Anle138b: a novel oligomer modulator for disease-modifying therapy of neurodegenerative diseases such as prion and Parkinson’s disease

In neurodegenerative diseases such as Alzheimer’s disease (AD), Parkinson’s disease (PD) and prion diseases, deposits of aggregated disease-specific proteins are found. Oligomeric aggregates are presumed to be the key neurotoxic agent. Here we describe the novel oligomer modulator anle138b [3-(1,3-benzodioxol-5-yl)-5-(3-bromophenyl)-1H-pyrazole], an aggregation inhibitor we developed based on a systematic high-throughput screening campaign combined with medicinal chemistry optimization. In vitro, anle138b blocked the formation of pathological aggregates of prion protein (PrPSc) and of α-synuclein (α-syn), which is deposited in PD and other synucleinopathies such as dementia with Lewy bodies (DLB) and multiple system atrophy (MSA). Notably, anle138b strongly inhibited all prion strains tested including BSE-derived and human prions. Anle138b showed structure-dependent binding to pathological aggregates and strongly inhibited formation of pathological oligomers in vitro and in vivo both for prion protein and α-synuclein. Both in mouse models of prion disease and in three different PD mouse models, anle138b strongly inhibited oligomer accumulation, neuronal degeneration, and disease progression in vivo. Anle138b had no detectable toxicity at therapeutic doses and an excellent oral bioavailability and blood–brain-barrier penetration. Our findings indicate that oligomer modulators provide a new approach for disease-modifying therapy in these diseases, for which only symptomatic treatment is available so far. Moreover, our findings suggest that pathological oligomers in neurodegenerative diseases share structural features, although the main protein component is disease-specific, indicating that compounds such as anle138b that modulate oligomer formation by targeting structure-dependent epitopes can have a broad spectrum of activity in the treatment of different protein aggregation diseases. Electronic supplementary material The online version of this article (doi:10.1007/s00401-013-1114-9) contains supplementary material, which is available to authorized users.

For the generation of the SAR-map substances from the DIVERSet libraries were subjected to a cluster analysis using the software package Benchware HTS DataMiner (DM; Tripos Inc., St. Louis, MO, USA). Since the set of all (20,000) compounds from the libraries would have been too large as a starting set for cluster formation using DM, the initial set of compounds was restricted to the set of primary hits (837 compounds) from the cell culture screening. Thus, clusters were built based on the active compounds only. Here, the DM program grouped structurally similar compounds into clusters thereby enabling the identification of potentially relevant new lead structures. In a second step, the thus established classification was applied to the rest of the library covering the compounds inactive in cell culture. Here, the DM program added the remaining (inactive) compounds to the generated clusters, if the employed measure indicated a high structural similarity. The result of the cluster analysis is displayed by DataMiner as a SAR-map, in which the substance clusters S are represented by symbols arranged in proximity according to structural similarity. The sizes, forms and colours of the symbols were allocated based on cluster-specific properties. The sizes of the symbols were selected proportional to the sizes |S| of the clusters, i.e. to the number of compounds C contained. The forms of the symbols were determined based on the cluster-local proportions P SIFT (S) = |{C S | a(C) ≥ a min }| / |S| of those compounds C in the respective clusters S, whose primary activity a(C) determined in the SIFT screening was above a selected threshold of a min = 0.25. Based on this, clusters whose proportion P SIFT (S) is above 50% are shown as stars, whereas the remaining clusters are shown as boxes. Analogously, the colours of the symbols encode the cluster-local proportions P CC (S) = |{C S | is primary hit}| / |S| of primary hits from the cell culture (CC) screening, whereupon the symbols of clusters with more than 50% active substances are coloured red and those of the remaining clusters are coloured grey. Hence, as pointed out, in the resulting SAR-map, large red stars symbolise clusters with a high proportions of SIFT and cell culture positive substances. Such clusters represent potential lead structures. Using DataMiner, clusters of interest were further analysed and a group of five neighbouring clusters was identified and termed DPP_1 through DPP_5 (shown in bold). The fact that these clusters are located close to each other indicates that they contain structurally similar compounds. In fact, all of them belong to the chemical compound class of 3,5-Di-Phenyl-Pyrazole (DPP) derivatives.

