Translational psychiatry is a branch of neurobiology that studies the mechanisms of mental illness and approaches to its treatment using animal models.

• How do you model schizophrenia on rats?
• How do we understand that a mouse’s emotional state reflects the clinical picture of depression?
• How do fish and worms help study addiction and develop new anti-anxiety drugs?
• These and other details of translational research are detailed in the article.

What is a translational study?

Translational research refers to the use of other animal species to study human diseases and disorders, but not all animal experiments are translational research. The other large group, basic research, also uses animals, but studies basic biological patterns rather than the mechanisms of disease pathogenesis. But often models of pathology are useful for basic research, and basic research discovers potential ways to treat disease, so separating the two groups is sometimes difficult.

Transmission (i.e., the transfer of information from one object to another) goes both ways:

• From bedside to bench – the reproduction of disease symptoms on model animals;
• From bench to bedside – testing new drugs on model animals and then on humans.

From bedside to bench

Before starting to fight any disease, it is necessary to gather as much information about it as possible and describe its clinical picture in as much detail as possible. Therefore, the first stage of translational biomedicine is clinical research. Doctors and scientists describe in detail psychiatric and neurological symptoms, deviations of biochemical markers in the blood, anatomical changes in organs and tissues, i.e. all possible signs of the disease.

Once we understand in sufficient detail what the disorder in question is (psychiatry tends to use this term instead of the more stigmatizing word “disease”), we can begin to reproduce its clinical picture on model animals.

Thus, the second stage of translation, from bedside to bench, is in vivo modeling. At this stage, through various manipulations with model animals, scientists try to reproduce the clinical picture of the disease or disorder as closely as possible. If they can successfully translate the disorder from humans to animals, the model can be used to better understand disease mechanisms and then to develop and test new drugs. After all, those molecular mechanisms whose breakdown leads to the development of disease are potential targets for drugs: in order to «fix» the body, one must first understand exactly where it «broke down».

In most cases, therapeutic targets are a variety of receptors, ion channels and transporters on the surface or inside cells, less often enzymes, transcription factors and other regulatory proteins.

From bench to bedside

Now we have reached the mechanisms underlying the disease and can start broadcasting backwards, that is, trying to create a drug that acts on the target molecules found. Most diseases are caused by a fairly broad spectrum of molecular abnormalities. They are often grouped around a particular functional system or anatomical structure: epilepsy is associated with an imbalance of excitation (glutamate) and inhibition (gamma-aminobutyric acid) in the brain; depression, schizophrenia, ADHD – with reduced or increased activity of monoamine structures. For many psychiatric disorders (schizophrenia, obsessive-compulsive – OCD, bipolar disorder and autism spectrum disorders – ASD) our understanding of the neurophysiological mechanisms is still rather superficial, which prevents the development of new more effective drugs.

Let’s imagine: we have chosen some molecular target and start searching for ways to chemically affect it. The first stage of translation is from bench to bedside – an “in silico” computer simulation. Special computer programs go through millions of possible chemical compounds and select from them those that can potentially bind to the target molecule, activate, suppress or somehow change its operation.

Next, promising candidates selected by computer modeling are synthesized chemically to begin testing on biological models. The next step on the road to a working drug is testing in “in vitro” models, that is, on cell cultures and tissue slices. In translational psychiatry and neuroscience, these are cultures of neurons and various glial cells or experiencing slices of different brain structures. For example, by methods of electrophysiological registration one can evaluate the effect of drugs on ionic currents through channels and receptors, by biochemical methods – activation of various intracellular signaling pathways.

If the candidate drug has shown good results on cells and survival slices, it moves on to the third stage – preclinical testing in in vivo models. Experimental animals are used to model the disorder in question and then try to treat it with the drug being tested. It is worth clarifying that not all experiments treating model animals are preclinical trials.

In fact, it is a highly regulated and thoroughly documented process (here’s an FDA regulation, for example) that is only begun after a candidate drug shows efficacy in preclinical experiments. In preclinical trials, model animals are injected with the drug being tested and its therapeutic effect evaluated depending on the dose and route of administration (oral, intraperitoneal, intravenous, intramuscular, subcutaneous).

Finally, if preclinical trials have shown both efficacy and safety of a potential drug, human clinical trials can begin. These consist of several phases:

• Phase 1 – safety evaluation in healthy volunteers
• Phase 2 – small group of patients
• Phase 3 – evaluation of efficacy in a large group of patients.
• Phase 4 – collecting information on side effects after the drug has been placed on the market.

