1. Introduction

The kynurenine (KYN) pathway plays an important role in the production of nicotinamide adenine dinucleotide (NAD+), nicotinamide adenine dinucleotide phosphate (NADP+) and other notable compounds from the cleavage of tryptophan (Trp) [1]. As a result of this diverse, multiple branched cleavage pathway, 95% of Trp is converted into kynurenic acid (KYNA) (

Figure 1

). KYNA is a well-known endogenous neuromodulator that controls the level of several neurotransmitters, such as glutamate, acetylcholine, dopamine and γ-aminobutyric acid (GABA). By being a non-competitive antagonist at the glycine site of glutamatergic N-methyl-D-aspartate (NMDA) receptor, KYNA is able to decrease excitotoxicity and serves as an endogenous neuroprotectant.

Figure 1.

.

Besides, KYNA is a weak antagonist on kainate and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamatergic and a few other receptors. Moreover, KYNA also acts as a free radical scavenger. All these effects complement the potent neuroprotective profile of KYNA ^[2]. An increasing amount of data regarding KYN metabolites suggests their indisputable role in neurophysiological diseases, such as ischaemia, headache, schizophrenia, epilepsy, Alzheimer’s disease (AD), Huntington’s disease (HD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS) and multiple sclerosis (MS). It has been shown that levels of KYNA drop in the plasma and in red blood cells of AD patients ^[3] as also in cerebrospinal fluid (CSF) of relapsing-onset MS patients during remission ^[4]. In the animal model of depression, decreased levels of KYNA were observed in the prefrontal cortex ^[5], while in HD patients, low levels of KYNA and reduced activity of kynurenine aminotransferase (KAT) were detected in the striatum and cortex ^[6][7][8]. These results arise the question whether exogenous application of KYNA could be beneficial in the treatment of neurological diseases. Unfortunately, KYNA has a very limited ability to pass the blood–brain barrier (BBB). By forming an active interface between the blood and the central nervous system (CNS), the BBB limits the permeability of several substances into the brain tissue. Formation of the BBB is one of the main functions of the neurovascular unit (NVU), endothelial cells of the brain being the most restrictive in the organism. Continuous tight junctions, efflux transporters, metabolic enzymes and other mechanisms contribute to the low permeability of the BBB, which not only restricts passage of harmful molecules, but also of several therapeutic agents ^[9]. Thus, pharmacological approaches are urgently needed to overcome the issue of low penetration of KYNA into the brain and to achieve its therapeutic potential ^[2][10]. There are three main possibilities that can be considered: first, administration of the BBB-penetrant KYNA precursor, KYN and its halogenated derivates. Fukushima and colleagues presented a comparative study of intraperitoneally administered natural isomer L-KYN and the unnatural D-KYN in plasma. Majority of L-KYN was rapidly metabolized by kynurenine-3-hydroxylase and kynureninase into 3-hydroxykynurenine and anthranilic acid and not into KYNA by KAT I and KAT II in the tissue, while D-KYN was metabolized into KYNA in the plasma and KYNA failed to enter to the brain tissue ^[11]. The second approach is to inhibit the efflux of KYNA from the brain by using probenecid ^[12][13] or inhibiting the conversion of KYN to quinolinic acid by using FCE 28833A ^[14]. Finally, analogues of KYNA can be developed to cross the BBB and to preserve the neuroprotective effect of KYNA in the brain. Beside structural modifications, penetration of KYNA through the BBB can be enhanced by encapsulating it into nanoparticles. Bovine serum albumin (BSA)/KYNA-core was encapsulated by polyallylamine hydrochloride (PAH) as a shell and it had significantly higher permeability than free KYNA ^[15]. Recent studies have shown that different KYNA derivatives may have as good neuroprotective properties as KYNA itself with possibly better pharmacokinetics. KYNA and its pharmacologically modified analogues proved to have antinociceptive effects at both first and second order sensory neurons in headache models ^[16]. In an animal model of migraine, administration of KYNA and 

