In new research, a team led by BBRF grantees provides a vivid example of how the longstanding effort to identify risk genes for psychiatric illness can pay off in tangible ways, informing research on the biological underpinnings of disorders such as schizophrenia and psychosis that could help generate novel diagnostics and therapeutics.

The leader of the team, Akira Sawa, M.D., of the Johns Hopkins University School of Medicine, is a 3-time BBRF grantee who has championed an approach to research called “reverse translation.” It is work that “always starts in the clinic—with our patients,” Dr. Sawa says. “We want to discover how our clinical observations can be explained in terms of biological mechanisms, or how they can be understood in a biological context.” The aim is to then translate this new scientific insight back into the clinic—in the form of biomarkers to aid diagnosis, new targets for improved treatments, or new treatments themselves.

In the newly reported research, Dr. Sawa, a member of BBRF’s Scientific Council and 2011 BBRF Distinguished Investigator and 2004 and 2002 Young Investigator, mobilized data he and colleagues had collected over a 5-year period from a large cohort of early psychosis patients at the Johns Hopkins Schizophrenia Center, which Dr. Sawa directs. For purposes of the study, “early psychosis” means individuals who experienced a first psychotic episode within 2 years of their recruitment to the study. Among the six additional BBRF grantees on the team was first author of their paper appearing in JAMA Psychiatry, Kun Yang, Ph.D., a 2019 BBRF Young Investigator.

Currently, psychosis and psychotic disorders are diagnosed mainly via clinical interviews and psychiatric evaluation. A “psychotic break,” or first episode of psychosis, is in many instances the initial event leading to a diagnosis of schizophrenia, which is made, sometimes, years after. The period immediately following—and arguably, immediately preceding—these first episodes is deemed extremely important by doctors, in large part because data indicates that those who receive treatment earliest have a better chance of a less-severe disease course than those treated later on.

One problem is that there is no standard, objective method of predicting the course of a first-episode psychosis patient. A key issue is whether such an individual will respond to antipsychotic medicines, or respond weakly or even not at all. As in many other instances in psychiatric care, medications must be prescribed on a trial-and-error basis. A further complication is that psychosis is “heterogeneous”— patients report a diversity of symptoms, and likely fall into a number of subsets that might conceivably be treated somewhat differently to achieve best results. Right now there is no objective way to do this.

Dr. Sawa’s team made use of the results of a recent comprehensive study of genome-based risk for schizophrenia. It was performed by the Psychiatric Genomics Consortium, a multinational academic partnership that pools data from hundreds of thousands of individuals to achieve statistical power that single-cohort studies cannot possibly match. In this case, in a study of over 300,000 people, 76,755 of whom were schizophrenia patients and the remainder controls, the Consortium identified commonly occurring DNA variations in 287 locations across the genome that are significantly more likely to occur in people with schizophrenia compared with healthy individuals.

Among these 287 genome-based risk factors for schizophrenia were DNA variations occurring in a gene called ACE, which encodes the angiotensin-converting enzyme. In the body, ACE converts the angiotensin I hormone to angiotensin II, which causes blood vessels to narrow and blood pressure to increase. Widely prescribed drugs called ACE inhibitors help to lower blood pressure in people with hypertension and improve survival following heart attacks.

What does this have to do with schizophrenia and psychotic disorder? That is precisely the kind of biological mystery posed by the identification of schizophrenia-related risk genes like the ACE gene.

This is where reverse translation comes into the picture. In their longstanding efforts to treat and observe patients, including early psychosis patients, at the Johns Hopkins Schizophrenia Center, Dr. Sawa’s team had access to a wide range of data types. These included genome data for patients that indicated genetic variants in the ACE gene; levels of ACE proteins in the cerebrospinal fluid (CSF) and blood serum; and the ability to analyze the activity of ACE enzymes in the serum samples.

For the present study, the researchers established a cohort of 200 individuals that included 122 of their early psychosis patients and 78 healthy controls. The typical participant was in their late 20s, with none older than 35 or younger than 18. Importantly (as it would turn out), within the patient group, about 1 in 4 was considered “treatment-resistant,” based on their medication history.

Analysis yielded several major findings. The most striking was that patients with early-stage psychosis had lower levels of ACE protein in both their CSF and serum. By leveraging the rich information from the Psychiatric Genomic Consortium results, the team discovered that lower ACE protein levels in serum correlated with an elevated genetic burden tied to the ACE gene. They believe this “may represent a primary molecular mechanism underlying pathology in psychotic disorders.” But how, exactly? Making this question even more pressing, the team did not observe any impairment of the function the ACE enzyme performs: that of helping to maintain healthy blood pressure. In other words, despite the lower levels of ACE proteins in early psychosis patients, the activity level of ACE enzymes in their serum did not change.

This led the team to explore the possibility that the ACE enzyme has other functions, in addition to its primary (“canonical”) function of regulating blood pressure, that might be impacting schizophrenia risk. There are other famous examples of proteins performing “non-canonical” roles; they are sometimes called “moonlighting proteins.” An important protein called PTEN, for example, that normally acts to remove phosphate groups from other molecules, has been found to have other functions that may be involved in autism pathology.

“It is conceivable that ACE plays a role in the pathophysiology of psychotic disorders, likely through its non-enzymatic, non-canonical function,” the team speculates. Here again, data from their patients enables them to advance a hypothesis. While the patients in the study cohort had reduced levels of ACE protein in their CSF and serum, and those with greater ACE-gene genetic burden had still less, it was also true that both of these relationships was even more intense in patients who were treatment-resistant. “Compared with patients with non-treatment resistant psychosis, those with treatment-resistant psychosis exhibited further reduced ACE protein levels and increased ACE-related genetic burden.”

It is possible, the team said, that the ACE protein has a “non-canonical” role in inflammation. Future experiments can now focus on this and other possible links between the protein and pathology that may be involved in early psychosis, and perhaps, some have suggested, Alzheimer’s disease. Further application of “reverse translation” may elucidate such a novel mechanism through investigations of animal models and human cell models. Meantime, ACE emerges from this study “as a promising candidate” as a noninvasive biomarker for psychotic disorder, and one, importantly, that can be measured on the basis of a simple blood draw. (This was why it was significant that ACE proteins levels were lower not only in CSF but also in serum samples from early psychosis patients.)

“These consistent changes across both biological fluids (CSF and blood) combined with genetic validation, position ACE as a uniquely promising biomarker candidate for psychotic disorders, especially in cases of treatment-resistance,” the team concluded.

The team also included: Lindsay Hayes, Ph.D., 2026 BBRF Young Investigator; Thomas W. Sedlak, M.D., Ph.D., 2006 BBRF Young Investigator; Jennifer M. Coughlin, M.D., 2013 BBRF Young Investigator; Frederick C. Nucifora Jr., DO, Ph.D., 2013 and 2010 BBRF Young Investigator; and Nicola G. Cascella, M.D., 2000 BBRF Young Investigator.