Prisoner in another world

In a short time period, modern genetic research has found numerous genes that contribute to the development of schizophrenia. By taking a closer look at how they are linked with the disease, scientists hope to understand the causes and find better remedies.

Professor Patrick Sullivan. Photo: Ulf Sirborn

Text: Fredrik Hedlund, first published in Medical Science issue 3, 2014.

"We can now show that there are 128 places in human DNA which are very probably of great significance for the risk of developing schizophrenia. Every one of them is a clue to the biology behind it. Our goal is to decode the genes and take us from genetic variation to molecules and then to cells," says Patrick Sullivan, professor of psychiatric genetics at the University of North Carolina and at the Department of Medical Epidemiology and Biostatistics, Karolinska Institutet.

He was recently recruited to the Swedish university part-time to develop the Psychiatry Genomics Institute at Karolinska Institutet, with the aim of understanding the underlying genetic causes of schizophrenia and producing a treatment that will enable those affected by the disease to live more normal lives. The exact causes of schizophrenia are still unclear, but as is the case with so many other diseases, there is a combination of heredity and environment.

Since the 1960s researchers have understood that there is a significant genetic factor because the disease is passed down in families, but also that there are some early environmental factors during the foetus stage that increase the risk. Premature birth, low birth weight due to growth problems during the embryonic stage, lack of oxygen during birth and mothers' infections during pregnancy are examples of factors that scientists have observed as increasing the risk of a child later developing schizophrenia. Much of this research has been done in Sweden where, due to the social security system with personal registration numbers and different registers, it was possible to follow up patients in a completely different way compared with other countries.

Genetic database

Professor Christina Hultman. Photo: Stefan Zimmerman

An internationally recognized pioneer in this work is Christina Hultman, professor of psychiatric epidemiology at the Department of Medical Epidemiology and Biostatistics, Karolinska Institutet. She has also compiled a genetic database with 5,800 schizophrenia patients and 7,100 healthy control subjects of the same age, sex and from the same counties as the patients. Thanks to the registers, she has been able to supplement the database with important information about non-genetic risk factors among the subjects.

"Every single one of my schizophrenia patients and controls has been asked if it is OK that we collect data from the Swedish health register, and the vast majority have said yes. This means we can obtain school results, family history, birth delivery data, medical prescriptions and so on - and obviously all the data is anonymised," she says.

Genetic research involves both patients and controls providing a blood sample; DNA is purified from the sample and is then analysed and compared. Genetic variations that often occur among the patients, but not in the control group, are those which are interesting. The basic principle of the research is thus very simple, but the genetic analysis is the difficult part. It is only thanks to the rapid pace of technological development in the area in the last ten years that this research is possible at all.

"We can now make analyses quickly and cheaply that would have been science fiction only ten years ago. ­Checking a million genetic markers for each trial subject is a routine task which costs SEK 600. When we started a few years ago, the same process cost 12 to 15 times as much," she says.

The costs of DNA analyses

It is fortunate that the costs of DNA analyses have decreased, because the researchers need to analyse many patients to get answers to their questions.

" We need to have very large groups of patients because the genetic effects are quite small on average. Usually when people hear about genetics, they think it is a question of a one/zero relationship - if you have the gene, you get the disease and you don't have it, you are safe. There are some diseases that work like that, but not schizophrenia. It is more complex," says Patrick Sullivan.

"­There are hundreds, maybe even thousands of genes that are involved. Schizophrenia is a polygenic disease," says Christina Hultman.

For this reason, they have started collaborating with other researchers and merged data from groups of patients in different parts of the world into something they call the Psychiatrics Genomic Consortium, which is lead by Patrick Sullivan. In this way, they have gathered around 36,000 patients and almost twice as many controls. This allows them to see more clearly which mutations and other genetic changes are significant for schizophrenia and which are not.

"­In our latest article we indicate 128 places where the genetic material is definitely different in the patients," says Christina Hultman. The fact that there are so many genes involved means that it takes more than just one mutation to get the disease.

"­Many of us have a number of risk genes for schizophrenia. Despite this fact, most of us do not develop the disease. This is because environmental factors are also important. It makes a difference, what people experienced as early as the embryonic stage, what they were subjected to later in life and what choices they make," she says.

Mix of diseases

As the genetic picture begins to clear, researchers have also discovered that diseases such as schizophrenia, bipolar disorder, depression, autism and ADHD have some risk genes in common.

Urban Ösby. Photo: Ulf Sirborn.

"It seems as if a number of these genes are not particularly disease-specific, says Urban Ösby, researcher in neurogenetics at the Department of Molecular Medicine and Surgery, Karolinska Institutet.