Supplement Figure 2:
Effects of treatment with DPP compound 10353_F11 on survival time of mice after intracerebral infection with RML scrapie. Treatment with compound 10353_F11 (chemical structure: see Suppl.- Fig.1) prolonged mean survival of mice infected intracerebrally with prion strain RML by eleven days (n=8, p < 0.05). The compound was administrated daily for 14 days from day 80 post infection (50 µl i.p., 10 mM compound in DMSO). Mean survival times are expressed in days + standard deviation.

DPP-derivatives tested for anti-prion activity in vivo
The table summarizes the effect of various compounds in regard to inhibition of prion propagation and prolongation of survival time in vivo. # relative inhibition of PrP Sc accumulation normalized to DMSO-treated group (0% inhibition) and PrP Sc level at start of treatment (100% inhibition). +  survival (prolongation of survival in days/treatment days up to mean survival of controls) * mice refused to eat peanut butter pellets a PrP Sc level in brain 120 days after i.c. infection and treatment for 40 days with 1 mg compound (oral, in peanut butter Models of likely 3-dimensional structures of compounds anle234b (a) and anle138b (b). Substitution of the bromine in the ortho-position (anle234b) abolished the inhibitory activity of anle138b which is substituted in the meta-position. The substitution tilts the phenyl ring so that the molecule is no longer planar indicating that a planar conformation is necessary for activity of inhibitory compounds. The 3-D structures were generated with the Molinspiration Galaxy 3D Structure Generator (www.molinspiration.com).

Supplement Figure 5:
Motor performance in prion infected mice at 160 d.p.i. is unimpaired in mice treated with anle138b. Mice were placed in an arena (265 x 150 x 420 mm). Movements were recorded with a monitoring system. Movements of the animals were tracked and analyzed with ImageJ 1.39. Representative paths for a DMSO-treated (left) or anle138b-treated (right) mouse at 160 days after intracerebral infection are shown.

Supplement Figure 6:
Statistical evaluation of survival data shown in Fig. 3a Comparison of Survival Curves

Supplement Figure 7:
Quantification of PrP C by immunoblotting of brain tissue from non-infected mice treated with anle138b (1 mg per day in DMSO/peanut butter) for 1 week. No reduction in PrP C level was observed in mice treated with anle138b when compared to control mice that received DMSO/peanut butter without anle138b. Error bars indicate standard error (n = 4 mice).

Inhibition of in vitro propagation of different prion strains by anle138b.
(A) Normal brain homogenates of C57BL/6 mouse and human seeded with a 100-fold dilution of infected brain homogenates respectively were mixed with different concentration of anle138b (0, 1, 3, 10, 30, 100 and 300 μM, final concentration). PMCA reactions were conducted 18 cycles for mouse substrate and 40 cycles for human substrate. The effect of anle138b was dose-dependent. Molecular weight markers are indicated on the right in kD. (B) The amount of PrP Sc was quantified densitometrically and normalized to the control reaction without compound. Similar dose-response curves were obtained for human prions (vCJD) and for the murine prion strain RML that was used in animal experiments. The EC 50 values for anle138b in the PMCA assay are 7.3 µM for RML prions and 7.1 µM for vCJD prions, respectively. Three independent experiments were performed. Results are presented as mean ± standard error.

Supplement Figure 9:
Dose-dependent effect of anle138b administration on PrP Sc levels in brain.
(A) C57/BL6 mice were inoculated intracerebrally with 30 µl of 1% brain homogenate (RML scrapie). Treatment was started at 80 days post infection with different amounts of anle138b applied orally mixed with DMSO/peanut butter. At 120 days post infection, animals were sacrificed and the amount of PrP Sc in the brain was quantified in comparison to animals sacrificed at day 80 post infection. Treatment with anle138b reduced PrP Sc accumulation in brain in a dose-dependent manner with an EC 50 of 0.21 mg/day. (B) In an independent experiment, different treatment schedules were compared using the same experimental approach as described in (A). Application of anle138b twice daily appears to be more efficient than one single dose/day. Error bars indicate standard error (n = 4 mice).

Supplement Figure 10:
Mixed application scheme resulting in a decrease of PrP Sc levels in anle138b-treated mice.
(A) Experimental protocol: Seven-week-old female C57/BL6 mice were inoculated intracerebrally with 30 µl of 1% brain homogenate (RML scrapie). Treatment was started at 80 days post infection with 0.84 mg compound (in 25 µl DMSO) per day applied by intraperitoneal injection for 14 days followed by 2 x 5 days of 1 mg compound (in 10 µl DMSO + 40 µl vegetable oil) applied orally by gavage. PrP Sc level in brain was measured before treatment and at 95 and 106 days post infection. (B) Change of PrP Sc levels after treatment with compounds anle138b and anle186b in comparison to controls. Treatment with anle186b leads to a slight reduction of PrP Sc accumulation. In the anle138btreated group, a decrease in PrP Sc levels can be observed indicating a virtually complete block of prion amplification. Error bars indicate standard error (n = 4; ** = p < 0.01).

