Jump to content
RemedySpot.com

Cholesterol & ALzheimers

Rate this topic


Guest guest

Recommended Posts

--- Jeff Novick <jnovick@...> wrote:

Hi All,

Being in a rush, the below pdf-available excerpts are the best available

version.

http://my.webmd.com/content/article/113/110683?src=RSS_PUBLIC

Nature Cell Biology, Oct. 9, 2005 advance online edition.

Alzheimer's disease: Understanding the lipid connection

Letter by Grimm et al.

Varying cholesterol levels in mouse models has been shown to affect progress of

Alzheimer's disease. Hartmann and colleagues now show that Ab peptide, the

amyloid

peptide that is produced normally but accumulates markedly during Alzheimer's,

controls lipid metabolism by acting on two key enzymes in the pathway.

Marcus O. W. Grimm1, Heike S. Grimm1, s J. Pätzold1, Eva G. Zinser1,

Riikka

Halonen1, Marco Duering1, Jakob-A. Tschäpe1, Bart De Strooper2, Ulrike Müller3,

Jie

Shen4 & Tobias Hartmann1

Amyloid beta peptide (Aß) has a key role in the pathological process of

Alzheimer's

disease (AD), but the physiological function of Aß and of the amyloid precursor

protein (APP) is unknown1, 2. Recently, it was shown that APP processing is

sensitive to cholesterol and other lipids3, 4, 5, 6, 7, 8, 9, 10.

Hydroxymethylglutaryl-CoA reductase (HMGR) and sphingomyelinases (SMases) are

the

main enzymes that regulate cholesterol biosynthesis and sphingomyelin (SM)

levels,

respectively. We show that control of cholesterol and SM metabolism involves APP

processing. A42 directly activates neutral SMase and downregulates SM levels,

whereas A40 reduces cholesterol de novo synthesis by inhibition of HMGR

activity.

This process strictly depends on gamma-secretase activity. In line with altered

A40/42 generation, pathological presenilin mutations result in increased

cholesterol

and decreased SM levels. Our results demonstrate a biological function for APP

processing and also a functional basis for the link that has been observed

between

lipids and Alzheimer's disease (AD).

One of the main characteristics of Alzheimer's disease (AD) is the accumulation

of

amyloid beta peptide (A), which is associated with neurodegeneration1, 2.

A production is initiated by the gamma-secretase cleavage of amyloid precursor

protein (APP), which results in production of the APP C-terminal fragment C99.

This

fragment is further cleaved by gamma-secretase to generate A40 and A42. The

active

centre of the gamma-secretase complex is formed by presenilin-1 or -2 (PS1 or

PS2)11. As the gamma-secretase cleavage site is centred within the transmembrane

domain, membrane composition may considerably influence Aß generation12. Such

proteolytic events of regulated intra-membrane proteolysis (RIP) are involved in

numerous cellular regulatory events, including lipid homeostasis13.

So far, the physiological role of APP and its proteolytic fragments, including

A,

have mainly been conceived as being brain specific. However, the ubiquitous

production of APP, Aß and PS indicates that their functions are also ubiquitous.

Recently, strong evidence has accumulated that links AD and Aß generation with

lipid

homeostasis5, and it has been suggested that cellular lipid metabolism controls

APP

processing5. Studies have shown that Aß production is sensitive to cholesterol

levels and lipid trafficking. In vitro enhanced cholesterol levels lead to

increased

Aß production14. Furthermore, cholesterol depletion reduces gamma-secretase

activity

and results in reduced Aß generation3, 4, 6, 7, 8, 9. Controversially, one

report

found a correlation between mildly decreased cholesterol levels and increased Aß

production in Chinese hamster ovary cells15. In vivo, Aß production and

deposition

are tightly associated with changes in cholesterol levels6, 10.

Lipid levels vary in response to changes in diet, and physical or neuronal

activity.

Similarly, Aß levels change in response to several conditions, including

cholesterol

and other factors16, 17. Brain cholesterol levels increase during the

progression of

AD18 and hypercholesterolaemia is an early risk factor in the development of

AD19.

Consequently, cholesterol-lowering drugs are being considered as a potential

treatment for AD16, 20, 21. Nevertheless, the cellular mechanisms that link

lipids,

Aß generation and AD are poorly understood.

