Oxysterols and Learning
This article from Dave Russell’s group, by Kotti et al., describes the effects caused by knockout of the gene for 24(S)-cholesterol hydroxylase (Cyp46).

Cholesterol levels are regulated by two processes: synthesis and catabolism. Removal of cholesterol in the brain is accomplished largely by oxidation at the 24(S) position, to produce the oxysterol termed 24(S) hydroxycholesterol. In contrast, removal of cholesterol from the rest of the body is mainly accomplished by oxidation of either the 27 position (to produce 27 hydroxycholesterol) or by oxidation at the 7α position to produce bile acids.

Cyp46 has attracted a great deal of attention in the field of AD for several reasons. First, the primary role of 24 hydroxycholesterol in removing cholesterol emphasizes its importance in understanding the neurochemistry of the brain. Second, levels of 24(S) hydroxycholesterol are increased by neurodegenerative processes. Third, 24(S) hydroxycholesterol regulates Aβ secretion. Finally, some papers report that polymorphisms in Cyp46 are...  Read more

Oxysterols and Learning
This article from Dave Russell’s group, by Kotti et al., describes the effects caused by knockout of the gene for 24(S)-cholesterol hydroxylase (Cyp46).

Cholesterol levels are regulated by two processes: synthesis and catabolism. Removal of cholesterol in the brain is accomplished largely by oxidation at the 24(S) position, to produce the oxysterol termed 24(S) hydroxycholesterol. In contrast, removal of cholesterol from the rest of the body is mainly accomplished by oxidation of either the 27 position (to produce 27 hydroxycholesterol) or by oxidation at the 7α position to produce bile acids.

Cyp46 has attracted a great deal of attention in the field of AD for several reasons. First, the primary role of 24 hydroxycholesterol in removing cholesterol emphasizes its importance in understanding the neurochemistry of the brain. Second, levels of 24(S) hydroxycholesterol are increased by neurodegenerative processes. Third, 24(S) hydroxycholesterol regulates Aβ secretion. Finally, some papers report that polymorphisms in Cyp46 are associated with AD. However, upon meta-analysis, none of the Cyp46 polymorphisms studied to date appear to show significant association with AD; this meta-analysis can be found in Alzgene.

The current manuscript by Kotti et al. explores the biology of Cyp46 knockout mice. Despite knockout of a principal catabolic pathway for brain cholesterol, the levels of cholesterol are not elevated in the Cyp46 knockout. Kotti et al. demonstrate that this is due to a compensatory decrease in the synthetic pathway. Although decreased synthesis of cholesterol maintains constant levels of cholesterol, it has other deleterious side effects because some of the intermediates in the cholesterol synthetic pathway are required for proper neuronal function. As a result, the animals are, well…to put it bluntly, stupid. They don’t learn. Kotti et al. explore the nature of the learning deficit and find that it is likely due to inadequate levels of geranylgeranyl phosphate, which is an intermediate in the cholesterol synthetic pathway that is required for palmitoylation. Palmitoylation, also known as prenylation and isoprenylation, allows small GTPases to be targeted to membranes, which means that proteins such as Rac and Rho show reduced activity (Fig. 1). The loss of small GTPase function results in reduced LTP activity, which Kotti et al. propose accounts for the deficits in learning exhibited by these mice.

View Figure 1

These results have a number of important implications. They emphasize the interconnected link between cholesterol metabolism and neuronal function. Neuronal function is highly dependent on lipid and steroid metabolism, and these mice exemplify this dependence. This work also emphasizes an unexpected but potentially important role that oxysterols might play in learning and memory. It is possible that oxysterols might provide an important avenue for modifying these processes. The final important point is one emphasized by Kotti et al. Although statins also reduce cholesterol metabolism, one should not jump to the conclusion that statins will eliminate memory. Patients taking statins rarely report loss of memory or other forms of cognition, which is probably because the degree of blockade produced by statins in the brain is insufficient to inhibit palmitoylation. The reason is that the enzymes mediating palmitoylation have a lower IC50 for their substrates than do HMG-CoA synthase and reductase. Hence, the palmitoylation pathway operates normally until cholesterol metabolism is inhibited by over 90 percent.