Supplement Figure 11:
HPLC chromatogram of mouse brain homogenate for quantification of anle138b. The high resolution mass spectrum (ESI + mode) identifies the present compound clearly as anle138b from the mass and the two equally populated isotopes 79 Br and 81 Br and the two additional small peaks shifted by one mass unit originating from the 13 C isotope for 16 carbons in the molecule. The tissues were defreezed at 4C prior to use. It was homogenized twice in 5 ml of acetonitrile at maximum speed for 3 minutes using a homogenizer (IKA ULTRA-TURRAX Tube drive workstation, Germany). The homogenate was ultrasonicated at 30C for 5 minutes and centrifuged at 5000g for 10 minutes. An aliquot (100 ul) of supernatant was injected into HPLC system. Briefly, analytical high performance liquid chromatography (HPLC) was performed using a Waters HPLC system with a Waters 996 Photodiode Array Detector. All separations involved a mobile phase of 0.1% trifluoroacetic acid (TFA) (v/v) in water (solvent A) and 0.1% TFA in acetonitrile (solvent B). HPLC was performed using reversed-phase (RP) column Eurospher RP 18, 100 Å, 5µm, 250  4.6 mm at flow rates of 1 mL/min with a gradient of solvent B from 0% to 100% in 50 minutes. The effluent was monitored for UV absorption at 260 nm. Samples were quantified using peak area ratio of compounds to external standard." The reason for the background free detection of anle138b in the HPLC chromatogram is the following: 1) A level of anle138b in the brain is high and anle138b has a high solubility in acetonitrile (solvent that we use for extraction).
2) Many other small organic molecules (e.g. amino acids, lipids, dopamine) that are present in the brain have a lower concentration and/or a lower solubility in acetonitrile and/or a low extinction coefficient at 260 nm. Therefore several small peaks observed in the chromatogram could reflect the presence of other UV-active endogenous compounds from the brain.
3) Proteins and other macromolecules are insoluble in acetonitrile.

Supplement Figure 12:
Results for compounds tested in regard to inhibition of formation of Fe 3+ /DMSO-induced α-syn oligomers by SIFT assay (see Materials and Methods). Shown are the results for all newly synthesized compounds that are related to the DPP lead structure as they contain two phenyl rings linked by a central five-membered ring containing nitrogen. In addition, the structures of the two control compounds 293G02 and baicalein are shown.

Effect of compounds in an MPTP in vivo mouse model of Parkinson´s disease
Mice were treated with MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine, 30 mg/kg bodyweight daily) by i.p. injection on days 1-5 to induce degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNpc). Animals (8-15 per experimental group) were treated with anle138b (5 mg daily) or vehicle by oral application (gavage, compound in 12.5 µl DMSO mixed with 487.5 µl olive oil, 6 h before the MPTP injections) on days 0-12. Loss of neurons compared to control mice and MPTP-treated mice that were treated with vehicle only (DMSO/olive oil) was quantified on day 12 (i.e. seven days after last MPTP application). For quantification of tyrosine hydroxylase (TH)-positive neurons in the SNpc, 50 µm sections were immunostained with an anti-TH-antibody (1:1000; Zymed). Every second section through the SNpc was analyzed using Stereo investigator software (MicroBrightfield, Colchester, VT, USA). Immunostained cells were counted by the optical fractionator method using a 20x objective. Stereological counts were performed blindly by two independent investigators.

A B
A) A significant loss of dopaminergic neurons in the substantia nigra can be observed in a sub-acute MPTP mouse Parkinson model compared to non-MPTP treated control animals. This effect is significantly ameliorated by treatment with anle138b. Shown is the mean and SEM. B) MPP+ levels were measured in striatal brain lysates by HPLC. Anle138b does not affect the level of MPP+ found in the brain (p=0.58). Shown is the mean and SD for n=3-4 animals.
Representative examples of immunohistochemistry for TH-positive cells in substantia nigra.

