![]() |
|
|

Departments of Biochemistry and
*
Pathology, and
Division of Animal Resources, Emory University, Atlanta, GA 303223050 and
**
Department of Biology, Williams College, Williamstown, MA 01267
4To whom correspondence should be addressed.
| ABSTRACT |
|---|
|
|
|---|
KEY WORDS: sphingolipids diet disease cancer functional foods
| INTRODUCTION |
|---|
|
|
|---|
| Structures of sphingolipids |
|---|
|
|
|---|
|
- (or, in the
case of the ceramides of skin, the
-) carbon atom. More complex
sphingolipids have a polar headgroup at position 1, as illustrated by a
few examples in Figure 1| Occurrence and functions |
|---|
|
|
|---|
|
,25-dihydroxycholecalciferol
and a growing list of agonists and toxins (and toxic insults, such as
-radiation) (for reviews see Kolesnick 1998
(TNF-
) usually activates only sphingomyelinase, which
results in ceramide accumulation. These differences have profound
effects on the behavior of the cells because sphingosine 1-phosphate is
a potent mitogen and an inhibitor of apoptosis (Cuvillier et al. 1998| Sphingolipids in food |
|---|
|
|
|---|
Table 1
summarizes the amounts of sphingolipids in food, estimated as closely
as possible from the available literature. The amounts vary
considerably, from the low micromoles per kilogram in fruits and some
vegetables to ~2 mmol/kg (12 g/kg) in dairy products, egg and
soybeans. It should be noted that the studies from which we calculated
these amounts were designed in large part to elucidate the chemical
structures of specific classes of sphingolipids rather than to quantify
the sphingolipid content. Many utilized indirect measurements such as
the phosphorous content of sphingomyelin (Blank et al. 1992
, Zeisel et al. 1986
), the hexose content of
cerebrosides (Walter et al. 1971
, Whitaker 1996
) or the total lipid nitrogen content (Gaillard
1968a
and 1968b
); a few employed HPLC or
gas chromatography (GC) to characterize individual molecular species
(see Cahoon and Lynch 1991
, Whitaker 1996
, Zeisel 1994
, for examples). Thus,
depending on the procedures that were employed, the studies provided
information about the content of an individual sphingolipid class
(usually selected because it was the major species) or the sum for a
group of compounds. Except for milk (Jensen 1995
,
Keenen and Patton 1995
), little is known about variation
in sphingolipid amounts over season (day of lactation, in the case of
milk), losses during food preparation and other aspects of food
chemistry. As far as we are aware, this is the first collation of data
on the sphingolipid content and types in food, and there is clearly a
need for further analyses.
|
The items in Table 1
cover almost 80% (by weight) of the foods
consumed in the United States; the remainder is comprised of caloric
sweeteners (9%) (which do not contain sphingolipids), other vegetables
and fruits (12%) and miscellaneous (3%). Therefore, using these data,
an approximation of the yearly consumption of sphingolipids from each
source was prepared. Dairy products appear to be major sources,
followed by meat and fish, eggs, and vegetables; the contribution from
vegetables was the most difficult to estimate from available data.
Yearly per capita intake of sphingolipids from the foods in Table 1
is
154 mmol, which is equivalent to ~116 g. If fruits and vegetables
contribute the higher estimates for these categories (based on the
average content of known fruits and vegetables, as described in
footnotes 7 and 8 of Table 1
), this adds another 28 mmol, for a total
of 181 mmol (139 g) per year. Based on a yearly per capita food
consumption of 873 kg, sphingolipids constitute from 0.01 to 0.02% of
the diet (by weight). This amount (0.30.4 g/d) provides few "fat
calories" but is comparable, instead, to lipids such as cholesterol
and tocopherols (Ensminger et al. 1994
). Consumption
could be higher than this estimate, for the reasons enumerated above,
and could vary considerably among individuals who consume foods that
are particularly rich in sphingolipids.
Structural variation of sphingolipids in food.
