HU231449B1 - Method for identifying drug candidate compounds - Google Patents

Description

PROCEDURE FOR THE IDENTIFICATION OF MEDICINAL COMPOUNDS

The subject of the invention is a new preclinical method that can be used to test the efficacy of candidate drug molecules, with which endpoints relevant to autism spectrum disorders can be measured in an automated living cage.

The invention makes it possible to behaviorally separate autistically behaving rodents from controls faster than conventional methods. The invention also makes it possible to perform efficacy testing of drug candidate compounds with therapeutic potential in the treatment of the central symptoms of autism on animals characterized in this way.

BACKGROUND OF THE INVENTION

Autism spectrum disorder (ASD) is a complex, highly challenging, prevalent neurodevelopmental condition, the frequency of which reaches 1% in both children and adults {Brugha et al., Arch. Gen. Psychiatry. 2011, 68:459-465: Murphy et al., Neuropsychiatr. Dis. Treat. 2016, 12:1669-1686). The disease can be characterized by two central groups of symptoms, these are socio-communication defects and repetitive, stereotypical behaviors and thinking {Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, pp. 50-59). The unsatisfied medical needs of ÁSD are huge, as there is currently no drug therapy option approved by the authorities to treat the central symptoms.

Several methods are available for measuring autistic-like behavior in experimental animals, such as three-chamber social interaction, washing, ultrasonic vocalization, dominance or social play tests {Silverman et al., Nat. Port. Neurosci. 2010, 490-502). However, a major disadvantage of these methods is that in order to perform them, the experimental animal must be removed from its home cage and adapted to a new environment during the observation period. During this, the animals are exposed to an excessive amount of human influence, which causes them unnecessary stress and this can have serious consequences for the animals' natural behavior. Stress can cause increased anxiety, which can lead to a significant distortion of the results.

ffl It can cause further problems if the measurements are carried out on autistic animals, because such '(0 Ή ÍN ÍN ÍN £N

In SZTNH-100355415 cases, autistic animals may react to the experimental conditions with increased sensitivity and irregular behavior, which behavior may even endanger the evaluation of the entire test.

With the help of automated living cages, the experimental individual can reduce the stress resulting from human proximity to a minimum. When such a living cage is used, the animals stay in their usual environment during the measurements, so that the social structure between them remains intact throughout. In this essentially undisturbed environment, it is possible to manifest more natural behaviors, which creates much more sensitive measurement conditions than traditional experimental methods.

There is a significant difference between traditional behavioral methods and automated home cages in terms of the length of behavioral observation and the amount of behavioral data obtained. While traditional behavioral observations are typically short-term (5 minutes to half an hour), longer, typically several days or even several weeks, 24 hours a day observation is also possible in the automated living cage. While traditional methods can typically only obtain 10-100 data points per observation, an automated cage observation can provide 10,000 to 100,000 data sets, which clearly enables a deeper behavioral analysis.

It is a known fact that some chemical substances, such as prenatal exposure to ethyl alcohol or known drugs (e.g. thalidomide) is a significant non-genetic risk factor for ASD (Edward et al., Neuropsych. Rév. 2011 21:73-80; Hallene et al., Neuroscience 2006, 142:267- 283). Treatment with such substances during pregnancy in rodents also leads to autistic symptoms in the offspring. Autistic behaviors include defects in social interaction, excessive repetitive behavioral elements, reduced vocal communication, and increased anxiety. These models have a well-established etiopathology and show great similarities with human ASD, which is why these disease models have a high translational value.

The validation of the disease models of ÁSD faces difficulties, since there is no medicine approved by the authorities for the central symptoms, which is called could be used as a gold standard. Validation in such a situation is only possible if we use compounds as reference substances that have produced at least signs of efficacy p _{and s} in human clinical trials. Such a substance can be, for example, the hormone oxytocin, which has been shown in several human studies to improve certain symptoms of ÁSD (Andari et al., Proc. Natl. Acad. Sci. USA. 2010, 107:4389-4394 ; Domes et al., Biol. Psychiatry 2013, 74:164-171; Gordon et al., Proc. Natl. Acad. Sci. USA. 2013, 110:20953-20958; Guastella et al., Biol. Psychiatry 2010, 67:692-694). Similarly, it has been proven in animal experiments that oxytocin can improve the autistic-like features of ÁSD model animals when given to rodents (e.g. Teng et al., Neuropharmacology 2013, 72:187-96). Based on all of this, oxytocin can be used to validate a specific disease model, to verify that clinical effectiveness can be expected from other medicinal products tested in the given model.