Supplement Figure 15:
Quantitative analysis of α-synuclein deposition in 69-week old transgenic mice.
As described in the Materials and Methods section, anle138b treatment was tested against placebo treatment with the vehicle (DMSO/peanut butter) in {(Thy1)-h[A30P]α-syn} mice on a genetic background of C57/Bl6. Treatment with anle138b and placebo, respectively, was initiated at the age of eight weeks. During the first two weeks of treatment, 2 mg of anle138b dissolved in 10 µl DMSO mixed with 200 µl peanut butter were given. After two weeks of treatment, the dose was increased to 5 mg in 10 µl DMSO/200 µl peanut butter. At the age of 33 weeks, the dose was increased to 2x5 mg per day. Four animals matched in regard to sex and litter were sacrificed per experimental group at the age of 69 weeks for analysis of pre-terminal histopathological changes. For histopathological and immunohistochemical investigation, formalin-fixed brain tissue was used. Pathological deposits of human α-syn were detected by the anti-human-α-syn antibody 15G7 (see also Fig. 8d). For quantitative analysis, the stained area was quantified in blinded sections at the level of the brainstem by image analysis software cell D 2.5 (Olympus/Soft Imaging Systems GmbH, Münster, Germany). The observed difference was statistically significant (t-test, p< 0.05).

Supplement Figure 16:
"Raw data" for figure 4b (all blots are from different mice): "Raw data" for figure 8f (all blots are from different mice):

Supplement Figure 17:
The amount of total alpha-synuclein was quantified by western blot analysis of brain homogenates from 69-week old mice treated with anle138b and DMSO-treated control mice. For every mouse, 10 µl of 10% brain homogenates (1 mg of brain) per lane were prepared for western blotting. Each group contained three mice. Samples were boiled with 2 x loading buffer at 100 o C for 10 min followed by separating on 15% SDS-PAGE gel. Proteins were transferred to PVDF membrane and detected by monoclonal antibody 15G7 (1:1000). No difference in level of total synuclein was found.

Supplement Figure 18:
Preclinical studies regarding acute toxicity and mutagenicity of anle138b. These experiments were outsourced to LPT Laboratory of Pharmacology and Toxicology, Hamburg, Germany. Studies were carried out according to the 'Good Laboratory Practice' Regulations.

a) acute toxicity study in mice (NMRI, Charles River)
2000 mg anle138b were mixed with 2 mL DMSO and allowed to stand 5 minutes in a 40°C water bath, followed by addition of olive oil to a total volume of 40 mL, followed by 5 minutes in the water bath and vortexing for 20 seconds. The administration volume was 40 mL/kg b.w., per oral. Observations were performed before and immediately, 5, 15, 30 and 60 min, as well as 3, 6 and 24 hours after the administration. All animals were observed. During the 14-day follow-up period changes of skin and fur, eyes and mucous membranes, respiratory and the circulatory functions, autonomic and central nervous system and somatomotor activity, as well as behavioural pattern were observed at least once a day until all symptoms had subsided, thereafter each working day. Attention was also paid to possible tremors, convulsions, salivation, diarrhoea, lethargy, sleep and coma.
Observations on mortality were made at least once daily to minimize loss of animals during the study. Individual body weights were recorded before administration of the test item and thereafter in weekly intervals up to the end of the study. Changes in weight were calculated when survival exceeded one day.
At the end of the experiment all animals were sacrificed, dissected and inspected macroscopically. All gross pathological changes were recorded. No microscopic examination was performed as no pathological findings were noted at necropsy of the animals at the end of the 14-day observation period.
Under the present test conditions, a single oral administration of 2000 mg anle138b/kg b.w. to mice did not reveal any signs of toxicity and no mortality. All animals gained the expected body weight throughout the whole study period. No macroscopical changes were noted at necropsy.

