The invention concerns a compound inhibiting butyrylcholinesterase (BChE), the compound for use in the treatment of a disease, a use of that compound and a method for synthesizing that compound. The BChE can be human butyrylcholines terase (hBChE).

The most established theory for development of Alzheimer ^' s disease (AD) is the cholinergic hypothesis. Briefly, patients in AD suffer from an extensive loss of cho linergic neurons and cholinergic activity caused by reduced activity of choline acet- ylase. A rise of acetylcholine level by inhibiting its enzymatic hydrolysis by acetyl cholinesterase (AChE) proved to slow down the progression of AD and had posi tive effect on patients ^' cognition and overall mental abilities in numerous clinical studies, both animal and human. Modern drugs for the treatment of AD target the two cholinesterases AChE and BChE. Among the two reversible and selective AChE inhibitors donepezil and galantamine, there is also the carbamate based drug molecule rivastigmine, targeting both enzymes in a pseudo irreversible man ner. Clinical trials showed that the approved drug molecules described above lack pharmacological action and effectivity, especially in later stages of AD, whereas they slow down the progression of dementia in early stages.

It has been found that BChE activity rises in response to reduced AChE activity in later stage of AD and that BChE takes over the hydrolytic function of AChE. There fore, inhibition of AChE is not sufficient to enhance acetylcholine (Ach) levels in this stage. BChE inhibitors could prove to be effective in the treatment of scopola mine induced amnesia and in an AD mouse model in vivo indicating their value for innovative treatments of dementia both from Alzheimer and other types. Significant progress on new cholinesterase inhibitors targeting both enzymes together has been reported in the last years, but inhibitors often lack selectivity and long term binding towards BChE. From the doctoral thesis of Edgar Sawatzky“Design and Synthesis of Selective Butyrylcholinesterase (BChE) Inhibitors for the Development of Radiotracers to In vestigate the Role of BChE in Alzheimer’s Disease” selective carbamate based BChE inhibitors having a tetrahydroquinazoline carrier scaffold are known. Com- pounds dual-targeting the Catalytic Active Site (CAS) and the Peripheral Anionic Site (PAS) of BChE such as

were found to be potent and selective BChE inhibitors that lack long term binding. The postu- lated compound

could not be synthesized.

The problem to be solved by the present invention is to provide a compound being an alternative potent and selective BChE inhibitor, the compound for use in the treatment of a disease, a use of that compound and a method for synthesizing that compound.

The problem is solved by the features of claims 1 , 9, 10 and 11. Embodiments are subject-matter of claims 2 to 8 and 12 to 14.

According to the invention a compound according to formula is provided, wherein 3 < n < 1 1 , wherein R consists of a heterocycle comprising one tertiary amino group providing the binding of R to the rest of the molecule and at least three carbon atoms, wherein all carbon atoms of the heterocycle are unsubstituted or wherein R is a fused bicy- clic compound comprising the heterocycle and one further cyclic compound. Con trary to the above-mentioned postulated compound, synthesis of the compound according to the invention could be performed in good yield without any problems. The inventors found that the compound is a potent BChE inhibitor showing pro- cognitive and antineuroinflammatory properties in vivo in an animal model of AD.

In vitro experiments showed that the properties are dependent on the length of the linker formed by the n methylenegroups.

Inhibition of BChE activity was shown to be selecitve vis-a-vis AChE. Furthermore, the inhibition was shown to be pseudo-irreversble and particularly effective when n is in the range of 6 to 10, in particular when n is 6, 7 or 8 and in particular when n is 6. The inhibition is pseudo-irreversible because the carbamate moiety of the inhibitor is transferred onto the serine hydroxyl group in the catalytic center of the BChE. As a covalent bond is formed during the inhibition, the inhibitor is bound relatively fixed to the enzyme. However, due to chemical instability, this inhibition is not completely irreversible. Enzyme activity is recovered over time by carbamate hydrolysis. Duration of action of the compound on the enzyme can be controled by chemical modification and design of the carbamate moiety. Especially a linker of 6 carbon atoms between carbamate group and a basic hexacylic amine has been shown to promote the long term binding to the enzyme.

The whole heterocycle may be unsubstituted. This simplifies synthesis of the com pound and enhances its enzyme inhibiting activity.

In an embodiment the heterocycle is a 5-, 6-, 7- or 8-membered ring. The further cyclic compound may be an aromatic cyclic compound. It can be a car- bocyclic compound or a heterocyclic compound. In an embodiment the further cy clic compound is benzene. The fused bicyclic compound may be unsubstituted.

