Introduction to schizophrenia
Schizophrenia is a psychiatric disorder that is not only debilitating to the patient but also a heavy disease burden for the patient’s family and society. It affects around 1% of the U.S. population and can manifest as early as 20 years old. While there are many proposed causes for schizophrenia, it is a highly heterogeneous disorder that cannot be pinpointed to a single etiology or tackled by a single treatment yet.
Schizophrenia is a difficult disorder to diagnose and treat because of its diverse spectrum of symptoms that often overlap with other psychiatric disorders. Patients can present positive, negative and cognitive symptoms in different forms and severity. The positive symptoms include delusions and hallucinations, which tend to relapse after the first episode. The negative symptoms include impaired motivation, speech, and social interactions. The cognitive symptoms relate to poor performance in a variety of cognitive functions, which often differ among individuals.
The current treatment generally consists of antipsychotic drugs and therapies. However, despite half a century’s effort in refining the antipsychotic medications for schizophrenic patients, antipsychotic drugs are mainly effective for preventing positive symptoms, but not negative and cognitive symptoms. In addition, unwanted side effects and treatment-resistance are common. Since developing effective therapeutic interventions are dependent on understanding the basic etiology of the disorder, the two most pressing challenges faced by pre-clinical neuroscientists are to uncover how schizophrenia develops and how treatments can be targeted to different symptoms. To this end, animal models have served an indispensable role, which will be the focus of this article.
The early theories on the pathophysiology of schizophrenia involve abnormalities in neurotransmission, which involve a deficiency or an excess of neurotransmitters including dopamine, glutamate and serotonin. Among them, the dopamine circuits were the most studied as dopamine D2 receptor antagonists were the first effective antipsychotic medications, leading to the well-known dopamine hypothesis to explain the positive symptoms. Glutamate antagonists induced schizophrenia-like symptoms, and medications targeting both dopamine and serotonin receptors turned out more effective than dopamine-only medications.
With more recent large scale genomic and epidemiology studies, it is now widely accepted that schizophrenia likely arises due to multiple developmental risk factors including genetics and environmental insults. Various genes have been linked to schizophrenia, including DISC1, Neuroregulin 1 and Dysbindin 1, which are also the same genes critical for brain development. Nevertheless, no two schizophrenic cases are the same and much is still to be learned.
Research techniques for studying schizophrenia
Epidemiological studies aim to draw correlations between environmental factors and schizophrenia, especially those important in early neurodevelopment such as pregnancy and birth complications. To date, schizophrenia has been associated with a wide range of environmental risk factors including birth season, immigration, and socioeconomic status. While such information may be useful for directing the attention of researchers, observational epidemiology cannot provide any insights into the underlying causation and has little explanatory power.
Human brain imaging
Neuroimaging is the only non-invasive approach currently available for humans that can relate schizophrenic symptoms to altered brain structures. For example, changes to the prefrontal cortex has been linked to some cognitive deficits of schizophrenic patients. The volume of grey matter seems to reduce with schizophrenia progression, but it is possibly associated with antipsychotic treatment rather than the disorder itself. So far, no anatomical or functional abnormalities revealed via neuroimaging are specific for schizophrenia.
In vivo studies using animal models
Although recent epidemiological and neuroimaging studies have shed light on potential psychosocial risk factors and neural correlates for schizophrenia, they only offer correlative evidence. The study of animal models at the level of neural circuits are necessary to drive mechanistic discoveries.
When it comes to a disorder like schizophrenia that involves human-specific cognitive functions, some may question the validity of animal models. However, the use of animal models is inevitable due to ethical limitations in human research. In addition, genome-editing approaches such as CRISPR are useful for introducing genetic risk factors identified in human studies to animal models, which could provide important insight into where, when and how certain genes may affect brain development and schizophrenia development. Lastly, animal models are essential for testing and developing effective therapeutics, which may lead to human clinical trials and ultimately promising medications for the patients.
To date, rodents are the most commonly used animal models for brain disorders and are considered our current best non-primate mammalian models. A number of behavioral changes in rodents have been suggested to mimic schizophrenia in humans, such as impairments in social interactions, working memory and locomotion. The next two sections will evaluate the animal models and behavioral assessments currently used in schizophrenia research.
Different mouse models of schizophrenia
In the past few decades, over 50 animal models have been characterized. Due to the complexity of schizophrenia, none of the models can accurately model all of the symptoms. However, this is not a reason not to use animal models, as schizophrenia also never presents in the same way in different patients. Therefore, knowing what clusters of symptoms or neural correlates of schizophrenia are altered in each animal model may hold the key to answering your particular research questions.
