Transcription factor 4 (TCF4) and schizophrenia: integrating the animal and the human perspective

Abstract

Schizophrenia is a genetically complex disease considered to have a neurodevelopmental pathogenesis and defined by a broad spectrum of positive and negative symptoms as well as cognitive deficits. Recently, large genome-wide association studies have identified common alleles slightly increasing the risk for schizophrenia. Among the few schizophrenia-risk genes that have been consistently replicated is the basic Helix-Loop-Helix (bHLH) transcription factor 4 (TCF4). Haploinsufficiency of the TCF4 (formatting follows IUPAC nomenclature: TCF4 protein/protein function, Tcf4 rodent gene cDNA mRNA, TCF4 human gene cDNA mRNA) gene causes the Pitt-Hopkins syndrome-a neurodevelopmental disease characterized by severe mental retardation. Accordingly, Tcf4 null-mutant mice display developmental brain defects. TCF4-associated risk alleles are located in putative coding and non-coding regions of the gene. Hence, subtle changes at the level of gene expression might be relevant for the etiopathology of schizophrenia. Behavioural phenotypes obtained with a mouse model of slightly increased gene dosage and electrophysiological investigations with human risk-allele carriers revealed an overlapping spectrum of schizophrenia-relevant endophenotypes. Most prominently, early information processing and higher cognitive functions appear to be associated with TCF4 risk genotypes. Moreover, a recent human study unravelled gene × environment interactions between TCF4 risk alleles and smoking behaviour that were specifically associated with disrupted early information processing. Taken together, TCF4 is considered as an integrator ('hub') of several bHLH networks controlling critical steps of various developmental, and, possibly, plasticity-related transcriptional programs in the CNS and changes of TCF4 expression also appear to affect brain networks important for information processing. Consequently, these findings support the neurodevelopmental hypothesis of schizophrenia and provide a basis for identifying the underlying molecular mechanisms.

Fig. 1

Different bHLH transcription factors direct central nervous system (CNS) development at embryonic stages and may be involved in adult brain plasticity. Inhibitory bHLH factors (HES1, ID1) and proneural factors ATOH1, ASCL1 and NEUROG1,2 as well as E-proteins TCF3 and TCF12 are involved in early developmental stages. The temporal expression patterns and mutational analyses of the neurogenic differentiation factors (NEUROD1,2 and 6) and inhibitors of differentiation ID2 and ID4 suggest instead a function in later stages of neuronal differentiation and in the adult CNS. The spatiotemporal expression pattern of Tcf4 overlaps substantially with all other bHLH factors involved in brain development. Moreover, TCF4 is capable of forming hetero-dimers with most involved neuron expressed bHLH factors although direct evidence is thus far only available for NEUROD1 and -2 (as indicated by a solid line in contrast to dashed lines). It should be noted that this schematic drawing is thought to be an overview representation not claiming detailed spatial and temporal expression domains of single genes (for citations, see main text)

Fig. 2

Phenotypical comparisons reveal different TCF4 gene dosage dependences in mice (a) and humans (b) in neurodevelopment related diseases including schizophrenia. Gain-of-function and loss-of-function analyses in mice and corresponding risk alleles and mutations in humans suggest that TCF4 expression differences are tolerated in a narrow range (range depicted in light blue). Exceeding critical thresholds increases disease risks (depicted in grey). Slightly increased postnatal expression of Tcf4 has been found to cause schizophrenia (SZ)-associated symptoms in mice. Phenotypic consequences of increased Tcf4 expression during embryonal stages are not yet known (a). Indirect evidence from human post-mortem brain and blood sampling suggests that elevated expression may be associated with SZ and bipolar disease (BD). The critical period of enhanced TCF4 expression in humans is unknown (b). The tolerance range for reduced gene dosage effects might potentially be higher in mice compared to humans, since heterozygous animals appear to be largely unaffected, although a thorough behavioural phenotyping has not so far been performed. Thus, it is unknown if reduced gene dosage in mice may cause SZ-like symptoms. The analysis of null mutants is hampered by perinatal lethality, but structural deficits in brain development have already been described, although not thus far representing Pitt-Hopkins-like symptoms (a). Loss-of-function of TCF4 (haploinsufficiency and mosaic deficiency) causes severe neurodevelopmental diseases including PTHS and possibly other autism-like syndromes. Given many examples of inverted-U-shape relationships of gene dosage with disease severity in autism-related neurodevelopmental diseases, it appears possible that slightly reduced expression levels of TCF4 may be implicated in SZ (b) (for citations, see main text). SZ schizophrenia, BD bipolar disorder, PTHS Pitt-Hopkins Syndrome, NDD neurodevelopmental disorder, MR mental retardation

Fig. 3

Brain structures involved in postulated deficits of information processing in mice and men. Behavioural and neuropsychological phenotypes obtained in mice (a) and human subjects (b) suggest a function of TCF4 in brain networks that are important for cognition (bold lines) and sensory processing (dotted grey lines). Deregulation of TCF4 expression levels during development interferes with proper functional connectivity within corresponding brain networks (for citations see main text). ACx auditory cortex, Amy amygdala, Hi hippocampus, NAc nucleus accumbens, PCx prefrontal cortex, PPN pedunculopontine nucleus, VTA ventral tegmental area, VP ventral pallidum