View all comments by Benjamin Wolozin

Normal learning and LTP require a constant source of isoprenoids
CNS roles for sterols and isoprenoids have attracted the attention of Alzheimerologists seeking to understand the mechanisms of action of ApoE isoforms and statins, each of which appears to modulate brain amyloidosis and the risk for Alzheimer’s. Russell and colleagues have explored these interrelationships either by using cholesterol-24-hydroxylase knockout mice (24-OH-ase KO), or by pharmacologically manipulating either HMG CoA reductase activity or levels of the isoprenoid geranylgeraniol.

24-OH-ase KO mice excrete cholesterol more slowly than normal, and tissues (including brain) compensate by suppressing the mevalonate pathway, which, in turn, causes isoprenoid levels to fall. The 24-OH-ase KO mice exhibit severe deficiencies in spatial, and associative and motor learning, and brain slices from these mice show defective induction of LTP. Acute application of HMG CoA reductase inhibitors (statins) leads to similar effects on LTP that are induced too rapidly to be attributable to changes in...  Read more

Normal learning and LTP require a constant source of isoprenoids
CNS roles for sterols and isoprenoids have attracted the attention of Alzheimerologists seeking to understand the mechanisms of action of ApoE isoforms and statins, each of which appears to modulate brain amyloidosis and the risk for Alzheimer’s. Russell and colleagues have explored these interrelationships either by using cholesterol-24-hydroxylase knockout mice (24-OH-ase KO), or by pharmacologically manipulating either HMG CoA reductase activity or levels of the isoprenoid geranylgeraniol.

24-OH-ase KO mice excrete cholesterol more slowly than normal, and tissues (including brain) compensate by suppressing the mevalonate pathway, which, in turn, causes isoprenoid levels to fall. The 24-OH-ase KO mice exhibit severe deficiencies in spatial, and associative and motor learning, and brain slices from these mice show defective induction of LTP. Acute application of HMG CoA reductase inhibitors (statins) leads to similar effects on LTP that are induced too rapidly to be attributable to changes in cholesterol levels. Consistent with this formulation, Russell and colleagues demonstrate that both the LTP abnormalities in 24-OH-ase KO mice and those that follow acute statin treatment can be reversed with geranylgeraniol but not with cholesterol.

The genetic or pharmacological dissection of cholesterol-mediated versus isoprenoid-mediated effects in the CNS is a challenging task. The new data indicate that a constant source of geranylgeraniol is required for synaptic function. Geranylgeraniol exerts actions on vesicle biology via the Rho/ROCK pathway, a mechanism consistent with the role in LTP revealed by Russell and colleagues. The new data dovetail well with the established role for the Rho/ROCK pathway to modulate vesicle trafficking and, hence, control growth cone pathfinding, as well as the polarization of neuronal processes into axons and dendrites (Arimura and Kaibuchi, 2005). The molecular basis of Rho/ROCK regulation of neuronal function apparently involves the key role that this pathway plays in linking signal transduction to neuronal development, polarity, and plasticity. Inasmuch as this regulation can occur at the synapse, these relationships probably explain the cause of the LTP disturbance associated with cholesterol 24-OH-ase deficiency or acute isoprenoid depletion disrupting vesicle biology by perturbation of Rho/ROCK signaling.

This is the first clear evidence that a constant supply of isoprenoids is required for neuronal function and that acute isoprenoid deficiency is associated with perturbations in normal CNS physiology. Statins are also known to regulate APP metabolism via ROCK1 (Pedrini et al., 2005), suggesting that isoprenoid levels control both neuronal function and APP metabolism, thereby placing these lipids at an interesting signaling crossroads with respect to Alzheimer disease (see ARF comment by Steve Paul on Pedrini et al.). Further work will be required to elucidate the relevant molecular targets of Rho/ROCK that are required for normal synaptic function and to determine what roles (if any) these interconnected pathways play in the molecular pathogenesis of Alzheimer disease.

