL, ganglion cell layer; IS/OS, inner segment/outer segment; INL, inner nuclear layer; ONL, outer nuclear layer.

Treatment With MCTI-566 Is Photoreceptor Neuroprotective Both In Vitro and In Vivo

Our previous work has shown that small molecule activation of PKM2 results in robust neuroprotection against extrinsic apoptotic signaling.^16,30 To ensure our medicinal chemistry efforts to enhance the physicochemical properties of MCTI-566 did not alter the clinical utility of this compound, in vitro experiments in 661W cells were performed to assess the ability of MCTI-566 to prevent Fas-ligand (FasL)-mediated apoptosis. In line with our previous studies that used ML-265 or other related PKM2 activators,^16,30 MCTI-566 improved 661W cell viability (Fig. 5A) and reduced caspase 3/7 (Fig. 5B) and caspase 8 (Fig. 5C) activity in the presence of FasL. Fas signaling and caspase activation are critical to PR apoptosis after experimental retinal detachment.^52,53,62 To this end, we previously showed that pharmacological activation of PKM2 with ML-265 or other small molecule PKM2 activators circumvents PR apoptosis in vivo after experimental retinal detachment.^16,30 To determine whether MCTI-566 dissolved in an aqueous solution maintains this PR neuroprotective effect, mice received an IVT injection of either MCTI-566 or BSS at the time of retinal detachment. The entire retina was then harvested and processed for flow cytometry to quantitate TUNEL-positive cells.⁴⁵ The percentage of TUNEL-positive cells in the retina three days after experimental retinal detachment was significantly decreased by more than twofold for the two highest concentrations of MCTI-566 tested (Figs. 5D, 5E).

Figure 5.

MCTI-566 protects against apoptotic cell death both in vitro in the murine 661W PR-like cell line and in vivo after experimental retinal detachment in mice. (A) Cell viability, (B) Caspase 3/7 activity, and (C) Caspase 8 activity were measured in 661W cells pretreated with MCTI-566 (200 nM or 2 µM) or vehicle (DMSO) for two hours before treatment with 500 ng/mL FasL Cell viability was measured after 48 hours of treatment. Caspase 3/7 and 8 activity was measured after eight hours of treatment. N = 3 per treatment. (D) Representative contour plots of flow cytometry to determine TUNEL-positive cells in retinas three days after experimental retinal detachment. Animals received IVT injection of either vehicle (BSS) or MCTI-566 (final vitreous concentration of 1, 10 or 100 µM) at the time of experimental retinal detachment in mice. The green gate denotes TUNEL-positive cells. (E) Quantification of the percent TUNEL-positive cells in the retina. N = 8–10 animals per group. Statistical differences in A, B, and C are based on an unpaired two-tailed Student's t-test. Statistical differences in E) were determined using a one-way ANOVA with Dunnett's post-hoc test for multiple comparisons where *P < 0.05, **P < 0.01, and ***P < 0.001. Data are presented as mean ± standard error of the mean.

MCTI-566 Increases Pyruvate Kinase Activity and Prolongs PR Survival in the rd10 Mouse

To determine the therapeutic potential of pharmacologically activating PKM2 in the context of IRD, we turned to the rd10 mouse. Interestingly, as compared to age-matched, wild-type mouse retina, the rd10 retina has significantly reduced total PK activity at p14 (Fig. 6A), which is prior to the onset of PR degeneration,⁵⁹ and systemic delivery of MCTI-566 increased retinal PK activity in the rd10 mouse at p21 (Fig. 6B) without altering the expression of PKM2 or PKM1 (Figs. 6C, 6D). As such, MCTI-566 is likely promoting the PKM2 tetramer, as observed in the co-crystal structure (Fig. 1B), to enhance its activity rather than increasing the total amount of protein present in the retina. Importantly, rd10 mice treated with a single IP injection of MCTI-566 (50 mg/kg) displayed preservation of total retinal thickness and outer nuclear layer (ONL) thickness as compared to rd10 mice treated with vehicle when assessed in vivo using OCT (Figs. 6E, 6F). Representative ex vivo histology also shows preservation of retinal structure following MCTI-566 treatment in rd10 animals (Fig. 6G). Notably, the ability of MCTI-566 to prolong PR survival in the rd10 mouse is consistent with a prior study that showed treatment of ex vivo rd10 retinal explants with the tool compound, ML-265, reduced the percent of TUNEL-positive cells in the ONL.³¹

Figure 6.

