Evaluation of Fluorinated Cromolyn Derivatives as Potential Therapeutics for Alzheimer’s Disease

Abstract

Background:

Cromolyn is an anti-neuroinflammatory modulator with a multifactorial mechanism of action that has been shown to inhibit amyloid-β (Aβ) aggregation and enhance microglial uptake and clearance of Aβ.

Objective:

We report the effects of fluoro-cromolyn derivatives on microglial cell toxicity and microglial clearance of Aβ₄₂.

Methods:

Microglial cell toxicity for cromolyn derivatives were determined in naive BV2 microglial cells. Microglial clearance assays were performed with Aβ₄₂ in naive BV2 microglial cell line and single cell clone BV2 line expressing CD33^WT. PET imaging was performed for three F-18 analogs in a rhesus macaque.

Results:

All compounds but derivative 8 exhibited low microglial cell toxicity. Cromolyn 1 and derivatives 2, 4, and 7 displayed an increased uptake on Aβ₄₂ in naïve BV2 microglial cells. Derivative 4 increased Aβ₄₂ uptake in a dose-dependent manner and at 75μM resulted in a one-fold increase in Aβ₄₂ uptake in BV2-CD33^WT. PET imaging for three [¹⁸F]cromolyn analogs revealed the order of brain tracer penetration to be 4a > 10 > 2a. Tracer 4a exhibited enhanced uptake in areas of high perfusion (putamen, grey matter, and cerebellum) and lower signal in areas of lower perfusion (caudate, thalamus, and white matter).

Conclusion:

Substantial uptake of Aβ₄₂ in both naïve BV2 and BV2-CD33^WT cells observed with 4 indicate conversion of microglial cells from a pro-inflammatory to an activation state favoring Aβ phagocytosis/clearance. These findings suggest that a fluoro-cromolyn analog could reduce fibril-prone Aβ₄₂ in vivo and thereby serve as a therapeutic for the treatment and prevention of AD.

Keywords

Aβ phagocytosis Alzheimer’s disease therapy amyloid microglial PET imaging

INTRODUCTION

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder that afflicts over 35 million people throughout the world, including 5.8 million Americans age 65 and older [1]. Currently there is no effective disease-modifying therapy available to slow or stop the destruction of neurons that cause AD symptoms [2]. Present biochemical and pathological evidence support the concept that an imbalance between production and clearance of soluble amylo-id-β (Aβ₄₂) resulting in abnormal accumulation of Aβ oligomers are responsible for the synaptotoxic effects that lead to neuronal stress, abnormal tau phosphorylation, synapse collapse, and memory im-pairment [3, 4]. Soluble Aβ monomers aggregate to form toxic higher order oligomers (dimers, trimers, up to dodecamers), which readily lead to extracellular deposition of fibrillar amyloid neuritic plaques that are resistant to degradation. Aβ oligomer accumulation drives tauopathy and neurodegeneration [5], all which combine to activate a microglial-driven neu-roinflammatory response. Microglia are critical for Aβ phagocytosis; however, chronic inflammatory activation compromises microglial clearance of Aβ contributing to disease progression prior to clinical symptoms [6, 7]. Hence, interfering with the assembly of Aβ peptides from monomer to oligomeric spe-cies and fibrils, or promoting their clearance from the brain, are targets of anti-Aβ directed therapies in AD [8 –10].

Microglia are the primary immune cells of the central nervous system (CNS) that play an important role in removal of cellular debris and aberrant protein debris. Microglial cells respond to the presence of AD pathological lesions (plaques and tangles) by changing their morphological characteristics, expre-ssing numerous cell surface receptors, and surroun-ding the lesions [11 –13]. Microglial cells can assume activation states in a spectrum ranging from phagocy-tic/neuroprotective to pro-neuroinflammatory. Under normal physiological conditions, microglia secrete anti-inflammatory cytokines and supportive growth factors for neuroprotection and neuro repair, while also clearing proteinaceous and cellular debris. Under pathological conditions, such as amyloid plaque acc-umulation, tauopathy and neurodegeneration in the AD brain, classically-activated microglia release to-xic pro-inflammatory cytokines that induce gliosis dramatically amplify neurodegeneration. This pro-cess leads to more neuronal cell death and debris and accelerates disease progression. Research eff-orts have focused on converting microglia from the pro-neuroinflammatory activation state to the neuro-protective/anti-neuroinflammatory activation state in which the toxic effects are reduced and phagocytic activity toward Aβ is enhanced [14 –17]. This is a potential avenue for new therapeutic approaches to treating and preventing AD.

Cromolyn (1, Fig. 1), also called cromolyn sodium, disodium cromoglycate, or hydroxy-cromolyn, is an FDA-approved drug originally introduced for the treatment of asthma [18]. Cromolyn is a synthetic chromone dimer derived from Khellin, a plant extract (Ammi visnaga) that was used to treat inflammatory diseases in ancient times [19]. This drug was shown to have lower toxicity in pure form compared to Khellin. Cromolyn is often characterized as a ‘mast cell stabilizer’ since it blocks allergen-induced bronchospasm presumably by inhibiting the release of inflammatory mediators from mast cells [20]. There is evidence that cromolyn also inhibits the inflammatory response of other cell types and thus it has been used for the treatment of diseases thought to involve mast cell activation, including allergic rhinitis, allergic conjunctivitis, and mastocytosis [21]. The underlying mechanism of action of cromolyn is not fully known; while cromolyn stabilizes mast cells, this mechanism does not fully explain its effect as an asthma medication [22]. Cromolyn is not restricted to antigen-evoked secretion, and some [23] have speculated that cromolyn may be a “membrane stabilizer” while others have suggested that it may facilitate closure of calcium channels in the mast cell membrane [24, 25]. Interestingly, cromolyn was found to inhi-bit the inflammatory processes in vivo in mast cell-deficient mice and to inhibit the activation of multiple hematopoietic cell types, including phagocytic cells other than mast cells in vitro [26].