References:
Kotti TJ, Ramirez DM, Pfeiffer BE, Huber KM, Russell DW. Brain cholesterol turnover required for geranylgeraniol production and learning in mice. Proc Natl Acad Sci U S A. 2006 Mar 7;103(10):3869-74. Abstract

View all comments by Samuel Gandy

Kotti et al., present an elegant study in which they use behavioral and electrophysical and biochemical analysis of wild type and 24-hydroxylase knockout (KO) mice to delineate a novel role for geranylgeraniol, an immediate precursor of the isoprenoid geranylgeranyl pyrophosphate (GGPP), in LTP and learning. The authors show that although the KO mice are free from any unusual brain anatomical defects and exhibit unchanged short-term plasticity, defects in spatial, associative and motor learning, in addition to significant impairments of LTP are apparent in these mice.

Under regular homeostasis, brain cholesterol is converted to 24(S)-hydroxycholesterol for excretion. 24-hydroxylase drives this conversion, in order to maintain a tightly regulated balance of brain cholesterol levels. In the absence of this enzyme, subtle increases in brain cholesterol initiate a feedback loop and decrease de novo brain cholesterol synthesis. However, inhibition of this biosynthetic pathway not only limits cholesterol biosynthesis but also affects the production of isoprenoids moieties, such...  Read more

Kotti et al., present an elegant study in which they use behavioral and electrophysical and biochemical analysis of wild type and 24-hydroxylase knockout (KO) mice to delineate a novel role for geranylgeraniol, an immediate precursor of the isoprenoid geranylgeranyl pyrophosphate (GGPP), in LTP and learning. The authors show that although the KO mice are free from any unusual brain anatomical defects and exhibit unchanged short-term plasticity, defects in spatial, associative and motor learning, in addition to significant impairments of LTP are apparent in these mice.

Under regular homeostasis, brain cholesterol is converted to 24(S)-hydroxycholesterol for excretion. 24-hydroxylase drives this conversion, in order to maintain a tightly regulated balance of brain cholesterol levels. In the absence of this enzyme, subtle increases in brain cholesterol initiate a feedback loop and decrease de novo brain cholesterol synthesis. However, inhibition of this biosynthetic pathway not only limits cholesterol biosynthesis but also affects the production of isoprenoids moieties, such as GGPP.

In a series of well-planned experiments, the authors extend the study to investigate the metabolic basis of the LTP impairment in the KO mice. By inhibiting the cholesterol biosynthetic pathway in vitro, by using statins as pharmacological tools, Kotti et al. reveal that the observed LTP impairment is specifically related to decreases in geranylgeraniol rather than depression of cholesterol levels.

Recently, a handful of papers have begun to reveal the potential importance of isoprenoids in molecular pathways of relevance to Alzheimer- disease pathophysiology. Although this field is currently in a state of flux with findings being controversial and hard to interpret, it is clear that alterations in isprenoid levels can affect APP metabolism, inflammation, tau hyperphosphorylation and synaptic plasticity (reviewed in Cole and Vassar, 2006). During isoprenylation, the attachment of isoprenoids to target proteins, such as small GTPases, is essential for correct cellular functioning. Indeed, Kotti and colleagues hypothesize that the disruption of LTP observed in the KO mice may be related to aberrant changes in AMPA receptor trafficking, which may result from decreases in the isoprenylation of key G-proteins required for the recruitment, and therefore activation, of AMPA receptors to the cell-surface. No doubt further studies are underway to investigate this line of thinking further. This study nicely combines both in-vivo and in-vitro work to examine the effects of reducing 24-hydroxylase, and its downstream targets, in memory and learning. Although technically challenging, it would also be of interest to try and determine if in-vivo application of geranylgeraniol could rescue the behavioral deficits in the KO mice.

In conclusion, the findings of the current study support the notion that isoprenoid moieties could be of importance in key cellular mechanisms, which change during the progression of diseases such as AD. Furthermore, this study highlights the benefits of using statins as pharmacological tools in delineating molecular mechanisms in vitro and correctly states that the chances of observing such effects in vivo in humans following statin use are highly unlikely. Sarah Cole (s-cole4@northwestern.edu) Robert Vassar (r-vassar@northwestern.edu)

View all comments by Robert Vassar