To investigate the role of PS in lipid regulation, we used mouse embryonic

fibroblast cells that were devoid of PS1 and PS2 (MEF PS1/2-/-)11 and

expression-level-matched, A-producing PS re-transfected MEF PS1/2-/- control

cells

(MEF PS1r and MEF PS2r, respectively) (see Supplementary Information, Fig. S1).

The

absence of PS resulted in increased cellular levels of cholesterol and

sphingomyelin

(SM), but not phosphatidylcholine (PC) or total lipid-associated phosphate (Fig.

1a). The catabolic sphingomyelinase (SMase) activities were downregulated,

anabolic

SM-synthase activity was upregulated and the SM-to-ceramide ratio was increased

in

PS-deficient cells in accordance with the altered neutral SMase (nSMase) and

SM-synthase activities (Fig. 1a, see Supplementary Information, Fig. S2a).

Absence of PS1 is lethal and, as PS2 is able to replace PS1 (for example, in

nSMase

activity, see Fig. 1c), brains from conditional PS1 knockout mice, crossed to

PS2

knockout mice, were used to evaluate PS function in lipid homeostasis in vivo.

In

these viable mice, PS1 remains expressed in some neurons and other cells22 and

brain

Aß levels in these mice are reduced to 20 & #8722;50%, depending on age and Aß

species23. As before, absence of PS in neurons resulted in increased cholesterol

and

SM levels, decreased SMase activity, but unchanged PC levels (Fig. 1b).

Dependence on gamma-secretase proteolytic activity was investigated using the

gamma-secretase activity inhibitor L-685,458. This treatment reduced nSMase

activity

in MEF PS1r and PS2r cells to similar levels as those of PS1/2-/- cells (Fig.

1c).

By contrast, this inhibitor treatment had no effect on MEF cells that were

devoid of

PS (Fig. 1c). Treatment of wild-type mouse neurons decreased nSMase activity and

increased SM levels (Fig. 1c). Accordingly, in inhibitor-treated PS-expressing

cells, SM levels were increased. Cholesterol levels increased in MEF PS1/2-/-

cells

(Fig. 1a), whereas no significant difference was observed between

inhibitor-treated

MEF PS1r and MEF PS1/2-/- cells (see Supplementary Information, Fig. S2b).

It was established previously that cholesterol influences Aß production. To

investigate Aß putative influence of SM on Aß production, we inhibited nSMase

activity in neurons. This resulted in increased SM levels and decreased

intracellular and secreted Aß (Fig. 2a). This was confirmed in COS7 cells using

additional inhibitors and direct SM exposure (see Supplementary Information,

Fig.

S3a, 3b). Reduced gamma-secretase cleavage causes C99 accumulation, which was

observed with increased SM levels (see Supplementary Information, Fig. S3c).

Our results implicate the involvement of a gamma-secretase substrate and that

its

cleavage is essential in PS-mediated lipid regulation. MEF APP/APLP2-/- and MEF

wild-type cells were analysed to evaluate whether APP or APLP2 might be involved

in

this regulatory cascade. SM, nSMase activity and cholesterol levels responded to

the

absence of APP/APLP2 in the same way as they did to the absence of

PS/gamma-secretase activity (Fig. 2b). The same was observed in APP-/- mouse

brain

(Fig. 2b), suggesting that an in vivo function of APP exists in lipid

homeostasis.

Unlike in wild-type cells, inhibition of gamma-secretase activity did not alter

cholesterol in MEF APP/APLP2-/- cells (Fig. 2c). This seems to rule out the

existence of other gamma-secretase substrates that are able to replace the

APP/APLP2

function in cholesterol regulation in these cells. The combined necessity of APP

and

gamma-secretase activity in this process indicates that a proteolytic APP

fragment

is involved in lipid homeostasis (Fig. 2b, c). The most prominent peptide that

is

derived from APP gamma-secretase cleavage is A. To determine a possible function

of

Aß in cholesterol and SM regulation, we analysed key enzymes and metabolites in

the

respective pathways.