Fig. 1

Cromolyn compounds

Numerous studies indicate the involvement of chronic neuroinflammation in the pathogenesis of AD [2, 27]. The sustained activation of microglia and other immune cells in the brain release a variety of proinflammatory and toxic products, including reactive oxygen species, nitric oxide, and cytokines which exacerbate both amyloid and tau pathology. We have been investigating the potential utility of cromolyn as a therapeutic for slowing down AD progression and have reported that cromolyn inhibits Aβ aggregation, enhances microglial recruitment to plaques in transgenic AD mice, and promotes phagocytosis and clearance of Aβ [8, 28]. Cromolyn was found to efficiently inhibit the aggregation of Aβ monomers into higher order oligomers and fibrils in vitro, without affecting Aβ production; in vivo, the levels of soluble Aβ were decreased by over 50%after only 1 week of daily intraperitoneally administered cromolyn in APP/PS1 AD mice [8]. This reduction in Aβ levels was largely attributed to the ability of cromolyn to promote microglial recruitment to Aβ plaques and to induce microglial phagocytosis of Aβ [28].

In a study of the effects of cromolyn on a the SOD1 ^G93A transgenic mouse model of familial amyo-trophic lateral sclerosis, cromolyn treatment significantly delayed the onset of neurological symptoms, increased motor neuron survival in the lumbar spinal cord, and decreased the expression of pro-inflam-matory cytokines/chemokines in the lumbar spinal cord and plasma. Together, these findings suggest that cromolyn sodium provides neuroprotection by decreasing neuroinflammatory response to disease pathology [29]. These neuroprotective mechanisms strongly suggest that cromolyn acts directly in the brain. To better understand the mechanism of action and to develop derivatives with enhanced brain penetration several fluoro-cromolyn derivatives have been investigated. Here, we report on the effect of fluoro-cromolyn derivatives on Aβ uptake by microglial cells, as well as microglial cell toxicity. We also describe the synthesis of F-18 labeled cromolyn analogs and their PET imaging pharmacokinetics in nonhuman primate brain.

MATERIALS AND METHODS

All experiments involving nonhuman primates were performed in accordance with the U.S. Department of Agriculture (USDA) Animal Welfare Act and Animal Welfare Regulations (Animal Care Blue Bo-ok), Code of Federal Regulations (CFR), Title 9, Cha-pter 1, Subchapter A, Part 2, Subpart C, §2.31. 2017. Experiments were approved by the Animal Care and Use Committee at the Massachusetts General Hos-pital. The animal used in this study was an adult male Rhesus macaque (13 years old). Prior to each study, animals were sedated with ketamine/xylazine (10/0.5 mg/kg IM) and were intubated for maintenance anesthesia with isoflurane (1-2%in 100%O₂). A venous catheter was placed for infusion of the radiotracer and, where applicable, an arterial catheter was placed for sampling of the arterial input function. The animal was then positioned on a heating pad on the bed of the scanner for the duration of the study.

A GE PETtrace 16.5 MeV cyclotron (GE Healthcare, Waukesha, WI, USA) was used for [¹⁸F]fluoride production by the ¹⁸O(p,n)¹⁸F nuclear reaction to irradiate ¹⁸O-enriched water (Isoflex Isotope, San Francisco, CA). Dynamic PET sinograms were performed on a GE Discovery MI (GE Healthcare) PET/CT scanner.

Chemistry

Hydroxy-cromolyn 1 (cromolyn sodium) was purchased from Quotient Sciences (Nottingham, UK) or Millipore Sigma (St. Louis, USA). All cromolyn derivatives were synthesized and fully characterized (Supplemental Material). Log P values for 4a and 10 were determined according to the following procedure. An aliquot (10μL, 74 kBq) of the F-18 tracer was added to a test tube containing 2.5 mL of octanol and 2.5 mL of phosphate buffer solution (pH 7.4). The test tube was mixed by vortex for 2 min and then centrifuged for 2 min to fully separate the aqueous and organic phase. The samples taken from the octanol layer (0.1 mL) and the aqueous layer (1.0 mL) were saved for radioactivity measurement. An additional aliquot of the octanol layer (2.0 mL) was transferred to a new test tube containing 0.5 mL of octanol and 2.5 mL of phosphate buffer solution (pH 7.4). The previous procedure (vortex mixing, centrifugation, sampling, and transfer to the next test tube) was repeated until six sets of aliquot samples had been prepared. The radioactivity of each sample was measured using PerkinElmer Wizard2 2480 gamma-counter. The log P values of each set of samples were calculated as Log D7.4 = Log (radioactivity of octanol layer×10 / radioactivity of PBS layer).

Preparation of cromolyn derivatives for microglial cell-based assays

Cromolyn derivatives, 3 and 4, displayed lower solubility in DMSO in comparison to 1, 2, 5, and 6–8. Therefore, we generated 25 mM stock solutions for all the compounds except for 3 and 4. The stock solutions for 3 and 4 were prepared at 5 mM and 7.5 mM, respectively.

