(Received for publication, May 8, 1996, and in revised form, October 9, 1996)
From the Niemann-Pick type C disease (NPC) belongs to the
group of lysosomal storage diseases characterized by an accumulation of
cholesterol and sphingomyelin. Using a mutant mouse strain, enzymatic
markers for lysosomes, mitochondria, microsomes, and peroxisomes were investigated in the liver and brain. Aside from lysosomal changes, we
found a sizable decrease of peroxisomal Niemann-Pick type C disease is characterized by the accumulation
of sphingomyelin and cholesterol in the lysosomes of several organs
(1). The genetic defect responsible for this disease is not yet known,
but the gene has been localized to chromosome 18 (2). The accumulation
of sphingomyelin is considered to be an event secondary to the
malfunction of cholesterol transport from the lysosomes to the
cytoplasm. Experiments on human fibroblasts from these patients have
demonstrated that the endocytosed low density lipoprotein cholesterol
is not discharged from the lysosomes and therefore attenuated
down-regulation of the key enzyme of cholesterol synthesis,
hydroxymethylglutaryl-CoA reductase, occurs (3). Using a mutant mouse
model of this disease, it was found that other mevalonate pathway
lipids are also affected (4). The concentration of dolichyl phosphate
is greatly increased in all organs, and the dolichol transport from the
endoplasmic reticulum to the lysosomes is inhibited. There is also a
relative increase of the higher isoprene number species of both the
dolichol and dolichyl phosphate fractions.
Not only are the contents of the mevalonate pathway lipids and
sphingomyelin modified in the disease, several other lipids are also
changed, particularly in the liver. Bis(monoacylglycerol)phosphate, glycosylceramide, and sphingoid bases are reported to be elevated (5-7). In the brain, abnormalities are more restricted and mainly affect glycolipids. A severalfold increase of glucosylceramide and
lactosylceramide and an increase of
GM21 and GM3
gangliosides have been found. In the liver of mutant mice the amount of
sterol carrier protein 2, a protein suggested to be involved in
cholesterol transport and biosynthesis, is decreased by 80% (8). Most
of the various lipid accumulations have been associated with lysosomes,
and the levels of several lysosomal enzymes were found to be altered
(9).
The extensive modifications in cholesterol and dolichol metabolism and
transport raise several questions concerning the biochemical etiology
of this disease. These two lipids are synthesized not only in
microsomes but also in peroxisomes and are subjected to independent
regulations (10-12). In this study, we have therefore investigated the
possibility that peroxisomes are also modified in Niemann-Pick type C
disease, which is generally considered to be elicited by an enhanced
lysosomal storage of lipids. The analyses led to the novel finding that
in this condition, in contrast to the lysosomal disorders, a deficiency
of peroxisomal function is an early event.
In all experiments, a mutant BALB/c mouse model of
NPC was used. The parental strain, from which the mutant mice were
derived, served as a control. The controls were age-matched to the
diseased mice. Since NPC is an autosomal recessively inherited disease, only 25% of the mice develop pathological symptoms. The most prominent symptom is ataxia, which appears around 48-50 days of age. The mice
showing ataxia were selected for investigation within 3 days.
After
decapitation, tissues were quickly removed and homogenized in 0.25 M cold sucrose (1 g/10 ml) with an ultraturrax blender at
20,000 rpm for 1 min, which effectively disrupts all tissues. This
homogenate was centrifuged at 310 × g for 10 min. The
supernatant was centrifuged at 2800 × g for 15 min to
obtain the heavy mitochondrial/lysosomal fraction (13). The resulting
supernatant was centrifuged at 25,300 × g for 15 min,
and the pellet was called the light mitochondrial/lysosomal fraction.
The supernatant above the latter pellet was centrifuged at 105,000 × g for 60 min to obtain the microsomal fraction.
Mouse liver peroxisomes were isolated on a Nycodenz gradient as
described previously (14).
Peroxisomal fatty acyl-CoA oxidase activity
was determined using lauroyl-CoA, palmitoyl-CoA, or arachidonyl-CoA as
substrates. The H2O2 produced was quantified
fluorometrically by following the horseradish peroxidase-catalyzed
oxidation of 4-hydroxyphenyl acetic acid into
6,6 The protein content was determined according to the method of Lowry
et al. (21) with bovine serum albumin as the standard.