R can be any of the following moieties:

and Boc ^v , wherein Boc is a tert-butyloxycarbonyl group. In a very effec tive compound according to the invention R is . It is particularly specific vis-a-vis inhibition of AChE ac- tivity and effective as inhibitor of BChE activity, in particular hBChE activity when n

is 6, 7 or 8. In the compound most effective in vivo R was G _", ^l and n was 6.

The invention further concerns a compound according to the invention for use in the treatment of Alzheimer's Disease.

A further aspect of the invention concerns the use of the compound according to the invention for inhibition of butyrylcholinesterase (BChE) in vitro.

Another aspect of the invention concerns a method for synthesizing a compound according to the invention comprising the following steps: a) Reacting a primary dihaloalkane comprising 4 to 10 methylene groups with a sodium or potassium salt of phthalimide to give the corresponding haloal- kylated phthalimide, b) reacting the haloalkylated phthalimide with an amine and H-R or activating the haloalkylated phthalimide in a Finkelstein reaction and then reacting the resulting iodoalkylated or fluoroalkylated phthalimide with H-R to yield an R- alkylated phthalimide, c) subjecting the R-alkylated phthalimide to hydrazinolysis or hydrolysis to yield alkylated amines of R, d) activating the alkylated amines of R into a carbamate by use of 4-nitro- phenyl chloroformate and e) reacting the carbamate with the tetrahydroisoquinazoline derivative accord ing to formula to yield the compound.

Steps a) and c) form a Gabriel-synthesis. The introduction of the residue R in step b) can be achieved in two different ways: In the first approach, the amine H-R may be stirred together with the haloalkylated phthalimide and triethyl amine in dry Di- methylformamide (DMF) to yield the R-alkyl phthalimide. In the second approach, a two-step, one pot reaction may be used by activating the haloalkylated phthalimide in a Finkelstein reaction first, and then adding the respective amine H- R to alkylate it. In some cases the second approach may result in an improved yield vis-a-vis the first approach.

Step e) may be performed by addition of an alkali hydride, in particular sodium hy dride, and the tetrahydroisoquinazoline derivative to the carbamate. In an embodiment the salt of the phthalimide is the potassium salt. The primary di- haloalkane can be a primary dibromoalkane. In this case the haloalkylated phthalimide resulting from step a) is bromoalkyl phthalimide. R may be

y

Boc ^v , wherein Boc is a tert-butyloxycarbonyl group.

The hydrazinolysis may be performed by heating the R-alkylated phthalimide in hydrazine and ethanol. The hydrolysis may be an acidic or an alkaline hydrolysis.

The reaction mixture resulting from step d) which reaction mixture comprises the carbamate may be directly subjected to step e). Directly means that the carbamate is not purified from the reaction mixture before performing step e). Step e) may be performed by addition of an alkali hydride, in particular sodium hydride, and the tetrahydroisoquinazoline derivative to the reaction mixture.

When R is

Boc

step e) may be followed by a deprotection under acidic conditions resulting in com pound

without cleaving the main pharmacophoric carbamate and the cyclic aminal struc ture.

Embodiments of the invention:

Fig. 1 shows a reaction scheme for the synthesis of a compound required for the synthesis of the compound according to the invention.

Fig. 2 shows a reaction scheme for the synthesis of the compound according to the invention.

Fig. 3 shows a diagram of the results of determination of ICso values of com pounds according to the invention having different length of the linker.

Fig. 4 shows a diagram of the results of determination of inhibition of BChE activity over time depending on the concentration of a compound ac cording to the invention.

Fig. 5 shows a diagram of the results of determination of regeneration of

BChE activity over time of different compounds according to the inven tion.

Fig. 6 shows a protocol of the experimental treatment and testing of mice.

Figs. 7a-h show the effect of the compounds on Ap25-3s-induced spontaneous al ternation deficits in mice.

Figs. 8a-d show the effect of the compounds on Ap25-3s-induced passive avoid ance impairments in mice.

Synthesis of the target compounds can be divided into two parts followed by the coupling of the two building blocks in the last step. The synthesis of the tetrahydroisoquinazoline scaffold is illustrated in Fig. 1 and known from Darras et al. , ACS Med. Chem. Lett. 2012, 3, 914-919 and Sawatzky, E. et al. , Tetrahedron Lett. 2014, 55, 2973-2976. Briefly, 3-anthralinic acid 10 was reacted with triphos- gen, followed by methylation to give anhydride 11. Benzyl protection of the phenolic alcohol group and subsequent ring fusion with dihydroisoquinoline followed by debenzylation under hydrogenation conditions gave dihydroquinazolinone 12. Re duction with LiAIFU gave tetrahydroisoquinazoline 13.