Translational psychiatry studies the mechanisms of psychiatric disorders on animals and trials of psychiatric drugs. It is the modeling of psychiatric disorders that will be discussed below.

What animals are used for experiments?

At first glance, it is not an idle question – how to choose among all the biological diversity of animals suitable for translational research? In fact, the choice of model organisms is not too great: most of them belong to rodents.

• The house mouse (Mus musculus) is the main subject of translational biomedicine. According to 2019 annual reports from the line ministries, mice were used in 68% of studies in Germany and 61% of studies in the UK. The main advantages of mice over other laboratory animals, including rats, are low cost (smaller size and faster reproduction) and the possibility of genetic modification.

• The gray rat (Rattus norvegicus) long shared the title of the main model animal with the house mouse, until the advent of genetic engineering methods allowed mice to get far ahead.
  
  The main advantage of rats over mice is their much higher intelligence, which is especially important in neurobiology. Another advantage of rats is their larger size, which, of course, makes them more expensive, but greatly simplifies many experimental manipulations.
  
  For example, perinatal hypoxic-ischemic lesions (disorders of blood supply to the brain as a result of ligation of the cerebral arteries) can be modeled on newborn rats with the proper skills, but on tiny mice – unreal. And such models are incredibly important for the development of treatments for cerebral palsy and other hypoxic encephalopathies.

• Voles (genus Microtus) burst into neurobiology in the early 1990s, when two articles by Thomas Insel’s laboratory were published one after another in the most prestigious scientific journals. They showed the key role of first oxytocin and then vasopressin in the formation of parental pairs in monogamous (prairie and pine) and polygamous (mountain and meadow) vole species.

• Guinea pigs have been mentioned in a large comparative analysis of the vomeronasal organ in mammals, but are mainly found in translational cardiology and dermatology. Jungar hamsters due to their seasonal activity are a good model to study the role of epiphysis and suprachiasmatic nucleus of hypothalamus in regulation of diurnal and seasonal rhythms.

• The zebrafish (Danio rerio) is a small aquarium fish rapidly gaining popularity in neurobiological research. Of course, the brain and behavior of fish are arranged incomparably simpler than in humans, primates and rodents, but the cheapness and ease of operation make Danio rerio an indispensable object for primary screening of potential drugs. To date, models of depression and anxiety, RAS and ADHD have been developed for these fish.

• Insects are also used in translational research. Drosophila melanogaster has been the protagonist of genetics since the author of the chromosomal theory, Thomas Morgan. In neurobiology, too, they are used mainly to study the molecular genetic mechanisms of behavior and mental disorders. In Drosophila, anxious behavior can be assessed by the tendency to stay close to the wall of the box. They can also clearly distinguish aggressive behavior, which allows, for example, the role of serotonin to be studied by pharmacological and genetic manipulations.

• The honeybee (Apis mellifera) is actively used in studies of decision-making mechanisms and motivational behavior related to reward or punishment. Biogenic amines, relatives of serotonin and dopamine, also play an important role in these processes in them.
• The honeybee (Apis mellifera) is actively used in studies of decision-making mechanisms and motivational behavior related to reward or punishment. Biogenic amines, relatives of serotonin and dopamine, also play an important role in these processes in them.

• Apes, of course, are evolutionarily closest to humans, so the structure and operation of their brains are as similar to humans as possible. At the same time, research on them is extremely expensive – apes are slow to reproduce, eat fruit, require large specially equipped living quarters, veterinary care and a ruthless multitude of bureaucratic procedures. Therefore, primates are used either to study higher mental functions in basic research or for final drug testing before human clinical trials.

What experimental models are used?

All models of mental disorders can be classified according to the type of influence we exert on the animal to reproduce the symptoms. The models are conveniently grouped into three groups – physical, chemical, and genetic.

Physical models

Physical models – are mostly stressor models of anxiety and depressive disorders. The results of a recent study suggest that models of acute physical stress are better suited for modeling anxiety disorders and PTSD, and milder models of psychological stress are better suited for modeling depression.

1. Model of acute stress – a predator is used as a factor of acute stress (for example, a python eats a rat in front of his relatives), immobilization in a cylindrical pen or placement in water without an opportunity to get out.
2. Chronic stress models – in clinical practice, the key factors in the development of depression are everyday stress (work, family, community) and feelings of loneliness (unhappy love, lack of friends, disrespect from colleagues).
3. Chronic unpredictable mild stress is caused by constantly changing traumatic events: loud music, bright lights or reversed daylight, deprivation of water or food, mixing animals from different cages, wet bedding and the like. An animal accustomed to a monotonous daily routine becomes depressed from such an unpredictable life.
4. Social isolation is achieved by placing the animal in a solitary cage for several weeks, which for social mice and rats is a severe stressor. During the coronavirus pandemic of 2020-2021, humanity felt the full force of social isolation, which predictably led to an increase in the incidence of depression.