Due to the significant unmet medical need mentioned above, there is a need to develop preclinical methods that can be used to measure autism-relevant endpoints and which can ultimately be used to test the efficacy of new drug candidates.

SUMMARY OF THE INVENTION

In the course of our investigations, we came to the surprising result that the group behavior of rodents can be determined more advantageously compared to traditional behavioral pharmacology methods based on a behavioral method that can be measured in an automated living cage. We were able to influence the group behavioral deviation determined in autistic animals with oxytocin, an agent used in ASD and under clinical development in humans. Therefore, the difference in group dynamics that we detected can be used industrially to test the effectiveness of drug candidates intended for the treatment of autism.

DETAILED DESCRIPTION OF THE INVENTION

During our studies, we observed and recorded the dynamics of the group behavior of the experimental rats. During our experiments, we observed that the pattern of group behavior of the animals is clearly different in the case of the normal control and their autistic peers. The group dynamic network was created with FlowR and iGraph software. During the calculation, we set up a connection network between the individuals of the group. This relationship is based on the order of the animals' visits to the drinking bottle, _taking into account the following order. Each animal can be characterized by a number, which determines the 4© ft ft ft ft relationship strength with other individuals. Based on the values of the individuals, the so-called we calculate a global reaching centrality index (global reaching centrality, GRC), which characterizes the given group of animals well and makes it easy to compare different groups of animals. By observing the behavior pattern and with the help of the group dynamic network created from the observation results, the group of animals behaving autistically can be easily recognized and defined.

According to a certain embodiment of the invention, this recognition enables the autistic behavior of group-housed rats to be recognized quickly and reliably with the help of one or two days of observation in an automated living cage, which would take several weeks to determine using traditional behavioral pharmacological methods. According to a certain embodiment of the invention, it is also possible that the natural behavior of animals that already behave autistically cannot be affected by any additional environmental factors or human intervention, so that the behavior of these animals can be well monitored, the behavior shown can be evaluated reliably, and the otherwise disturbing environmental or errors resulting from human intervention can be completely excluded.

According to a certain embodiment of the invention, the above analysis method also provides the opportunity to characterize the effect of pharmacological interventions in the animal model showing autistic symptoms with the help of the global reach centrality index.

DEFINITIONS

"Rodent" refers to mouse or rat species.

“Core symptoms of autism” include sociocommunicative symptoms and excessively repetitive, restricted behaviors/thinking as defined in the DSM-5 Diagnostic and Statistical Manual 5th Edition (2013).

"Autistic behavior" includes social, communicative, and repetitive behavioral defects in rodents.

An "automated living cage" is a living cage that can be controlled by remote control and the rodents living in the community can be subjected to stress-free experimental observation.

2222163

The "group dynamic network" is the connection network between the animals participating in the experiment, which is determined based on the order of follow-up of the members of the group.

The "global access centrality index" is a characteristic of the network formed by the animals in the cage, so it is a stamp at the group level. Each individual is connected to the others if there is strong following or strong avoidance between them (the sign is different), and the index is a measure of this network as a directed graph.

The prenatal exposure ÁSD model

Total exposure to certain substances is known to increase the risk of ÁSD in humans (Christensen et al., JAMA 2013, 309:1696-1703). Similarly, exposure during intrauterine life also causes an autistic-like phenotype in rodents (Kim et al., Tox. Lett. 2011, 201:137-142). The autistic phenotype includes the following behavioral elements: ultrasonic communication defect in juvenile rodents, play behavior disorder in adolescent rodents, hyperactivity, excessive repetitive behaviors, and social interaction defects in adult animals. Since the pathophysiological origin and symptomatology of prenatal exposure models are very similar to the human condition, these models are given a particularly high translational value.

