Thomas: A Ban: In historical perspective. Peralta, Cuesta and their associates’ findings on the highest familiality of Leonhard’s classification in polynosologic study


Larry Stein’s comment

Is schizophrenia the price that Homo sapiens pays for abstract problem-solving?     Are the negative/deficit symptoms, in particular, under strong genetic regulation—and are they caused largely by excessive synaptic pruning in adolescence?


I read your fine Perspective on the psychiatric classification/biological research issue as well as the valuable Peralta et al. 2015 reference with great interest.   As you have long argued, better psychiatric classification will be necessary both for the development of improved drugs and for biological understanding.  You nicely demonstrate that Leonhard's classification has advantages over other diagnostic systems for neurobiological and genetic research.

In the reports which you summarize, however, I was surprised to note an apparent discrepancy between the pharmacological and genetic findings in schizophrenia.  The disparity is most salient when Leonhard's classification is employed.  Specifically, I refer here to Leonhard's delineations of unsystematic schizophrenia and systematic schizophrenia.  According to the pharmacological studies of Astrup (1959) and Fish (1964) cited in your review, individuals in Leonhard's unsystematic schizophrenia category respond more reliably to chlorpromazine-like drugs than do those in the systematic schizophrenia category.  On the other hand, Peralta et al (2015) show convincingly that unsystematic schizophrenia exhibits a smaller familiality estimate (h2 = 0.39) than systematic schizophrenia (h2 = 0.78).  Thus, contrary to expectation, the pharmacology and genetics go in opposite directions in Leonhard's scheme—i.e., the “less-familial” unsystematic schizophrenia diagnosis is more responsive to drugs than is the “more-familial” diagnosis of systematic schizophrenia.

If you agree that this observation seems paradoxical, an explanation may be sought in a parallel distinction, which contrasts the “positive” and “negative” symptoms of schizophrenia.  Following Crow (1980), it is recognized that dopamine-blocking antischizophrenic drugs are effective primarily against schizophrenia's “positive/psychotic” symptoms and are relatively ineffective against the “negative/deficit” symptoms.   (Accordingly, only the drug-sensitive positive symptoms are thought to be associated with brain dopamine hyperactivity).  The negative symptoms are largely irreversible as well as drug-resistant.  If these irreversible negative symptoms are featured in Leonhard's definition of systematic schizophrenia (as I presume them to be), then it follows that the genetically-important category of systematic schizophrenia in fact will be relatively resistant to the dopamine-blocking drugs.

If correct, this argument indicates that the most important genetic factors underpinning schizophrenia are not directly associated with a disordered brain dopamine substrate (although some dopamine involvement is likely to be contributory).  Rather, the critical genetic risk factors for schizophrenia more probably involve some other, but currently obscure, substrate—i.e., the unknown brain circuitries whose dysfunction underlies the negative/deficit symptoms.  Understanding the disordered brain physiology in schizophrenia, it was hoped, might best be approached by turning the problem around and first deciphering the genetics.  This hope has long been frustrated, but now a paper online in Nature may have given the genetic risk factors for schizophrenia their first substantial exposure (Sekar et al 2016).   According to the accompanying News & Views commentary of Dhindsa and Goldstein (2016), “Sekar et al. present a remarkable genomic and neurobiological study that...finally gives us the first real inroad into the molecular aetiology of schizophrenia".

A huge (>60,000 cases) genome-wide association study reported in Nature two years ago provided the starting point (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014).  Of 108 schizophrenia risk loci carefully identified, by far the strongest was the major histocompatibility complex (MHC), a region on chromosome 6 related to immunity. Within the MHC locus, the peak was near a pair of genes (C4A and C4B) which code for complement component 4.  This C4 protein is a critical component of the immune system pathway for the recognition of pathogens and cellular debris—but which (if any) of the hundreds of genes in the notoriously challenging MHC region might actually be relevant to schizophrenia was unknown.  Sekar et al. make ingenious use of RNA expression data and other evidence to establish the potential relevance of C4, despite the enormous molecular complexity of its many diverse alleles.  Remarkably, they are able to show that the various alleles of the C4 genes associate with schizophrenia in proportion to their ability to generate C4A expression.  In other words, the problematic structural heterogeneity of the many C4 variants is nicely reconciled by their convergence at a common, functionally homogeneous, output—C4A expression—to predictably increase the risk of schizophrenia. 

Significantly, in the brain, the complement cascade has been implicated in synaptic elimination and remodeling. Sekar et al. confirm in a mouse model that C4 expression, in fact, is upregulated during periods of synaptic pruning, and they further hypothesize that a similar elevation of C4A expression in schizophrenia results in excessive synaptic pruning.  Sekar et al. thus provide detailed genetic support for Feinberg's (1982/83) prescient and closely-related neurodevelopmental model of schizophrenia.  According to Feinberg, a profound reorganization of human brain function takes place in adolescence, when adult problem-solving “power” begins to appear.  Because a reduction in cortical synaptic density occurs at the same time (Huttenlocher 1979), the cognitive and behavioral changes are explicitly attributed by Feinberg to an increase in synaptic pruning.  Why are fewer synapses better than many?   Feinberg explains that too many competing neural connections in the brains of children may impede information processing, and that the selective pruning of erroneous and irrelevant synapses gradually shapes the adult capacity to solve highly abstract and. complex problems.  But, according to Feinberg, these advantages of pruning can be taken too far.  Genetic variants or mutations that cause excessive synaptic pruning at a critical neurodevelopmental stage may explain the typical manifestation of schizophrenia in late adolescence and early adulthood.




Astrup C. The effects of ataraxic drugs on schizophrenic subgroups. Related to experimental findings. Acta Psychiatr Scand 1959; 34 (supplement 136): 388-93.

Crow TJ.  Positive and negative schizophrenic symptoms and the role of dopamine.The British Journal of Psychiatry 1980; 137 (4) 383-6.

Dhindsa RS, Goldbstein DB. From genetics to physiology last. 2016. Dol: 10.1038/nature16874.

Feinberg I. Schizophrenia: caused by a fault in programmed synaptic elimination during adolescence? J Psychiatr Res 1982/1983; 17: 319–34.

Fish F. The influence of the tranquilizers on the Leonhard schizophrenic syndromes. Encephale  1964; 53: 245-9.

Huttenlocher PR Synaptic density in human frontal cortex—developmental changes and effects of aging. Brain Res 1979; 163: 195-205.

Leonhard K. Aufteilung der endogenen Psychosen. Berlin: Akademie Verlag; 1957.

Peralta V, Goldberg X, Ribeiro M, Sanches-Torres AM, Fananas L, Cuesta MJ. Familiality of psychotic disorders: A polynosologic study in multiple families. Schizophrenia Bulletin Advance Access 2015 doi: 101093/schbul/sbv192.

Schizophrenia Working Group of the Psychiatric Genomics Consortium (2014). Biological insights from 108 schizophrenia-associated genetic loci. Nature 2014; 511, 421-7..

Sekar A, Bialas AR, de Rivea H, et al. Schizophrenia risk from complex variation of complement component 4.  Nature 2016


Larry Stein

April 14, 2016