Analysing SM metabolism in MEF APP/APLP2-/- cells showed a decrease in nSMase

activity compared with MEF wild-type cells (Fig. 2b). When exposed to

A-containing

conditioned media from human SH-SY5Y cells, the nSMase enzymatic activity was

partially restored in MEF APP/APLP2-/- cells (Fig. 2d, left). To investigate

whether

this might be caused by A, cell homogenates of MEF APP/APLP2-/- cells were

exposed

to synthetic A1 & #8722;40/A1 & #8722;42 at a physiological ratio24, resulting in

enhanced nSMase activity (Fig. 2d, right). To determine whether SMase is a

molecular

target for regulation by Aß and to evaluate whether this interaction might be

direct, purified human placenta nSMase was exposed to synthetic Aß in a

biochemical

assay that was free of other cellular components. Covering the physiological

concentration range of A24, the effect reached a maximum at peptide levels that

are

typically found in cerebrospinal fluid. The effect decreased thereafter,

probably

due to augmented peptide aggregation (Fig. 2e). An effect of Aß on the other

assay

components was excluded (see Supplementary Information, Fig. S4a); A42 & #8722;1

(inverse A) was inactive and aggregated Aß had only residual stimulatory

activity,

presumably due to remnant non-aggregated peptide. An excess of A40 reduced

nSMase

activation by A42, indicating A42 specificity (see Supplementary Information,

Fig.

S4b).

To evaluate the involvement of oxidative stress in nSMase activation by Aß

peptides,

the production of reactive oxygen species (ROS) was investigated. No significant

differences in ROS were observed when wild-type and rescue cells were compared

with

the respective knockout cells (Fig. 2f). Incubation of MEF PS1/2-/- with Aß in

the

concentration range used above did not result in a concentration-dependent

increase

of ROS (Fig. 2e).

To investigate the regulation of cholesterol metabolism, MEF APP/APLP2-/- cells

were

treated with A, which reversed the effects that had been induced by the absence

of

APP/APLP2 (see Supplementary Information, Fig. S4c). To ascertain the mechanism

of A

in cholesterol homeostasis, MEF APP/APLP2-/- cells were treated with the two Aß

species separately and de novo synthesized cholesterol was analysed. In contrast

to

nSMase activity with A40, which showed only a minor effect, de novo cholesterol

synthesis responded already to low levels of A40 and partially restored

wild-type

levels (Fig. 2g, left).

To identify the enzyme in the cholesterol biosynthetic pathway targeted by A,

MEF

APP/APLP2-/- cells were incubated with the hydroxymethylglutaryl-CoA reductase

(HMGR) substrate HMG-CoA, or with mevalonate, the HMGR product. Increased de

novo

cholesterol synthesis was observed upon incubation with HMG-CoA, but not with

mevalonate (Fig. 2g, right). Genotype-specific alterations in cellular

cholesterol

release could be excluded (Fig. 2h). This identifies HMGR as a target in the

regulation of cholesterol de novo synthesis by APP/A.

The different effects of A40 (HMGR inhibition) and A42 (SMase activation) allow

for

a specific prediction of their activities in a cell-biological context. A change

in

A40/42 ratio should lead to altered lipid levels. Familial PS mutations (PS-FAD)

strongly increase A42 and mildly decrease A40 levels, without a relevant impact

on

the total A levels. Expression of PS-FAD mutations resulted in increased

cholesterol

and decreased SM levels, which was further confirmed by increased nSMase

activity

(Fig. 3). As predicted by the previous experiments, changes in the A40/42 ratio

alter lipid homeostasis.

In summary, our in vivo and in vitro data are consistent with a model (Fig. 4)

in

which the cascade of gamma-secretase, APP processing and A peptides downregulate

cholesterol and sphingolipid levels, which, in turn, regulate A generation. The

molecular targets for A in the cholesterol and SM metabolic pathways are HMGR

and

nSMase, respectively.

The role of A in lipid homeostasis implies that its precursor APP has a function

in

lipid homeostasis, which is supported by altered lipid homeostasis in APP-/-

mice.