The sphingolipid backbones, fatty acids and headgroups vary
considerably with the type of food. Most foods of mammalian origin
(beef, milk or poultry, for example) have a wide spectrum of complex
sphingolipids (sphingomyelins, cerebrosides, globosides, gangliosides
or sulfatides) that are comprised of many different headgroup
components (phosphocholine, glucose, galactose,
N-acetylglucosamine, N-acetylgalactosamine,
N-acetylneuraminic acid, fucose and other carbohydrates) and
ceramide backbones (d18:1
4, d18:0 and t18:0,
with amide-linked fatty acids of 1630 carbon atoms in length,
some of which have an
-hydroxyl group) (Merrill and Sweeley 1996
). For example, milk contains (per L) 39119 mg of
sphingomyelin, 611 mg of glucosylceramide, 6.515 mg of
lactosylceramide and ~11 mg of gangliosides (~913 mg
GD3, 1.2 mg GD1b1, 0.7 mg
GM2, 0.3 mg GM3 and 0.001
mg GM1) (see Jensen 1995
for a review); the lipid
backbones of milk sphingomyelin have mainly sphingosine
(d18:1t4, with smaller amounts of sphinganine and
other chain length homologs) and 16:0, 22:0, 23:0 and 24:0 as the major
fatty acids (Morrison 1969
, Schmelz et al. 1996
, Zeisel et al. 1986
).
In contrast, the complex sphingolipids of plants are mainly
cerebrosides (mono- and oligohexosylceramides) with glucose (Glc, the
most common hexose), galactose (Gal), mannose (Man) and
inositol.7
Examples are as follows: wheat grain has glycosphingolipids with
primarily Glc, but also, Man-Glc,
[Man]2-Glc and
[Man]3-Glc headgroups, and has the sphingoid
base backbones d18:1
4,
d18:1
8, d18:2
4,8,
t18:0 and t18:1
8 with 14:026:0 fatty acids
(most as
-hydroxy fatty acids) (Fujino et al. 1983
);
rice grain has Glc, Man-Glc, Glc-Glc,
[Man]2-Glc, Glc-Man-,
[Man]3-Glc with d18:0,
d18:1
4 and d18:2
4,8
sphingoid bases and 16:024:0 fatty acids (including some
-hydroxy
fatty acids) (Fujino et al. 1985
); spinach leaves
sphingolipids are comprised of Glc, Cellobiose and
Glc-[Man]2-Glc with d18:0,
d18:1
8, d18:2
4,8,
t18:0, t18:1
8 with 16:024:0 fatty acids
(Ohnishi et al. 1983
); soybean has a single cerebroside,
Glc ceramide (Cer), with d18:0, d18:1
4,
d18:1
8, d18:2
4,8,
t18:0, t18:1
8 and 16:026:0 fatty acids
(including
-hydroxy and
,ß-dihydroxy fatty acids)
(Ohnishi and Fujino 1982
);8
bell pepper and tomato also have mainly GlcCer with
d18:2
4,8, d18:1
8,
t18:1
8 sphingoid bases and 16:024:0
(including
-hydroxy-) fatty acids (Whitaker 1996
).
This structural variability and lack of reference material pose special
problems for analyses of sphingolipids in food. Recent developments in
the analysis of sphingolipids by GC/HPLC/mass spectroscopy (MS) and
tandem mass spectrometry are making it feasible to accurately identify
and quantify complex sphingolipids (for additional information, see
Adams and Ann 1993
, Murphy 1993
).
| Sphingolipid digestion and utilization |
|---|
|
|
|---|
Sphingomyelin and cerebrosides undergo little cleavage in the stomach,
but are hydrolyzed in all subsequent regions of the small intestine and
colon of rats and mice (Nilsson 1968
and 1969b
,
Schmelz et al. 1994
). The luminal contents of rat small
(and large) intestine contain substantial sphingomyelinase,
glucoceramidase and ceramidase activities (Nilsson 1969a
and 1969b
). Not all of the ingested sphingolipids are
hydrolyzed and absorbed, however. Nilsson (1968)
reported that ~25% of an administered dose of sphingomyelin was
excreted in feces, of which 10% was the intact molecule, 8090% was
ceramide and 36% was free sphingosine. There is a direct correlation
between the amount of sphingomyelin that is fed vs. the amount found in
the colon (Nyberg et al. 1997
). Germ-free mice show
a drastically reduced hydrolysis of sphingomyelin, which suggests that
intestinal microflora are major contributors to sphingolipid turnover
in the lower bowel (Duan et al. 1995 and 1996
). Similar
studies with cerebrosides (Nilsson 1968
) found that 43%
was excreted, with 4070% as the intact molecule and 2560% as
ceramide. Less is known about human metabolism of sphingolipids, but
human pancreatic juices contain a taurocholate-dependent neutral
sphingomyelinase (Chen et al. 1992
), and an alkaline
sphingomyelinase has been detected in human bile (Nyberg et al. 1996
).