Chloroform/methanol/water in a ratio of 1:1:0.3 (v/v)
was used for the extraction of lipids from tissue homogenates. The
pelleted residue was re-extracted with chloroform/methanol (2:1, v/v), and the two supernatants were pooled and the non-lipid residue was used
for DNA quantification according to Burton (22). The lipid extract was
placed onto a Silica-SepPak cartridge, and the neutral lipids were
eluted with chloroform and phospholipids with methanol. All major
phospholipid classes were separated by high performance liquid
chromatography (HPLC) (23). The phospholipid fraction was dissolved in
hexane/2-propanol (3:2, v/v) and injected onto a Zorbax SIL (25 cm × 4.6 mm inner diameter, 5-6 µm) column maintained at 34 °C.
Solvent A was hexane/2-propanol (3:2, v/v), and solvent B was
hexane/2-propanol/water (56.7:37.8:5.5, v/v). The flow rate was 1.5 ml/min, and the absorbance of the eluant was monitored at 205 nm. The
different phospholipid classes were collected and quantified by
measuring lipid phosphorus as described previously (24).
The diacyl, alkenylacyl, and alkylacyl subclasses of PE were separated
by HPLC after their conversion to 1,2-diradylglycerol derivatives, as
described previously (25). The polar head group was removed with
phospholipase C (Bacillus cereus, Sigma),
and the 1,2-diradylglycerol mixture obtained was acetylated with
pyridine and acetic anhydride. Separation of the acetyl diradylglycerol subclasses was achieved on a Zorbax SIL (25 cm × 4.6 mm inner diameter, 5-6 µm) column by an isocratic system consisting of cyclopentane/hexane/t-butylmethyl ether/acetic acid
(73:24:2:0.01, v/v). The column was maintained at 36 °C, and the
flow rate was 1.5 ml/min. The quantities of alkenylacyl, alkylacyl, and
diacyl subclasses were determined by gas chromatography of the fatty acid methyl esters, using heptadecanoic acid as an internal
standard.
Lipids were methylated in 14%
boron trifluoride in methanol (26), and the resultant fatty acyl methyl
esters and dimethyl acetals were extracted with n-pentane/5
M NaOH (4:1, v/v). The upper pentane phase was analyzed by
GC on a Shimadzu GC-9A (Kyoto, Japan) equipped with a flame ionization
detector. Hydrogen was used as the carrier gas, and the column was a
fused silica capillary with a 0.18-µm cyanosiloxane 60 film as the
stationary phase (30 m × 0.25 mm inner diameter). The temperature
was programmed from 130 °C to 190 °C, and total time per run was
45 min. Fatty acyl methyl esters and dimethyl acetals were identified
by comparison of their retention times with known standards.
Quantification was achieved by integrating the peaks on a Shimadzu
Chromatopac C-R3A.
For electron microscopy, peroxisomes
were stained with 3,3 Several enzyme activities, known to be
associated with specific organelles, were analyzed in the liver
homogenates of mice exhibiting clinical signs of NPC disease,
i.e. in the initial phase, around 48-50 days of life. The
specific activity of NADPH-cytochrome c reductase, an enzyme
present only in the endoplasmic reticulum, was identical in diseased
mice and the age-matched controls (Table I). The
activity of cytochrome oxidase, a component of the inner mitochondrial
membrane, was also unaltered. On the other hand, three enzymes
associated with peroxisomes exhibited reduced activity in comparison
with the controls. There was a moderate decrease of urate oxidase and
catalase activities (20-30%), and the
Enzyme activities in liver homogenate
Department of Biochemistry,
NINDS, National Institutes of
Health, Bethesda, Maryland 20892
-oxidation of fatty acids
and catalase activity in the brain and liver. Isolated peroxisomes displayed a significant decrease of these enzyme activities.
Furthermore, the only phospholipid change in brain was a decreased
content of the plasmalogen form of phosphatidylethanolamine, and
the dimethylacetal pattern was also modified. The electron
microscopical appearance of peroxisomes did not display any large
changes. The defect of peroxisomal enzymes was already present 18 days
before the onset of the disease. In contrast, the lysosomal marker
enzyme increased in activity only 6 days after appearance of the
symptoms. The events of the studied process have previously been
considered to be elicited by a lysosomal deficiency, but this study
demonstrates disturbances similar to those in a number of peroxisomal
diseases. It appears that the peroxisomal impairment is an early event
in the process and could be a factor in the development of Niemann-Pick type C disease.
