Ciproxifan, a histamine H3 receptor antagonist, reversibly inhibits monoamine oxidase A and B

Ciproxifan is a well-investigated histamine H3 receptor (H3R) inverse agonist/antagonist, showing an exclusively high species-specific affinity at rodent compared to human H3R. It is well studied as reference compound for H3R in rodent models for neurological diseases connected with neurotransmitter dysregulation, e.g. attention deficit hyperactivity disorder or Alzheimer’s disease. In a screening for potential monoamine oxidase A and B inhibition ciproxifan showed efficacy on both enzyme isoforms. Further characterization of ciproxifan revealed IC50 values in a micromolar concentration range for human and rat monoamine oxidases with slight preference for monoamine oxidase B in both species. The inhibition by ciproxifan was reversible for both human isoforms. Regarding inhibitory potency of ciproxifan on rat brain MAO, these findings should be considered, when using high doses in rat models for neurological diseases. As the H3R and monoamine oxidases are all capable of affecting neurotransmitter modulation in brain, we consider dual targeting ligands as interesting approach for treatment of neurological disorders. Since ciproxifan shows only moderate activity at human targets, further investigations in animals are not of primary interest. On the other hand, it may serve as starting point for the development of dual targeting ligands.


Results
IC 50 determinations for human MAO. The IC 50 values of ciproxifan for human membrane-bound MAO (hMAO) were measured spectrophotometrically using kynuramine (KYN) and benzylamine (BZA) as MAO B substrates, while for MAO A only KYN was used. We found IC 50 values for ciproxifan in a micromolar range    Table 2).

Reversibility of human MAO inhibition.
In order to determine whether ciproxifan shows a reversible or irreversible inhibition type, dilution experiments using the spectrophotometric assay were performed, where hMAOs were preincubated with ciproxifan (10 × IC 50 ). After preincubation probes were diluted 100-fold, measured at saturated substrate conditions and the remaining enzyme activity was compared to that of MAO preincubated without ciproxifan. For both hMAO isoforms no considerable decrease in enzyme activity after preincubation with ciproxifan compared to control (set to 100%) were observed, suggesting a reversible inhibition type ( Table 2). The remained enzyme activities for hMAO A and hMAO B preincubated with ciproxifan were 107.7 ± 3.4% and 91.4 ± 9.7%, respectively. In order to verify the test procedure, l-deprenyl was tested in the same manner showing decreased remaining enzyme activity of hMAO B (51.1 ± 2.9%) after preincubation.

Discussion
Ciproxifan is frequently used as the reference histamine H 3 receptor (H3R) antagonist in rodent models for neurological diseases like cognition 6 , Alzheimer's disease 7 or sleep-wake disorders 10,17 , because of its explicit high affinity and efficacy in rodent H3R (K i < 1.0 nM) which is about 30-to 100-fold lower than that at the human H3R. In our study we showed an additional property of ciproxifan. It inhibits human and rat MAO A and MAO B reversibly in a micromolar concentration range with a slight preference for MAO B.
We consider a combined activity pattern of ligands at H3R and MAO as new interesting approach for the treatment of neurological diseases. These are often associated with a neurotransmitter dysregulation, assumed to be adjustable by H3R blockade 12 . Neurotransmitter levels could be also modulated by inhibition of their degradation. MAOs are enzymes involved in oxidative deamination of neurotransmitters in neuronal cells after reuptake from the synaptic cleft. Therefore, deactivation of MAO A or MAO B is a principle well established in therapy of neurological disorders like depression and Parkinson's disease 18 . Additionally, MAOs are thought to promote oxidative stress when highly expressed in neuronal tissues. This can force increased neuronal cell death, a condition observed in Alzheimer's disease 19 . Thus, we hypothesize that reversible or more probably irreversible MAO inhibitors and H3R inverse agonists/antagonists can have overlapping pharmacological utilities, suggesting ligands, interacting with both targets, as promising candidates for treatment of neurodegenerative diseases. Concerning ciproxifan, which displays IC 50 values for both hMAO isoforms only in a micromolar range, we found its inhibition far too low for therapeutic efficacy in humans, which is anyway limited by its low affinity at human H3R. For example, safinamide, a MAO B selective reversible inhibitor most recently approved as first add-on treatment of Parkinson's disease, is active in submicromolar concentration ranges (IC 50 = 0.048-0.112 μ M for hMAO B; IC 50, MAO B / IC 50, MAO A < 0.001 [20][21][22]    point for investigation of dual targeting ligands in the future. Since ciproxifan is frequently used in rodent models for several neurological diseases, its MAO inhibition should be taken into consideration in retrospect or in the future using rodent models as possible accompanying effect. Ciproxifan has about four orders of magnitude higher activity at rH3R compared to the in-cell target rMAO. However, with a calculated log P value of 2.76 24 we assume good membrane penetration by ciproxifan 23 . Additionally, it could be shown that ciproxifan can reach brain concentrations up to approximately 10 μ M when applied i.p. to rats 17 , evidencing that partial rMAO inhibition is at least conceivable under test conditions. Taken together, the moderate and reversible MAO inhibitory properties of ciproxifan are interesting newly described properties which may interact with some previous animal screening data at high ciproxifan dosages. It can still be taken as a reference H3R antagonist, but its MAO A and MAO B inhibitory properties may be considered in retrospect on high dosage screenings.
Nevertheless, ciproxifan may serve as starting point for the design of dual targeting ligands combining H 3 receptor inverse agonism/antagonism and MAO inhibition, possibly favourable in treatment various neurological diseases. Since ciproxifan inhibited rat brain MAO also in micromolar concentration ranges, its activity has to be take into consideration in the future as a possible accompanying effect, when using it in rat models for neurological diseases. In retrospect, some of its effects explored in different species were probably contributed by its MAO A/B inhibitory properties.