Uptake of sphingolipids.
Much of the sphingosine (and, perhaps, ceramide) that is derived from
hydrolysis of complex sphingolipids is rapidly taken up by intestinal
cells and degraded to fatty acids (via fatty aldehydes) or
reincorporated into complex sphingolipids that remain associated
primarily with the intestine (Nilsson 1968
,
Schmelz et al. 1994
). When sphingoid-baselabeled
sphingolipids are fed to rats, a small amount of the radiolabeled
sphingoid base is found in lymph, blood and liver, which implies that
some component(s) of dietary sphingolipids are transported through the
mucosa and appear in systemic circulation (Nilsson 1968
,
Schmelz et al. 1994
). Chylomicrons may be involved in
sphingolipid transport because intestinal lymph contains ~1 nmol/mL
of sphingolipid (~40% of which is ceramide) (Merrill et al. 1995
).
Transport of sphingolipids via serum lipoproteins.
Sphingolipids are components of serum lipoproteins, with the greatest
amounts in LDL followed by VLDL > HDL (Merrill et al. 1995
). Sphingomyelin is the major sphingolipid of LDL and HDL,
whereas VLDL contain mainly ceramide. Small amounts of free sphingoid
bases are present in blood (Wang et al. 1992
),
associated with albumin and circulating cells (both erythrocytes and
leukocytes) (Wilson et al. 1988
); sphingosine
1-phosphate is also found in plasma and serum, but appears to be
derived from platelets (Yatomi et al. 1995
). The latter
finding is intriguing because endothelial cells have a high affinity
receptor (Edg-1) for sphingosine 1-phosphate (Van Brocklyn et al. 1998
).
Cellular metabolism of sphingolipids and regulation of sphingolipid biosynthesis by diet.
An in-depth discussion of cellular sphingolipid metabolism lies
beyond the scope of this review, but can be found elsewhere
(Merrill and Sweeley 1996
, Merrill et al. 1997
). All organs appear to be capable of de novo sphingolipid
biosynthesis (Merrill et al. 1985
, Nagiec et al. 1996
), and there is no evidence that consumption of dietary
sphingolipids is required for growth under normal conditions.
Nonetheless, exogenous sphingolipids are required for the growth of
mammalian cells with defects in serine palmitoyltransferase
(Hanada et al. 1992
), the initial enzyme of sphingolipid
biosynthesis, which establishes that sphingolipids are necessary for
normal cell function.
De novo sphingolipid synthesis is subject to some degree of feedback
regulation. Incorporation of radiolabeled serine into sphingolipids is
partially suppressed by LDL (Chatterjee 1998
,
Verdery and Theolis 1984
) or sphingoid bases
(Merrill 1983
, van Echten et al. 1990
) at
the level of serine palmitoyltransferase expression (Mandon et al. 1991
) and involving sphingoid base 1-phosphates (van Echten-Deckert et al. 1997
). Therefore, it is possible that the
sphingoid base backbones that are recovered from dietary sphingolipids
affect tissue sphingolipid biosynthesis.
| Dietary sphingolipids and cancer |
|---|
|
|
|---|
Effects of sphingolipids on colon carcinogenesis.
Normal intestinal cells undergo rapid turnover, except in cancer in
which there is loss of normal growth arrest and apoptosis. Therefore,
digestion of sphingolipids to ceramide and sphingosine might reduce the
risk of colon cancer if, as shown in Figure 3
, uptake of these compounds induces growth arrest, differentiation
and/or apoptosis (perhaps by by-passing a defect in
sphingomyelinase that was noted by Dudeja et al. 1986
to
be one of the earliest biochemical changes detected in colon cancer).