Generation of the single cell clone BV2 microglial line stably expressing human CD33 (BV2-CD33^WT)

To generate a single cell clone BV2 line stably expressing human CD33, a well of a 6-well-plate containing naive BV2 cells were transiently transfected with the pcDNA3.1 plasmid encoding human wild-type CD33 (CD33^WT) using Lipofectamine-Plus (Life Technologies), according to the manufacturer’s instructions. The following day the media was cha-nged. Two days post transfection, we started the process of isolating a clonal population of BV2 cells st-ably expressing CD33^WT. For this purpose, cells were trypsinized and centrifuged. Afterwards, cells were washed with phosphate buffer saline (PBS) and centrifuged. The cell pellet was resuspended in FACS sorting buffer that consists of 2%heat-inactivated fetal bovine serum, 2%B27 supplement (Gibco) in PBS. Subsequently, cells were incubated in FACS sorting buffer containing Fc block (1μg/ml, anti-mouse CD16/32, clone 93, Biolegend) for 10 min on ice. Afterwards, cells were labeled with either anti-CD33 antibody-FITC (mouse IgG1, clone HIM3-4, BD Pharmingen) or IgG1-FITC (BD Pharmingen) as control for 30 min on ice. Cells were rinsed with FACS sorting buffer and centrifuged. Then, cells were gently resuspended in FACS sorting buffer and filtered into polystyrene filter top tubes (BD Falcon) for FACS sorting.

BV2 cells expressing CD33^WT were visualized based on CD33 expression, using FACS ARIA (BD Biosciences). The anti-CD33 antibody-FITC bound to CD33 expressed on the cell surface, while cells treated with the isotype antibody IgG1 served as control. Single cells were sorted based on the expression levels of CD33 detected in the FITC channel, using FACS ARIA. Single cells were sorted into 96-well plates containing cell growth media: DMEM (Lonza), 10%heat-inactivated fetal bovine serum, 2 mM L-glutamine, 1%penicillin/streptomycin, and geneticin at 400μg/ml (G418 sulfate, Life Technologies). The pcDNA3.1-CD33^WT vector contains the neomycin resistance gene for selection of stable cell lines using geneticin. After one week, half of the cell growth media was replaced with fresh media in each well every three days. After an additional week, each single cell clone was trypsinized and transferred into one well in 24-well plates for cell expansion. Upon reaching confluency, each single cell clone was trypsinized and plated into two wells in 6-well plates. Once confluent, cells from one well were frozen in heat-inactivated fetal bovine serum containing 10%dimethyl sulfoxide (Sigma Aldrich), while cells in the second well were lysed in RIPA lysis buffer (EMD Millipore) supplemented with protease inhibitors (Roche) and phosphatase inhibitors (Thermo Fisher Scientific). We obtained several cell clones that we screened for CD33 expression by using western blotting and anti-human CD33 antibody (mouse monoclonal, clone PWS44, Leica Biosystems). Finally, the single cell clone line expressing CD33^WT (BV2-CD33^WT) was identified and confirmed by western blotting. The corresponding frozen cell vial was rapidly thawed and plated in cell growth media in a 10 cm culture dish to allow cell expansion.

Cromolyn compounds microglial toxicity and Aβ uptake assays in microglia

Scheme 1

Radiosynthesis of 4a, 2a, and 10.

Naive BV2 microglial cell line or single cell clone BV2 line stably expressing CD33^WT (BV2-CD33^WT) were plated in 12-well plates at the density of 5x10E5 cells for naïve BV2 and 7x10E5 cells for the BV2-CD33^WT cell clone, in proliferating media: DMEM containing 10%heat-inactivated fetal bovine serum, 2 mM L-Glutamine and 1%penicillin/streptomycin (Life Technologies). On the following day, cells were treated with DMSO (control) or cromolyn derivatives at different concentrations in proliferating media for 3 h. 1, 2, 5, and 6–8 were tested at 10, 50, 100, and 150μM, while 3 and 4 were assessed at 5, 25, 50, and 75μM due to solubility limit in DMSO. Afterwards, cells were washed twice with PBS and treated with cromolyn derivatives or DMSO (control) in the presence of soluble untagged Aβ₄₂ peptide (400 nM) in DMEM media for 2 h. At the end of the treatment, cell media was collected and compound toxicity was assessed with CytoTox-ONE^™ lactate dehydrogenase (LDH) assay. LDH is a cytosolic enzyme present in many different cell types that is released after cell death and disruption and is a well-established indicator of cellular toxicity. Remaining cells in the plate were washed three times with cold PBS and lysed with RIPA buffer (EMD Millipore) supplemented with protease inhibitors (Roche) and phosphatase inhibitors (Thermo Fisher Scientific). Protein concentrations in the lysate supernatants were determined using the Pierce™ BCA protein assay kit and 2-3μg/well of total protein from each lysate was analyzed for Aβ₄₂ uptake using the Aβ₄₂ ELISA kit from Wako. Aβ₄₂ levels were normalized to total protein concentration. Toxic compound concentrations were excluded from the Aβ₄₂ analysis and further studies.