Referring to Fig. 1 reagents and reaction conditions were as follows: (i) 1 . CO(CCI ₃ )2, THF, 70°C, 3-5 h; 2. Methyl iodide, DIPEA, DMAc, 40°C, 24h; (ii) 1 . Benzyl bromide, K2CO3, DMF, 40°C, 4h; 2. Dihydroisoquinoline, DMF, 120°C, 18 h; 3. H2, Pd/C, MeOH, 50°C, 3 h; (iii) LiAIFU, THF, reflux, 3 h.

To investigate how a peripheral binding site of BChE can be addressed with carba mate based inhibitors, different alkylene linkers and residues were introduced into the carbamate part of the inhibitors by using Gabriel synthesis as illustrated in Fig. 2 in combination with Table 1 . Therefor, different dibromo alkanes 14a-f of length from 2 to 10 methylene groups were reacted in each case with potassium phthalimide in DMF to give the alkylated phthalimides 15a-f in high yields. For the introduction of the amine residues two different strategies were applied:

In a first approach, the respective amine, such as morpholine or 1 ,2,3,4-tetrahydro- sioquinolin, was stirred together with the respective bromoalkyl phthalimide 15a-f and triethyl amine in dry DMF to yield the alkylated amine derivatives 16a-l.

In a second approach, a two-step, one pot reaction was used by activating the bro- moalkane 15d in a Finkelstein reaction first, and then adding the respective amine to alkylate it, which gave access to improved yields of 16m-q.

Subsequent hydrazinolysis in ethanol under reflux conditions gave respective amines 17a-q in high yields. Amines were then activated into respective carbamates 18a-q by using 4-nitrophenyl chloroformate, which carbamates were then transferred onto the tetrahydroisoquinazoline scaffold 13 to yield the respective car bamate based BChE inhibitors 19a-q.

Deprotection of Boc-protected piperazine 19q under acidic conditions yielded an- other prospective inhibitor 19r without cleaving the main pharmacophoric carbamate and the cyclic aminal structure.

Referring to Fig. 2 reagents and reaction conditions were as follows: (i) Potassium phthalimide, DMF, 24 h, r.t. (ii) a) R-H, NEts, DMF, 24 h, 100°C or b) Nal, NEts, THF, o.n. reflux, 60-91 % (iii) N2H4Ή2O, ethanol, 18 h, reflux, 90-100% (iv) 4-Nitro- phenylchloroformate, NEts, DCM, 1 -6 h, r.t., 20-86%; (v) NaH, DCM (dry), 1 -48 h, r.t., 30-80%; (vi) 1.25 M HCI in methanol, ethanol (dry), 4.5 h, 0°C -> r.t., 96%; (vii) 1. 4-Nitrophenylchloroformate, NEt3, DCM, 1-6 h, r.t., no purification, 2. NaFI, 13, DCM (dry), 1 -48 h, r.t., 40-60%. For substitution pattern of n and R see table 1. Pathway vii is an alternative pathway to separate pathways iv and v. In pathway vii the reaction mixture resulting from pathway iv is directly - i. e. without purification of carbamates 18a-q - used for reacting the carbamate contained in the reaction mix ture with the tetrahydroisoquinazoline derivative 13. Table 1. Overview on synthesized compounds according to Fig. 2

n R n R

For evaluation of enzyme inhibition, Ellman ^' s colorimetric assay was conducted as described in Ellman, G. L., Arch. Biochem. Biophys 1958, 74, 443-450 and Ellman, G. L. et al., Biochem. Pharmacol. 1961 , 7, 88-95. Therefor, 5,5'-dithiobis-(2-nitro- benzoic acid) (DTNB) solution, respective Cholinesterase (ChE) and inhibitor were incubated together and enzyme activity was measured after 20 min. The long incu bation time was needed to allow the carbamate transfer between ChE and inhibitor.

The carbamate residue was altered systematically by changing residues in the ter- minal position of carbamate moiety and different flexible alkylene linkers were intro duced. Obtained enzyme inhibition constants were reviewed with respect to bench mark heptyl carbamate 7 described in Darras, F. H. et al., ACS Med. Chem. Lett. 2012, 3, 914-919. As compound 7 shows low nanomolar inhibition of human BChE (/?BChE) and approximately 1000-fold higher inhibition of human AChE (/?AChE), it serves as a perfect lead compound for further optimizations because of its selectiv ity, high potency and plasma stability. Nevertheless, it must be noted that IC50 values only serve as a first indicator of inhibitor affinity and selectivity. Due to the chemical reaction of carbamate transfer the enzyme inhibition is strongly time dependent. Ki netic investigations indicated that 20 min is usually sufficient to allow carbamate transfer.