To test this hypothesis, sphingomyelin was purified from powdered
milk9
and fed to female CF1 mice that had been treated with
1,2-dimethylhydrazine (DMH) to induce colon tumors (Dillehay et al. 1994
). The controls were fed a standard AIN76A diet, which
is composed of defined ingredients that contain very low amounts of
sphingolipids. Sphingomyelin supplementation at 0.1% of the diet
(wt/wt) had no effect on weight gain of the animals, but reduced the
number of aberrant colonic crypt foci (an early marker of colon
carcinogenesis) by ~70% and, with longer feeding, reduced the number
of adenocarcinomas (the latter was only marginally significant,
P = 0.08, perhaps due to the small number of animals
used in this study).
|
Structure-function relationships between sphingolipids and their effects on colon carcinogenesis.
As already noted, the sphingolipids of food vary in both the lipid
backbones and headgroups. To evaluate whether the sphingosine backbone
is required, N-palmitoylsphingomyelins with sphingosine or
sphinganine as the backbone were synthesized and fed to
DMH-treated CF1 mice (Schmelz et al. 1997
).
Dihydrosphingomyelin (with sphinganine) was more effective than
sphingomyelins (with the sphingosine backbone) in the reduction in
aberrant crypt formation. These findings are noteworthy because
ceramide signaling usually requires the 4,5-trans-double
bond (Bielawska et al. 1993
); therefore, the inhibition
of aberrant colonic crypt formation by dietary (dihydro)sphingomyelins
appears to be due to the free sphingoid base (sphingosine or
sphinganine) rather than ceramide.
The efficacy of glycosphingolipids in reducing the formation of
adenocarcinomas has not yet been determined. However, ganglioside
GM1 is at least four- to eightfold more potent
than sphingomyelin (Dillehay et al. 1994
), and milk
glucosylceramide, lactosylceramide and ganglioside
GD3 are comparable to sphingomyelin (Schmelz et
al., unpublished observations) in suppressing aberrant colonic crypt
formation. Thus, both sphingomyelin(s) and glycosphingolipids affect
this early stage of colon carcinogenesis.
Sphingolipids and human colon cancer.
Neither human clinical trials nor epidemiologic studies have yet
evaluated whether sphingolipids influence human colon cancer.
Nonetheless, sphingosine and ceramide induced apoptosis in a human
adenocarcinoma cell line, HT29 cells (Schmelz et al. 1998
), and we have recently found that sphingolipids reduce
tumor number in Min mice (Schmelz et al. unpublished observations),
which have a genetic defect similar to that found in human familial
adenomatous polyposis (which arises from a defective APC gene).
Mutation of the APC gene is also found in up to 60% of sporadic human
colon cancers (Powell et al. 1992
). In addition,
sphingomyelinase activity is decreased in human colorectal carcinoma
(Hertervig et al. 1997
), as has been seen in colon
carcinogenesis in rodents (Dudeja et al. 1986
). On the
basis of these findings, it is plausible that dietary sphingolipids
influence human colon cancer risk.
Studies of anticancer activity in other cell types.
Sphingolipids are growth inhibitory and cytotoxic for numerous
transformed cell lines in culture (Merrill et al. 1996
,
Stevens et al. 1990
), and inhibit the transformation of
C3H10T1/2 cells by both
-irradiation (Borek et al. 1991
) and chemical carcinogens (Borek and Merrill 1993
) with phorbol esters as the promoter. Sphingoid bases and
their analogs inhibit the growth and metastasis of human and mouse
tumor cells in athymic and euthymic mice (Endo et al. 1991
, Sadahira et al. 1992
), inhibit the
induction of ornithine decarboxylase in mouse skin (Enkvetchakul et al. 1989
), and increase skin cancerfree survival under
some application protocols (Birt et al. 1998
).
Therefore, dietary sphingolipids may affect cancers at sites other than
the colon. It should be borne in mind that sphingosine is mitogenic in
a few instances (Zhang et al. 1990
), apparently via its
conversion to sphingosine 1-phosphate (Zhang et al. 1991
); therefore, effects in vivo should be evaluated carefully
and thoroughly.
Epidemiologic relationships between diet and cancer in view of sphingolipids.