PET imaging

A CT scan was acquired for attenuation correction of the PET images. Emission PET data collection was initiated after bolus injection of F-18 cromolyn derivatives (207.2 MBq) into the lateral saphenous vein of a rhesus macaque followed by a saline flush. Dynamic PET imaging was performed for up to 180 min with arterial blood sampling. Total volume of distribution (vt) was estimated using compartmental modeling (one- and two-tissue configurations) and graphical analysis techniques (Logan regression methods). Time bins were 6×10, 8×15, 6×30, 8×60, 8×120, and 30×300 s for 18F-TS274 and 6×10, 8×15, 6×30, 8×60, 8×120, and 18×300 s. Imaging started at the same time as bolus injection. Dynamic PET sinograms were reconstructed using the VPFXS reconstruction with 34 subsets and 3 iterations. Final reconstructed images had voxel dimensions of 256×256×89 and voxel sizes of 1.7×1.7×2.8 mm. The GE discovery MI scanner used for these studies has a spatial resolution around 5 mm at the center of the field of view. Dynamic PET sinograms were reconstructed using the VPFXS reconstruction with 34 subsets and 3 iterations. Final reconstructed images had voxel dimensions of 256×256×89 and voxel sizes of 1.7×1.7×2.8 mm. The GE discovery MI scanner used for these studies has a spatial resolution around 5 mm at the center of the field of view.

Whole blood and plasma time-activity curves were obtained by linear interpolation of radioactivity concentrations (kBq/cc) measured from the arterial blood samples. Blood samples were drawn at different time points during the entire experiment. Arterial samples of 1–3 mL were acquired every 30 s immediately following radiotracer injection during the first 5 min and decreased in frequency to every minute, 2 min, 5 min, 15 min, and 30 min toward the end of the scanning duration. Metabolite analysis was performed on standard and the plasma samples drawn at 5, 10, 15, 30, 60, 90, 120, and 180 min. These samples were assayed with column-switching radio-HPLC to determine the fraction of total radioactivity that was attributable to intact radiotracer. Plasma protein binding was measured in blood drawn immediately prior to tracer injection.

Radiofluorination

5, 5’- [(2-[¹⁸F]Fluoro- 1, 3-propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid diethyl ester] (4a)

A [¹⁸F]fluoride solution (50 mCi) was trapped onto a QMA Sep-Pak Cartridge (Sep-Pak plus light, Waters, Milford, MA) and then released by a solution of Kryptofix-2.2.2. (6 mg) and potassium carbonate (2 mg) in acetonitrile and water (1 mL, v/v 7:3) into a sealed Wheaton 5 mL reaction vial. The [¹⁸F] flu-oride solution was azeotropically dried with the aid of nitrogen gas at 120°C for 10 min. The K¹⁸F/Kry-ptofix complex was dried three times at 120°C by the addition of 1 mL of acetonitrile followed by evapo-ration of the solvent using a nitrogen flow. A solution of 1,3-bis[tolylsulfonyl)oxy]-2-[(trifluoromethyl)-sulfonyl]oxy-propane (4 mg, 14) in anhydrous ace-tonitrile was added to the vial and fluorination was performed at 80°C for 10 min. The resultant 2-[¹⁸F] fluoropropane 1,3-ditosylate solution (90%, radio-TLC) was passed through a silica gel SepPak using methylene chloride into a vial containing K₂CO₃ (10 mg) and ethyl 5-hydroxy-4-oxo-4H-chromene-2-carboxylate (10 mg) (Sigma-Aldrich). After solvent removal, anhydrous DMSO (0.8 mL) was added and the mixture was heated for 10 min at 130°C. Once cooled to 25°C, 1 mL of 5%HCl, followed by 2 mL of 50/50 acetonitrile 0.1 M ammonium formate was added and the mixture was filtered (Millex-LCR 0.45μm). F-18 cromolyn diester 4a was purified by HPLC (Phenomenex Luna C18, 250×10 mm, 50:50 acetonitrile/0.1 M ammonium formate (pH 6.0). The fraction containing 4a was diluted with water (30 mL) and loaded onto a C-18 Sep-Pak (Waters, Milford, MA). Finally, the cartridge was washed with water (5 mL) and 4a was eluted with ethanol (1 mL), diluted with saline to a final 10%ethanol/saline solution and filtered (0.22u Millex-GV). Synthesis was complete within 90 min and chemical purity was greater than 95%(5 mCi, 20±5%EOB, n = 10). The molar act-ivity, radiochemical identity and purity of injected 4a were determined by an analytical HPLC system using an analytical column (Waters, XBridge, C18, 3.5μ, 4.6×150 mm) with a mobile phase of acetonitrile/0.1M aqueous ammonium formate solution (60/40, v/v) at a flow rate of 1 mL/min and UV absorption at λ= 254 nm.

5, 5’- [(2-[¹⁸F]Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid] (2a)

Following the procedure for the synthesis of 4a, water (1 mL) was added in place of HCl at the com-pletion of the final step and the mixture was heated at 80°C for 10 min. Once cooled to 25°C, 1 mL of 50/50 acetonitrile 0.1 M ammonium formate was added and the mixture was filtered (Millex-LCR 0.45μm). F-18 cromolyn diacid 2a was purified by HPLC (Phenomenex Luna C18, 250 x 10 mm, 10:90 acetonitrile/0.1 M ammonium formate). The fraction containing F-18 cromolyn diacid was diluted with water (30 mL) and loaded onto a C-18 Sep-Pak and the cartridge was washed with water (5 mL) and activity was eluted with ethanol (1 mL), diluted with saline to final a 10%ethanol/saline solution and filtered (0.22u Millex-GV). Synthesis was complete within 90 min and chemical purity was greater than 95%. (14±5%EOB, n = 10).