Besides, KYNA is a weak antagonist on kainate and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamatergic and a few other receptors. Moreover, KYNA also acts as a free radical scavenger. All these effects complement the potent neuroprotective profile of KYNA [2]. An increasing amount of data regarding KYN metabolites suggests their indisputable role in neurophysiological diseases, such as ischaemia, headache, schizophrenia, epilepsy, Alzheimer’s disease (AD), Huntington’s disease (HD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS) and multiple sclerosis (MS). It has been shown that levels of KYNA drop in the plasma and in red blood cells of AD patients [3] as also in cerebrospinal fluid (CSF) of relapsing-onset MS patients during remission [4]. In the animal model of depression, decreased levels of KYNA were observed in the prefrontal cortex [5], while in HD patients, low levels of KYNA and reduced activity of kynurenine aminotransferase (KAT) were detected in the striatum and cortex [6,7,8]. These results arise the question whether exogenous application of KYNA could be beneficial in the treatment of neurological diseases. Unfortunately, KYNA has a very limited ability to pass the blood–brain barrier (BBB). By forming an active interface between the blood and the central nervous system (CNS), the BBB limits the permeability of several substances into the brain tissue. Formation of the BBB is one of the main functions of the neurovascular unit (NVU), endothelial cells of the brain being the most restrictive in the organism. Continuous tight junctions, efflux transporters, metabolic enzymes and other mechanisms contribute to the low permeability of the BBB, which not only restricts passage of harmful molecules, but also of several therapeutic agents [9]. Thus, pharmacological approaches are urgently needed to overcome the issue of low penetration of KYNA into the brain and to achieve its therapeutic potential [2,10]. There are three main possibilities that can be considered: first, administration of the BBB-penetrant KYNA precursor, KYN and its halogenated derivates. Fukushima and colleagues presented a comparative study of intraperitoneally administered natural isomer L-KYN and the unnatural D-KYN in plasma. Majority of L-KYN was rapidly metabolized by kynurenine-3-hydroxylase and kynureninase into 3-hydroxykynurenine and anthranilic acid and not into KYNA by KAT I and KAT II in the tissue, while D-KYN was metabolized into KYNA in the plasma and KYNA failed to enter to the brain tissue [11]. The second approach is to inhibit the efflux of KYNA from the brain by using probenecid [12,13] or inhibiting the conversion of KYN to quinolinic acid by using FCE 28833A [14]. Finally, analogues of KYNA can be developed to cross the BBB and to preserve the neuroprotective effect of KYNA in the brain. Beside structural modifications, penetration of KYNA through the BBB can be enhanced by encapsulating it into nanoparticles. Bovine serum albumin (BSA)/KYNA-core was encapsulated by polyallylamine hydrochloride (PAH) as a shell and it had significantly higher permeability than free KYNA [15]. Recent studies have shown that different KYNA derivatives may have as good neuroprotective properties as KYNA itself with possibly better pharmacokinetics. KYNA and its pharmacologically modified analogues proved to have antinociceptive effects at both first and second order sensory neurons in headache models [16]. In an animal model of migraine, administration of KYNA and 

N

-(2-

N

,

N

-dimethylaminoethyl)-4-oxo-1H-quinoline-2-carboxamide (SZR-72; also called KYNA-a, KYNAA1 or KYNAA2), decreased electrical trigeminal ganglion (TRG) stimulation-induced PACAP mRNA expression in nucleus trigeminalis caudalis (NTC) [17]. In the inflammation model of TRG activation, overexpression of mitogen-activated protein kinase (MAPK) and NFκB was observed, which could be reduced by intraperitoneally administered KYNA and SZR-72 [18]; furthermore, systematic injection of KYNA and SZR-72 also abolished the elevated levels of c-fos and glutamate in NTC neurons [19]. Greco and colleagues demonstrated the antihyperalgesic effect of KYNA and SZR-72 by attenuating the nitroglycerin-induced neuronal activation in NTC [20]. Beside migraine, excitotoxicity is also known to be involved in the pathology of HD. In a transgenic mouse model of HD, administration of SZR-72 completely prevented the atrophy of the striatal neurons, prolonged the lifetime of mice and ameliorated their hypolocomotion [21]. In order to find new KYNA analogues with better CNS penetration, C-3 aminoalkylated KYNA derivatives bearing amide side-chains have been recently synthesized [22]. In the present study, we present a systematic study on the BBB permeability of the compounds.

2. Novel Strategies and Discussion

2. Results and Discussion

As KYNA has poor CNS penetration, novel strategies are needed to take advantage of its important neuroprotective effects. Among the several possible approaches [9], two strategies seem to be the most suitable in this respect. First, the use of vector-mediated drug delivery through the brain endothelium, especially nanocarriers, which are admittedly among the most promising tools to overcome the BBB [9]. Our research consortium has been successfully working in this field as well [15]. Second, the synthetic approach, i.e., chemical modification of KYNA to obtain compounds with similar biological effects but significantly improved ability to cross the BBB-forming cerebral endothelial cells. Here, we applied this second strategy and we designed new compounds (aminoalkylated KYNA amide derivatives). The compounds investigated and thus the modifications carried out on the KYNA skeleton were chosen based on a previous study on C-3 substituted KYNA derivatives [23]. Through the use of different secondary amines (piperidine, pyrrolidine) and also morpholine, we aimed to investigate the BBB penetration altering effect of the aminoalkyl function formed during the modified Mannich reactions. We first assessed the permeability of the newly synthesized compounds using an in vitro model system, which mimics the in vivo anatomical structure of the BBB (

Figure 2) and is suitable for drug testing ^[24][25].

4) and is suitable for drug testing [25,26].

Figure 24.