The risk of colon cancer has been associated with diet; however,
identification of the responsible factors remains controversial
(Kim and Mason 1996
). Because the sphingolipid content
of food has not been considered in any of these analyses, this might
explain some of this confusion. Some foods that are rich in
sphingolipids (such as dairy products and soy) have received attention
from cancer researchers for some time. For example, dairy products
reduce the incidence of aberrant crypts (Abdelali et al. 1995
, Nelson et al. 1987
) in rats, reduce
aberrant colonic epithelial cell proliferation and restore a more
normal differentiation profile in humans (Holt et al. 1998
) and are correlated with a reduced risk of human colon
cancer (Glinghammar et al. 1997
, Van der Meer et al. 1997
). These effects may reflect the calcium and vitamin D
in dairy products; however, case-control and cohort studies
concerning calcium intake and colon carcinogenesis have been
inconclusive (Giovannucci and Willet 1994
, Kim and Mason 1996
, Pence et al. 1996
, Potter et al. 1993
). It is possible that the presence of sphingolipids
may help explain some of the benefits of dairy products and other
foods.
| Other potential relationships between diet, sphingolipids and disease |
|---|
|
|
|---|
|
Associations between sphingomyelin and cholesterol have intrigued
researchers for decades (Barenholz and Thompson 1980
,
Ikonen 1997
, Merrill and Jones 1990
,
Slotte and Bierman 1988
, Vandenheuvel 1965
). At least one molecular explanation for the cellular
association of these lipids is their colocalization in
"microdomains" such as caveolae (Harder and Simons 1997
) that are thought to be enriched in membrane receptors and
transporters, especially those linked to the plasma membrane via
glycosylphosphatidylinositol (GPI)-anchors (Bilderback et al. 1997
, Fiedler et al. 1994
). Depletion of either
sphingomyelin or cholesterol disrupts these microdomains and the
functioning of proteins associated with them, as has been seen in the
loss of folate transport in Caco-2 cells treated with inhibitors of
cholesterol or sphingolipid biosynthesis (Stevens and Tang 1997
).
Sphingomyelin affects many aspects of cholesterol transport and
metabolism (and vice versa), as indicated in Figure 4
, including the
following: cholestrol efflux from cells (Jian et al. 1997
, Yancey et al. 1995
, Zhao et al. 1996
); the conversion of cholesterol to bile acids, cholesterol
esters and other metabolites (Boldin and Jonas 1996
,
Rye et al. 1996
, Subbaiah and Liu 1993
;
the regulation of ß-hydroxyl-ß-methyl glutarate (HMG)-CoA reductase
activity (Gupta and Rudney 1991
); and, proteolysis of
sterol regulatory element binding proteins (Scheek et al. 1997
). Induction of sphingomyelin turnover as part of cell
signaling (in response to TNF-
) increases cholesterol esterification
(Chatterjee 1994
), which provides a relatively
unexplored link between cell signaling events and cholesterol
homeostasis.
Cholesterol and other lipids can also alter sphingomyelin metabolism
(Leppimaki et al. 1998
). An inhibitor of cholesterol
synthesis, 25-hydroxycholesterol, stimulates sphingomyelin synthesis in
Chinese hamster ovary cells (Ridgway 1995
). In vivo,
diets supplemented with cholesterol (Geelen et al. 1995
,
Nikolova-Karakashian et al. 1992
) affect tissue
sphingomyelin content and metabolism. Feeding of different oils to
experimental animals (Bettger et al. 1996
) influences
the fatty acid composition of sphingomyelin; and essential fatty acid
deficiency reduces the formation of the skin ceramides (Wertz 1992
).