We assume that additional gamma-secretase substrates and cleavage products

contribute to lipid homeostasis. This is supported by partial reversal of the

lipid

phenotype in cells following addition of A (Fig. 2g). Multiple gamma-secretase

products exist. Our data identify the involvement of A, but do not rule out the

involvement of other fragments. In the MEF APP/APLP2-/- cells, lipid homeostasis

could not be altered by inhibition of gamma-secretase proteolytic activity. This

identifies that APP/APLP2 in the MEF model are essential substrates. In

addition, it

rules out that, in the absence of APP/APLP2, other putative gamma-secretase

substrates have a detectable effect on SM and cholesterol homeostasis.

The gamma-secretase-dependent regulation of lipid homeostasis is likely to be

influenced by additional factors, as indicated by the other proteases that are

involved in APP RIP and gamma-secretase-independent regulatory cycles13.

In agreement with the ubiquitous presence of APP, A and PS25, 26, the mechanism

described here was found in cells of different origin.

Stimulation of SMase by addition of synthetic A peptides in the absence of any

other

modulating factors in a cell-free assay indicates that A42 directly interacts

with

nSMase. This is consistent with the presence of an nSMase luminal/extracellular

domain and the subcellular localizations of nSMases and A. Physiological A42

concentrations were not previously investigated in this context, but high

concentrations of A42 or fibrillary A42 were reported to cause oxidative stress,

to

affect lipid levels, but not cellular cholesterol content, and to stimulate

nSMase

in vitro by oxidative stress27, 28. This may provide an alternative explanation

for

the direct interaction of A with SMase. ROS effects were previously observed to

start occurring at A concentrations, coinciding with the highest A42 levels used

here. However, at this peptide concentration, nSMase activity already declined

in

our system. We observed, at the comparatively low A concentrations used here,

only

marginal differences in ROS levels. This indicates that either low A levels

produce

ROS levels below our detection limit or that, in agreement with previous work28,

low

A concentrations do not trigger sufficient ROS production to explain the

pronounced

increase in SMase activity observed here. Reduced nSMase activity in the

presence of

additional A40 (see Supplementary Information, Fig. S4b) and in the absence of a

dose-dependent response of ROS to A indicates that a ROS-independent interaction

is

more likely. This may represent a remarkable difference between the

physiological

and pathological function of A at different peptide concentrations.

Properties of synthetic peptides may not reflect the in vivo situation. The

cellular

experiments with PS-FAD mutations were carried out on an endogenous APP

background,

without synthetic peptides or protein overexpression. As these experiments

confirmed

the conclusions drawn from the use of synthetic peptides, we have to assume that

the

experiments with synthetic peptides gave sufficiently valid mechanistic

information.

Cholesterol, gamma-secretase and A represent factors in a regulatory cycle:

Cholesterol enhances gamma-secretase activity, thereby yielding more A. A40

suppresses HMGR activity, leading to decreased cholesterol levels. The absence

of

APP/APLP2 influences cholesterol de novo synthesis at the level of HMGR but not

downstream of it, clearly identifying HMGR as the target. In principle, the

structure of HMGR would allow a direct interaction with A, but it is currently

unknown whether or not this actually occurs.

After the discovery of the lipid-controlled RIP mechanism of the sterol

regulatory

element binding protein (SREBP) with site-1 protease and site-2 protease (S2P),

our

findings characterize a second proteolytic system in lipid homeostasis. Although

there are no sequence homologies, the functional analogies to the system

presented

here are striking13. Both systems act via HMGR, but an important difference is

that

APP processing reduces, whereas SREBP processing upregulates, cholesterol de

novo

synthesis. In a further analogy, the example of S2P suggests that not all

gamma-secretase substrates are processed lipid-dependently29.

The effect of PS-FAD mutations in SM homeostasis is inversed compared with

cholesterol homeostasis, validating the stronger activation of nSMase by A42 and

the

downregulation of de novo cholesterol synthesis by A40. Notably, this

differential

regulation rules out SM homeostasis as a causative mediator of PS-dependent

cholesterol levels.

Active involvement of APP cleavage, A, gamma-secretase activity and the effect

of

pathological PS-FAD mutations in lipid homeostasis provides a functional context

for

APP processing that has direct relevance for AD and may provide a rational basis

for

therapy.

Methods ...

Al Pater, PhD; email: old542000@...