These models were established early in schizophrenia research, based on the neurotransmitters that were shown to be altered in schizophrenia. They involve drug administration to induce schizophrenia-like symptoms, but have poorly elucidated modes of action.
Amphetamine is a dopamine agonist, used to induce dopamine hyperactivity and psychotic-like symptoms. Amphetamine-induced symptoms could be alleviated with antipsychotics, supporting its validity in modeling the positive symptoms of schizophrenia.
The downside of the amphetamine model is that it fails to induce negative or cognitive symptoms in animals, limiting its construct validity. Some schizophrenic patients are presented with predominantly negative symptoms and do not respond to antipsychotic treatments. It is expected that the amphetamine model would not be useful for modeling this subset of schizophrenic patients.
Anesthetic Drug Phencyclidine (PCP) model
PCP acts as a glutamate antagonist, and induces delusions and hallucinations in humans. In rodents, both acute and long-term PCP administration models have been characterized and produce different neurochemical and behavioral effects. Notably, the chronic PCP model can lead to positive symptoms as well as some negative and cognitive deficits. It has also shown translational relevance to non-human primates and humans, making it a superior model than the amphetamine model.
The main criticism for the PCP model is that the intervention starts in adult animals, disregarding the critical developmental time window which may dispose many schizophrenic patients to later developing the symptoms.
The neurodevelopmental hypothesis of schizophrenia is one that has gained significant support in recent years of schizophrenia research. Since then, many developmental models have been established, involving lesions, environmental manipulations or drug administration during a critical perinatal or early postnatal time window.
Neonatal ventral hippocampal (NVH) lesions
Unlike the traditional pharmacological based animal models, lesioning the ventral hippocampus during a critical neonatal stage acts as a developmental insult that leads to prolonged consequences on neural circuits. It is the most thoroughly characterized neurodevelopmental model for studying the origin of schizophrenia.
A major caveat with such lesion models is that the induced structural damage may be far greater than schizophrenic individuals, who usually present unremarkable brain structures. Nevertheless, postpubertal emergence of dopamine-related dysfunctions which are also responsive to antipsychotic treatments demonstrate their validity to some extent.
Perinatal immune activation model
Since an adverse prenatal environment such as prenatal exposure to infection was linked to schizophrenia by epidemiology, rodent models based on exposure to virus-like agents such as synthetic RNA and bacterial endotoxin were established to explore a potential causal link.
Depending on the precise timing of immune challenge, the perinatal immune activation models can lead to different structural and behavioral dysfunctions, which provide insight into the underlying mechanisms of different symptom types. However, researchers still need to take care when extrapolating the results to humans due to the differences in the timing of brain development.
Methylazomethanol (MAM) model
MAM administration on gestational day 17 introduces a developmental disruption leading to histological, neurophysiological and behavioral deficits that resemble schizophrenia in humans. Some schizophrenia-like behavioral defects include sensorimotor gating deficits, working memory deficits, and social withdrawal.
The controllable time of insult is an advantage of the model as gestational day 17 targets later developing paralimbic and temporal cortices, areas involved in the histopathology in human patients. However, like many other schizophrenia models, MAM-induced phenotype does not necessarily reflect the etiology of schizophrenia, but nonetheless provide valid psychopathology to study.
pre- or perinatal hypoxia model
Infants who suffered from hypoxia are at increased risk to develop schizophrenia in adulthood. To study this phenomenon, a number of animal models challenged by hypoxia at different developmental times were developed. Depending on the duration and severity of hypoxia, animals display a variety of neurological symptoms including altered brain anatomy and neurotransmission. It is worth noting that, currently, the hypoxia models are poorly standardized, producing variables and sometimes contradicting findings.
Developmental Vitamin D (DVD) model
The developmental vitamin D (DVD) deficiency hypothesis stems from the observation that schizophrenia risk is higher in people who are born in winter or spring or live further from the equator. These environments are less exposed to sunlight, thereby reducing vitamin D production. Rat models have been developed in which rodent dams were exposed to a diet completely deficient in vitamin D from pre-conception to birth.
Disruption of brain development, gene dysregulation and altered dopaminergic system were reported. However, some characteristic schizophrenia-like behaviors such as sensorimotor gating are not affected.
The concept that genetics may also have a large part to play in schizophrenia came from twin studies estimating an 80% heritability for schizophrenia. A range of candidate genes have been identified in genome-wide studies, most of which are in fact involved in neural functions, making them prime targets for further investigations in animal models.
Disrupted in schizophrenia 1 (DISC1) model
DISC1 is an important structural protein for many facets of neuronal development and also one of the earliest genes suspected to cause schizophrenia. While DISC-1 mutant mouse brains show similar structural changes to schizophrenic patients, there are mixed reports on their behavioral outcomes. Some of the discrepancies may arise in how different transgenic strains are produced, and further model validation and investigations are required.