These interactions suggest that sphingomyelin may influence
atherosclerosis, either directly or by affecting other risk factors
such as cholesterol. Additional observations that also support this
possibility are as follows: 1) sphingomyelin affects LDL
binding and utilization by cells in culture (Chatterjee 1993
); 2) hydrolysis of LDL sphingomyelin by an
extracellular sphingomyelinase that is enriched in atherosclerotic
lesions alters the aggregation state of the particle and promotes foam
cell formation by macrophages (Marathe et al. 1998
,
Schissel et al. 1996a
and 1996b
); 3) oxidized
lipoproteins have been reported to stimulate the growth of vascular
smooth muscle cells (Augé et al. 1996
) and human
blood monocytes (Kinscherf et al. 1997
) via triggering
of the sphingomyelin signaling
pathway;104) there is an elevation of sphingomyelin in aortic lesions
in which this lipid can account for 70% of the total phospholipid
(Barenholz and Gatt 1982
); a substantial portion of the
sphingomyelin found in arteries and atherosclerotic lesions appears to
arise from synthesis in the arterial tissue accompanied by decreased
turnover (Eisenberg et al. 1969
, Zilversmit et al. 1961
); and 5) the ratio of sphingomyelin to
phosphatidylcholine increases fivefold in VLDL from
hypercholesterolemic rabbits (Rodriguez et al. 1976
).
There are also interesting associations between glycosphingolipids and
atherosclerosis (see Chatterjee 1998
, Prokavoza and Bergelson 1994
).
Short-term (Imaizumi et al. 1992
) and long-term
(Kobayashi et al. 1997
) feeding experiments with rats
have indicated that sphingolipids reduce plasma cholesterol, a risk
factor for atherosclerosis. Plasma total cholesterol was 30% lower for
rats fed semipurified diets supplemented with a mixture of
sphingomyelin and glycosphingolipids (1% of the total diet) plus 4%
soybean oil for up to two generations, compared with rats fed 5%
soybean oil (plasma triacylglycerols were not different).
Unfortunately, the supplement contained additional components
(including cholesterol) that may have also contributed to these
results. More in vivo studies of this association are clearly
warranted.
Sphingolipid signaling may play a role in some of the progressive loss of cell function that accompanies aging.
Changes in sphingomyelin content with aging have been seen in many
tissues, including calf liver (Jenkins and Kramer 1988
),
rat brush border membranes (Levi et al. 1989
), human
aorta (Eisenberg et al. 1969
) and heart myocytes
(Yechiel and Barenholz 1986
). As noted earlier in this
review, ceramide can inhibit cell growth and induce apoptosis
(Hannun and Obeid 1995
), and has been implicated as a
mediator of senescence in a cell culture model for aging (Lee and Obeid 1997,
Venable et al. 1995
). Therefore,
modulation of sphingolipid metabolism by the diet could affect aging
via this signaling pathway(s).
Sphingolipid signaling is likely to be involved in the mechanism of action of a substantial number of other components of the diet.
A growing list of nutritional factors can modulate this signaling
pathway by affecting sphingomyelinase activity, such as
1
,25-dihydroxycholecalciferol (Okazaki et al. 1989 and 1990
), unsaturated fatty acids (Robinson et al. 1997
) and cellular levels of glutathione (Liu and Hannun 1997
). Dietary (n-3) polyunsaturated fatty acids (PUFA) have
been reported to suppress the formation of ceramide (and
diacylglycerol) (Jolly et al. 1997
). Furthermore,
sphingolipid signaling pathways are involved in the regulation of
important enzymes, such as some isoforms of cytochrome P450
(Merrill et al. 1999
, Nikolova-Karakashian et al. 1997
).
"Bioactive" sphingolipid metabolites (e.g., sphinganine or
ceramide) can be produced by aberrant induction of sphingolipid
biosynthesis (Fig. 4)
, as has been shown in the toxicity of palmitate
for cells in culture when uptake by mitochondria is blocked genetically
or by inhibitors (Paumen et al. 1997
). The toxicity was
attributed to sphingolipid biosynthesis because it was selective for
palmitic acid (Paumen et al. 1997
) (serine
palmitoyltransferase activity is highly dependent on cellular levels of
serine and fatty acyl-CoA, with a high degree of selectivity for
palmitoyl-CoA; Merrill et al. 1988
) and was prevented by
inhibition of serine palmitoyltransferase. Zucker diabetic fatty (ZDF)
rats exhibit loss of ß cells by apoptosis and have been shown to have
elevated ceramide; incubation of islets from prediabetic and diabetic
ZDF rats with fatty acids increased ceramide and apoptosis
(Shimabukuro et al. 1998b
). Therefore, these authors
concluded that ß cell apoptosis is induced by de novo ceramide
formation. Overexpression of serine palmitoyltransferase can also
induce apoptosis, as has recently been reported for obese prediabetic
fa/fa rats (Shimabukuro et al. 1998a
) and
associated with induction of apoptosis in pancreatic ß cells. These
studies suggest that perturbation of intermediary metabolism (perhaps
by many means) affects sphingolipid biosynthesis; when intermediates of
this pathway accumulate, there can be profound effects on cell
behavior.