Table 2. Inhibition of h BChE and /?AChE by carbamate based inhibitors 7, 9, 19a-r and binding of these inhibitors to h BChE and /?AChE

(pICso ± SEM)

R n

7 CHs 6 6.4 * 3.8 mM 594

(8.19 ± 0.02) (5.42 ± 0.05) *

9 F 7 5.2 * 3.6 pM 692

(8.28 ± 0.02) (5.44 ± 0.05) *

19a X 2 7772 17% ^a >10

(5.11 ± 0.04)

19b X 3 937.8 13% ^a >100

(6.03 ± 0.02)

19c 4 50.6 13% ^a >1000

(7.30 ± 0.02)

19d 6 49.3 45% ^a >1000

(7.31 ± 0.02)

19e 8 19.8 22 % ^a >1000

(7.07 ± 0.06)

10 11.5 25% ^a >1000

(7.94 ± 0.07)

2 1720 10% ^b >10

(5.76 ± 0.05)

3 1 160 12% ^b >10

(5.94 ± 0.05)

4 191.7 16% ^b >100

(6.72 ± 0.06)

6 26.7 9% ^b >1000

(7.57 ± 0.05)

8 32.6 11 % ^b >1000

(7.49 ± 0.03)

10 41.8 15% ^b >1000

(7.38 ±0.04)

6 7.4 44% ^c >1000

(8.13 ± 0.04)

6 13.3 18% ^c >1000

(7.88 ± 0.06)

6 34.3 0.88 mM 26

(6.47 ± 0.04) (6.06 ± 0.03)

6 83.7 1.2 mM 14

(7.08 ± 0.04) (5.91 ± 0.03)

6 123.4 15% ^c >100

(6.1 ± 0.04)

6 80.6 16% ^c >100

(7.09 ± 0.03) a100 mM; ^b 10 mM ^c 25 mM

‘According to Sawatzky, E. et al. , ChemMedChem 2016, 1 1 , 1540-1550 Table 2 shows that inhibition of h BChE and /?AChE by carbamate based inhibitors 7, 9 and 19a-r and binding of these inhibitors to /?BChE and /?AChE depend on linker length n and residue R.

To investigate the influence of length of carbon chain linker between carbamate group and residue possibly targeting a second binding site located near the catalyt- ical active serin in the center of /?BChE, a set of compounds 19a-f having a morpho line derivative residue and a further set of compounds 19g-l having a tetrahydroiso- quinoline derivative residues were investigated. The compounds in each of the two sets differ from each other by the number n of carbon atoms in the carbon chain linker between carbamate group and residue. Fig. 3 shows the ICso values obtained with /?BChE and the two compound sets having morpholine 19a-f (symbol "+") and tetrahydroisoquinoline derivatives 19g-l (symbol "x") as residues. Fig. 3 shows a strong dependency of the ICso values on the number n of carbon atoms in the carbon chain linker. From the ICso values it can be concluded, that short linker length influ ence enzyme inhibition in a negative manner. Elongation of the linker to 4 methylene groups as seen for compounds 19c and 19i lead to a significant increase of inhibi tory potency showing two-digit nanomolar ICso values comparable to known inhibi tors 7 and 9. Further elongation of linker length improves the activity of the morpho line compounds slightly, while for the tetrahydroisoquinoline compounds a linker length of 6 methylene groups (19j) is optimal and the affinity decreases with a longer linker. Nevertheless, ICso of heptyl carbamate 7 cannot be achieved with polar res idues such as morpholine, piperidine, pyrrolidine and piperazine, but approximately 10-fold lower activity of the compounds 19d, 19o and 19p still proves the finding of new very potent and even more selective BChE inhibitors. Most ICso values of the investigated compounds on /?AChE could not be determined due to their poor water solubility and low affinity.

From the binding data presented here, it can be concluded that /?BChE tolerates a huge variety of residues, when the linker is longer than 4, better 6 methylene groups. Both very polar groups like morpholine and sterically demanding groups like tetrahydroisoquinoline and benzimidazole do not alter the inhibition of h BChE in a significant manner. These binding data do not indicate the presence of a second binding site in /?BChE that can be addressed with the inhibitors.

As carbamate based inhibitors (C-X) show a pseudo irreversible manner of binding, the ICso values must be seen critically because two chemical reactions determine them: Carbamate (C) transfer onto the enzyme (E) and carbamate hydrolysis. So, the Kc values, k3 and k4 rate constants according to below formula were investigated for some representative compounds (c.f. Fig. 3 and Fig. 4).