Neureguline-1 (NRG1) model
The Neureguline-1 (NRG1) gene was associated with schizophrenia in human genomic studies. Subsequently, genetic NGR1 knockout mouse models were created. Behavioral assessments have shown robust deficits that resemble positive, negative and cognitive symptoms of schizophrenia. For example, impaired sensorimotor gating processing in a schizophrenia-like manner was revealed by prepulse inhibition in a NRG-overexpression mouse model. The mice also demonstrated social interaction deficits, reflected in the loss of preference for being in a chamber with novel over familiar mouse.
Other genetic models
Other genes implicated in genome-wide association studies that have been developed into schizophrenic animal models include Dysbindin, Reelin, G72, DAO, SRR. While some of these single gene-based models provided important insight into brain development and disorders, many others did not prove successful in modeling a complex and heterogeneous disorder like schizophrenia.
Behavioral assessments for schizophrenia models
In addition to considering the pros and cons of different schizophrenic rodent models to ask your research question in, choosing the appropriate behavioral assessments are equally crucial for getting an interpretable answer. A number of behavioral phenotypes have been characterized in rodent models, including impaired motor and sensory function, working memory and social behaviors.
Prepulse inhibition (PPI) of the acoustic startle reflex
Multiple schizophrenic rodent models have shown abnormalities in startle responses, measured by a pre-pulse inhibition test (PPI). It assesses the reflexive response to loud acoustic stimuli, which can indicate sensorimotor gating defects and/or anxiety.
PPI is a very useful behavioral assessment as it is fully translational. It is applicable to both rodents and humans and deficits are indeed commonly reported in schizophrenic patients. Therefore, an impaired sensorimotor gating function is considered a core symptom of the disorder.
Morris water maze (MWM)
The Morris Water Maze (MWM) measures spatial learning and memory based on an animal’s desire to escape a pool of water by escaping to a hidden platform.
In NVH lesion, PCP, Reelin knockout models, rodents have shown impaired working memory, which rendered them unable to find the location of the hidden platform.
Radial Arm Maze (RAM)
The Radial Arm Maze (RAM) is a widely used behavioral test for spatial memory and learning, which tasks the animals to retrieve food at the end of each arm that they have not already explored. Failure for the animals to remember which arms they have been to may indicate deficits in their spatial learning and memory functions.
In some schizophrenic rodent models including the NVH lesion, DISC-1 and dysbindin knockout models, RAM was used to show learning and memory impairments.
The T Maze is widely used in neuroscience for assessing spatial learning and memory. The maze contains a left and right arm for rodents to explore and seed food rewards. After a successful retrieval of food from one of the arms, the rodent will naturally explore the un-visited arm, a behavior dependent on working memory.
Deficits in delayed alternation in the T-maze has been reported in chronic PCP and NVH lesion models.
The Y maze is built on the concept of a T maze, which is used to test spatial recognition, memory and cognition. It is often a preferred choice as an additional arm in the design makes turning more gradual and learning faster.
Subchronic PCP, MAM and NRG1 knockout models have shown altered behaviors in Y mazes.
Novel object recognition task
The novel object recognition task is a visual learning and memory task based on an innate curiosity to explore novel objects that is shared between rodents and humans. It requires both object recognition and cognitive abilities.
A number of schizophrenic models have shown deficits in novel object recognition, including PCP and dysbindin knockout mice.
In summary, there is currently no and will unlikely ever be a so called “one size fits all” animal model to fully recapitulate the full spectrum of schizophrenic symptoms. However, researchers should not be discouraged as most of the existing models do replicate changes in schizophrenic brains in some ways or another, which still hold great potentials of revealing the cellular and molecular mechanisms underlying brain development and disorders. For example, the pharmacological models have played important roles in the development of existing therapeutics, and the neurodevelopmental and genetic models have been unraveling ever more new insights into brain development and how it can go awry.
With at least 50 animal models currently available for studying schizophrenia, researchers have more than enough options to choose how they wish to begin. Having weighed the pros and cons of different schizophrenic models and justified the choice before carrying out experiments are crucial.
- Patel, K. R., Cherian, J., Gohil, K. & Atkinson, D. schizophrenia: overview and treatment options. P T 39, 638-645 (2014).
- Owen, M. J., Sawa, A. & Mortensen, P. B. schizophrenia. Lancet 388, 86-97, doi:10.1016/S0140-6736(15)01121-6 (2016).