PKM2 activation is neuroprotective in the rd10 mouse. (A) Pyruvate kinase activity was measured in retina from p16 WT and p14 rd10 mice. N = 8 eyes per group. (B) Pyruvate kinase activity was measured in rd10 mouse retinas at p21, 7 days after a single IP injection of vehicle (β-hydroxycyclodextrin) or 50 mg/kg MCTI-566. N = 6–8 eyes per group. (C) Representative Western blot images for PKM1 and PKM2 in retinas from rd10 mice at p21 after treatment with vehicle or MCTI-566 (50 mg/kg) at p14. N = 4 animals per group. (D) Quantitation of Western blot results with fold change calculated in relation to vehicle-treated animals. N = 4 animals per group. (E) Representative OCT images from WT or rd10 mice at p21 after a single IP injection of vehicle or MCTI-566 (50 mg/kg) at p16. GCL, ganglion cell layer; INL, inner nuclear layer. (F) Quantitation of OCT data in Efor total retinal thickness, ONL thickness, and inner retinal thickness (IRT) measured in WT or rd10 mice. N = 6–18 eyes per group. (G) Representative hematoxylin and eosin-stained retinal sections at p21 from rd10 mice after single (IP) injection of vehicle or MCTI-566 (50 mg/kg) at p14. Scale bar: 50 µm. Statistical differences in A, B, D, and F are based on an unpaired Student's t-test where *P < 0.05 and ***P <0.001. Data are presented as mean ± standard error of the mean.

MCTI-566 Modulates the Inflammatory Response in the rd10 Retina

In the retina, PKM2 is expressed mainly by PRs as part of aerobic glycolysis;⁸ yet, PKM2 has also been linked to the activation of immune cells, including MNPs.^{25,33,35,37,63} MNP activation and invasion of the subretinal and outer retinal space has been implicated in retinal degenerative diseases, as well as in the rd10 mouse model of IRD.^23,64–66 To this end, immunofluorescence examination of rd10 retinas at p21 show PKM2 expressed mainly within the inner segments of PRs; however, cells in the outer retinal and subretinal space were identified that stained for both PKM2 and Iba1, a marker of MNPs, such as microglia and monocyte-derived cells (Fig. 7A, Supplementary Figure S3).²³ Of note, Iba1-positive cells in the inner retina did not appear to co-stain with PKM2. To begin to characterize the impact of MCTI-566 on the inflammatory response, qRT-PCR was performed on retina from p21 rd10 animals treated with MCTI-566 or vehicle to profile the expression of 84 immune-related genes using the Mouse Inflammatory Cytokines and Receptors RT² Profiler PCR Array (Qiagen). Five genes were found to be statistically significantly altered: Ccr10, Cx3cl1, Mif, Tnfsf11, and Tnfsf13 (Figs. 7B, 7C).

Figure 7.

PKM2 activation alters inflammatory cytokine expression and impacts immune cell populations in the rd10 retina. (A) Representative immunofluorescence images of rd10 retina at p21 stained with IBA1 (green), PKM2 (red), and DAPI (blue). The merged imaged shows colocalization of IBA1 and PKM2 indicated by the white arrows. Scale bar: 50 µm. (B) Volcano plot of immune-related genes modulated in retinas of cyclic-reared rd10 mice treated with MCTI-566 (50 mg/kg) at p14 and harvested at p21 as compared to vehicle-treated. (C) The qRT-PCR analysis of significantly altered genes (P < 0.05) shown in C. Data is represented as fold change of MCTI-566-treated rd10 retina compared to vehicle-treated rd10 retina after normalization with Actb, Gapdh and Hsp90ab1. N = 4 animals per group. Flow cytometric analysis of cells isolated from retinas of cyclic-reared rd10 mice after single intraperitoneal injection with vehicle (β-hydroxycyclodextrin) or MCTI-566 (50 mg/kg) at p14 and harvested at p21. (D) Representative contour plots of CD11b+/Iba1+ positive cells (red gates), which represent mononuclear phagocytes and quantified in E. (F) Representative contour plots of CD45+/CD11b+/Ly6C+ cells (green gates), which represent inflammatory monocytes and quantified in F. Events reported were normalized to the number of live cell events per sample. N = 4 animals per group. Statistical differences in C, E, and G are based on an unpaired Student's t-test where *P < 0.05 and ***P < 0.001. Data are presented as mean ± standard error of the mean.