The mechanism of enzyme inhibition can be divided into three steps: In a pre-equi librium, the carbamate based inhibitor C-X is bound to the enzyme E and a complex (EC-X) is formed reversible and can be characterized with K _c . Then the carbamate C is transferred onto the enzyme E and binds covalently to form E-C with the rate constant k3 while carrier scaffold X is released. Then E-C is hydrolyzed and enzyme activity is recovered and the amine C ^' is released with the rate constant k4, also called decarbamoylation rate.

For the determination of K _c and k3 a similar experimental setup was used compared to the evaluation of ICso values, whereas enzyme activity was measured at various time points after addition of inhibitor.

Thereby, inhibition with the carbamate-based inhibitors showed a first order kinetic with the apparent first order rate constant kobs, where A is the enzyme activity at time t, Ao the enzyme activity at t = 0 and A~ the enzyme activity at t = . A = A ₀ e ^_k ° ^bs + A _¥ (I)

Plotting the enzyme activity (in %) against the incubation times, lead to nonlinear time-dependent inhibition curves for the respective inhibitor concentrations whereby kobs can be determined with an exponential fit correspondent to equation (I). Fig. 4 shows the time-dependent inhibition of relative enzyme activity (rel. enz. act.) by different concentrations of compound 19j on /iBChE.

Furthermore, a hyperbolic curve is obtained, plotting the rate constant k ₀ bs against the inhibitor concentration [I], which is described by equation (II), with the substrate concentration [S] and the Michaelis-Menten constant KM.

This equation (II) can be rearranged to the reciprocal of kobs, to get a linear plot:

Since the substrate butyrylthiocholine iodide is not added directly to the enzyme and the inhibitor, but after the incubation of enzyme with inhibitor, [S] can be set to zero, leading to Michaelis-Menten like equation IV.

1 _ K _c 1 1

(IV)

ko _b s k ₃ [I] k ₃

Relating to the linear equation (IV), Kc can be established from the slope and car- bamoylation rate constant k3 from the y-intercept.

As a chemical reaction is taking place, Kc must be seen as a‘real’ measure for quantification of enzyme inhibitor interactions. Due to the high costs in measurements on h BChE only some representative compounds were chosen to evaluate also in kinetic studies based on the results from IC50 measurements (table 3). Table 3. Kinetic values for carbamoylation on h BChE for selected representative compounds

X

19r 6 121 ± 116 0.32 ± 0.13

HN^I

‘According to Sawatzky, E. et al. , ChemMedChem 2016, 1 1 , 1540-1550 From the carbamoylation rate constants k3 it can be concluded, that linker and res idue do not have a big influence on the carbamate transfer from the carrier scaffold to the enzyme. No additional effect of a peripheral binding site on the binding of inhibitor could be found. The only exception is tetrahydroisoquinoline compound 19g probably forcing the inhibitor into a different binding mode because of its geometry with a very bulky residue close to the carrier scaffold. This confirms the results of the IC50 measurements.

From this set of compounds, compound 19j must be noted with special interest as its Kc value is even smaller than the IC50 value which is rarely the case. This might indicate a peripheral binding as K _c characterizes the equilibrium of reversible binding (Fig. 4), whereas IC50 also includes pseudo-irreversible binding (carbamoylation). Nevertheless, it must be noted that IC50 gives an inhibitor concentration, whereas Kc is a dissociation constant and therefore cannot be interpreted as concentration. Therefore, it is remarkable to reach very low K _c values in the one-digit nanomolar range, showing that inhibitors 19j, m and n show a very high affinity to /?BChE and outreach lead compound 7 in this regard.

To further investigate the presence and nature of a peripheral binding site, also the hydrolysis of carbamate and the resulting release of the inhibitor from the enzyme was investigated by measuring the time dependent recovery of enzyme activity. To measure the decarbamoylation rate k4, dilution experiments were performed with /?BChE following the procedure described in Sawatzky, E. et al. , ChemMedChem 2016, 1 1 , 1540-1550 and in Bartolucci, C. et al., Biochem. J. 2012, 444, 269-277. A high concentration of enzyme and inhibitor where incubated for 1 h. The concen tration of inhibitor was chosen, such that > 85% of /?BChE was inhibited through carbamate transfer. Then the mixture was diluted with buffer 1 : 1000 to a concentra tion at which the enzyme is not further carbamoylated. Due to the hydrolysis and release of the carbamate from the catalytic active serine enzyme activity was recov ered slowly. This regeneration of the enzyme was monitored at different time points after the dilution. Plotting the time dependent enzyme activity against time showed a first order exponential curve, with the associated equation (V).