In addition to stress models of anxiety and depression, rodents are used to investigate mechanisms of circadian rhythms, sleep and its disturbances. Sleep deprivation or changes in daylight hours can be used for this purpose.

Chemical models

Chemical models use different substances to provoke the symptoms of the simulated disorder.

1. Pharmacological models are the bulk of chemical models. By now mankind has advanced very far in understanding the molecular mechanisms of a huge number of psychiatric disorders. And if we know exactly what breaks down in the development of a given disorder, we can aim to break it down in a model animal and reproduce the pathology in it.
   
   For example, we know a lot about the role that the brain’s dopamine system plays in the pathogenesis of schizophrenia, ADHD, and the manic phase of bipolar disorder. So by activating it with amphetamine, we get a hyper-excitable animal with symptoms of these disorders. The most common RAS model, administration of valproic acid (a common antiepileptic drug) to a pregnant female, was obtained in the opposite way: toxicological studies of the drug revealed side effects in the offspring, which were turned into a specific model.

1. Drug models – it is obvious that chemical addictions can only be studied by injecting the appropriate substances into model animals. Thus scientists study the mechanisms of drug, nicotine and alcohol addiction formation and look for new methods of their pharmacological treatment.
   
   The most common method is to self-inject the study drug by pushing the lever. In this way, the animal is motivated and remembers what it has to do to make it feel good. And the rat will push the lever for a new “dose” until it dies of hunger and thirst.
2. Immune models – the crucial role of neuroinflammatory processes is increasingly being discussed in the context of a wide variety of psychiatric disorders.

Inflammatory processes during intrauterine development play an important role in the pathogenesis of schizophrenia. Therefore, maternal immune activation (administration of interferon or bacterial lipopolysaccharide to pregnant females) is widely used as the most common model of schizophrenia not only in rodents, but also in rhesus macaques.

Genetic models

Genetic models are genetic changes in model animals that result in congenital symptoms of the disorder, an increase or decrease in the predisposition to develop symptoms when exposed to known risk factors.

• Breeding. Long before the advent of genetic engineering methods, humans had learned through long selection to breed and line animals with a wide variety of characteristics, including those related to behavior. The classic Krushinsky-Molodkina rats with their propensity to convulse under the influence of a loud sound played a huge role in understanding the mechanisms of epilepsy.
• Mutant mice. The first successes of genetic engineering made it possible to selectively “break” certain genes, which, according to our understanding, play an important role in the pathogenesis of this or that disorder. This is a favorite method of scientists in general – to understand the function of some protein in the body, the most reliable way is to turn off its synthesis and see what happens. Such animals are called “knockouts by the corresponding gene.”

For example, knockouts for various dopamine system proteins serve as models of schizophrenia and ADHD, and oxytocin knockouts serve as models of RAS, with the first oxytocin-knockout voles appearing in 2019. The vast majority of mutant models have been made in mice – genetically modifying rats is still difficult and expensive. But some transgenic rat lines are still being used to study stress and a number of neurodegenerative diseases.

However, each gene performs a huge number of functions, and they also differ in different cells and tissues. Therefore, if you turn off these genes in the whole organism, there will be too many effects, including lethal ones. For example, the neurotrophic factor BDNF is necessary for the functioning of all neuronal structures, including the spinal cord and peripheral nerves, so a total knockout of BDNF is fatal. One can either turn off only one copy of the gene (the role of neurotrophic factor BDNF in depression is often studied in heterozygotes), or localize the mutation in space and time.

How do you test a drug in these experimental models?

Now that we are familiar with such a wide range of translational models of various mental disorders, let us look at them through the lens of the Wilner criteria, to see if they really reflect the stated pathologies. By what parameters can this be assessed?

In models of mental disorders, of course, behavioral disorders come to the forefront – they are assessed in special behavioral tests. The tests are validated by predictive validity, i.e., by the effect of known drugs.

1. Models of anxiety disorders. Mice and rats are animals that avoid open spaces. Therefore, in an elevated cross-shaped maze, fear motivation will cause them to sit in the dark sleeve and curiosity motivation will cause them to go out into the open.
   