__________________________________

Start your day with - Make it your home page!

http://www./r/hs

Link to comment
Share on other sites

--- Jeff Novick <jnovick@...> wrote:

Hi All,

Being in a rush, the below pdf-available excerpts are the best available

version.

http://my.webmd.com/content/article/113/110683?src=RSS_PUBLIC

Nature Cell Biology, Oct. 9, 2005 advance online edition.

Alzheimer's disease: Understanding the lipid connection

Letter by Grimm et al.

Varying cholesterol levels in mouse models has been shown to affect progress of

Alzheimer's disease. Hartmann and colleagues now show that Ab peptide, the

amyloid

peptide that is produced normally but accumulates markedly during Alzheimer's,

controls lipid metabolism by acting on two key enzymes in the pathway.

Marcus O. W. Grimm1, Heike S. Grimm1, s J. Pätzold1, Eva G. Zinser1,

Riikka

Halonen1, Marco Duering1, Jakob-A. Tschäpe1, Bart De Strooper2, Ulrike Müller3,

Jie

Shen4 & Tobias Hartmann1

Amyloid beta peptide (Aß) has a key role in the pathological process of

Alzheimer's

disease (AD), but the physiological function of Aß and of the amyloid precursor

protein (APP) is unknown1, 2. Recently, it was shown that APP processing is

sensitive to cholesterol and other lipids3, 4, 5, 6, 7, 8, 9, 10.

Hydroxymethylglutaryl-CoA reductase (HMGR) and sphingomyelinases (SMases) are

the

main enzymes that regulate cholesterol biosynthesis and sphingomyelin (SM)

levels,

respectively. We show that control of cholesterol and SM metabolism involves APP

processing. A42 directly activates neutral SMase and downregulates SM levels,

whereas A40 reduces cholesterol de novo synthesis by inhibition of HMGR

activity.

This process strictly depends on gamma-secretase activity. In line with altered

A40/42 generation, pathological presenilin mutations result in increased

cholesterol

and decreased SM levels. Our results demonstrate a biological function for APP

processing and also a functional basis for the link that has been observed

between

lipids and Alzheimer's disease (AD).

One of the main characteristics of Alzheimer's disease (AD) is the accumulation

of

amyloid beta peptide (A), which is associated with neurodegeneration1, 2.

A production is initiated by the gamma-secretase cleavage of amyloid precursor

protein (APP), which results in production of the APP C-terminal fragment C99.

This

fragment is further cleaved by gamma-secretase to generate A40 and A42. The

active

centre of the gamma-secretase complex is formed by presenilin-1 or -2 (PS1 or

PS2)11. As the gamma-secretase cleavage site is centred within the transmembrane

domain, membrane composition may considerably influence Aß generation12. Such

proteolytic events of regulated intra-membrane proteolysis (RIP) are involved in

numerous cellular regulatory events, including lipid homeostasis13.

So far, the physiological role of APP and its proteolytic fragments, including

A,

have mainly been conceived as being brain specific. However, the ubiquitous

production of APP, Aß and PS indicates that their functions are also ubiquitous.

Recently, strong evidence has accumulated that links AD and Aß generation with

lipid

homeostasis5, and it has been suggested that cellular lipid metabolism controls

APP

processing5. Studies have shown that Aß production is sensitive to cholesterol

levels and lipid trafficking. In vitro enhanced cholesterol levels lead to

increased

Aß production14. Furthermore, cholesterol depletion reduces gamma-secretase

activity

and results in reduced Aß generation3, 4, 6, 7, 8, 9. Controversially, one

report

found a correlation between mildly decreased cholesterol levels and increased Aß

production in Chinese hamster ovary cells15. In vivo, Aß production and

deposition

are tightly associated with changes in cholesterol levels6, 10.

Lipid levels vary in response to changes in diet, and physical or neuronal

activity.

Similarly, Aß levels change in response to several conditions, including

cholesterol

and other factors16, 17. Brain cholesterol levels increase during the

progression of

AD18 and hypercholesterolaemia is an early risk factor in the development of

AD19.

Consequently, cholesterol-lowering drugs are being considered as a potential

treatment for AD16, 20, 21. Nevertheless, the cellular mechanisms that link

lipids,

Aß generation and AD are poorly understood.