He researches mainly in the co-morbidity of mental diseases and is also part of the Psychiatrics Genomic Consortium, searching for genetic differences and similarities in mental diseases. Studies by the Consortium show a strong correlation between schizophrenia and bipolar disease, and a slightly weaker correlation with depression. He believes that the current classification of diseases is not necessarily the correct one.

"Today's clinical classification is based on symptoms and treatment, but they do not necessarily correspond to the biological and genetic causes. We must accept that current clinical boundaries may not match those of biology," he says. He explains that there is a big discussion within the research community on whether schizophrenia and bipolar disorder are really different diseases or not. "It is perhaps a kind of sliding scale, on which the boundaries are a little random between the two diseases," he says.

Patrick Sullivan also has similar beliefs.

"Personally, I believe it is possible to define extremes in which classic schizophrenia never has any bipolar symptoms, and classic bipolar disorder never has any schizophrenia symptoms. Such cases exist, but they are rare. Most patients have a diverse mix of both. The symptoms may also vary over a lifetime," he says.

"Better knowledge of the genetic cause-and-effect relationships could explain how the diseases are really linked, and what is what. But there are more benefits in understanding the underlying genetics and how it affects the biological processes that result in these mental diseases. One of these is about the possibility of predicting the outcome of new-onset psychosis.

Some patients with psychosis

Doctors now know that some patients with psychosis will later be diagnosed as schizophrenic, others will be diagnosed as bipolar and the remainder will recover and be healthy, but they cannot say which individuals will have each diagnosis. The problem is that the three groups should be given different treatments, and waiting to see which treatment is required worsens the long-term prognosis.

"Can genetics help us and guide us towards what should be done clinically in these complex cases?" asks Patrick Sullivan rhetorically.

The most important gain in understanding the genetics behind schizophrenia is, of course, that it can lead to new methods of treatment that really address the fundamental problem and not just the symptoms. By finding out which proteins are being encoded by the genes that are mutated in schizophrenia, researchers may eventually discover what processes in the brain are affected and how they are affected, and in this way determine the best treatment for the disease.

But with so many genes, it is a huge amount of work to trace them all to their proteins and then to the biological processes. Patrick Sullivan is convinced that they will succeed, though, and his recipe for speeding up the process is to involve many different researchers.

"One of the key factors in rapid progress is "Team science". If we can bring together researchers in the same area and get them to focus on the same issues, it will accelerate important discoveries. We want to make a real impact on the treatment of schizophrenia that will help patients," he says.

He even sees the possibility of using genetic information at the individual level sometime in the future.

"We are trying to see if we can understand the genome of a specific person. If we succeed in doing that, it could result in diagnosis and treatment for individuals based on their specific genetic profile. We call it precision medicine. It is the ultimate goal," he says.

From biology to genetics

If Christina Hultman, Patrick Sullivan and Urban Ösby move from the inside out - from mutations in genes towards biological processes in the body, then Sophie Erhardt, docent in pharmacology at the Department of Physiology and Pharmacology, Karolinska Institutet, moves in the other direction. In 1997 she started with the fact that a substance classed as an illegal drug, phencyclidine or more widely known as PCP or "angel dust", can induce schizophrenia-like psychosis in healthy people and exacerbates symptoms in schizophrenics.

portrait of Sophie Erhardt

She wondered whether there were any natural body chemicals that acted like PCP. It emerged that some other researchers a few years earlier had found kynurenic acid in the brain of healthy people. Kynurenic acid is a neuro-active substance with an effect similar to PCP. Sophie Erhardt was able to show that cerebrospinal fluid taken from patients with schizophrenia actually contained raised levels of kynurenic acid. She was also able to show later that a raised level of kynurenic acid in rats increased dopamine activity, something that had been proposed as a contributing factor to schizophrenia as early as the 1960s.

"In the United States I learned to use animal models for schizophrenia and then I was able to show that if kynurenic acid levels were raised in mice and rats they exhibited schizophrenia-like behaviour," she says.

The next question was why was the kynurenic acid level was raised. It emerged that the production of kynurenic acid was triggered by the body's immune response. Sophie Erhardt and her research team went on to examine a group of schizophrenia onset patients and were searching for the body's immune molecules called cytokines. It was found that a cytokine called interleukin (IL)-1beta was significantly higher in these patients.

At the time, no-one knew whether IL-1beta had any connection with kynurenic acid, but Sophie Erhardt and her colleagues have recently found that additional IL-1beta increases the production of kynurenic acid in both human cells and laboratory animals. But why are IL-1beta levels raised?