5, 5’-(2-[¹⁸F] Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid di-tert-butyl ester] (10)

A sealed Wheaton 5 mL reaction vial containing 50 mCi of fluorine–18 in 1 mL of ¹⁸O-enriched water, Kryptofix-2.2.2. (6 mg), and potassium carbonate (2 mg) was heated at 120°C and solvent was evaporated with the aid of nitrogen gas. The K¹⁸F/Kryptofix complex was dried three times at 120°C by the addition of 1 mL of acetonitrile followed by evaporation of the solvent using a nitrogen flow. A solution of the tosylate cromolyn t-butyl diester 12 (5 mg) in DMSO (0.7 mL) was added to the vial and fluorination was performed at 120°C for 10 min. Once cooled to 25°C, 2 mL of 50/50 acetonitrile 0.1 M ammonium formate was added and. F-18 cromolyn t-butyl die-ster 10 was purified by HPLC (Phenomenex Luna C18, 250×10 mm, 50:50 acetonitrile/0.1 M ammon-ium formate). The fraction containing F-18 cromolyn diester 10 was diluted with water (30 mL) and loaded onto a C-18 Sep-Pak and the cartridge was washed with water (5 mL) and activity was eluted with eth-anol (1 mL) into a vial, diluted with saline to final 10%ethanol/saline solution and filtered (0.22u Millex-GV). Synthesis was complete within 90 min and che-mical purity was greater than 95%. (10%EOB)

RESULTS

Cromolyn microglial cell toxicity

Naive BV2 microglial cells were treated with DMSO (control) or hydroxy-cromolyn 1 or cromolyn derivatives (2–8) for 3 h at different concentrations. Afterwards, cells were incubated with soluble un-tagged Aβ₄₂ peptide and DMSO or cromolyn derivatives for 2 h. At the end of the treatment, cell media was collected, and compound toxicity was assessed with the lactate dehydrogenase (LDH) assay. All compounds but one exhibited very low microglial cell toxicity at 10 to 150μM concentrations (Fig. 2). BV2 microglial cells treated with the cromolyn derivative 8, originally designed as a prodrug to deliver 1 through the skin [30], exhibited increased toxicity at 100 and 150μM in comparison to cells treated with the vehicle (DMSO) (see Fig. 4). Toxic compound concentrations were excluded from the Aβ₄₂ analysis and further studies.

Fig. 2

Microglial toxicity of cromolyn derivatives compared to DMSO.

Fig. 4

Effect of cromolyn derivatives on Aβ₄₂ uptake in microglial BV2-CD33^WT activated cells.

Cromolyn derivative effect on microglial uptake of Aβ₄₂

To explore the effects of cromolyn derivatives on Aβ₄₂ uptake, we employed a microglial cell-based assay. Specifically, naïve BV2 microglial cell cultures were treated with cromolyn derivatives 1–8 at different concentrations ranging between 5 and 150μM for 3 h. Subsequently, cells were incubated with soluble Aβ₄₂ and cromolyn derivatives for 2 h. After incubation, cells were collected for ELISA analysis to measure Aβ₄₂ levels. Cromolyn 1 (hydroxy cromolyn) and cromolyn derivative 2 led to a modest increase of Aβ₄₂ uptake at 100 and 150μM and 10 and 100μM, respectively in BV2 cells. Interestingly, the cromolyn derivative 6 led to a robust inhibition of Aβ₄₂ uptake in BV2 microglial cells (Fig. 3). Remarkably, we found that the cromolyn derivative 4 is the most effective and at 75μM leads to a significantly increased uptake of Aβ₄₂ in BV2 microglial cells in comparison to vehicle treatment.

Fig. 3

Effect of cromolyn derivatives on Aβ₄₂ uptake in naïve BV2 microglial cells.

We previously showed that the AD risk factor CD33 inhibits uptake and clearance of Aβ₄₂ in mic-roglial cell cultures, a process that requires the sia-lic acid-binding V-Ig domain of CD33 [31]. Here, we generated a single cell clone BV2 line stably expressing human wild-type CD33 (BV2-CD33^WT). The effect of cromolyn derivatives on microglial uptake and clearance of Aβ₄₂ was investigated in the single cell clone BV2-CD33^WT microglial line. BV2-CD33^WT cells were treated with DMSO (control) or cromolyn derivatives at different concentrations ranging between 5 and 150μM. Results for compounds 3, 4, and 7 impacted microglial uptakes of Aβ₄₂ at the concentrations tested in comparison to DMSO. We also showed that cromolyn derivative 6 inhibits negative uptake of Aβ₄₂ in BV2-CD33^WT cells as compared to DMSO treatment (Figs. 3 4). Using the GraphPad Prism 7 software, the IC₅₀ for fluoro-cromolyn diethyl ester 4 was 54.7μM in BV2-CD33^WT cells.

Radiofluorination

Radiosynthesis of cromolyn 2a and 4a was ach-ieved in a two-step process due to susceptibly of the of the diethyl ester precursor to hydrolysis at reaction temperatures above 120°C. Fluorine-18 labeling of the cromolyn diethyl ester tosylate or mesylate using either potassium carbonate or potassium bic-arbonate and Kryptofix-2.2.2 or using tetrabutyla-mmonium bicarbonate in DMSO at temperatures between 120°C-140°C gave 0 to 5%yield of 4a. In contrast, the two-step method provided 4a in an overall yield of 20%, consistently. This was possible due to the selective displacement of the triflate moiety in 1, 3-bis[tolylsulfonyl)-oxy]-2-[(trifluoromethyl)-sulfonyl]oxypropane (14) with fluorine-18 at 80°C in acetonitrile. Radiofluorination of the tosylate of cromolyn di-tert-butyl ester (12) to provide 10 did not undergo substantial hydrolysis during labeling conditions with potassium carbonate/Kryptofix-2.2.2.