Next, we sought to determine the impact of MCTI-566 treatment on the inflammatory cell response in the rd10 retina. Whole retinas from p21 rd10 animals that received a single IP injection of either vehicle or MCTI-566 at p14 were harvested and processed for flow cytometry. After excluding debris, cell clumps, and dead cells, Iba1 was plotted against CD11b to identify those cells positive for both markers, likely representing the majority of MNPs in the retina (Fig. 7D).²³ Systemic treatment with MCTI-566 significantly decreased the CD11b⁺/Iba1⁺ cells in the rd10 retina (Fig. 7E). CD11b⁺ cells were further defined by plotting Lys6C versus CD45. Cells positive for CD11b/CD45/Lys6C represent inflammatory monocytes⁵⁶ and were also significantly decreased with MCTI-566 treatment (Figs. 7F, 7G).

Discussion

Here, we show that MCTI-566 binds to a unique site on PKM2 to increase its activity near 200% with nanomolar potency in vivo, and this target engagement is durable, increasing PK activity in the retina for at least 90 days after a single administration. Moreover, the clinical utility of MCTI-566 is improved when compared to previous small molecule PKM2 activators, because its increased solubility allows for intravitreal delivery of an aqueous formulation devoid of organic solvents to avoid toxic effects in the retina. This pharmacokinetic and safety profile, coupled with its potent PKM2 activation and ability to improve PR survival in multiple models, makes MCTI-566 a promising therapeutic candidate for the treatment of retinal degenerative diseases.

Current evidence suggests PKM2 enzyme activity is connected to its oligomeric state. The endogenous allosteric activator, FBP, alone promotes tetramerization of PKM2 by stabilizing the C-C′ interface (dimer-dimer interface) to increase its activity.^34,67 Interestingly, previously developed PKM2 activators, such as ML-265 and DASA-58, have only been shown to partially shift the oligomeric state of PKM2 to the tetramer when incubated with the apoenzyme. However, when exposing PKM2 to FBP briefly prior to incubation with the small molecule activator, a stabilization of the tetramer was noted. To this end, the co-crystal structures of these activators with PKM2 demonstrate PKM2 co-purifies with FBP, and when assessing PK activation in the absence of FBP, DASA-58 demonstrated a micromolar AC₅₀ compared to a low nanomolar AC₅₀ in the presence of FBP. Therefore it has been suggested that binding of these PKM2 activators along the A-A′ interface or the monomer-monomer interface of one of the dimers, requires FBP binding, and once the small molecule is bound, it stabilizes the tetramer and possibly prevents the release of FBP.³⁴

In support of this biochemical mechanism, MCTI-566 co-crystalizes with PKM2 in the tetrameric state and is bound in the same pocket across the A-A′ interface as ML-265 and DASA-58 (Figs. 1B, 1C). Furthermore, this structure has partial FBP occupancy. Also, while MCTI-566 has an AC₅₀ = 83 nM with the human recombinant enzyme,³⁰ the dissociation constant (K_d) for the interaction between MCTI-566 and the PKM2 apoenzyme was in the low micromolar range (Table). In contrast, the K_d and AC₅₀ of FBP are both in the nanomolar range.⁶⁸ AC₅₀ depends on the experimental conditions utilized in the assay, and denotes the concentration of the drug that provides 50% maximal activity. K_d, on the other hand, is an equilibrium constant, that reports on the thermodynamic stability of the interaction.⁶⁹ So, while these values represent different biochemical principles, it is possible that the MCTI-566 K_d is weaker than expected because of minimal FBP present and as a result, an oligomeric equilibrium that favors the monomer. Future experiments that assess the dissociation constant of the MCTI-566-PKM2 interaction in the presence of FBP are necessary to investigate this possibility. Additionally, MCTI-566 may not need to occupy all possible binding sites to achieve its maximal effect, resulting in K_d being significantly greater than AC₅₀.⁷⁰

