Getting our AKT together in schizophrenia?

AKT1 is a protein kinase that functions as a central element in many pathways involved in the control of cell growth and apoptosis. Genetic and biochemical data now provide evidence that AKT1 may be a susceptibility gene for schizophrenia.

Since Bleuler¹ first introduced the term in 1911, schizophrenia has been the subject of controversy. The concept of schizophrenia has changed substantially over time, and its boundaries in relation to other types of psychosis (bipolar disorder, for example) are far from clear. Discussion of the etiology of schizophrenia has shifted away from the mother-infant relationship to brain imaging and molecular genetics. But one finding has been replicated repeated: schizophrenia is familial, occurring at a higher frequency among relatives of individuals with the disorder. Coupled with the fact that the concordance rate is much higher in monozygotic twins than in dizygotic twins, this is compelling evidence of a genetic causation. With the advent of molecular approaches, identification of genes associated with schizophrenia has become the 'holy grail' in the field, and on page 131 of this issue, Effat Emamian and colleagues² identify AKT1 as a susceptibility gene for schizophrenia.

Genes and schizophrenia

More than 20 genome-wide scans have been published and many more candidate genes investigated for a possible association with schizophrenia. What constitutes a 'real' positive association is a topic of heated debate.

There are nearly insurmountable difficulties in establishing convincing proof. First, there is the diagnostic problem. Diagnosis is based on a wide range of symptoms in such areas as inferential thinking, language, social behavior, volition and emotional expression. At one end of the spectrum are cognitively intact individuals describing a complex delusional system; at the other end are withdrawn, socially isolated individuals who are unable to express and take care of their own needs. Variability in diagnosis is inevitable. Second, the complex mode of inheritance suggests that there is no simple one-to-one relationship between genotype and phenotype. Third, schizophrenia is a disease of the brain. Unlike Alzheimer disease, no specific brain pathology has been described for schizophrenia, and the observed abnormalities are subtler and not restricted to a particular area in the brain. Measurements in post-mortem material are confounded by several factors, such as cause of death and previous medication. Fourth, the development of animal models has been hampered by the lack of comparability between the animal brain and the human brain. Not surprisingly, then, a central issue in schizophrenia research has been the difficulty in showing that a particular gene is causally related to the disease. Researchers are coming to realize that proof cannot be based on genetics alone, and any positive finding must be corroborated by additional evidence.

The evidence for AKT1

Emamian et al.² start from the general hypothesis that genetically induced alterations in the expression of protein kinases may be related to schizophrenia and should be detectable in lymphocytes. They screened for differences in protein levels of various protein kinases and phosphatases between individuals diagnosed with schizophrenia and controls. Significant differences were found for only one of the kinases studied, AKT1. They followed up this finding by comparing post-mortem brains of schizophrenic individuals with those of controls and found lower levels of AKT1 in the hippocampus and frontal cortex of schizophrenic individuals. No differences were found in the two other known isoforms of AKT, AKT2 and AKT3, highlighting AKT1 as a candidate gene for association with schizophrenia.

The authors next tested directly for association between sequence variants in AKT1 and schizophrenia. After finding a significant positive association in a sample of 265 families with a schizophrenic proband, they went on to show that the core haplotype conferring a higher risk to schizophrenia was also associated with lower AKT1 levels in lymphocytes of controls. It is unlikely, however, that the haplotype accounts entirely for the differences in AKT1 levels between schizophrenic individuals and control samples.

But are there consequences of a decrease in AKT1 levels? To address this question, Emamian et al.² carried out two experiments. First, they found that phosphorylation of glycogen synthase kinase-3β (GSK3β), which is activated by AKT1, was significantly lower in schizophrenic brains. Second, they found differences in prepulse inhibition (PPI) in mice deficient for AKT1. PPI is the phenomenon in which a weak prepulse stimulus attenuates the response to a subsequent startling stimulus. Impairment of PPI has been observed in individuals with schizophrenia. The startle response can be measured easily in rodents³. A reduction of the PPI after administration of amphetamine², which can produce psychotic-like symptoms in humans, was observed in mice deficient for AKT1, but not in wild-type mice.

A multifunctional protein kinase

The door is open for speculation as to how AKT1 affects brain development and schizophrenia. AKT1 was first cloned as the cellular homolog of the v-act oncogene and is also known as protein kinase B. AKT1 has been extensively researched and is the subject of several thousand papers⁴. It has been linked to several signaling pathways mediating multiple responses (Fig. 1) and may be regulated by a large number of protein-protein interactions⁴.

Figure 1: AKT is linked to several cellular pathways.

Many different mediators interact with AKT, leading to changes in cell metabolism, cell survival and apoptosis.

In relation to schizophrenia, activation of AKT has been linked to the GABAergic⁵ and glutaminergic system⁶. Alterations in the GABA and glutamate systems have been described in post-mortem studies of schizophrenia for over two decades⁷. The two neurotransmitters are intrinsically interconnected. It has been postulated that schizophrenia may be the result of an imbalance between the inhibitory GABA and the excitatory glutaminergic neurons⁷. Neurons from AKT1-deficient mice are more sensitive to toxicity of kainate, a glutamate receptor agonist⁶. In addition, environmental stress factors (ultraviolet or hyperosmotic stress, for example) can lead to downregulation of AKT. Decreased AKT activity is related to post-ischemic neuronal damage, whereas increased activity can lead to reduced cell death⁸.

But GABA and glutamate are only two of many molecules suggested to be associated with schizophrenia and with AKT. No doubt the reported association between AKT1 and schizophrenia will lead to many new avenues of research. Learning more about the different roles of the three isoforms will be a crucial first step. Because of the many cellular pathways that affect AKT activity, defining the true relationship between AKT1 and schizophrenia will be challenging.