21 Feb 2005

Having just brought you some news on the nurture half of the developmental equation (see ARF related news story), we flip the coin and examine a new study demonstrating the power of genetics in mental health. An article published in the February 17 Neuron finds that significant consequences for rat behavior arise from changes in the expression level of a single protein. Such dosage effects of none other than Aph-1b, a component of the γ-secretase complex, underlie variations in some rats that can, in fact, mimic aspects of schizophrenia.

Gerard Martens and colleagues at Radboud University in Nijmegen, the Netherlands, with collaborators at Organon Laboratories in Newhouse, Scotland, and Katholieke University in Leuven, Belgium, searched for the molecular basis of a curious phenomenon among outbred Wistar rats. Some of these rats respond to dopamine agonists such as apomorphine or amphetamine by gnawing furiously on whatever is handy. As coauthors Alexander Cools and Bart Ellenbroek had found in studying these rats over the past 15 years, these "apomorphine-susceptible" (APO SUS) rats differ from their "unsusceptible" (UNSUS) brethren in many ways. For example, deviating from normal rat cautiousness, the APO SUS rats boldly explore new environments and objects. Accompanying this behavior are distinct endocrine responses, such as persistently high plasma release of adrenocorticotropic hormone (ACTH) and corticosteroids. The overall behavioral phenotype of this APO SUS rat, combined with the observation that amphetamine psychosis mimics schizophrenia somewhat, has led some researchers to propose it as a model for this complex disease.

In the present study, first author Marcel Coolen and colleagues set out to identify the genes behind these behavioral differences. Based on recent trends in research on the genetics of common diseases, they would have expected to find a complex situation involving several genes. Instead, they found differences in expression of only a single gene, that for γ-secretase component Aph-1b (see ARF related news story).

Whereas humans have just one Aph-1b gene, rats harbor two in close proximity to each other. In Wistar rats, however, the situation is not cut-and-dried. The researchers found that a genomic rearrangement has produced a third version, a chimera of the other two Aph-1b genes. Coolen and colleagues found that outbred Wistar rats can have several combinations of these three genes—i.e., some rats have both original genes plus the chimeric Aph-1b on each chromosome, while others are left with only a chimeric gene on each chromosome. Others are intermediate, with two copies per chromosome.

It turns out that the APO UNSUS rats are the lucky ones, with six Aph-1b copies that produce six times as many Aph-1b mRNAs as the two chimeric copies found in the genomes of the lesser fortunate APO SUS rats. (Current methods do not allow accurate quantification of the actual protein.) Intermediate numbers of gene copies produced intermediate amounts of Aph-1b mRNA.

Coolen and colleagues then looked at what effect these genetic differences have on the protein's activity. After all, the γ-secretase complex is involved in the cleavage of a large number of proteins critical in neurodevelopment and beyond. The authors report that the number of Aph-1b genes did not substantially alter mRNA and protein expression levels of Aph-1b's γ-secretase partners (presenilin-1 or 2, nicastrin, or PEN-2) or its cousin Aph-1a. Even so, there were clear effects on γ-secretase activity. In the pons/medulla and olfactory bulb, where Aph-1b is the more plentiful variant, having only one Aph-1b gene significantly reduced γ-scretase cleavage of APP. By contrast, in regions where Aph-1a is present in higher amounts (e.g., cerebellum or hypothalamus), reduced Aph-1b expression did not appear to limit γ-secretase activity.

Behavioral testing revealed a similar effect, with different profiles of behavior segregating with different Aph-1b genetic profiles. "Thus, a subtle imbalance in the expression of a single gene product that is involved in a wide variety of developmental signaling pathways may well constitute the molecular basis of a complex phenotype that is generally believed to have a multifactorial background," the authors conclude.—Hakon Heimer

Q&A with authors Gerard Martens and Bart De Strooper.

Q: Did you find the differential expression of Aph-1b in the two groups of rats at both time points you examined (postnatal days 9 and 60)?
GM: Since Aph-1b was not on the P9 chips, we did not initially detect the difference using this Affymetrix chip. However, later quantitative PCR clearly established that at both time points Aph-1b expression was different.

Q: You are proposing that gene dosage effects could explain all facets of some disorders, mentioning specifically schizophrenia. Where does that leave the various environmental factors proffered by epidemiologic research into schizophrenia (paternal age, intrauterine influenza exposure, etc.)?
GM: This genetic background does not preclude the influence of environmental factors on the final outcome of the phenotype. Moreover, the genetic rearrangement event may well have been caused by some environmental influence. The story is thus that both a genetic effect and an (early) environmental factor are at the basis of such disorders. Indeed, one could say that epigenetic phenomena may multiply the gene dosage effect.

Q: Are these results relevant to Alzheimer disease, given that humans only have one copy of Aph-1b? Should one be more cautious in interpreting any results regarding Aph-1b in mouse/rat models of AD?
GM: The Aph-1b gene encodes a component of the Alzheimer-linked γ-secretase enzyme. However, this enzyme is also involved in multiple (neuro)developmental pathways. Thus, γ-secretase may be associated with not only neurodegenerative diseases (such as Alzheimer), but also neurodevelopmental disorders (such as schizophrenia). This provides a new perspective on the Alzheimer-linked γ-secretase enzyme.
BDS: The amazing thing is that a very complex "psychiatric" phenotype in rats could be brought back to a very specific mutation in the genome. This mutation is specific since it could be reproduced by repeating the whole breeding selection. Thus, a quite complex behavioral syndrome can be explained by a single genomic mutation. Furthermore, the functional consequences are a less active γ-secretase complex in certain brain areas, revealing the role of γ-secretase in the fine-tuning of normal behavior.

In rodents, a gene duplication has occurred, resulting in very closely related Aph-1b and Aph1c genes that are genetically closely linked. They are both very similar to the human Aph-1b. We therefore recently proposed that the double Aph-1b/c should be considered as the model for Aph1b in human (see ARF related news story). Indeed, one should be cautious when interpreting single Aph-1b or Aph-1c knockout effects because they cover only part of the Aph-1b functions. Once the two are considered together, they are probably as good a model for human Aph-1b function as rodent presenilin-1 is for human presenilin-1. Thus it seems to me that the caution to be taken is identical to that with any other protein when extrapolating from mouse to human: Fundamental biological functions are likely conserved, but one should always be aware that a mouse is only a model organism.

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

1. Perinatal Soup—Early Pathogen or Toxin Exposures Leave Brain Vulnerable 14 Feb 2005
2. Aph1A KO Stalls Embryo Development, Supports γ-Secretase Variance 24 Jan 2005

Further Reading

No Available Further Reading