Schizophrenia Research Forum - A Catalyst for Creative Thinking

What Lies Between—Extracellular Matrix and Schizophrenia

8 February 2010. When searching for clues to schizophrenia in the brain, it's important to look in the seemingly open spaces between neurons. That is the message from two new studies of postmortem brain tissue that implicate certain extracellular matrix (ECM) molecules in schizophrenia. One study, published in the Archives of General Psychiatry on February 1 focuses on ECM molecules known as chondroitin sulfate proteoglycans (CSPGs) and the glial cells containing them. The other, to be published in PNAS online, centers on reelin and its effect on neurons. Both make the case that—though somewhat removed from the critical goings-on of neurons—these modest molecules matter.

From their place in between neurons, a variety of ECM molecules have a hand in building and maintaining the brain. Not only do they form a scaffold that provides brain cells a structurally sound place to sit, they also guide cells to their proper locations during development. Perturbing these molecules can derail neuronal migration, miswire circuits, and even change how neurons respond to experience—consistent with the disruptions in connectivity and synaptic function suspected in schizophrenia. Polymorphisms in the gene encoding one type of CSPG (PTPRZ1) and in the gene for reelin (RELN) have previously been associated with the disorder. Both studies aimed to check the integrity of these ECM molecules in schizophrenia.

Disrupted nets
Sabina Berretta and colleagues at McLean Hospital, Belmont, Massachusetts, and Harvard Medical School focused on CSPGs, a collection of part-protein, part-carbohydrate molecules that are secreted by neurons and glia into the extracellular space. There, CSPGs interlock with other ECM molecules to form scaffolds or weblike structures called perineuronal nets (PNN) that tightly surround individual neurons and are thought to restrict neuronal plasticity. Looking in the amygdala and nearby entorhinal cortex, memory and emotion centers known to be affected in schizophrenia, the researchers counted the number of glia cells staining positively for CSPGs in brain tissue from 11 people with schizophrenia, 11 with bipolar disorder, and 15 controls.

Led by first author Harry Pantazopoulos, the researchers found a massive increase of CSPG-positive glial cells in schizophrenia—ranging from a 419 percent (P <.001, cortical nucleus of the amygdala) to a 1,560 percent increase (P = .03, layer III of the lateral entorhinal cortex) over control values, depending on the specific brain areas compared. This increase was not due to a multiplication of glia related to brain degeneration, because the number of glia cells, marked by positive staining for glial fibrillary acidic protein (GFAP) was not significantly different between schizophrenia and controls. Instead, the researchers propose that either the number of cells producing CSPGs went up or the amount produced by the cells was boosted. Consistent with this, they measured increases in mRNA expression for the individual CSPG molecules—versican, neurocan, brevican, aggrecan, and phosphocan (PTPRZ1)—in schizophrenia tissue relative to controls. Brain tissue from bipolar subjects did not show this anomalous increase in CSPG-containing glial cells.

On the flipside, there were fewer PNNs in schizophrenia—a 62 percent decrease (P = .01) relative to controls in the lateral nucleus of the amygdala and a 65 percent decrease (P = .004) in the lateral entorhinal cortex. This was not because there were fewer neurons available to be enwrapped by PNNs: the total number of neurons positive for parvalbumin—a marker for interneurons that are specifically targeted by PNNs—was not significantly different between schizophrenia and controls. The researchers looked for, but did not find, explanations for these CSPG anomalies in a number of factors that differed between the subjects, including postmortem time interval, duration of illness, or antipsychotic treatment.

Reelin back on the map?
While this study focused on CSPGs and glia, the second study, led by Erminio Costa (recently deceased, see
SRF related news story) at the University of Illinois in Chicago, focused on the well-known ECM molecule reelin and its potential effects on neurons in the cerebellum, a part of the brain involved in coordinating movement and in cognition. Just a decade ago, reelin was considered a highly promising lead in schizophrenia pathology, thanks in part to work from the group of Costa and his colleague Alessandro Guidotti. The field has shown less interest in the molecule in recent years, though the gene RELN retains a positive meta-analysis in SchizophreniaGene at the time this article was written.