Table 4. Kinetic values for decarbamoylation from /?BChE of selected compounds

R n k ₄ [h- ¹ ] ti _/2 [E-C] [h]

7 CHs 6 0.62 ± 0.04 * 1.12 ± 0.07 *

9 F 7 0.81 ±0.04 * 0.86 ± 0.04 *

4 0.48 ± 0.06 1.44 ± 0.16

6 0.042 ± 0.004 16.50 ± 1.44

8 1.58 ± 0.15 0.44 ± 0.04

4 0.028 ± 0.001 24.76 ± 0.85

6 0.025 ± 0.001 27.73 ± 1.07

8 0.83 ± 0.06 0.84 ± 0.06

6 0.45 ± 0.04 1.55 ± 0.13

6 0.82 ± 0.03 0.85 ± 0.03

6 0.98 ± 0.04 0.71 ± 0.03 6 0.129 ± 0.006 5.37 ± 0.24 19r 6 1.44 ± 0.07 0.48 ± 0.02

‘According to Sawatzky, E. et al. , ChemMedChem 2016, 1 1 , 1540-1550

In case of the morpholine compounds a significant difference is observed comparing compounds 19c and 19d with 4 and 6 methylene groups in the linker, respectively. While the kinetic of hydrolysis of compound 19c is comparable to the heptyl carba mate reference compound 7, morpholine compound 19d shows a significantly re duced speed of enzyme regeneration with a half-life time ti/2 = 16.3 h compared to 1.4h for compound 19c and 1.1 h for reference compound 7. For a linker length of n = 8 in compound 19e the half-life time is decreased significantly again, so that a linker of 6 methylene groups was considered best. This novel finding indicates the presence of a second binding site because the enzyme is recovered considerably slower and since due to the alkylene linker attached to the carbamate moiety formed with the serine, no difference in chemical reactivity of the carbamate is expected. The time dependent enzyme regeneration from carbamoylated state [E-C] from compounds 19d( ^* ), 19c(+), 7 (x) is shown in Fig. 5

In contrast to the morpholine compounds the tetrahydroisoquinoline compounds did not show this behavior in such a pronounced manner. Nevertheless, two tested in hibitors showed a very long duration of inhibition. Compound 19i and 19j inhibit the enzyme for a long duration, while a linker length of n = 8 (19k) decreases the dura tion of inhibition as it was observed and described above. This finding could be ex plained by the fact, that the residue is bigger and can target a peripheral binding site also with shorter linkers.

The finding of a slower decarbamoylation of carbamate inhibitors from /?BChE could be explained both by a specific binding of the residues to a second binding region and/or unspecific blocking of water molecules from the serine in the CAS slowing down the chemical reaction of hydrolysis. Consequently, to distinguish between those two explanations more inhibitors with a linker length of n = 6 methylene groups were synthesized, including very different residues with basic and non-basic, ali phatic and aromatic, polar and non-polar properties.

In contrast to our expectation the aromatic compounds with benzimidazole 19m and imidazole residues 19n did not show improved duration of enzyme inhibition, despite their very promising binding and affinity data. This leads to the conclusion that a specific interaction between the transferred carbamate residue is generally possible, but aromatic residues do not address the PAS specifically. In case of an unspecific steric inhibition of carbamate hydrolysis it can be expected to observe a high stability of the carbamoylated enzyme inhibitor complex (E-C) (c.f. Fig. 4) when bulky resi dues like benzimidazole are introduced into the inhibitor.

In analogy to AChE inhibitors basic residues proved to be more promising for a binding to a PAS. Despite their comparably slightly deteriorated binding properties, most compounds bearing a basic amine function and a linker length of six methylene groups showed a decelerated hydrolysis behavior compared to the reference com pound 7.

This leads to the conclusion that both /?AChE and /?BChE possess a second binding site, which can be addressed with carbamate inhibitors and affects their duration of binding dramatically.

Generally, also size and shape of residue seems to play a role for deceleration of carbamate hydrolysis, whereas hydrolysis of benzimidazole carbamate 19m (ti/2 = 1 .55 ± 0.13 h) is slower than imidazole carbamate 19n (ti/2 = 0.85 ± 0.03 h). The accelerated hydrolysis of pyrrolidine compound 19o (ti/2 = 0.71 ± 0.03 h) in compar ison to the bigger piperidine carbamate 19p (ti/2 = 5.37 ± 0.24 h) is in agreement with this result.

Another explanation for this finding could be found in the conformation of the resi dues. The best compounds from the half-life aspect, all share a tert-piperidine moi ety (tetrahydroisoquinoline, morpholine and piperidine) and exhibit a chair conformation. In this conformation, the most potent compounds 19d, 19i and 19j all possess an electron rich system (oxygen atom or benzene ring, respectively) adding another affinity, which seems to further enhance carbamate stability compared to piperidine compound 19p.