To investigate the role of PS in lipid regulation, we used mouse embryonic

fibroblast cells that were devoid of PS1 and PS2 (MEF PS1/2-/-)11 and

expression-level-matched, A-producing PS re-transfected MEF PS1/2-/- control

cells

(MEF PS1r and MEF PS2r, respectively) (see Supplementary Information, Fig. S1).

The

absence of PS resulted in increased cellular levels of cholesterol and

sphingomyelin

(SM), but not phosphatidylcholine (PC) or total lipid-associated phosphate (Fig.

1a). The catabolic sphingomyelinase (SMase) activities were downregulated,

anabolic

SM-synthase activity was upregulated and the SM-to-ceramide ratio was increased

in

PS-deficient cells in accordance with the altered neutral SMase (nSMase) and

SM-synthase activities (Fig. 1a, see Supplementary Information, Fig. S2a).

Absence of PS1 is lethal and, as PS2 is able to replace PS1 (for example, in

nSMase

activity, see Fig. 1c), brains from conditional PS1 knockout mice, crossed to

PS2

knockout mice, were used to evaluate PS function in lipid homeostasis in vivo.

In

these viable mice, PS1 remains expressed in some neurons and other cells22 and

brain

Aß levels in these mice are reduced to 20 & #8722;50%, depending on age and Aß

species23. As before, absence of PS in neurons resulted in increased cholesterol

and

SM levels, decreased SMase activity, but unchanged PC levels (Fig. 1b).

Dependence on gamma-secretase proteolytic activity was investigated using the

gamma-secretase activity inhibitor L-685,458. This treatment reduced nSMase

activity

in MEF PS1r and PS2r cells to similar levels as those of PS1/2-/- cells (Fig.

1c).

By contrast, this inhibitor treatment had no effect on MEF cells that were

devoid of

PS (Fig. 1c). Treatment of wild-type mouse neurons decreased nSMase activity and

increased SM levels (Fig. 1c). Accordingly, in inhibitor-treated PS-expressing

cells, SM levels were increased. Cholesterol levels increased in MEF PS1/2-/-

cells

(Fig. 1a), whereas no significant difference was observed between

inhibitor-treated

MEF PS1r and MEF PS1/2-/- cells (see Supplementary Information, Fig. S2b).

It was established previously that cholesterol influences Aß production. To

investigate Aß putative influence of SM on Aß production, we inhibited nSMase

activity in neurons. This resulted in increased SM levels and decreased

intracellular and secreted Aß (Fig. 2a). This was confirmed in COS7 cells using

additional inhibitors and direct SM exposure (see Supplementary Information,

Fig.

S3a, 3b). Reduced gamma-secretase cleavage causes C99 accumulation, which was

observed with increased SM levels (see Supplementary Information, Fig. S3c).

Our results implicate the involvement of a gamma-secretase substrate and that

its

cleavage is essential in PS-mediated lipid regulation. MEF APP/APLP2-/- and MEF

wild-type cells were analysed to evaluate whether APP or APLP2 might be involved

in

this regulatory cascade. SM, nSMase activity and cholesterol levels responded to

the

absence of APP/APLP2 in the same way as they did to the absence of

PS/gamma-secretase activity (Fig. 2b). The same was observed in APP-/- mouse

brain

(Fig. 2b), suggesting that an in vivo function of APP exists in lipid

homeostasis.

Unlike in wild-type cells, inhibition of gamma-secretase activity did not alter

cholesterol in MEF APP/APLP2-/- cells (Fig. 2c). This seems to rule out the

existence of other gamma-secretase substrates that are able to replace the

APP/APLP2

function in cholesterol regulation in these cells. The combined necessity of APP

and

gamma-secretase activity in this process indicates that a proteolytic APP

fragment

is involved in lipid homeostasis (Fig. 2b, c). The most prominent peptide that

is

derived from APP gamma-secretase cleavage is A. To determine a possible function

of

Aß in cholesterol and SM regulation, we analysed key enzymes and metabolites in

the

respective pathways.