Time-mated female Wistar rats (Janvier, France) were kept on a soy-free diet and treated intraperitoneally with 300 mg/kg 2-propylpentanoic acid sodium salt in a treatment volume of 2.5 ml/kg, in physiological saline, on day 12 of their pregnancy. At the same time, animals used as controls received a volume of physiological saline solution calculated based on their body weight. At the age of 11 weeks, ten individuals were selected from the offspring born (n=10 control, n=10 treated) and placed in two automated living cages. Under isoflurane anesthesia, a microchip was implanted in the animals one week before the beginning of the measurements (UNO PICO-ID ISO transponder, UNO BV, The Netherlands). The rats were kept on a 12/12-hour light/dark cycle (dark starts at 4:00 p.m.), at a temperature of 22 ± 2°C and a humidity of 40-50%.

Automated living cages

The system makes it possible to examine the spontaneous behavior of group-housed rodents and their behavior in different experimental situations (Gerlai, Trends Neurosci. 2002, 25:506-509). The system has four corners for registration. Drinking water only in these IP#

Φ is accessible in corners, behind doors that can be operated by remote control. When a rodent enters

PM

PM W PM into the corner, an antenna detects the signal of the transponder implanted in the animal and records the fact of the visit. The design of the corners allows only one animal to visit at a time. Each corner has two drinking bottles with a left and right arrangement. Based on the registration of the visits to the corners, the activities of the animals can be monitored using a software. The most important parameters measured are the number of visits to corners, the number of nose pokes in corners, the number of licks while drinking, and the duration of all these parameters.

Competition test in the automated living cage

Animals in automatic cages were allowed to drink for two hours after 22 hours without access to water to be sufficiently motivated. During the 2-hour access period, you could only drink from one bottle, so the competition was strong. The competition was repeated a total of four times without giving materials, which established the reference between the autism model group and the control. The differences in the behavior of the two groups were also checked using traditional methods. Subsequently, 30 minutes before the competition, the autism model animals were treated with a reference substance (oxytocin; Richter Gedeon Nyrt.) three times, while the control animals were treated with a vehicle (physiological saline solution) 30 minutes before the start of the observations. We tested the effect of oxytocin on the differences between the two groups.

Background of network analysis

The big advantage of the automatic cage is that we can analyze not only the behavior of individual animals, but the characteristics of the group as a whole, since the animals live relatively undisturbed in the cage for a long time. When characterizing the community, we started from the group structure. The social structure can be described by the extent to which some animals follow each other in behavior compared to chance. In this case, the behavior means access to the drinking area. We calculate an adjacency matrix based on the animals in the cage, which is directed and expresses how well the animals in the rows exhibit following behavior in the direction of the animals in the columns. To do this, for each animal's visit to a corner, we calculate the number of visits of all other animals to that corner. The number of follow-ups is compared with the ratio of visits to the corners, and the resulting chi-square values represent the social distance between two animals.

ft h

This creates a directed graph whose vertices are the animals in the cage and whose edges point from the following animal to the followed animal. The weight of the edges is given by the chi-square value, which is 0 if the observed tracking does not differ from the expected value, positive if the animals follow each other more often than expected and negative if less.

To estimate the group structure, we took positive and strong ties into account, that is, we filtered the graph for positive weights and kept edges with a weight greater than their average. Mones et al. (PLoS ONE 7.3 2012, e33799) we calculated a global access centrality index for the graphs. The global reach centrality of a weighted directed graph can be calculated based on the local reach centrality values, as the average of the difference between the largest local centrality of the graph and the local centrality of each vertex. The local access centrality ^na Sy expresses how much of the graph can be reached from a vertex through the vertex's neighbors.

Implementation of the network analysis

The calculation took place in R and python, which means that it is fully open source, in a controlled environment. We calculated the adjacency matrix from the series of visits using the R statistical program environment script. By properly transforming the adjacency matrix, we created a pandas data table, and the tuples transformed from it were entered as edges of a materialized directed graph class object. The global access centrality calculation was performed using the networkx library. Finally, the obtained results were plotted in the R statistical program environment using ggplot, and the graphs were displayed using the iGraph package. This index has not yet been used to characterize group dynamics between rats. The results are presented at two levels. On the one hand, at the level of the global reach centrality index, as a measure of the community in the cage, it can be compared between the treatments. On the other hand, we display the networks themselves.