"We are now focusing on two genes that undergo changes during schizophrenia or psychosis and which lead to an increase in IL-1beta levels," she says.

A number of genes

This is where the different researchers' paths cross. In an article from 2009, Patrick Sullivan and Christina Hultman pointed out a number of genes that they believed were important for the development of schizophrenia.

"The gene they ranked at the top was identical to the gene that we had identified as the most interesting for us," says Sophie Erhardt.

It has not yet been proven that the so-called kynurenic acid hypothesis is correct and that it is disturbances in this system which lie behind at least some cases of schizophrenia. Sophie Erhardt and her research team have indeed found that laboratory mice deprived of the most interesting gene have increased levels of IL-1beta and kynurenic acid in the brain, they have more dopamine released and suffer from a deterioration in cognition. All of these changes are similar to what happens with schizophrenia.

"I think it is an interesting track. It is one of several current hypotheses," says Christina Hultman.

"The kynurenic acid hypothesis has come up several times and I agree that it seems to be one of the more promising hypotheses. We have it on our radar screen too," says Patrick Sullivan.

They take up other hypotheses that also appear promising, such as disturbances in the brain's signalling system via calcium, for example, or NMDA receptors.

"I believe that in the next ten years, with the help of genetic research, we will be able to narrow down the different hypotheses and understand what biological effects they have," says Christina Hultman.

At the same time, Sophie Erhardt continues with her research. She is now trying to connect the experimental results with clinical studies. Among other things, she and other researchers have demonstrated that patients with bipolar disorder who have episodes of psychosis have higher levels of IL-1beta and kynurenic acid. Along with psychiatrists, she has also started to monitor patients with new episodes of psychosis prior to treatment, in order to see what happens to the levels of immunological markers and kynurenic acid over time.

These patients will be followed in a long-term study, during which their risk of developing cardiovascular disease will also be studied. Animal studies have shown that another metabolite in the kynurenic metabolism pathway protects against cardiovascular disease.

"Our hypothesis is that when the production of kynurenic acid increases, it is at the cost of this protective metabolite and patients then become more susceptible to cardiovascular disease," says Sophie Erhardt.

And it is cardiovascular disease in particular that affects patients with schizophrenia out of all proportion.

Cardiovascular disease common

"Patients with schizophrenia have 15-20 years shorter life expectancy on average compared with the normal population. And the main reason for this is increased mortality through cardiovascular disease," says Urban Ösby.

He has done research into co-morbidity in schizophrenia and has several explanations for the higher rate of death through cardiovascular diseases.

"The disease in itself brings about a lower quality of life, but we also note that there is significant under-treatment of cardiovascular disease. In a study not yet published, we also show that the treatment of cardiovascular risk factors has been neglected in this group."

Urban Ösby believes that it is partly that the disease makes patients a little unpredictable, so that they do not always come to appointments or take the medicines they have been prescribed. But more significant is the organisational division in healthcare between body and soul, between somatics and psychiatry.

"These patients are cared for through psychiatry, but psychiatry has no competence in dealing with high blood pressure, or disruptions in blood sugar levels or blood fats. Somatic health care is not organised or tuned into this side of things," he says.

Another factor is that anti-psychotic medication often results in weight gain, which in turn can contribute to the development of risk factors for cardiovascular disease such as high blood pressure and blood sugar disruptions. In other words, schizophrenia patients are more susceptible to cardiovascular risks but receive poorer treatment than the normal population.

"It is likely that doctors in primary health care underestimate just how large this increased risk is, with the result that treatment is less robust than it should be," says Urban Ösby.

He is now making a study to examine how schizophrenia patients can best be treated for their cardiovascular risk factors. But he also hopes that psychiatry and primary health care will improve their cooperation and strengthen it.

"In the relatively short-term, there is a lot we can do that would certainly have an impact on patients' quality of life as well as their life expectancy," says Urban Ösby.

Increased suicide risk

The second major threat to life expectancy among schizophrenia patients is the risk of suicide. According to a study of Swedish patients in 2000, the risk of a schizophrenia patient dying through suicide was 15 times higher than the normal population. A more recent study of Swedish patients published last year shows that the risk had increased to 22 times.

"We are now making a new study, which points out that this difference will grow even further," says Urban Ösby.

Through his studies he can confirm that the difference between the normal population and schizophrenia patients regarding morbidity from cardiovascular disease and suicide is increasing. Which of course is not good.

"This indicates that the gap between schizophrenia patients and the normal population is increasing. You hope that health care will get better over time, not worse. But this is the situation," he says.

Text: Fredrik Hedlund, first published in Medical Science issue 3, 2014.