PET imaging

The aim of this PET study was to measure brain exposure and regional distribution of selected F-18 labeled cromolyn analogs administered intravenously in a monkey with an intact BBB. Pharmacokinetics of three F-18 analogs (diethyl ester 4a, di-tert-butyl ester 10, and diacid 2a) were determined by dynamic PET imaging in a rhesus macaque monkey. Arterial blood sampling and radio-metabolite analysis was also performed. PET imaging revealed the order of brain tracer penetration to be diethyl ester 4a>di-tert-butyl ester 10>diacid 2 (Figs. 5 6). The brain distribution of diacid 2a matched the cerebral blood volume, suggesting negligible BBB penetration at tracer levels.

Fig. 5

PET imaging data comparison of [¹⁸F]fluoro-cromolyn diacid 2a (top right), [¹⁸F]fluoro-cromolyn di-tert-butyl ester 10 (bottom left) and [¹⁸F]fluoro-cromolyn diethyl ester 4a (bottom right) in monkey.

Fig. 6

Time-activity curves from PET imaging data of [¹⁸F]fluoro-cromolyn diacid 2a, [¹⁸F]fluoro-cromolyn di-tert-butyl ester 10 and [¹⁸F]fluoro-cromolyn diethyl ester 4a in monkey.

PET images obtained with [¹⁸F]fluoro-cromolyn diethyl ester 4a showed enhanced uptake in areas of high perfusion (putamen, grey matter and cerebellum) and lower signal in areas of lower perfusion (caudate, thalamus, and white matter) (Fig. 5). Brain penetration was immediate, reaching maximum at 2 min (2.3 SUV) and washout was slow, 2 SUV at 20 min, 1.5 at 60 min (Fig. 6). Blood sampling for this study revealed ¹⁸F-diacid cromolyn as the only metabolite (20–140 min). Hydrolysis to form [¹⁸F] fluoro-cromolyn diacid 2a appears to take place in the blood as evidenced by ex vivo stability tests. The t-butyl ester derivative 10 also metabolized to the ¹⁸F-diacid, albeit, more slowly.

DISCUSSION

Our results indicate that fluoro-cromolyn derivatives represent potential therapeutic agents for treating and preventing AD by reducing pro-neuro-inflammatory activation of microglia, enhancing mic-roglial phagocytosis and promoting clearance of Aβ. Several fluoro-cromolyn and non-fluoro-cromolyn derivatives were investigated in this initial screening to better understand the structure-activity relationship of cromolyn to microglial uptake and clearance of Aβ₄₂ and to develop derivatives with enhanced brain penetration. Replacement of OH in cromolyn by F and esterification of the acid or reduction to the alcohol are expected to result in enhanced lipophilicity and thus greater brain permeable compounds. Moreover, fluorine derivatives offer the added advantage of monitoring brain penetration by labeling with flourine-18 and using PET imaging. Additionally, cromolyn derivatives containing labile substituents such as ethyl ester, t-butyl ester, acetate or methyl pivalate ester could act as lipophilic transient derivatives or prodrugs to deliver cromolyn 1, 2, 5, or 6 to the brain. Substitution of the hydroxyl group in 1 for fluoride did not significantly alter microglial uptake of Aβ₄₂ based on comparable results obtained with 2 (Fig. 3). However, substitution of the hydroxyl group for fluoride in diethyl ester 3 did show an increase in microglial uptake of Aβ₄₂ based on comparable results obtained with 4 at 75μM (Fig. 4). Interestingly, replacing carboxyl groups with hydroxymethyls as in 5 and 6 negatively impacted microglial uptake of Aβ₄₂ indicating the possible significance of this moiety in the binding mechanism. Although, it must be noted that triacetate 7 derived from hydroxymethyl 5 exhibited a less negative impact of microglial uptake of Aβ₄₂ in BV2 microglial cells (Fig. 3) and had a positive impact of microglial up-take of Aβ₄₂ in BV2-CD33 activated microglial cells at 100μM (Fig. 4). This increase microglial uptake of Aβ₄₂ may be due to the presence of acetyl groups on 7. Remarkably, fluoro-cromolyn diethyl ester 4 increased microglial uptake of Aβ₄₂ in a dose-dependent manner and at 75μM resulted in a one-fold increase in Aβ₄₂ uptake in BV2-CD33^WT. The IC₅₀ for fluoro-cromolyn diethyl ester 4 was found to be 54.7μM in BV2-CD33^WT cells. These results suggest that fluoro-cromolyn diethyl ester 4 promotes microglial uptake and clearance of Aβ₄₂ by pharmacologically converting microglial cells from neurotoxic/pro-inflammatory activation to a more beneficial neuroprotective/pro-phagocytic activation state.