With brain tissue from 13 schizophrenia, 17 bipolar disorder patients, and 24 controls, the researchers measured the density of Purkinje cells—the large interneurons with cell bodies arranged in orderly rows—in the anterior lobe of the cerebellum. First author Ekrem Maloku and researchers found about one fewer Purkinje cell per millimeter in both schizophrenia and bipolar patients than the density found in controls, similar to what has been reported in other regions of the cerebellum. This modest change amounted to a 20 percent decrease in density, which could disrupt the tight arrangement of circuitry within the cerebellum.

Because of reelin's role in neuronal migration and maturation, and because reelin is abnormally low in the cerebellum of schizophrenia and bipolar patients, the researchers next asked whether abnormal reelin levels might also be associated with the decrease in Purkinje cell density. When looking at the cerebellum in a subset of cases displaying low Purkinje cell density—three brains for schizophrenia and three for bipolar disorder—the researchers found an accompanying reduction of reelin mRNA in granule cells compared to three controls. Granule cells form glutamatergic synapses onto the elaborate dendritic arbors of Purkinje cells; if granule cells are not secreting enough reelin, this could interfere with Purkinje cell migration or synapse maturation. Consistent with this, decreasing reelin levels in genetically engineered mice resulted in a 10 percent decrease in Purkinje cells than were found in wild-type mice.

Though future studies will have to work out the precise effects of the ECM molecule abnormalities detected in these reports, it's clear that these molecules can touch many aspects of brain development and function. This means ECM molecules may provide a way to account for the myriad brain anomalies documented in schizophrenia that cut across different brain regions, cell types, and synapses, including GABAergic, glutamatergic, and dopaminergic circuits.—Michele Solis.

Pantazopoulos H, Woo TW, Lim MP, Lange N, Berretta S. Extracellular matrix-glial abnormalities in the amygdala and entorhinal cortex of subjects diagnosed with schizophrenia. Arch Gen Psychiatry. 2010 Feb; 67: 155-166. Abstract

Maloku E, Covelo IR, Hanbauer I, Guidotti A, Kadriu B, Hu Q, Davis JM, Costa E. Lower number of cerebellar Purkinje neurons in psychosis is associated with reduced reelin expression. Proc Natl Acad Sci USA. 2010 Feb.

Comments on Related News

Related News: Obituary—Erminio Costa, M.D. (1924-2009), University of Illinois at Chicago

Comment by:  Artyom Tikhomirov
Submitted 2 December 2009
Posted 3 December 2009

My condolences. May he rest in peace. Let his outstanding work be furthered by his pupils and thus let the memory of him live on.

View all comments by Artyom Tikhomirov

Related News: Obituary—Erminio Costa, M.D. (1924-2009), University of Illinois at Chicago

Comment by:  Lorna Role
Submitted 22 December 2009
Posted 23 December 2009

Erminio will be missed by many, including me. My first encounter with Erminio was as a graduate student attending my first scientific meeting— a Catecholamine Gordon conference at Asilomar. His enthusiasm, firey disposition and commitment to science was an inspiration! My condolences to his family and special thoughts to Sandro ...

View all comments by Lorna Role

Related News: Research Roundup: PV Interneurons and Neural Circuit (Dys)Function

Comment by:  Takao Hensch
Submitted 15 February 2014
Posted 15 February 2014

Our Conte Center is focused on the transcriptome, connectome, and plasticity of PV cells as the neurodevelopmental basis for mental illness. Their maturational state dictates the degree of plasticity in developmental critical periods, and now we know, from Donato et al., in adult learning. Once plasticity is opened by PV cells' function, it closes when they mature ("high PV" state), including the tightening of the perineuronal nets (PNNs) around them.

In schizophrenia, PV cells may remain in the "low PV" weak PNN state for some time longer than normal, suggesting, interestingly, that developmental plasticity may be prolonged (i.e., neural circuits fail to stabilize when they normally should). PV cell maturation may potentially be controlled by Otx2 secreted from the choroid plexus, which would link enlarged ventricles to impaired PV cells in the brain in schizophrenia (see Spatazza et al., 2013).


Spatazza J, Lee HH, Di Nardo AA, Tibaldi L, Joliot A, Hensch TK, Prochiantz A. Choroid-plexus-derived Otx2 homeoprotein constrains adult cortical plasticity. Cell Rep. 2013 Jun 27;3(6):1815-23. Abstract

View all comments by Takao Hensch