   The ratio of time in open and closed arms increases under the influence of anxiolytics, allowing it to be used as a measure of an animal’s anxiety level.
2. Models of depression. Unfortunately, not all models of depression can be transferred from humans to laboratory animals, but some do succeed.
   
   Anhedonia (decreased motivation to pleasure) can be assessed in a sucrose test. Of the two bottles of regular and sweet water, the healthy rat prefers the tasty one, while the depressed one drinks from both equally because she doesn’t care: “What’s the point of drinking sweet water if I’m going to die anyway?” Another important symptom of depression – despair and decreased faith in one’s own strength – is reproduced in the tests of forced swimming and hanging by the tail. In them, desperate animals stop climbing up the wall of a vessel with water or pulling themselves up on their tails in an attempt to get out. The tests are actively used to evaluate the effectiveness of antidepressants.
3. Models of attention deficit and hyperactivity disorder. To assess hyperactivity the task is elementary – we launch the animal into a confined space (the classic “open field” test) and measure how much it will run in 5 minutes.
   
   Assessing attention is more difficult – the rat needs to focus on the task and not be distracted by extraneous stimuli. This can be organized in instrumental learning tasks (press a lever with a certain sequence of signals, distracted – pressed incorrectly).
   
4. Models of obsessive-compulsive disorder. The main symptom of OCD is repetitive actions designed to reduce anxiety (washing hands, checking off appliances). It is impossible to transfer them directly to animals, but intensive repetitive actions can be regarded as analogues of compulsive rituals in people. The classic method is the marble burying test, in which healthy rodents bury some of the marbles and calm down, while animals in the OCD model cannot stop until they have buried every last marble.
5. Models of schizophrenia. Schizophrenia is a highly complex disorder with a wide variety of symptoms, of which a specific patient may exhibit only some. These symptoms are grouped into three groups: positive (hallucinations and delusions), negative (social avoidance) and cognitive (memory and attention disorders). It is not usually possible to reproduce all symptoms in one model because of the complex and largely unclear mechanisms underlying them, so different models of schizophrenia focus on different symptoms.
   
   Obviously, we cannot reproduce hallucinations and delusions in animals, but general motor hyperactivity is induced by psychostimulants and suppressed by neuroleptics, which allows it to be considered a reflection of positive symptoms in patients (predictive validity). Negative social symptoms are assessed in the communication tests described above. Various mazes (spatial memory), training with positive (sweet food) or negative (electric shock) reinforcement (associative memory) or a test of new object recognition (figurative memory) are used to assess memory. Cognitive tests for various forms of memory are also used in studies of many other mental disorders – depression, ADHD, and ASD.
6. Models of bipolar disorder. Bipolar disorder is still one of the biggest mysteries of psychiatry. We still know very little about both the factors of manic and depressive phase changes and the role of specific genes in the pathogenesis of bipolar disorder (although twin analyses estimate an 80% contribution of genotype).

Mood stabilizers, such as lithium, are primarily used for therapy, but their mechanism of action is not yet understood. Under such conditions, it is impossible to create a working model of bipolar disorder, so translational studies can only study the two phases separately, which does not bring us any closer to understanding their relationship.

Conclusion

Translational models of mental disorders have undoubtedly contributed enormously to our understanding of their mechanisms and the development of psychopharmacological drugs of the widest profile. In addition to the accumulation of an enormous body of new knowledge, technological advances and new research methodologies have played an important role in the development of translational psychiatry.

A strong impetus to our understanding of the molecular mechanisms of psychiatric disorders came from the emergence of first genetically modified model animals and then from optogenetics. In this incredibly popular method nowadays, genetic engineering can be used to insert light-sensitive ion channels borrowed from bacteria into neuronal membranes.

By bringing a thin light guide to a particular structure, light can activate and inhibit different types of neurons in it (according to the neurotransmitter used). So scientists can, for example, study the role of corticoliberiberergic neurons of the amygdala in anxious behavior or dopaminergic neurons of the adjoining nucleus in motivational behavior and drug addiction formation. Just imagine – we send a light signal to the brain, and the rat freezes in fear or starts pushing the lever more actively – this is real modern research.

There is also an opposite method of calcium imaging – genetically engineered intracellular calcium ion sensors allow us to see the activation of individual neurons in a free-moving mouse under a microscope. In this way, we can see which neurons are related to anxiety, which are related to sexual behavior, and which are related to aggression.

It is hoped that new genetic models and technical methods will finally allow us to find effective treatments for schizophrenia, obsessive-compulsive, bipolar, and other mental disorders.