PET imaging with three [¹⁸F]fluoro-cromolyn derivatives revealed the order of brain tracer penetr-ation to be diethyl ester 4a > di-tert-butyl ester 10 > diacid 2 (Figs. 5 6). Brain uptake corresponded to lipophilicity values where the measured logP value for the diethyl ester was 2.2 compared to the di-tert-butyl with a measured value of 3.2 that is slightly above the range for ideal blood-brain barrier (BBB) penetration. BBB penetration is optimal when LogP values are in the range of 1.5–2.7 with a mean value of 2.1 [31]. The brain distribution of diacid 2a matched the cerebral blood volume, suggesting negligible BBB penetration at tracer levels. However, it must be noted that administration of hydroxy-cromolyn 1 in typical drug doses by either oral inhalation (17.1 mg) or i.p. injection (25 mg/kg) results in detectable brain uptake [7, 32]. PET images obtained with [¹⁸F]fluoro-cromolyn diethyl ester 4a showed enhanced uptake in areas of high perfusion (putamen, grey matter, and cerebellum) and lower signal in areas of lower perfusion (caudate, thalamus, and white matter) (Fig. 5). Brain penetration was immediate, reaching maximum at 2 min and washout was slow. Metabolite analysis of both diethyl 4a and di-tert-butyl 10 derivatives revealed that these compounds hydrolyze to diacid 2a. These findings indicate that [¹⁸F]fluoro-cromolyn diethyl ester 4a readily enters the brain whereupon it hydrolyzes to [¹⁸F]fluoro-cromolyn diacid 2a and is accumulated in the brain. This can be a method to deliver 2a into the brain following i.v. administration of [¹⁸F]fluoro-cromolyn diethyl ester 4a to determine regional distribution. Unfortunately, for use as a therapeutic drug, the rate of hydrolysis or short half-life of 4 in vivo needs to be addressed. Development of a controlled-release formulation or oral inhalation delivery of cromolyn 4 for optimal bioavailability is one approach. Another strategy is to screen cromolyn derivatives and prodrugs that are less prone to hydrolysis but still maintain optimal therapeutic performance.

In summary, these results indicate that fluoro-cromolyn derivatives represent potential therapeutic agents for treating and preventing AD by reducing pro-neuroinflammatory activation of microglia and simultaneously promoting clearance of Aβ via microglial phagocytosis.

Footnotes

ACKNOWLEDGMENTS

This study was funded by the Gordon Center for Medical Imaging NIH Grant: P41EB022544 TDR3.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

(2020) 2020 Alzheimer’s disease facts and figures. Alzheimers Dement 16, 391–460.

Kinney

, Bemiller

, Murtishaw

, Leisgang

, Salazar

, Lamb

(2018) Inflammation as a central mechanism in Alzheimer’s disease. Alzheimers Dement (N Y) 4, 575–590.

Selkoe

, Hardy

(2016) The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol Med 8, 595–608.

Larson

, Lesné

(2012) Soluble Aβ oligomer production and toxicity. J Neurochem 120(Suppl 1), 125–139.

Choi

, Kim

, Hebisch

, Sliwinski

, Lee

, D’Avanzo

, Chen

, Hooli

, Asselin

, Muffat

, Klee

, Zhang

, Wainger

, Peitz

, Kovacs

, Woolf

, Wagner

, Tanzi

, Kim

(2014) A three-dimensional human neural cell culture model of Alzheimer’s disease. Nature 515, 274–278.

Rogers

, Lue

(2001) Microglial chemotaxis, activation, and phagocytosis of amyloid beta-peptide as linked phenomena in Alzheimer’s disease. Neurochem Int 39, 333–340.

Yamanaka

, Ishikawa

, Griep

, Axt

, Kummer

, Heneka

(2012) PPARγ/RXRα-induced and CD36-mediated microglial amyloid-β phagocytosis results in cognitive improvement in amyloid precursor protein/presenilin 1 mice. J Neurosci 32, 17321–17331.

Hori

, Takeda

, Cho

, Wegmann

, Shoup

, Takahashi

, Irimia

, Elmaleh

, Hyman

, Hudry

(2015) FDA approved asthma therapeutic agent impacts Aβ in the brain in a transgenic model of Alzheimer’s disease. J Biol Chem 290, 1966–1978.

Jamasbi

, Wade

, Separovic

, Hossain

(2016) Amyloid beta (Aβ) peptide and factors that play important roles in Alzheimer’s disease. Curr Med Chem 23, 884–892.

10.

Yasumoto

, Takamura

, Tsuji

, Watanabe-Nakayama

, Kmamura

, Inoue

, Nakamura

, Inoue

, Kimura

, Yano

, Nishijo

, Kiuchi

, Teplow

, Ono

(2019) High molecular weight amyloid β1-42 oligomers induce neurotoxicity via plasma membrane damage. FASEB J 3, 9220–9234.

11.

Thériault

, ElAli

, Rivest

(2015) The dynamics of monocytes and microglia in Alzheimer’s disease. Alzheimers Res Ther 7, 41–51.

12.

Wilcock

(2012) A changing perspective on the role of neuroinflammation in Alzheimer’s disease. Int J Alzheimers Dis 2012, 495243.

13.

McGeer

, McGeer

(2010) Targeting microglia for the treatment of Alzheimer’s disease. Expert Opin Ther Targets 19, 497–506.

14.

Mandrekar

, Landreth

(2010) Microglia and inflammation in Alzheimer’s disease. CNS Neurol Disord Drug Targets 9, 156–167.

15.

Rubio-Perez

, Morillas-Ruiz

(2012) A review: Inflammatory process in Alzheimer’s disease, role of cytokines. ScientificWorldJournal 2012, 756357.

16.

Akiyama

, Barger

, Barnum

, Bradt

, Bauer

, Cole

, Cooper

, Eikelenboom

, Emmerling

, Fiebich

, Finch

, Frautschy

, Griffin

, Hampel

, Hull

, Landreth

, Lue

, Mrak

, Mackenzie

, McGeer

, O’Banion

, Pachter

, Pasinetti

, Plata-Salaman

, Rogers

, Rydel

, Shen

, Streit

, Strohmeyer

, Tooyoma

, Van Muiswinkel

, Veerhuis

, Walker

, Webster

, Wegrzyniak

, Wenk

, Wyss-Coray

(2000) Inflammation and Alzheimer’s disease. Neurobiol Aging 21, 383–421.

