By Ewen Callaway of Nature magazine

Before committing suicide at the age of 22, an anonymous man with schizophrenia donated a biopsy of his skin cells to research. Reborn as neurons, these cells may help neuroscientists to unpick the disease he struggled with from early childhood.

Experiments on these cells, as well as those of several other patients, are reported today in Nature. They represent the first of what are sure to be many mental illnesses 'in a dish', made by reprogramming patients' skin cells to an embryonic-like state from which they can form any tissue type.

Recreating neuropsychiatric conditions such as schizophrenia and bipolar disorder using such cells represents a daunting challenge: scientists do not know the underlying biological basis of mental illnesses; symptoms vary between patients; and although psychiatric illnesses are strongly influenced by genes, it has proved devilishly hard to identify many that explain more than a fraction of a person's risk.

"All of us had been contacted by patients asking 'when can I get my stem cells to solve my schizophrenia'. It's not as simple as that," says Russell Margolis, a psychiatrist and neurogeneticist at Johns Hopkins University in Baltimore, Maryland, who was not involved in the study. "It's an additional piece to the puzzle as opposed to the answer."

Cocktail recipe

Since researchers reported that cocktails of particular genes can be used to reprogram human cells to an embryonic-like state, scientists have learned to coax these 'induced pluripotent stem cells' (iPSCs) into various cell types and used them to try to work out what goes awry in disease. So far, iPSC-derived models have been published for about a dozen diseases, from rare heart conditions to hereditary blood diseases.

Fred Gage, a neuroscientist at the Salk Institute for Biological Studies in San Diego, California, and his team created iPSC models from the cells of the 22-year-old man mentioned above, as well as those of two brother-sister pairs, all of whom had either schizophrenia or related conditions such as schizoaffective disorder.

When the authors transformed the iPSCs into neurons, they noticed that the patient-derived cells made fewer connections, or synapses, with other neurons in the same dish than did neurons from people without psychiatric disorders. However, tests showed that the patients' neurons conducted electrical pulses just as well as normal neurons did.

Interestingly, the antipsychotic medication loxapine, used to treat schizophrenia, boosted the number of synapses formed by the patient-derived neurons to normal levels.

Four other antipsychotic drugs had no consistent effect, although Gage notes that all of the drugs benefited cells from at least one patient. His team also reported differences in gene expression between the neurons of patients with schizophrenia and those of healthy people, including changes in genes related to synapse function and others previously implicated in genetic studies of the disease.

Michael Owen, a psychiatric geneticist at Cardiff University, UK, agrees that synapses are a reasonable place to look for differences between neurons from people with schizophrenia and those of healthy individuals. However, he says it is a logical leap to conclude that such differences underlie schizophrenia.

Moreover, differences between cells derived from patients with mental illnesses and those of healthy people could reflect changes brought about by the process of creating iPSCs and not the disease itself, warns Kwang-Soo Kim, a stem-cell scientist at McLean Hospital in Belmont, Massachusetts. This could be problematic in mental illness, in which the differences between healthy and disease cells may be slight, Kim says.

So far, many of the iPSC models published are for diseases resulting from mutations in a single gene. Mental illnesses couldn't be more different. A recent study of more than 3,000 people with schizophrenia suggested that thousands of genetic variations contribute to the disease. Equally problematic is the fact that one patient's form of schizophrenia may have different genetic and environmental causes from another's, says Owen. "These disorders are not really disorders. There's no such thing as schizophrenia. It's a syndrome. It's a collection of things psychiatrists have grouped together."

Model targets

Stephen Haggarty, a chemical neurobiologist at Massachusetts General Hospital in Boston, is tackling the genetic complexity of mental illness head-on. His team is creating neurons from patients with specific mutations implicated in schizophrenia, bipolar disorder and other conditions. Scientists do not know what most mutations linked to schizophrenia do to a cell, and iPSC models offer a way to find out, says Haggarty.

Despite these challenges, iPSC models of mental illness may be the best hope for identifying the fundamental defects that underlie these diseases - and ways to reverse them. Most antipsychotic drugs target the same dopamine receptor, and iPSC models could be "a way to find new treatments that are not more of the same", says Margolis.

Evan Snyder, a stem-cell biologist studying mental illness at the Sanford-Burnham Medical Research Institute in San Diego, says it will be a long slog before scientists identify differences in the neurons of psychiatric patients that are relevant to their disease. "We'd like to think that one can model a complex disease like schizophrenia in a dish, but we have to realize this is the ultimate in reductionism."

Nonetheless, he and other scientists are optimistic that, with enough scientists creating iPSCs from enough patients, real insight into the confounding diseases will follow. "We've got to start somewhere," says Snyder.

This article is reproduced with permission from the magazine Nature. The article was first published on April 13, 2011.