Animals
-dihydroxy-(1,1
-biphenyl)-3,3
-diacetic acid (15). Palmitoyl-CoA
oxidation was monitored by following the reduction of NAD
spectrophotometrically at 340 nm. KCN was used to inhibit mitochondrial
-oxidation (16). Catalase activity in the liver was measured
spectrophotometrically at 240 nm by following the disappearance of
H2O2 (17). Another catalase assay with higher
sensitivity was used for brain samples. After incubation, the remaining
H2O2 was quantified by adding titanium oxide
sulfate, and the yellow peroxytitanium sulfate complex was measured at 405 nm (18). Urate oxidase was measured by following the oxidation of
urate as a decrease in absorbance at 292 nm (18). Cytochrome oxidase
and NADPH-cytochrome c reductase were determined as
described earlier (18, 19). Acid phosphatase was measured
spectrophotometrically at 400 nm, by determining the
Na-K-tartrate-inhibited cleavage of p-nitrophenyl phosphate
into p-nitrophenol (20).
-diaminobenzidine (27). The samples were fixed in
a mixture containing 3% glutaraldehyde, 0.2 M sodium
cacodylate chloride, and 0.25 M sucrose (pH 7.2) at 4 °C
overnight. Postfixation was performed in 1% osmium tetroxide in 0.15 M sodium cacodylate chloride, pH 7.2, for 90 min at
20 °C. The fixed samples were dehydrated and embedded in epoxy
resin.
Enzyme Composition
-oxidation enzyme,
lauroyl-CoA oxidase, activity was decreased by 50%.
Enzyme
Liver
Control
NPC
NADPH-cytochrome c
reductasea
3.31 ± 0.41
3.37
± 0.33
Cytochrome oxidaseb
0.133
± 0.015
0.154 ± 0.016
Urate oxidasec
0.070
± 0.005
0.057 ± 0.006*
Catalased
0.406
± 0.055
0.297 ± 0.028*
Lauroyl-CoA
oxidasee
0.891 ± 0.120
0.437 ± 0.028***
a
nmol cytochrome c reduced/(min × mg
protein).
b
µmol cytochrome c oxidized/(min × mg
protein).
c
µmol urate/(min × mg protein).
d
mmol H2O2/(min × mg protein).
e
nmol/(min × mg protein).
Differential centrifugation was performed to obtain a heavy mitochondrial/lysosomal fraction and a light mitochondrial/lysosomal fraction, as well as microsomes. The mitochondrial fractions contained the peroxisomes. The peroxisomes present in the two fractions exhibited different properties (Table II). The specific activities of both catalase and lauroyl-CoA oxidase were greatly decreased in the heavy mitochondrial fraction of the diseased mice, while the decrease was quite limited in the light mitochondrial fraction. The specific activity of acid phosphatase was doubled in the two mitochondrial fractions from diseased mice, which is in agreement with the finding that lysosomes are affected in NPC disease. There was also a 3-fold increase of acid phosphatase activity in the microsomes, which may be explained by the slow sedimentation of a fraction of lipid-filled lysosomal vesicles remaining in the microsomal fraction. Both cytochrome oxidase and NADPH-cytochrome c reductase activity remained constant when compared to control mice, indicating that neither the mitochondria nor the endoplasmic reticulum are affected by the disease.
|
In order to verify the defects at the organelle level, peroxisomes were isolated using Nycodenz gradients (Table III). Using this procedure, the fraction obtained has a very low cross-contamination with other organelles but the recovery is low and prevents extensive investigation. The isolated peroxisomes exhibited 40-50% decreased urate oxidase, catalase, and lauroyl-CoA oxidase activities when the isolated fractions from NPC mice were compared to the control.
|
Brain homogenates were also analyzed for peroxisomal activities. Catalase activity, also in this tissue, was found to be decreased (Table IV). Since brain peroxisomes preferentially oxidize long chain fatty acids, arachidonyl-CoA oxidase was measured and was found to be decreased by as much as 70% in diseased animals.
|
The peroxisomal
-oxidation of FA in the liver homogenates was determined by
measuring H2O2 production in the presence of various FA-CoA species (Fig. 1). The acyl-CoA oxidase
activity measured with octanoyl-, lauroyl-, myristoyl-, palmitoyl-, and arachidonyl-CoA as substrates was decreased by half or more as compared
to the control, indicating a generalized impairment of peroxisomal
-oxidation.