b) acute toxicity study in rats (CD, Charles River)
2000 mg anle138b were mixed with 2 mL DMSO and allowed to stand 5 minutes in a 40°C water bath, followed by addition of olive oil to a total volume of 40 mL, followed by 5 minutes in the water bath and vortexing for 20 seconds. The administration volume was 40 mL/kg b.w., per oral. Observations were performed before and immediately, 5, 15, 30 and 60 min, as well as 3, 6 and 24 hours after the administration. All animals were observed. During the 14-day follow-up period changes of skin and fur, eyes and mucous membranes, respiratory and the circulatory functions, autonomic and central nervous system and somatomotor activity, as well as behavioural pattern were observed at least once a day until all symptoms had subsided, thereafter each working day. Attention was also paid to possible tremors, convulsions, salivation, diarrhoea, lethargy, sleep and coma.
Observations on mortality were made at least once daily to minimize loss of animals during the study. Individual body weights were recorded before administration of the test item and thereafter in weekly intervals up to the end of the study. Changes in weight were calculated when survival exceeded one day.
At the end of the experiment all animals were sacrificed, dissected and inspected macroscopically. All gross pathological changes were recorded. No microscopic examination was performed as no pathological findings were noted at necropsy of the animals at the end of the 14-day observation period.
Under the present test conditions, a single oral administration of 2000 mg anle138b/kg b.w. to rats did not reveal any signs of toxicity and no mortality. All animals gained the expected body weight throughout the whole study period. No macroscopical changes were noted at necropsy.

c) Mutagenicity Study in the Salmonella typhimurium Reverse Mutation Assay (AMES test)
Anle138b was examined in the 5 Salmonella typhimurium strains TA 98, TA 100, TA 102, TA 1535 and TA 1537 in two independent experiments, each carried out without and with metabolic activation (a microsomal preparation derived from Aroclor 1254-induced rat liver).

d) in vitro assessment of the clastogenic activity in cultured human peripheral lymphocytes
Test samples of anle138b were assayed in an in vitro cytogenetic study using human lymphocyte cultures both in the presence and absence of metabolic activation by a rat liver post-mitochondrial fraction (S9 mix) from Aroclor 1254 induced animals. Anle138b was dissolved in dimethyl sulfoxide (DMSO). The test was carried out employing 2 exposure times without S9 mix: 4 and 24 hours, and 1 exposure time with S9 mix: 4 hours. The experiment with S9 mix was carried out twice. The harvesting time was 24 hours after starting of exposure. The incubation procedure took place in the dark. The study was conducted in duplicate. Mitomycin C and cyclophosphamide were employed as positive controls in the absence and presence of metabolic activation, respectively.

Tests without metabolic activation (4-and 24-hour exposure)
The mean incidence of chromosomal aberrations (excluding gaps) of the cells treated with anle138b at concentrations from 312.5 to 2500 or 78.13 to 312.5 µg/mL medium (4-h or 24-h exposure) in the absence of metabolic activation ranged from 1.5% to 4.0%. The results obtained are considered to be within the normal range of the solvent control where a mean incidence of chromosomal aberrations (excluding gaps) of 1.5% or 1.0% was observed after a 4-hour and 24-hour exposure, respectively. Only at the pronounced cytotoxic concentration of 625 µg/mL medium (24-hour exposure, only 92 of 200 metaphases could be evaluated) a marginal, though not significant increase to 4.3% was noted in the number of aberrations. It is known that high cytotoxicity causes artefacts in the form of aberrations in in vitro chromosomal tests. Hence, the increase at the concentration of 625 µg anle138b/mL medium (24-hour exposure) is considered as artefact and not test item-related.

Test with metabolic activation (4-hour exposure)
The mean incidence of chromosomal aberrations (excluding gaps) of the cells treated with ANLE138B at concentrations from 312.5 to 2500 or 312.5 to 1250 µg/mL medium in the presence of metabolic activation in the first and second experiment, respectively, ranged from 1.5% to 3.5%. The results obtained are considered to be within the normal range of the solvent control where a mean incidence of chromosomal aberrations (excluding gaps) of 1.5% was observed after a 4-hour exposure. Only at the pronounced cytotoxic concentration of 2500 µg/mL medium (the second experiment, only 130 of 200 metaphases could be evaluated) an increase to 4.6% (significant at p  0.05) was noted in the number of aberrations. It is known that high cytotoxicity causes artefacts in the form of aberrations in in vitro chromosomal tests. Hence, the increase at the concentration of 2500 µg anle138b/mL medium in the second experiment is considered as artefact and not test item-related. No item-related polyploidy or endoreduplication was noted in the experiments with or without metabolic activation. Thus, under the present test conditions, anle138b tested up to cytotoxic concentrations in the absence and in the presence of metabolic activation employing two exposure times (without S9) and one exposure time (with S9) revealed no indications of mutagenic properties with respect to chromosomal or chromatid damage. In the same test, Mitomycin C and cyclophosphamide induced significant damage.