Methods
All authors confirmed that all methods were carried out in accordance with relevant guidelines and regulations. . For optimal enzyme activity all assays were carried out under potassium phosphate buffered conditions (50 mM, pH = 7.4). Initial velocities were determined spectrophotometrically by a microplate reader at 30 °C by following product formation of 5-hydroxyquinoline and benzaldehyde at 316 nm and 250 nm, respectively, over a period of at least 30 minutes (interval of 20-30 seconds). Initial velocities, expressed as mAU min −1 , were obtained from the linear phase of product formation (see Supplementary Information, Fig. 1). Data were analysed using GraphPad PRISM version 6. For IC 50 determinations initial velocities for each experiment were normalized (expressed as percentage), plotted against inhibitor concentrations and fitted using the non-linear regression "log inhibitor vs. response (three parameters)". IC 50 values were determined in at least three independent experiments, each performed at least in duplicates.

Spectrophotometric IC
Radiometric IC 50 determination using rat brain MAO. Wistar male rats were sacrificed, the brains were quickly excised from the skulls, cleaned of residual meninges and frozen on dry ice. For each experiment the crude homogenates prepared from pooled brain from three rats were used. Enzyme activity of MAO A or MAO B was measured using radioactive substrate (PerkinElmer/NEN): serotonin (5-[2-14 C]-hydroxytryptamine binoxalate) or β -[ethyl-1-14 C]-phenyl-ethyl-amine hydrochloride (PEA), respectively, with the procedure described by Fowler and Tipton 1981 25 with some modification according to Gómez et al. 26 . IC 50 values were determined using six different concentrations of ciproxifan between 10 −9 M and 5 × 10 −4 M at fixed concentration of substrate (200 μ M serotonin or 20 μ M PEA). For IC 50 determinations values were normalized (expressed as percentage), plotted against inhibitor concentrations and fitted using the non-linear regression "log inhibitor vs. response (three parameters)". IC 50 values were determined in three independent experiments, each performed in duplicates.

Reversibility of human MAO inhibition. Reversibility of hMAO inhibition by ciproxifan was assessed
by dilution experiments. Supplied MAOs (5 mg mL −1 in potassium phosphate 100 mM, sucrose 0.25 M, EDTA 0.1 mM, glycerol 5%) were preincubated with inhibitor (10 × IC 50 ) or water at 30 °C for 15 minutes in a water bath. The inhibitor volume represented one-tenth of the total preincubation volume. After preincubation, probes were 100-fold diluted with potassium phosphate buffer (100 mM, pH = 7.4) to yield a final concentration of 12.5 ng μ L −1 for the enzyme. Enzymatic conversion rates were determined in the presence of kynuramine for hMAO A and benzylamine for hMAO B at saturating substrate concentrations (10 × K M ) under potassium phosphate buffer conditions (50 mM, pH = 7.4). Spectrophotometric measurements were carried out as described for IC 50 determinations over a period of at least 30 minutes for hMAO A and hMAO B, respectively (interval of 20-30 seconds). Initial velocities, obtain from the linear phase of product formation, of hMAO preincubated with inhibitor (10 × IC 50 ) were compared to hMAO preincubated with water (control) to define reversible or irreversible inhibition mode (see Supplementary Information, Fig. 2). Initial velocities, expressed as mAU min −1 , were normalized and given as percent of control.