Analysing SM metabolism in MEF APP/APLP2-/- cells showed a decrease in nSMase

activity compared with MEF wild-type cells (Fig. 2b). When exposed to

A-containing

conditioned media from human SH-SY5Y cells, the nSMase enzymatic activity was

partially restored in MEF APP/APLP2-/- cells (Fig. 2d, left). To investigate

whether

this might be caused by A, cell homogenates of MEF APP/APLP2-/- cells were

exposed

to synthetic A1 & #8722;40/A1 & #8722;42 at a physiological ratio24, resulting in

enhanced nSMase activity (Fig. 2d, right). To determine whether SMase is a

molecular

target for regulation by Aß and to evaluate whether this interaction might be

direct, purified human placenta nSMase was exposed to synthetic Aß in a

biochemical

assay that was free of other cellular components. Covering the physiological

concentration range of A24, the effect reached a maximum at peptide levels that

are

typically found in cerebrospinal fluid. The effect decreased thereafter,

probably

due to augmented peptide aggregation (Fig. 2e). An effect of Aß on the other

assay

components was excluded (see Supplementary Information, Fig. S4a); A42 & #8722;1

(inverse A) was inactive and aggregated Aß had only residual stimulatory

activity,

presumably due to remnant non-aggregated peptide. An excess of A40 reduced

nSMase

activation by A42, indicating A42 specificity (see Supplementary Information,

Fig.

S4b).

To evaluate the involvement of oxidative stress in nSMase activation by Aß

peptides,

the production of reactive oxygen species (ROS) was investigated. No significant

differences in ROS were observed when wild-type and rescue cells were compared

with

the respective knockout cells (Fig. 2f). Incubation of MEF PS1/2-/- with Aß in

the

concentration range used above did not result in a concentration-dependent

increase

of ROS (Fig. 2e).

To investigate the regulation of cholesterol metabolism, MEF APP/APLP2-/- cells

were

treated with A, which reversed the effects that had been induced by the absence

of

APP/APLP2 (see Supplementary Information, Fig. S4c). To ascertain the mechanism

of A

in cholesterol homeostasis, MEF APP/APLP2-/- cells were treated with the two Aß

species separately and de novo synthesized cholesterol was analysed. In contrast

to

nSMase activity with A40, which showed only a minor effect, de novo cholesterol

synthesis responded already to low levels of A40 and partially restored

wild-type

levels (Fig. 2g, left).

To identify the enzyme in the cholesterol biosynthetic pathway targeted by A,

MEF

APP/APLP2-/- cells were incubated with the hydroxymethylglutaryl-CoA reductase

(HMGR) substrate HMG-CoA, or with mevalonate, the HMGR product. Increased de

novo

cholesterol synthesis was observed upon incubation with HMG-CoA, but not with

mevalonate (Fig. 2g, right). Genotype-specific alterations in cellular

cholesterol

release could be excluded (Fig. 2h). This identifies HMGR as a target in the

regulation of cholesterol de novo synthesis by APP/A.

The different effects of A40 (HMGR inhibition) and A42 (SMase activation) allow

for

a specific prediction of their activities in a cell-biological context. A change

in

A40/42 ratio should lead to altered lipid levels. Familial PS mutations (PS-FAD)

strongly increase A42 and mildly decrease A40 levels, without a relevant impact

on

the total A levels. Expression of PS-FAD mutations resulted in increased

cholesterol

and decreased SM levels, which was further confirmed by increased nSMase

activity

(Fig. 3). As predicted by the previous experiments, changes in the A40/42 ratio

alter lipid homeostasis.

In summary, our in vivo and in vitro data are consistent with a model (Fig. 4)

in

which the cascade of gamma-secretase, APP processing and A peptides downregulate

cholesterol and sphingolipid levels, which, in turn, regulate A generation. The

molecular targets for A in the cholesterol and SM metabolic pathways are HMGR

and

nSMase, respectively.

The role of A in lipid homeostasis implies that its precursor APP has a function

in

lipid homeostasis, which is supported by altered lipid homeostasis in APP-/-

mice.

