Penn Medicine researchers are the first to discover two distinct neuroanatomical subtypes of schizophrenia after analyzing the brain scans of over 300 patients.
The first type showed lower widespread volumes of gray matter when compare to healthy controls, while the second type had volumes largely similar to normal brains.
The findings, published Thursday in the journal Brain, suggest that, in the future, accounting for these differences could inform more personalized treatment options.
“Numerous other studies have shown that people with schizophrenia have significantly smaller volumes of brain tissue than healthy controls.
However, for at least a third of patients we looked at, this was not the case at all — their brains were almost completely normal,” said principal investigator Christos Davatzikos, PhD, the Wallace T. Miller Professor of Radiology in the Perelman School of Medicine at the University of Pennsylvania.
“In the future, we’re not going to be saying, ‘This patient has schizophrenia,’ We’re going to be saying, ‘This patient has this subtype’ or ‘this abnormal pattern,’ rather than having a wide umbrella under which everyone is categorized.”
Schizophrenia is a poorly understood mental disorder that typically presents with hallucinations, delusions, and other cognitive issues — though symptoms and responses to treatment vary widely from patient to patient.
Up until now, attempts to study the disease, by comparing healthy to diseased brains, has neglected to account for this heterogeneity, which Davatzikos says has muddled research findings and undermined clinical care.
To better characterize the distinct brain differences within the schizophrenia patient population, Davatzikos established a research consortium that spanned three continents — the United States, China, and Germany.
The international cohort of study participants included 307 schizophrenia patients and 364 healthy controls, all of whom were 45-years-old or younger.
Davatzikos and engineering colleagues then analyzed the brain scans using a machine learning method developed at Penn called HYDRA (Heterogeneity Through Discriminative Analysis).
The approach helps to identify “true disease subtypes” by limiting the influence of confounding variables, such as age, sex, imaging protocols, and other factors, according to the study authors.
“This method enabled us to sub-categorize patients and find how they differed from the controls, while allowing us, at the same time, to dissect this heterogeneity and tease out multiple pathologies, rather than trying to find a dominant pattern,” Davatzikos said.
After applying this machine learning method to the brain images, the researchers found that 115 patients with schizophrenia, or nearly 40 percent, did not have the typical pattern of reduced gray matter volume that has been historically linked to the disorder.
In fact, their brains showed increases of brain volume in the middle of the brain, in an area called the striatum, which plays a role in voluntary movement.
When controlling for differences in medication, age, and other demographics, the researchers could not find any clear explanation for the variation.
“The subtype 2 patients are very interesting, because they have similar demographic and clinical measures with subtype 1, and the only differences were their brain structures,” said Ganesh Chand, PhD, a lead author and postdoctoral researcher in the radiology department at Penn.
There are a variety of antipsychotic medications available to manage the symptoms of schizophrenia, but how they will affect a particular patient — both positively or negatively — is often a shot in the dark, according to study co-senior author Daniel Wolf, MD, PhD, an associate professor of Psychiatry at Penn.
“The treatments for schizophrenia work really well in a minority of people, pretty well in most people, and hardly at all in a minority of people. We mostly can’t predict that outcome, so it becomes a matter of trial and error,” Wolf said.
“Now that we are starting to understand the biology behind this disorder, then we will hopefully one day have more informed, personalized approaches to treatment.”
As to why an entire subset of patients with schizophrenia have brains that resemble healthy people, Davatzikos is not willing to speculate.
In a large clinical study, 60 percent of patients with schizophrenia (subtype 1) had decreased gray matter volumes throughout the brain compared to healthy people, which is the typical pattern seen in those with this disorder.
However, researchers found that over a third of schizophrenia patients (subtype 2) did not present with this pattern.
These brains had increased volumes of gray matter in the basal ganglia, but were otherwise similar to healthy controls. The image is credited to Christos Davatzikos et al.
“This is where we are puzzled right now,” Davatzikos said. “We don’t know. What we do know is that
studies that are putting all schizophrenia patients in one group, when seeking associations with response to treatment or clinical measures, might not be using the best approach.”
Future research, he said, will provide a more detailed picture of these subtypes in relation to other aspects of brain structure and function, clinical symptoms, disease progression, and etiology.
Additional Penn authors include: Guray Erus, Aristedidis Sotiras, Erdem Varol, Dhivya Srinivasan, Jimit Doshi, Raymond Pomponio, Taki Shinohara, Ruben C. Gur, Raquel E. Gur, Russell T. Shinohara, Haochang Shou, Yong Fan, and Theodore D. Satterthwaite.