 The in vitro BBB model. (

A

) Schematic representation of the cross-section of a brain capillary, showing the cellular composition. (

B

) The in vitro BBB model consists of cerebral endothelial cells cultured on filter inserts having pericytes on the bottom surface. Filters containing endothelial cells and pericytes are emerged in a well containing astrocyte-conditioned media. Right panels show representative immunofluorescence staining of respective cellular markers.

KYNA analogues (SZR-100, SZR-101, SZR-104, SZR-105, SZR-106 and SZR-107) were applied in the top/blood compartment in a final concentration of 10 µM. Samples were collected from the bottom/brain compartment after 60 min. Sodium-fluorescein was used as a control compound, since it is a small molecular tracer for the indication of cellular junctions’ integrity and restricted paracellular transport between cerebral endothelial cells. All of the above-mentioned analogues crossed the BBB more efficiently than sodium-fluorescein. Furthermore, the permeability of SZR-104 and SZR-105 was significantly higher than the permeability of other analogues (

Figure 3).

5).

Figure 35.

 Penetration of KYNA analogues through the BBB. (

A

) Permeability of KYNA derivates after 60 min compared to sodium-fluorescein. N = 2, average ± SD (ANOVA and Bonferroni’s post hoc test). * 

p

 < 0.05 (SZR-104 and SZR-105: significant difference compared to all other groups; significant difference between SZR-106 and SZR-101 or SZR-107). (

B

) Permeability coefficients of KYNA analogues after 60 min compared to sodium-fluorescein.

SZR-104 not only had the highest permeability in our preliminary study, but previous results suggested that it had important biological effects. In a recent study, human lymphoma U-937 cells were infected with Staphylococcus aureus to induce cytokine production and then were treated with KYNA and its newly synthesized analogues (SZR-104, SZR-105 and SZR-109). The analogues reduced tumor necrosis factor-α (TNF-α) production and increased the mRNA expression of the anti-inflammatory Tumor necrosis factor-Stimulated Gene-6 (

TSG-6) ^[26]. In addition, an in vivo electrophysiological study has demonstrated that systemic administration of SZR-104 decreased population spike activity in the hippocampus, and additionally provided protection against pentylenetetrazol-induced epileptiform seizures ^[27]. Therefore, we continued our study and focused on the characterization of SZR-104. We compared its permeability with that of KYNA, xanthurenic acid and its analogue, 39B, which is under patent protection ^[23]. SZR-104 had a significantly higher permeability through the in vitro BBB model than KYNA, xanthurenic acid or 39B both at 30 min and at 60 min time points. Differences among permeability of KYNA, xanthurenic acid and 39B were not statistically significant (

) [27]. In addition, an in vivo electrophysiological study has demonstrated that systemic administration of SZR-104 decreased population spike activity in the hippocampus, and additionally provided protection against pentylenetetrazol-induced epileptiform seizures [28]. Therefore, we continued our study and focused on the characterization of SZR-104. We compared its permeability with that of KYNA, xanthurenic acid and its analogue, 39B, which is under patent protection [23]. SZR-104 had a significantly higher permeability through the in vitro BBB model than KYNA, xanthurenic acid or 39B both at 30 min and at 60 min time points. Differences among permeability of KYNA, xanthurenic acid and 39B were not statistically significant (

Figure 4). Our results indicate that SZR-104 has a much higher permeability through the BBB than free KYNA.

6). Our results indicate that SZR-104 has a much higher permeability through the BBB than free KYNA.

Figure 46.

 Penetration of SZR-104 through the BBB

. (A)

 Permeability of SZR-104 after 30 and 60 min compared to KYNA, xanthurenic acid and its analogue, 39B. N = 3, average ± SD. * 

p

 < 0.01 (SZR-104: significant difference compared to all other groups, in both time-points). 

(B)

 Permeability coefficients of SZR-104, xanthurenic acid and 39B after 30 and 60 min compared to KYNA.

3. Conclusions

Based on our systematic investigation, SZR-104 and SZR-105 both showed high BBB penetration (with the former having the highest apparent permeability) compared to the other derivatives tested. In line with this comparison, we hypothesize that aminoalkylation in C-3 facilitates BBB penetration with the morpholinomethyl functional group showing the best results. These results are in line with in vivo data showing that peripherally administered SZR-104 can reach sufficient concentration in the brain, since it is able to inhibit epileptiform activity ^[27]. In conclusion, SZR-104 is a promising neuroprotective candidate drug.

Based on our systematic investigation, SZR-104 and SZR-105 both showed high BBB penetration (with the former having the highest apparent permeability) compared to the other derivatives tested. In line with this comparison, we hypothesize that aminoalkylation in C-3 facilitates BBB penetration with the morpholinomethyl functional group showing the best results. These results are in line with in vivo data showing that peripherally administered SZR-104 can reach sufficient concentration in the brain, since it is able to inhibit epileptiform activity [28]. In conclusion, SZR-104 is a promising neuroprotective candidate drug.