We assume that additional gamma-secretase substrates and cleavage products

contribute to lipid homeostasis. This is supported by partial reversal of the

lipid

phenotype in cells following addition of A (Fig. 2g). Multiple gamma-secretase

products exist. Our data identify the involvement of A, but do not rule out the

involvement of other fragments. In the MEF APP/APLP2-/- cells, lipid homeostasis

could not be altered by inhibition of gamma-secretase proteolytic activity. This

identifies that APP/APLP2 in the MEF model are essential substrates. In

addition, it

rules out that, in the absence of APP/APLP2, other putative gamma-secretase

substrates have a detectable effect on SM and cholesterol homeostasis.

The gamma-secretase-dependent regulation of lipid homeostasis is likely to be

influenced by additional factors, as indicated by the other proteases that are

involved in APP RIP and gamma-secretase-independent regulatory cycles13.

In agreement with the ubiquitous presence of APP, A and PS25, 26, the mechanism

described here was found in cells of different origin.

Stimulation of SMase by addition of synthetic A peptides in the absence of any

other

modulating factors in a cell-free assay indicates that A42 directly interacts

with

nSMase. This is consistent with the presence of an nSMase luminal/extracellular

domain and the subcellular localizations of nSMases and A. Physiological A42

concentrations were not previously investigated in this context, but high

concentrations of A42 or fibrillary A42 were reported to cause oxidative stress,

to

affect lipid levels, but not cellular cholesterol content, and to stimulate

nSMase

in vitro by oxidative stress27, 28. This may provide an alternative explanation

for

the direct interaction of A with SMase. ROS effects were previously observed to

start occurring at A concentrations, coinciding with the highest A42 levels used

here. However, at this peptide concentration, nSMase activity already declined

in

our system. We observed, at the comparatively low A concentrations used here,

only

marginal differences in ROS levels. This indicates that either low A levels

produce

ROS levels below our detection limit or that, in agreement with previous work28,

low

A concentrations do not trigger sufficient ROS production to explain the

pronounced

increase in SMase activity observed here. Reduced nSMase activity in the

presence of

additional A40 (see Supplementary Information, Fig. S4b) and in the absence of a

dose-dependent response of ROS to A indicates that a ROS-independent interaction

is

more likely. This may represent a remarkable difference between the

physiological

and pathological function of A at different peptide concentrations.

Properties of synthetic peptides may not reflect the in vivo situation. The

cellular

experiments with PS-FAD mutations were carried out on an endogenous APP

background,

without synthetic peptides or protein overexpression. As these experiments

confirmed

the conclusions drawn from the use of synthetic peptides, we have to assume that

the

experiments with synthetic peptides gave sufficiently valid mechanistic

information.

Cholesterol, gamma-secretase and A represent factors in a regulatory cycle:

Cholesterol enhances gamma-secretase activity, thereby yielding more A. A40

suppresses HMGR activity, leading to decreased cholesterol levels. The absence

of

APP/APLP2 influences cholesterol de novo synthesis at the level of HMGR but not

downstream of it, clearly identifying HMGR as the target. In principle, the

structure of HMGR would allow a direct interaction with A, but it is currently

unknown whether or not this actually occurs.

After the discovery of the lipid-controlled RIP mechanism of the sterol

regulatory

element binding protein (SREBP) with site-1 protease and site-2 protease (S2P),

our

findings characterize a second proteolytic system in lipid homeostasis. Although

there are no sequence homologies, the functional analogies to the system

presented

here are striking13. Both systems act via HMGR, but an important difference is

that

APP processing reduces, whereas SREBP processing upregulates, cholesterol de

novo

synthesis. In a further analogy, the example of S2P suggests that not all

gamma-secretase substrates are processed lipid-dependently29.

The effect of PS-FAD mutations in SM homeostasis is inversed compared with

cholesterol homeostasis, validating the stronger activation of nSMase by A42 and

the

downregulation of de novo cholesterol synthesis by A40. Notably, this

differential

regulation rules out SM homeostasis as a causative mediator of PS-dependent

cholesterol levels.

Active involvement of APP cleavage, A, gamma-secretase activity and the effect

of

pathological PS-FAD mutations in lipid homeostasis provides a functional context

for

APP processing that has direct relevance for AD and may provide a rational basis

for

therapy.

Methods ...

Al Pater, PhD; email: old542000@...

__________________________________

Start your day with - Make it your home page!

http://www./r/hs

Link to comment
Share on other sites

Join the conversation

You are posting as a guest. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...