A missing protein in spinal fluid points to a new way to treat schizophrenia's cognitive symptoms

Northwestern University / Neuron

Schizophrenia medications can quiet hallucinations and dampen delusions. What they cannot do — what no approved drug does — is treat the cognitive symptoms that often determine whether a patient can hold a job, live independently, or maintain relationships. Disorganized thinking, impaired working memory, difficulty with planning and executive function: these are the symptoms that drive long-term disability, dependence on family, homelessness, and, in some cases, suicide.

A study published March 19 in Neuron offers a potential path forward. Researchers at Northwestern University's Feinberg School of Medicine identified a protein that is depleted in the cerebrospinal fluid of people with schizophrenia, created a synthetic version of it, and showed that injecting it into the brains of mice with a genetic model of the disorder corrected both abnormal brain circuit activity and the behavioral problems linked to cognitive dysfunction.

Finding the signal in spinal fluid

The research team, led by Peter Penzes, professor of neuroscience, pharmacology, and psychiatry at Feinberg, examined cerebrospinal fluid from more than 100 schizophrenia patients and healthy controls. They identified a previously unknown, freely circulating form of a brain protein called Cacna2d1 (calcium voltage-gated channel auxiliary subunit alpha2delta-1).

In patients with schizophrenia, levels of this circulating Cacna2d1 signal were significantly reduced compared to controls. The protein normally plays a role in regulating synaptic connections and neural circuit excitability. When it is depleted, brain circuits become overactive — a state consistent with the excitation-inhibition imbalance that neuroscientists have long suspected underlies schizophrenia's cognitive symptoms.

This is notable partly because psychiatry has very few biomarkers. Diabetes has blood sugar. Heart disease has cholesterol. Schizophrenia has a clinical interview. The identification of a measurable molecular signal associated with the disorder — particularly one tied to a specific biological pathway — opens the door to something psychiatry badly needs: objective diagnostic tools and targeted treatment selection.

SEAD1: from biomarker to therapeutic candidate

The team did not stop at diagnosis. They synthesized a version of the protein, which they named SEAD1, and tested it in mice carrying a genetic duplication (16p11.2 duplication) that models schizophrenia. This particular genetic variant is associated with a tenfold increase in schizophrenia risk in humans.

A single injection of SEAD1 into the animals' brains produced two results. First, it corrected the abnormal circuit activity — the overexcitation that had been running unchecked. Second, it reversed behavioral deficits linked to cognitive dysfunction in the mice. And it did so without causing sedation or reduced movement, side effects that plague many existing psychiatric medications.

"Our treatment reopens a crucial window to rewire connections in adult brains," said first author Marc Dos Santos, research assistant professor of neuroscience at Feinberg. "The lack of brain plasticity is believed to be a key factor in the development of symptoms in schizophrenia. Reforming synapses could also be beneficial for other mental disorders, such as depression."

The tandem approach: biomarker plus drug

What makes this work distinctive is the coupling of a biomarker with a therapeutic derived from the same biology. The Cacna2d1 signal in cerebrospinal fluid could identify patients with the specific molecular deficit that SEAD1 is designed to correct. Those patients would be the ones most likely to respond to the treatment.

This matters enormously for clinical trials. Psychiatric drug development has a dismal track record, not necessarily because the drugs do not work, but because the patient populations enrolled in trials are biologically heterogeneous. A drug that helps the 30% of patients with a specific molecular profile will look ineffective in a trial where it is given to everyone. By selecting patients based on a biomarker, the researchers can concentrate the signal and dramatically improve the odds of a trial succeeding.

Penzes described the vision: "The next step for us would be to develop a blood biomarker to identify a subset of schizophrenia patients who can respond to this treatment, and then we can give them this peptide — almost like a regular injection that you can give once a week."

From mouse brain to human patient: the distance remaining

The caveats here are substantial and worth stating plainly. This is a mouse study. The injection was delivered directly into the brain, not systemically. The team does not yet know how long the therapeutic effects last. And the 16p11.2 duplication mouse model, while genetically relevant, does not fully recapitulate the complexity of human schizophrenia, which involves multiple genetic and environmental risk factors interacting across decades of brain development.

Moving from a single intracranial injection in mice to a practical human therapy requires solving several hard problems: developing a formulation that crosses the blood-brain barrier or can be delivered peripherally, establishing dosing and duration, conducting toxicology studies, and running clinical trials in a patient population that is notoriously difficult to recruit and retain.

The blood biomarker that Penzes envisions — a simpler test than a spinal tap — does not yet exist. The current study measured Cacna2d1 in cerebrospinal fluid, which requires a lumbar puncture. That is acceptable for research but impractical for routine clinical screening.

Dos Santos acknowledged the uncertainty about duration: the team plans to study how long the effects persist in future experiments. The research group is now optimizing SEAD1 for future clinical trials in patients with 16p11.2 duplication syndrome specifically, which is a narrower population than schizophrenia as a whole.

Why cognitive symptoms matter more than clinicians have acknowledged

The focus on cognitive symptoms is itself a corrective. For decades, schizophrenia treatment has been defined by its ability to suppress positive symptoms — hallucinations, delusions, paranoia. The antipsychotics that accomplish this, from chlorpromazine in the 1950s to modern atypicals, transformed psychiatric care. But they left the cognitive core of the disorder largely untouched.

Schizophrenia affects roughly 0.5% of the global population — about two million people in the United States alone. For many of them, the inability to think clearly, plan, and organize is what prevents reintegration into society, even when hallucinations and delusions are controlled. A treatment that addresses this dimension of the disorder would fill a gap that has been open for more than half a century.

Whether SEAD1 or something derived from it will be that treatment remains to be seen. But the framework — a molecular biomarker linked to a targetable pathway and a candidate therapeutic — is the kind of foundation that drug development requires.