EXPLANATIONS OF FIGURES Fig. l. Global reach centrality indices in experimental groups undergoing different treatments.

.figure. ί*ϊ group structure formed during competition among rats treated prenatally with physiological saline, represented by a graph. The vertices of the graph (circles) in the automated living cage W H ÍN CN ÍN ÍN

2,222,163 animals living together (the individual identifiers of the animals can be seen in the circles). The edges (arrows) point from the following animal to the followed animal. The thickness of the arrow is related to the chi value, which describes how much tracking between any given two animals deviates from what is expected; the thicker, the stronger the tracking between the two animals compared to the expected value.

.figure. Among rats exposed to prenatal 2-propylpentanoic acid sodium salt, group structure formed during competition represented by a graph.

.figure. Prenatal 2-propylpentanoic acid among rats exposed to sodium salt, group structure formed during competition after oxytocin treatment represented by a graph.

The invention is illustrated in more detail with the following examples without limiting the scope of our invention to them.

1. EXECUTION EXAMPLE

As can be seen from Figure 1, there is a significant difference between the group showing autistic traits as a result of prenatal exposure and the control group in the community structural indicator, the global access centrality index. The GRC index can vary between 0 and 1, the higher it is, the more order appears in the group's structure. In the case of the control group, the global access centrality index was 0.75, which is relatively high. This means that the animals follow each other in a relatively specific order when drinking, the ones that are higher in the ranking get to the water sooner in the event of competition, as we would expect. In the case of the autistic model group, the global access centrality index is approx. It was 0.26, so it is much lower compared to the control group. The effect of the reference substance oxytocin can be clearly demonstrated on the order, the global reach centrality index rose to a value of around 0.59 when the members of the autistic model group received treatment half an hour before the competition.

2. EXECUTION EXAMPLE

By comparing Figures 2 and 3, it can be seen that the behavioral patterns of the prenatally vehicle-treated control group and the animals from the prenatal exposure model are radically different. While a complex pattern sequence during competition between control animals

TENDON

N

N is formed, until then the animals serving as ÁSD models can only create a group with a primitive, fragmented dynamic in a similar situation. Figure 4 shows that oxytocin treatment slightly changes this weak connection network. As a result of the oxytocin treatment, a group that behaves together is created, in which there is at least one animal that is followed by several animals, indicating that, although in a simple form, some kind of more complex sequence is formed as a result of the treatment.

Claims (3)

DEMAND POINTS 1. Live aggressively to identify compounds suitable for the treatment of the symptoms of human autism spectrum disorder (ASD) by measuring the autistic-like behavior of rodents, characterized by the fact that during the observation of the animals in an automated living cage, we set up a network of connections between the individuals of the group, which is based on the sequence of the animals' visits to the drinking bottle , then we perform network analysis in such a way that a) in the case of each animal, for each visit to a corner, we calculate the number of visits of every other animal to the given corner, then the number of follow-ups is compared with the proportion of visits to the drinking bottles in the corners and each animal is characterized by a serial number, which determines the strength of the relationship with other individuals, b) based on the received serial numbers, we calculate a global reach centrality index with a value between 0-1, which characterizes the order in the group structure, using statistical analysis, c) the control and ÁSD model animal groups are compared on the basis of this global access centrality index, and the group of animals behaving autistically is separated, d) the group of animals behaving autistically is treated with the compound to be tested, e) after the treatment, the global access centrality index value is determined again, then f) those compounds, as a result of which the global reach centrality index in the group of animals behaving autistically increased to a value close to that of the control animal group, are identified as active ingredients suitable for the treatment of human ÁSD symptoms. 2. The method according to claim 1, characterized by the fact that in step ab) an adjacency matrix is calculated from a series of visits with the help of a statistical program environment script. 3. The method according to any one of claims 1-2, characterized by the fact that in step c) the control group shows a global reach centrality value above 0.7 and the autistically behaving animal group below 0.3. wrote SZTNH-100381604