In vivo studies

Compounds 7, 19d, 19j and 13 were tested for their neuroprotectant and pro-cog nitive properties in an in vivo model of AD in mice as described before in Dolles, D. et al. , J. Med. Chem. 2018, 61, pages 1646-1663, Maurice, T. et al. , Brain Res. 1996, 706, pages 181-193 and Lahmy, V. et al., Neuropsychopharmacol. 2013,

38, pages 1706-1723.

AD-like cognitive dysfunctions were induced in male Swiss mice, 6 weeks old, weighing 31 -36 g by intracerebroventricular (icv) injection of the oligomerized Ab25- 35 peptide into the mouse brain on day 1 .

Compounds were weighed, dissolved in pure DMSO at a concentration of 1 mg/ml and diluted into final test concentrations with saline. The final percentage of DMSO in saline was 60% for all compounds. Vehicle solution used for control groups was 60% DMSO in saline. Compounds were injected intraperitoneally (ip) from day 1 to 7 and behavioral studies were conducted between day 8 and 10. Spatial working memory was evaluated using a spontaneous alternation test in a Y-maze on day 8. The Y-maze is made of grey polyvinylchloride. Each arm is 40 cm long, 13 cm high, 3 cm wide at the bottom, 10 cm wide at the top, and converging at an equal angle. Each mouse is placed at the end of one arm and allowed to explore the maze freely for 8 min. The sequence of arm entries (including possible returns into the same arm) was checked visually and noted down. If the mouse enters all three arms on consecutive occasion, this is defined as an alternation. Therefore, the to tal number of arm entries minus two is also the maximum number of alternations. The percentage of alternation was calculated as (actual alternations / maximum al ternations) x 100. Parameters for the evaluation of behavior are given as the percentage of alternation (memory index) and a total number of arm entries (ex ploration index). When an extreme behavior (Alternation percentage < 20% or > 90% or a number of arm entries < 10) was observed, animals were excluded from the calculations, which corresponded to a 4.7% attrition in this study.

Non-spatial long-term memory was analyzed using a step-through passive avoid ance test, with training on day 9 and retention on day 10. The setup for the experi ment consists of a two-compartment (15 x 20 x 15 cm high) polyvinylchloride box, whereas one compartment is white and illuminated with a bulb (60 W, 40 cm above the apparatus) and the other black with a cover and grid floor. The two compartments are separated by a guillotine door. On day 9, during the training session, each animal was placed in the white compartment with the door closed. After 5 s, the door was opened and the mouse was allowed to enter the dark com partment. When it had placed all its paws on the grid floor, the door was closed and a foot shock was delivered (0.3 mA) for 3 s using a scrambled shock genera tor (Lafayette Instruments, Lafayette, USA). The time spent to enter the dark com partment (step-through latency), and the level of sensitivity to the shock was eval uated (no sign=0, flincing reactions=1 , vocalizations=2). None of the treatment af fected the step-through latency or shock sensitivity in the present study (data not shown). The retention test was carried out on day 10. 5 s after the mouse was placed in the white compartment, the door was opened and the mouse was al lowed to explore the box. The time spent to enter the dark compartment (step- through latency) was measured up to 300 s. Animals showing latencies during the training and retention session lower than 10 s and shock sensitivity = 0 were con sidered as failing to learn the task and discarded from calculations. In this study, it corresponded to 0.6% attrition.

A protocol of the experimental treatment and testing of mice is illustrated in Fig. 6. The meaning of the abbreviations used in Fig. 6 is as follows: icv, intracerebroven- tricular injection; YMT, spontaneous alternation test in the Y-maze; ST-PA, step- through passive avoidance test. Statistical analyses

All values, except passive avoidance latencies, are expressed as mean ± S.E.M. Statistical analyses were performed on the different conditions using one-way ANOVA (F value), followed by the Dunnett’s post-hoc multiple comparison test. Passive avoidance latencies do not follow a Gaussian distribution since upper cut off times (300 s) are defined. Therefore, they were analyzed using a Kruskal-Wal- lis non-parametric ANOVA ( H value), followed by a Dunn's multiple comparison test p < 0.05 was considered as statistically significant.

Results

Repeated treatments with the compounds

None of the treatments (ip for the compounds partially solubilized in DMSO and icv for the peptide) affected significantly the mouse body weight gain during the week of treatment showing a good tolerability. Animals gained 2.2-4.5 g during the 7 days treatment period and particularly recovered quickly from the icv injection stress after day 1 .