Nevertheless, MCTI-566 displays nanomolar potency in the retina in vivo after IVT injection (Fig. 2A) with prolonged target engagement that is specific for PKM2 within PRs (Figs. 2B–F). MCTI-566 does not covalently bind to the enzyme. Rather, it is buried within its binding pocket that is distinct from the kinase active site and FBP binding site (Figs. 1B, 1C). As such, few molecules or metabolites may be present to compete with the drug for its binding site, resulting in MCTI-566 being released on the order of protein turnover, which has been noted to be low with half-lives greater than 60 days in PRs.⁷¹ MCTI-566 binding to melanin in the RPE, to act as an adjacent depot, may be another potential reason for the long retinal residence time of MCTI-566. Melanin binding of small molecules is an attractive strategy to prolong small molecule retention in the eye⁷² and does not predict ocular toxicity.⁷³ To this end, we previously showed that activating PKM2 in well-differentiated RPE with ML-265 is nontoxic in vitro,⁷⁴ and in this report, an aqueous formulation of MCTI-566 demonstrated no anatomic or functional safety signals after four weekly IVT injections (Fig. 4). Detailed pharmacokinetic experiments can be performed in albino animals in the future to assess the impact of melanin binding on MCTI-566 retention in the eye. Regardless, the prolonged residence of MCTI-566 in the retina following IVT injection is unique and attractive from a clinical standpoint as quarterly administration may be possible without the need for a sustained-release formulation depending on the degree of PK activity needed to produce a therapeutic effect, significantly reducing patient and physician treatment burden.⁷⁵

IVT injection is a well-accepted route of administration that directly delivers the therapeutic to the tissue of interest in retinal degenerative disease and may avoid significant systemic exposure as observed in the inability to detect MCTI-566 in the plasma after a single IVT injection in rabbits. However, as with any procedure, IVT injection has potential risks including retinal detachment, endophthalmitis, and intraocular inflammation depending on the drug product being administered.⁷⁶ Systemic administration may avoid these risks while reducing patient and physician burden that comes with frequent IVT injections. Yet, avoiding toxic off-target effects becomes a greater hurdle with systemic delivery, and it can be challenging to achieve therapeutic levels in the retina due to the ability of the drug to cross the blood retinal barrier as well as its bioavailability after oral administration. Here, we demonstrated that systemic administration of MCTI-566 given either orally or IP results in enhanced PK activity in the retina for a prolonged period, suggesting that the small molecule can traverse the blood retinal barrier to engage its target (Figs. 3A–C). From a preclinical standpoint, the ability to deliver MCTI-566 systemically avoids some of the technical challenges associated with using experimental murine models of retinal degenerative disease, namely the limited size of the eye and the need to intervene before development of the eye is complete in some models of rapid PR degeneration. From a clinical standpoint, the data herein suggest this class of PKM2 activators has the potential for oral administration, which is critical in the context of IRDs where patients may need lifelong therapy to prevent blindness. Of course, rigorous toxicology studies are necessary to ensure the long-term safety of this therapeutic strategy prior to translating it to the bedside.

