Widespread Synapse Loss Maps Schizophrenia's Brain Effects

Schizophrenia synapse loss reshapes the brain’s wiring

When researchers turned a high‑resolution PET tracer toward the brains of people diagnosed with schizophrenia, they saw a striking pattern: synaptic density was reduced across both hemispheres, but the left side carried a disproportionate share of the damage. This systematic loss of connections, now called schizophrenia synapse loss, maps onto the brain’s intrinsic circuitry and hints at why the disorder disrupts thought, perception, and emotion. The finding emerged from a multimodal imaging effort that combined synaptic vesicle glycoprotein 2A (SV2A) PET, structural MRI, and sophisticated network analyses. By charting where synapses disappear, scientists are beginning to trace the cascade that leads from genetic risk to the clinical picture that clinicians confront every day.

Mapping the deficit: how the imaging study was built

The Rutgers team recruited a balanced sample of individuals with schizophrenia, matched controls, and a small group of first‑degree relatives. Participants received a scan with the radioligand [¹¹C]UCB‑J, which binds to SV2A, a protein present on virtually every presynaptic terminal. Because SV2A density correlates with the number of functional synapses, the tracer provides a proxy for synaptic abundance in vivo. After correcting for partial‑volume effects and normalizing to a cerebellar reference region, the researchers generated voxelwise maps of binding potential across the cortex.

To move from raw images to a description of network‑level loss, they overlaid the PET data onto a standard connectome derived from diffusion‑weighted imaging. This allowed them to ask whether synapse loss respects the brain’s wiring architecture. Statistical maps revealed a widespread reduction (p < 0.05, family‑wise error corrected) that was strongest in left frontal and temporal cortices, regions traditionally implicated in language and executive function. The pattern was not random; it followed the brain’s modular organization, suggesting that disease processes travel along established pathways.

Why the left hemisphere shows the greatest vulnerability

The left‑dominant profile aligns with long‑standing observations that schizophrenia often impairs language‑related functions. The imaging data showed that the left inferior frontal gyrus (IFG) and superior temporal gyrus (STG) had the lowest SV2A binding, while homologous right‑side regions were comparatively spared. One plausible explanation comes from the brain’s chemical gradients. Neurotransmitter systems such as dopamine and glutamate are asymmetrically distributed, and the left prefrontal cortex receives denser dopaminergic input. Excessive dopamine signaling can destabilize glutamatergic synapses, making them more prone to pruning.

Another angle comes from developmental neurobiology. During adolescence, the left hemisphere undergoes a protracted period of synaptic refinement, a window that overlaps with the typical age of schizophrenia onset. If a genetic or environmental insult interferes with this fine‑tuning, the left side may bear the brunt of the resulting loss. The imaging study therefore provides a structural correlate for the “left‑hemisphere bias” that clinicians have described for decades.

Molecular underpinnings: microglia, complement C4, and synaptic pruning

Post‑mortem work has repeatedly shown that microglia—the brain’s resident immune cells—are more activated in schizophrenia. In a co‑culture model of human neurons and microglia‑like cells, researchers demonstrated that microglia can drive synaptic loss when exposed to schizophrenia‑linked signals (Breitmeyer et al., 2023). The model showed increased engulfment of synaptic boutons and a reduction in dendritic spine density, directly linking microglial activity to the synapse deficit observed in vivo.

A genetic candidate that may trigger this microglial response is the complement component 4 (C4) gene. Overexpression of C4, a risk allele identified in genome‑wide studies, leads to excessive tagging of synapses for removal. In a recent preprint, Phadke et al. (2023) reported that C4 overexpression impairs AMPA‑type glutamate receptor trafficking via an intracellular pathway involving SNX27, resulting in synaptic loss that does not depend on the classic complement receptor 3 route. This suggests that C4 can drive synapse elimination through multiple mechanisms, both extracellular (microglial pruning) and intracellular (receptor degradation).

Both lines of evidence converge on the idea that schizophrenia synapse loss is not a passive decay but an active, regulated process gone awry. The left‑frontal hotspot identified by PET may reflect a region where C4‑mediated tagging and microglial engulfment intersect with a high baseline of synaptic turnover.

Electrophysiology meets imaging: loss of synaptic gain

Electroencephalography (EEG) studies have long reported reduced auditory‑evoked gamma oscillations in schizophrenia, a phenomenon interpreted as diminished synaptic gain—the efficiency with which postsynaptic neurons respond to input. Computational modeling of simultaneous EEG and functional MRI data indicated a consistent loss of pyramidal‑cell synaptic gain across patients (Adams et al., 2021). While that work did not image synapses directly, its inference of reduced gain dovetails with the PET finding of fewer presynaptic terminals. Together, the modalities suggest that structural loss translates into functional deficits that can be captured by non‑invasive recordings.