Funding: This research was funded by the National Institutes of Health grant R01MH112070 and by the PRONIA project as funded by the European Union 7th Framework Program grant 602152.
Little is known about the molecular pathogenesis of schizophrenia, possibly because of unrecognized heterogeneity in diagnosed patient populations. We analyzed gene expression data collected from the dorsolateral prefrontal cortex (DLPFC) of post-mortem frozen brains of 189 adult diagnosed schizophrenics and 206 matched controls.
Transcripts from 633 genes are differentially expressed in the DLPFC of schizophrenics as compared to controls at Bonferroni-corrected significance levels. Seventeen of those genes are differentially expressed at very high significance levels (<10-8 after Bonferroni correction).
The findings were closely replicated in a dataset from an entirely unrelated source. The statistical significance of this differential gene expression is being driven by about half of the schizophrenic DLPFC samples, and importantly, it is the same half of the samples that is driving the significance for almost all of the differentially expressed transcripts.
Weighted gene co-expression network analysis (WGCNA) of the schizophrenic subjects, based on the transcripts differentially expressed in the schizophrenics as compared to controls, divides them into two groups.
“Type 1” schizophrenics have a DLPFC transcriptome similar to that of controls with only four differentially expressed genes identified.
“Type 2” schizophrenics have a DLPFC transcriptome dramatically different from that of controls, with 3529 expression array probes to 3092 genes detecting transcripts that are differentially expressed at very high significance levels.
These findings were re-tested and replicated in a separate independent cohort, using the RNAseq data from the DLPFC of an independent set of schizophrenics and control subjects. We suggest the hypothesis that these striking differences in DLPFC transcriptomes, identified and replicated in two populations, imply a fundamental biologic difference between these two groups of diagnosed schizophrenics, and we propose specific paths for further testing and expanding the hypothesis.
*-*-*-
Almost half a century ago, Fred Plum1 called schizophrenia “the graveyard of neuropathologists”, and in many ways the situation has not appreciably changed: In spite of decades of anatomic, histologic, and molecular inroads, little progress has been made elucidating the pathobiology of schizophrenia.
A longstanding hypothesis to explain this lack of progress is that schizophrenia is a heterogeneous disease and that meaningful results have been obscured in studies which pool data from biologically different patients. Two publicly available sources of molecular data were used to test that hypothesis.
The first dataset was generated by scientists in the Clinical Brain Disorders Branch of the Intramural Research Program at the National Institute of Mental Health (NIMH), under the direction of Dr. Daniel Weinberger; it consists of Illumina HumanHT-12 v4 expression array data from the dorsolateral prefrontal cortex (DLPFC) of post-mortem brains of almost a thousand patients with psychiatric disease (including schizophrenia and other diagnoses) and neurologically normal matched controls. Although those investigators have never published their analysis of that data, the data itself are publicly available (dbGaP study accession phs000979.v1.p1).
The second relevant dataset contains RNAseq data from post-mortem DLPFC collected by the CommonMind Consortium (CMC) and made publicly available through their website2.
We show first that the schizophrenics in the NIMH expression array dataset are clearly of two distinct types: “type 1” patients have a DLPFC transcriptome very similar to that of the controls, whereas “type 2” patients have a dramatically different DLPFC transcriptome with several thousand genes differentially expressed compared to the controls.
We then replicate that observation in the CMC RNAseq dataset, showing that the same genetic subsets define the same two patient subtypes in this unrelated cohort. We characterize the composition of the two subtypes, and then propose a specific set of targeted studies that can strengthen or weaken the findings identified here.
The neuroanatomy and pathogenesis of schizophrenia
A common hypothesis regarding the pathogenesis of schizophrenia is that some combination of genetic predisposition and environmental events around the time of birth leads to an alteration in the newborn brain which predisposes the patient to the development of schizophrenia. From that perspective, the observation that NPY is the most downregulated gene and that both TAC1 and VIP are highly downregulated in the “type 2” DLPFC is particularly interesting. Neuropeptide Y (the product of the gene NPY), substance P (produced by proteolytic processing of the TAC1 gene product), and VIP are all well recognized as anatomic markers for particular subsets of inhibitory neocortical interneurons.