17.

Liu

, Liu

, Hao

, Wolf

, Kiliaan

, Penke

, Rübe

, Walter

, Heneka

, Hartmann

, Menger

, Fassbender

(2012) TLR2 is a primary receptor for Alzheimer’s amyloid β peptide to trigger neuroinflammatory activation. J Immunol 188, 1098–1107.

18.

Altounyan

REC

(1967) Inhibition of experimental asthma by a new compound-disodium cromoglycate (Intal). Acta Allergol 22, 487–9.

19.

Mesbah

(1992) Determination of Khellin and Visnagin in Ammi visnaga fruits and in renal teas by HPLC. Egypt J Pharm Sci 33, 897–904.

20.

Zhang

, Finn

, Barlow

, Walsh

(2016) Mast cell stabilizers. Eur J Pharmacol 778, 158–168.

21.

Storms

, Kaliner

(2005) Cromolyn sodium: Fitting an old friend into current asthma treatment. J Asthma 42, 79–89.

22.

Rang

, Dale

(2003) Mechanism of Action 8. Pharmacology, Churchill Livingstone, Ed., Fifth Edition.

23.

Cox

JSG

(1977) Pharmacological and Biochemical Properties of Drug Substances, Goldberg ME, ed. American Pharmaceutical Association, Washington, D.C.

24.

Theoharides

, Sieghart

, Greengard

, Douglas

(1980) Antiallergic drug cromolyn may inhibit histamine secretion by regulating phosphorylation of a mast cell protein. Science 207, 80–82.

25.

Foreman

, Hallett

, Mongar

(1977) Site of action of the antiallergic drugs cromoglycate and doxantrazole. Br J Pharmacol 59, 473–474.

26.

Oka

, Kalesnikoff

, Starkl

, Tsai

, Galli

(2012) Evidence questioning cromolyn’s effectiveness and selectivity as a “mast cell stabilizer” in mice. Lab Invest 92, 1472–1482.

27.

Wyss-Coray

, Rogers

(2012) Inflammation in Alzheimer disease-a brief review of the basic science and clinical literature.. Cold Spring Harb Perspect Med 2, a006346.

28.

Zhang

, Griciuc

, Hudry

, Wan

, Quinti

, Ward

, Forte

, Shen

, Ran

, David

, Elmaleh

, Tanzi

(2018) Cromolyn reduces levels of the Alzheimer’s disease-associated amyloid β-protein by promoting microglial phagocytosis. Sci Rep 8, 1–9.

29.

Granucci

, Griciuc

, Mueller

, Mills

, Le

, Dios

, McGinty

, Pereira

, Elmaleh

, Berry

, Paganoni

, Cudkowicz

, Tanzi

, Sadri-Vakili

(2019) Cromolyn sodium delays disease onset and is neuroprotective in the SOD1^G93A Mouse Model of amyotrophic lateral sclerosis. Sci Rep 9, 17728.

30.

Bedor

, Zupan

, Selk

(1980) Improved delivery through biological membranes. VII. Dermal delivery of cromoglycic acid (cromolyn) via its prodrugs. Int J Pharm 7, 63–75.

31.

Pajouhesh

, Lenzb

(2005) Medicinal chemical properties of successful central nervous system drugs. NeuroRX 2, 541–553.

32.

Using oral inhalation of micronized cromolyn, the average Cmax in CSF, indicative of drug delivery to the brain, was found to be 0.24 ± 0.077 ng/ml. Elmaleh DR, Elmaleh, MD (2019) Dispenser for dry-powder inhalation devices. US20190321571 A1 2019-10-24.

Evaluation of Fluorinated Cromolyn Derivatives as Potential Therapeutics for Alzheimer’s Disease

Abstract

Background:

Objective:

Methods:

Results:

Conclusion:

Keywords

INTRODUCTION

MATERIALS AND METHODS

Chemistry

Preparation of cromolyn derivatives for microglial cell-based assays

Generation of the single cell clone BV2 microglial line stably expressing human CD33 (BV2-CD33WT)

Cromolyn compounds microglial toxicity and Aβ uptake assays in microglia

PET imaging

Radiofluorination

5, 5’- [(2-[18F]Fluoro- 1, 3-propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid diethyl ester] (4a)

5, 5’- [(2-[18F]Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid] (2a)

5, 5’-(2-[18F] Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid di-tert-butyl ester] (10)

RESULTS

Cromolyn microglial cell toxicity

Cromolyn derivative effect on microglial uptake of Aβ42

Radiofluorination

PET imaging

DISCUSSION

Footnotes

ACKNOWLEDGMENTS

References

Generation of the single cell clone BV2 microglial line stably expressing human CD33 (BV2-CD33^WT)

5, 5’- [(2-[¹⁸F]Fluoro- 1, 3-propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid diethyl ester] (4a)

5, 5’- [(2-[¹⁸F]Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid] (2a)

5, 5’-(2-[¹⁸F] Fluoro- 1, 3- propanediyl) bis(oxy)] bis[4- oxo-4H- 1- benzopyran- 2- carboxylic acid di-tert-butyl ester] (10)

Cromolyn derivative effect on microglial uptake of Aβ₄₂