Developmental Changes
Symptoms of the disease become apparent
around 48 days of age, and the mice die at around 70 days. The
peroxisomal and lysosomal activities in the liver were followed during
the time period after development of clinical symptoms. When the
symptoms first appeared, palmitoyl-CoA oxidation and lauroyl-CoA
oxidase as well as catalase and urate oxidase activities were greatly
diminished, but activities gradually recovered and almost reached the
level of control values by the end of the 3rd week (Fig.
2). In contrast, lysosomal acid phosphatase activity was
about the same as the control during the first few days and reached
maximal amplification 13 days later.
The experiment above suggested that the peroxisomal enzyme activities
may already be decreased before the onset of overt clinical symptoms.
Groups of mice pooled into affected and nonaffected were selected for
analysis of the liver (Table V). The average age of the
mice was 31 days, 18 days before the symptoms are expected to appear.
Mice exhibiting elevated levels of cholesterol and dolichyl phosphate
were considered as homozygous for the NPC mutation. This group of
animals exhibited a considerably decreased activity of peroxisomal
-oxidation, indicating that the peroxisomal deficiency is manifested
long before the appearance of the disease symptoms.
|
Hepatocytes contain morphologically well characterized peroxisomes with a rich proteinaceous content in the lumen. We have used a diaminobenzidine staining procedure, which detects peroxisomal catalase as a granular precipitation (data not shown). No obvious differences in appearance and distribution could be observed between control and NPC liver. Investigating a series of images from both control and diseased tissues, we could not see any great differences in the number or size of these organelles.
Phospholipids of the Brain and LiverThe major phospholipid
components of the mouse brain, PE and phosphatidylcholine, are present
in about equal amounts (Fig. 3A). The brain
is also relatively rich in phosphatidylserine and sphingomyelin, while
cardiolipin and phosphatidylinositol are smaller components. No
clear-cut difference in total amount of phospholipid could be observed
between control and NPC mouse brain. On the other hand, after
phospholipid class separation there was a small decrease of PE content.
PE consists of two main subclasses, the alkenylacyl and the diacyl
forms (Fig. 4). The alkylacyl PE, the precursor of the
alkenylacyl form, makes up only a few percent of the total. In the
diseased brain, the amount of alkenylacyl PE was decreased by one
third, while the levels of the two other forms were unchanged. Thus,
the decrease of PE in NPC brain is solely caused by a decrease of PE
plasmalogen. Some increase of sphingomyelin was also noticed, but this
change was not statistically significant.
In NPC liver, the total phospholipid level on DNA basis was increased from 12.89 ± 1.23 to 15.71 ± 1.05 mg of phospholipid/mg of DNA. This is explained by the severalfold elevation of the sphingomyelin content (Fig. 3B). The dominating phospholipid, phosphatidylcholine, as well as cardiolipin, phosphatidylinositol, and phosphatidylserine, all remained unchanged.
Fatty Acid CompositionThe fatty acid patterns of total lipids in the brain and liver are very different (Table VI). Dimethyl acetals (DMAs) are the aldehyde associated with carbon 1 on the plasmalogens. They are present in the liver in only minute amounts because of the very low level of plasmalogens in this tissue and therefore were not determined. In both organs, the FA composition of total lipids remained completely unchanged in the diseased state. In the brain, DMAs are restricted to a few components, consisting of 16:0, 18:0, and 18:1. There were significant changes in DMA distribution in the case of NPC brain, apparent as an increase of the 16:0 and a decrease of the 18:1 species.
|
Because of the observed modifications of the DMA pattern, we also analyzed the FA composition of PE subclasses (Table VII). The plasmalogen form was highly enriched in polyunsaturated FA, while the diacyl form was dominated by the 16:0 and 18:0 species. The FA composition of alkenylacyl and diacyl PE was very similar in both control and NPC mice. The DMA composition of the plasmalogen fraction of the PE was, as expected, similar to that found for the total lipids. In the diseased state, 16:0 was elevated, 18:1 decreased, and 18:0 remained unchanged.
|
The C-type Niemann-Pick disease represents a deviation from other forms of NP, since it is dominated by cholesterol accumulation in the lysosomes, caused by a lesion in the intracellular trafficking of endocytosed cholesterol. The murine model of the disease makes possible a detailed study of the enzyme modifications involved. Using this system, we have found impairments in peroxisomal activities, indicating the involvement of this organelle in the pathogenesis of this disorder. These observations reveal new aspects of the disease process, as they demonstrate that peroxisomal deficiencies occur long before the manifestation of the disease symptoms.