Neuropeptide Y is found in Martinotti cells, neurogliaform neurons, and a subset of the fast-spiking, parvalbumin-positive, basket cells13. The first two of those classes of cortical interneurons are well described. The Martinotti cell is a somatostatin-containing interneuron with an axonal plexus in layer 1, making synaptic contact with the spines of pyramidal neuron tuft dendrites. Neurogliaform neurons are non-VIP, 5HTR3A-positive, nitric oxide synthetase-positive neurons with short dendrites spreading radially in all directions and a wider, spherical, very dense axonal plexus. They are present in all layers of the cortex, but are especially prominent in layer 1 where they form the major neuronal component14. The NPY(+) basket cells are much less well characterized and ignored by many authors.
Substance P expression in the neocortex is largely restricted to a specific subclass of basket cells14. Given the down-regulation of both TAC1 and NPY in the DLPFC of schizophrenics, it is interesting to note that there is a reciprocal interaction between these neurons and the NPY-positive neurogliaform neurons22. There is, however, an immunohistochemical study using both light- and electron microscopy which describes a second class of large, intensely stained substance P-containing neurons which also express NPY23.
VIP is found in about 40% of the 5HT3aR-expressing interneurons. The majority of these neurons are layer 2/3 bipolar interneurons, but overall they are a heterogeneous class of neurons with a variety of morphologies and co-expressed markers14.
Our current understanding of the diversity of cortical interneurons is, however, far from complete and rapid advances in this field are expected with the availability of single-cell and single-nucleus RNAseq technology. If these interneurons in DLPFC are to blame for “type 2” schizophrenia, the diagnosis could relate either to a dearth of or an abnormality in these interneurons.
Forty-five percent of schizophrenics (“type 1”) have a relatively normal transcriptome in the DLPFC. This suggests that “type 1” schizophrenics have physiologically significant pathology elsewhere in their cortex, perhaps in the superior temporal or cingulate gyri. Identifying a cortical area where the transcriptome of the “type 1” but not “type 2” schizophrenics contains many differentially expressed genes would provide additional strong evidence for the physiologic importance of the distinction between “type 1” and “type 2” schizophrenics and potentially a major step forward in our understanding of the pathobiology of schizophrenia. (If further studies identify a cortical region with transcriptomic abnormalities in the “type 1” schizophrenics, it will be important to look for correlations between the clinical features of the schizophrenics and their molecular subtype. For example, if the “type 1” patients have molecular pathology in their superior temporal lobes, it would be important to know if those are also the patients with predominantly positive symptoms, including auditory hallucinations.)
Cytometry could test the first part of this hypothesis by comparing the number of NPY and TAC1 labeled neurons in the autoradiographic images of schizophrenic and normal DLPFC made public by the Allen Institute. A complementary approach would be to isolate an individual nucleus from DLPFC (as in the Nuc-Seq technique) and then perform quantitative rtPCR for NPY and TAC1.
This less expensive alternative to RNAseq would enable the study of a large enough sample of nuclei to generate meaningful data regarding these relatively rare interneurons. This represents a novel and potentially powerful new target for studies of schizophrenic etiology—and intimates the future possibility of predictive assays.
Because the current work provides a list of candidate genes, the initial screening of other cortical areas for alterations in the transcriptome of “type 1” schizophrenics could be an inexpensive qPCR-based study. This would be a potentially high-yield experiment. Fortunately, tissue from both the superior temporal and cingulate gyri from the specific patients included in this study is available from the Human Brain Collection Core of the NIMH intramural program.
Implications of increased statistical power and druggable targets
By analyzing the “type 1” and “type 2” schizophrenics separately, the subject pool is divided, yielding far fewer subject per group, and yet we showed a dramatic gain in statistical power to detect differentially expressed transcripts. Using all schizophrenics combined in a single group, 633 genes were identified as differentially expressed from controls. By contrast, once the heterogeneity of the schizophrenic population is recognized, the separate analysis of the two subtypes yielded more than 3200 genes: a five-fold increase in detection.
This increased statistical power and the scientific observations it makes possible are among the most scientifically and clinically important consequences of this work. An exhaustive review of the molecular biology of the differentially expressed genes and the possible implications of their differential expression in schizophrenic DLPFC is beyond the scope of this report. However, a cursory examination of the list of differentially expressed genes (Supplemental Table B3) reveals many potentially druggable targets.
Proteins known to be differentially expressed in DLPFC of the novel “type 1”/“type 2” populations identified here are targets of existing published PET probes, enabling the “type 1”/“type 2” distinction to be studied in diagnostics of living patients (see: hyperlink https://www.brainengineering.org/publications/2019/5/1/schizophreniaclinicaldiagnostic).
Source:
University of Pennsylvania