- Mouri, A., Nagai, T., Ibi, D. & Yamada, K. Animal models of schizophrenia for molecular and pharmacological intervention and potential candidate molecules. Neurobiol Dis 53, 61-74, doi:10.1016/j.nbd.2012.10.025 (2013).
- Winship, I. R. et al. An Overview of Animal Models Related to schizophrenia. Can J Psychiatry 64, 5-17, doi:10.1177/0706743718773728 (2019).
- Barch, D. M. & Ceaser, A. Cognition in schizophrenia: core psychological and neural mechanisms. Trends Cogn Sci 16, 27-34, doi:10.1016/j.tics.2011.11.015 (2012).
- Vita, A., De Peri, L., Deste, G. & Sacchetti, E. Progressive loss of cortical gray matter in schizophrenia: a meta-analysis and meta-regression of longitudinal MRI studies. Transl Psychiatry 2, e190, doi:10.1038/tp.2012.116 (2012).
- Fusar-Poli, P. et al. Progressive brain changes in schizophrenia related to antipsychotic treatment? A meta-analysis of longitudinal MRI studies. Neurosci Biobehav Rev 37, 1680-1691, doi:10.1016/j.neubiorev.2013.06.001 (2013).
- Arguello, P. A. & Gogos, J. A. Modeling madness in mice: one piece at a time. Neuron 52, 179-196, doi:10.1016/j.neuron.2006.09.023 (2006).
- Tseng, K. Y., Chambers, R. A. & Lipska, B. K. The neonatal ventral hippocampal lesion as a heuristic neurodevelopmental model of schizophrenia. Behav Brain Res 204, 295-305, doi:10.1016/j.bbr.2008.11.039 (2009).
- Jones, C. A., Watson, D. J. & Fone, K. C. Animal models of schizophrenia. Br J Pharmacol 164, 1162-1194, doi:10.1111/j.1476-5381.2011.01386.x (2011).
- Marcotte, E. R., Pearson, D. M. & Srivastava, L. K. Animal models of schizophrenia: a critical review. J Psychiatry Neurosci 26, 395-410 (2001).
- McGrath, J. et al. A systematic review of the incidence of schizophrenia: the distribution of rates and the influence of sex, urbanicity, migrant status and methodology. BMC Med 2, 13, doi:10.1186/1741-7015-2-13 (2004).
- Meyer, U., Feldon, J. & Fatemi, S. H. In-vivo rodent models for the experimental investigation of prenatal immune activation effects in neurodevelopmental brain disorders. Neurosci Biobehav Rev 33, 1061-1079, doi:10.1016/j.neubiorev.2009.05.001 (2009).
- Kesby, J. P., Burne, T. H., McGrath, J. J. & Eyles, D. W. Developmental vitamin D deficiency alters MK 801-induced hyperlocomotion in the adult rat: An animal model of schizophrenia. Biol Psychiatry 60, 591-596, doi:10.1016/j.biopsych.2006.02.033 (2006).
- Lodge, D. J., & Grace, A. A. (2009). Gestational methylazoxymethanol acetate administration: A developmental disruption model of schizophrenia. Behavioural Brain Research, 204(2), 306–312. https://doi.org/10.1016/j.bbr.2009.01.031
- Hefter, Dimitri, Marti, Hugo H., Gass, Peter, & Inta, Dragos. (2018). Perinatal Hypoxia and Ischemia in Animal Models of Schizophrenia. Frontiers in Psychiatry, 9, 106. https://doi.org/10.3389/fpsyt.2018.00106
- Schoenrock, S. A., & Tarantino, L. M. (2016). Developmental vitamin D deficiency and schizophrenia: the role of animal models. Genes, Brain and Behavior, 15(1), 45–61. https://doi.org/10.1111/gbb.12271
- Sullivan, P. F., Kendler, K. S. & Neale, M. C. schizophrenia as a complex trait: evidence from a meta-analysis of twin studies. Arch Gen Psychiatry 60, 1187-1192, doi:10.1001/archpsyc.60.12.1187 (2003).
- Papaleo, F., Yang, F., Paterson, C., Palumbo, S., Carr, G. V., Wang, Y., … Law, A. J. (2016). Behavioral, Neurophysiological, and Synaptic Impairment in a Transgenic Neuregulin1 (NRG1-IV) Murine Schizophrenia Model. The Journal of Neuroscience : the Official Journal of the Society for Neuroscience, 36(17), 4859–4875. https://doi.org/10.1523/JNEUROSCI.4632-15.2016
- Swerdlow, N. R., Braff, D. L. & Geyer, M. A. Animal models of deficient sensorimotor gating: what we know, what we think we know, and what we hope to know soon. Behav Pharmacol 11, 185-204, doi:10.1097/00008877-200006000-00002 (2000).