MCTI-566 is the third small molecule PKM2 activator that we have characterized, all of which have demonstrated a significant neuroprotective effect in the context of in vitro FasL -mediated 661W apoptosis and in vivo PR degeneration during experimental retinal detachment (Fig. 5).^16,30 Here, we have expanded the neuroprotective benefit of this strategy to include PR degeneration secondary to an IRD-associated gene.⁵⁹ Mutations in the subunits of phosphodiesterase 6 account for up to 8% of all cases of retinitis pigmentosa (RP),²⁹ and the rd10 mouse used in this report carries a missense mutation in the Pde6b gene, which encodes the β subunit of the phosphodiesterase 6 protein.⁷⁷ Phosphodiesterase 6 hydrolyzes cGMP, whose dysregulated signaling has been implicated in PR degeneration.⁷⁸ Previous studies suggest PKM2 regulates PDE6β expression⁶⁰ and increasing cGMP signaling in retinal explants reduces the expression of Pkm2.⁷⁹ Interestingly, in this report, we demonstrated decreased PK activity in the rd10 retina at an age prior to the commencement of PR degeneration, and enhancing retinal PK activity with MCTI-566 slowed early PR degeneration (Fig. 6). The impact of PKM2 activation on PDE6β expression, cGMP signaling, and retinal metabolism in the rd10 mouse model of RP warrants further investigation. However, the low activity PKM2 monomer/dimer has been shown to shuttle glycolytic intermediates into the pentose phosphate pathway and serine biosynthetic pathway to produce nucleotides.^{16,34,80–82} In contrast, pharmacologically inducing the tetramer to increase PKM2 activity in certain cell types has been shown to decrease both intracellular serine and the pentose phosphate pathway product ribose 5-phosphate, which are involved in nucleotide biosynthesis.^16,34 As such, MCTI-566-mediated PKM2 activation may reduce the nucleotide precursors necessary to maintain guanine pools, inhibiting cGMP-dependent PR death. Dysregulated cGMP signaling also results in the downstream activation of caspases, and cGMP has been shown to increase Fas and FasL expression.^83,84 Prior work has shown that Fas signaling contributes to PR apoptosis, immune cell activation, and inflammatory cytokines in the rd10 mouse.^64,65 To this end, increasing PK activity circumvents FasL-mediated apoptosis in vitro and reduces PR apoptosis after experimental retinal detachment (Fig. 5),^16,30 which is driven by Fas signaling and caspase activation.^52,53,62 Hence, pharmacologically activating PKM2 may alter the PR PDE6β-cGMP axis via a multitude of mechanisms to prolong cell survival.

Immune cells, including MNPs, increase the expression of PKM2 upon activation, and it is the low activity, monomer/dimer form of PKM2 that mediates many of the proinflammatory effects of immune cell activation.^25,33,37 Indeed, modulating PKM2 has been shown to impact the inflammatory response in a multitude of preclinical contexts.^{25,26,33,35–37} Inflammation and immune cell infiltration into the retina, including the ectopic accumulation of MNPs in the subretinal space, has been shown to play a role in retinal degeneration in multiple different preclinical models.^{5,7,20,21,56,85–88} The effect of PKM2 activation on this inflammatory response in the retina, however, has not been reported to date. To this end, we demonstrated that Iba1⁺ cells in the subretinal space of the rd10 retina express PKM2 unlike the Iba1⁺ cells in the inner retina (Fig. 7A). Activating PKM2 with MCTI-566 led to a notable increase in Cx3cl1 gene expression (Fig. 7B and C). Previous work has shown that a deficiency in Cx3cl1/Cx3cr1 signaling increases rod PR death in the rd10 mouse, and soluble CX3CL1 gene therapy improves PR survival in rd10 mice.^88–90 MCTI-566 treatment also led to a significant decrease in Mif that has been shown to impact PR survival after experimental retinal detachment.⁹¹ Importantly, MCTI-566 treatment significantly decreased the CD11b⁺/Iba1⁺ MNP population in the rd10 retina (Figs. 7D, 7E), as well as those considered inflammatory monocytes (CD11b⁺/CD45⁺/Lys6C⁺) (Figs. 7F, 7G). These data broaden the mechanistic targets of PKM2 activation to potentially include activated immune cells in the retina in addition to PRs.

In summary, this study demonstrates that MCTI-566 is a long-acting, well-tolerated, small molecule PKM2 activator that is PR neuroprotective in multiple models. A major research goal is to identify core and gene-specific pathophysiological mechanisms that can be therapeutically targeted to prevent vision loss associated with IRD. This report suggests PKM2 is not only a novel therapeutic target to slow PR degeneration in IRDs associated with Pde6b but also, to potentially promote PR resistance to stress in a multitude of settings. This strong preclinical package provides the foundation for future mechanistic and investigational new drug–enabling studies to accelerate the translation of this bona fide small molecule drug candidate to prevent vision loss in those afflicted with currently untreatable retinal degenerative diseases.