There was a sizable decrease of all peroxisomal enzymes determined in
the liver, i.e. -oxidation of various types of fatty acids + catalase and urate oxidase. The impairment of
-oxidation and
catalase was also expressed in the brain of these mice, indicating involvement of peroxisomal dysfunction in multiple organs. Furthermore, the only change in brain phospholipids was a 33% decrease of PE plasmalogens. The initial portion of plasmalogen synthesis, formation of the ether bond, takes place in peroxisomes (28), which explains why
only plasmalogen and not other phospholipids are affected. The dimethyl
acetal pattern of PE plasmalogens was also substantially modified,
and, consequently, the decrease in PE plasmalogens is probably caused
by an impaired peroxisomal synthesis rather than increased
breakdown.
Some peroxisomal diseases, such as Zellweger syndrome, are
characterized by accumulation of very long chain fatty acids (29). In
these cases, the peroxisomal -oxidation is completely missing. In
NPC mice, the
-oxidation is only partially deficient and no changes
in the lipid-bound fatty acids could be observed. Thus, it appears that
the residual
-oxidation capacity is sufficient enough, which,
together with the relatively short life span of the diseased mice,
prevents accumulation of very long chain fatty acids.
Electron microscopical investigations did not reveal great changes in the structure and the number of peroxisomes in the diseased mice. This finding is not surprising since in many identified peroxisomal diseases where partial defects of enzyme systems have been observed, the peroxisomal morphology remained unchanged (30). This fact, however, does not exclude the possibility that structural modifications at the organelle level exist. In our experiments, the peroxisomes present in the heavy mitochondrial fraction displayed considerable enzyme deficiencies in contrast to those present in the light mitochondrial fraction. This suggest that a unique population of this organelle may be responsible for the anomalies found in the disease process.
Isolated peroxisomes exhibit a 40-50% decrease of various peroxisomal enzyme activities, and it would be of considerable interest to perform subfractionation studies to investigate the possible heterogeneity in this fraction. Such a subfractionation study, however, has not yet been performed mainly because of the small number of peroxisomes in the liver and the extremely poor recovery afforded by the present methods employed.
It has previously been found that in the NPC mouse liver, sterol carrier protein-2 exhibits an 80% decrease in the postnuclear fraction (8). This protein is believed to be necessary in certain reactions, including the transformation of lanosterol to cholesterol (31), the activation of the microsomal cis-prenyltransferase (32), and the transport of steroids (33). In the liver, the major part of this protein is present in peroxisomes and the rest is a cytosolic component (34). The large decrease of the level of sterol carrier protein-2 in NPC lends further support to the participation of peroxisomes in the cellular pathogenic events of this disorder.
NPC is associated with a genetic defect on chromosome 18. The nuclear
hormone receptor superfamily has a member that is crucial for
peroxisome biogenesis, known as the peroxisome proliferator activated
receptor, PPAR. PPAR consists of three isoforms, which are activated by
distinct signaling pathways and regulated by various other nuclear
hormone receptors (35-37). For example, it has been shown that
glucocorticoids can stimulate synthesis of PPAR mRNA and that
this activation was mediated via the glucocorticoid receptor (38). Thus
it is plausible that the mutation on chromosome 18 could interfere with
peroxisomal activation.
The peroxisomal changes appear not only to be a part of an adaptation mechanism to the disease manifestation but already exist before the appearance of clinical symptoms. During the early phases of clinical presentation, when symptoms like ataxia first appear, peroxisomal enzyme activities are greatly decreased, while lysosomal enzyme markers appear less affected. Surprisingly, all peroxisomal enzyme activities begin to normalize during the late phase of the disease and begin to approach the control values. In contrast, the initially normal acid phosphatase increases continuously and is doubled 2 weeks after the appearance of the symptoms. These results demonstrate that peroxisomal modifications play a significant role in initial events and thereby are important in understanding the etiology of the NPC disease process.