Spatial working memory measure using the spontaneous alternation performance in the Y-maze

Figs. 7a-h show the effect of the compounds on Ap25-3s-induced spontaneous al ternation deficits in mice. Mice received Ab25-35 (9 nmol icv) or vehicle (V) solution (3 pi icv) on day 1 and compounds 7 (Figs. 7a, b), 19d (Figs. 7c, d), 19j (Figs. 7e, f) and 13 (Figs. 7g, h), in the 0.3-3 mg/kg ip dose range once per day (o.cf.) from day 1 to 7. Figs. 7a, c, e, g: spontaneous alternation performance; Figs. 7b, d, f, h: number of arm entries in the Y-maze test performed at day 8. Data show mean ± SEM. ANOVA: F _(4,79) = 2.61 , p < 0.05, n = 13-18 in (a); F _{( ,79)} = 1 .32, p > 0.05 in (b); F(4,61) = 4.06, p < 0.01 , n = 1 1 -13 in (c); F(4,61) = 1 .00, p > 0.05 in (d); F( 4,63) = 4.30, p < 0.01 , n = 1 1 -15 in (e); F(4,63) = 4.50, p < 0.01 in (f); F(4,63) = 3.07, p < 0.05, n = 1 1 -15 in (g); F _{(4, 63)} = 2.59, p < 0.05 in (h). ^* p < 0.05, ^** p < 0.01 vs. (V+V)- treated group; # p < 0.05, ## p < 0.01 , ### p < 0.001 vs. (Ap25-35+V)-treated group; Dunnett’s test.

Compound 7 attenuated the Ap25-3s-induced spontaneous alternation deficits at 1 and 3 mg/kg ip doses (Fig. 7a). However, the increases in alternation percentage remained non-significantly different from the Ap25-3s-treated group. Compounds 19d and 19j appeared very active since they were active at lower doses: 0.3 mg/kg ip for 19d and 1 mg/kg ip for 19j (Figs. 7c, e). Interestingly, the phenolic compound 13 lacking BChE inhibition was not active in this response (Fig. 7g). It must be noted that the Ab25-35 treatment failed to affect exploratory response, measured in terms of the number of arms entered during the session (Figs. 7b, d, f, h). Noteworthily, all compounds tended to slightly increase locomotor activity (Figs. 7b, d, f, h) and in a significant manner for compound 19j (Fig. 7f).

Long-term memory assessment using the passive avoidance test

Figs. 8a-d show the effect of the compounds on Ap25-3s-induced passive avoid ance impairments in mice. Mice received Ab25-35 (9 nmol icv) or vehicle solution (3 pi icv) on day 1 and compounds 7 (Fig. 8a), 19d (Fig. 8b), 19j (Fig. 8c) and 13 (Fig. 8d), in the 0.3-3 mg/kg ip dose range o.d. from day 1 to 7. Animals were trained in the passive avoidance test on day 9, and retention (step-through la tency) was analyzed on day 10. Data show median and interquartile range. Krus- kal-Wallis ANOVA: H = 12.3, p < 0.05, n = 16-18 in (a); H = 21 .1 , p < 0.05, n = 1 1 - 15 in (b); H = 13.5, p < 0.01 , n = 1 1 -14 in (c); H = 1 1 .8, p < 0.01 , n = 1 1 -12 in (d). * p < 0.05, ^** p < 0.01 , ^*** p < 0.001 vs. (V+V)-treated group; # p < 0.05, ## p < 0.01 vs. ^25-35+V)-treated group; Dunn’s test.

Compound 7 significantly prevented the Ab25-35-induced passive avoidance deficits at all dose tested (Fig. 8a). Both compounds 19d and 19j were effective at the 3 doses tested and particularly at the lowest 0.3 mg/kg ip dose (Figs. 8b, c). Compound 13 was not active, with only a few animals showed improved latency at the highest dose (Fig. 8d).

In summary, as compared to the parent compound 7, compounds 19d and 19j showed very promising effects on both investigated behavioral parameters. They particularly fully prevented the effects of Ab injection at a dose of 1 or 3 mg/kg ip. the in vivo results prove the penetration of the blood brain barrier, despite the high lipophilicity and molecular weight of compounds.

Abbreviations:

ACh, Acetylcholine; AChE, Acetylcholinesterase; Ab, amyloid b; AD, Alzheimer’s disease; ATC, acetyl thiocholine iodide; BChE, Butyrylcholinesterase; BTC, butyryl thiocholine iodide; eq, equation; DTNB, 5,5 ^' -dithiobis-(2-nitrobenzoic acid); icv, in- tracerebroventricular; ip intraperitoneal; o.d., once per day; V, vehicle, ST-PA, step-through passive avoidance test; YMT, spontaneous alternation test in the Y- maze;