Synaptic density PET in the broader neuropsychiatric context

Synaptic loss is not unique to schizophrenia; similar patterns appear in Alzheimer’s disease, where tau pathology spreads along network connections (Luan et al., 2025). However, the distribution in schizophrenia follows a different template, respecting the brain’s intrinsic connectivity rather than the hierarchical spread seen in neurodegeneration. This distinction matters for interpreting PET data: a uniform drop in SV2A binding would point to a global synaptic crisis, whereas the patterned loss observed here implicates network‑specific mechanisms.

Other mental‑health conditions show overlapping but distinct signatures. For example, functional MRI studies of early‑stage psychosis have identified altered spatial variation in connectivity that may serve as early biomarkers (Georgia State University press release, 2026). The convergence of PET and fMRI signatures could eventually allow clinicians to differentiate between schizophrenia, bipolar disorder, and major depression based on how and where synapse loss occurs.

Implications for therapeutic strategies

Understanding that synapse loss follows a predictable map opens several avenues for intervention. One approach targets the complement cascade. Small‑molecule inhibitors of C4 or downstream effectors could blunt inappropriate tagging of synapses, preserving connectivity in vulnerable regions like the left IFG. Another strategy focuses on microglial modulation. Drugs that shift microglia from a pro‑pruning to a neuroprotective phenotype are already in early trials for other neuroinflammatory disorders; repurposing them for schizophrenia could test whether dampening microglial activity restores SV2A binding.

On the synaptic side, enhancing AMPA‑receptor trafficking might counteract the intracellular loss mechanism described by Phadke et al. (2023). Compounds that stabilize GluR1 at the membrane or boost SNX27 function could theoretically rescue synaptic density. Finally, neuromodulatory techniques such as transcranial magnetic stimulation (TMS) applied to the left frontal cortex might promote synaptogenesis, a hypothesis that could be evaluated with repeat SV2A PET scans.

What the map does not yet tell us

Despite the richness of the imaging data, several questions remain. First, the cross‑sectional design cannot determine whether synapse loss precedes symptom onset or follows chronic illness. Longitudinal PET studies of high‑risk youths would be needed to resolve this chicken‑and‑egg problem. Second, the PET tracer measures presynaptic terminals but cannot distinguish between excitatory and inhibitory synapses; the balance between the two may be critical for network stability. Third, the sample size, while larger than many prior PET investigations, still limits the ability to detect subtle regional effects or interactions with medication status.

Finally, the left‑hemisphere bias raises the issue of handedness and language dominance, variables that were not systematically recorded in the original cohort. Future work that stratifies participants by these factors could clarify whether the observed pattern reflects a universal neurobiological feature of schizophrenia or a subgroup effect.

Where the field is heading

Researchers are already building on the current map. A recent preprint combined SV2A PET with single‑cell RNA sequencing of post‑mortem tissue, linking regional synapse loss to cell‑type‑specific expression of complement genes (Chopra et al., 2025). Such multimodal studies promise to bridge the gap between macro‑scale imaging and molecular pathology. In parallel, advances in ultra‑high‑field MRI may allow direct visualization of dendritic spines, offering a complementary view of synaptic health.

Beyond basic science, the ultimate goal is to translate these findings into clinical tools. If a PET scan can identify patients with pronounced left‑frontal synapse loss, clinicians might prioritize interventions that target language and executive function, such as cognitive remediation or targeted neuromodulation. Conversely, patients with a more diffuse pattern might benefit from systemic anti‑inflammatory approaches.

In sum, the systematic mapping of schizophrenia synapse loss reshapes how we think about the disorder’s architecture. By tying together imaging, genetics, and cellular biology, the work points to a cascade that begins with molecular tags, proceeds through microglial pruning, and ends in a patterned loss of connectivity that mirrors the clinical syndrome. The challenge now is to intervene at the right point in that cascade, before the brain’s wiring is irrevocably altered.

References

Adams. (2021). Computational Modeling of Electroencephalography and Functional Magnetic Resonance Imaging Paradigms Indicates a Consistent Loss of Pyramidal Cell Synaptic Gain in Schizophrenia. Biological Psychiatry. https://doi.org/10.1016/j.biopsych.2021.07.024
Breitmeyer. (2023). Regulation of synaptic connectivity in schizophrenia spectrum by mutual neuron-microglia interaction. Communications Biology. https://doi.org/10.1038/s42003-023-04852-9
Luan. (2025). Synaptic loss pattern is constrained by brain connectome and modulated by phosphorylated tau in Alzheimer’s disease. Nature Communications. https://doi.org/10.1038/s41467-025-61497-4
Phadke. (2023). C4 induces pathological synaptic loss by impairing AMPAR trafficking. bioRxiv. https://doi.org/10.1101/2023.09.09.556388
Chopra. (2025). Network-based Molecular Constraints on in vivo Synaptic Density Alterations in Schizophrenia. medRxiv. https://doi.org/10.1101/2025.03.22.25324465

Paper data via Semantic Scholar.