The implications for diabetes are especially provocative because other
interrelationships between sphingolipids and diabetes have been noted
as follows: free sphingoid bases inhibit insulin-induced glucose
uptake and oxidation by adipose cells (Robertson et al. 1989
); ceramide down-regulates GLUT4 gene transcription in
3T3-L1 adipocytes (Long and Pekala 1996
); and
sphingolipids may alter insulin action at the level of the cell
membrane (Candiloros et al. 1996
).
Perturbation of sphingolipid metabolism is the mechanism of action of mycotoxins and other fungal secondary metabolites.
A number of microorganisms produce secondary metabolites that disrupt
sphingolipid metabolism (Merrill and Sweeley 1996
); the
most thoroughly characterized of these are the fumonisins, which are
produced by Fusarium moniliforme and related fungi.
Fumonisins are common contaminants of maize and other foods and cause
equine leukoencephalomalacia, porcine pulmonary edema and various other
diseases of animals, including humans (Marasas 1995
).
Fumonisins inhibit ceramide synthase (Wang et al. 1991
),
which results in accumulation of sphinganine (and sometimes
sphingosine) and reduced formation of complex sphingolipids. As a
consequence of disruption of sphingolipid metabolism, fumonisins
inhibit progression through the cell cycle (Ciacci-Zanella et al. 1998
, Lee et al. 1998
) and induce apoptosis
(Riley et al. 1996
, Schmelz et al. 1998
).
Elevations in sphinganine can be detected in blood and urine of animals
that consume fumonisins and can be used as a biomarker for exposure
(Riley et al. 1994
, Wang et al. 1992
).
One of the other interesting inhibitors of sphingolipid metabolism is
ISP1 (also called myriocin), a potent inhibitor of serine
palmitoyltransferase (Miyake et al. 1994
). Long-term
treatment with ISP1 can be toxic. However, by preventing the
accumulation of sphingoid bases and ceramides, ISP1 protects cells
(Schmelz et al. 1998
) and animals (Riley et al. 1999
) from fumonisin toxicity. Thus, naturally occurring
inhibitors of sphingolipid metabolism can have both toxic and
protective effects, depending on the context in which they are
encountered.
The presence of sphingolipids in food may protect against bacteria toxins and infection.
Many microorganisms, microbial toxins and viruses bind to cells via
sphingolipids. These include cholera toxin (ganglioside
GM1) (Thompson and Schengrund 1998
),
verotoxin (globosides) (Bast et al. 1997
,
Farkas-Himsley et al. 1995
), Shiga-like toxin 2e
(globotriaosylceramide, Gb3) (Jacewicz et al. 1995
, Keusch et al. 1995
), and Clostridium
botulinum type B neurotoxin (to synaptotagmin II associated with
gangliosides GT1b/GD1a) (Nishiki et al. 1996
). Furthermore, many bacteria utilize sphingolipids to
adhere to cells, e.g., Escherichia coli (galactosylceramide)
(Blomberg et al. 1993
, Khan et al. 1996
,
Payne et al. 1993
), Hemophilus influenza
(gangliotetraosylceramide and gangliotriosylceramide) (Hartmann and Lingwood 1997
), Helicobacter pylori
(gangliotetraosylceramide, gangliotriaosylceramide, sulfatides and
GM3) (Huesca et al. 1996
, Kamisago et al. 1996
, Simon et al. 1997
, Wadstrom et al. 1997
), Borrelia burgdorferi (galactocerebroside;
Virulent strain 297: glucosylceramide, lactosylceramide and
galactosylgloboside) (Garcia Monco et al. 1992
,
Kaneda et al. 1997
), and Pseudomonas
aeruginosa and Candida albicans
(asialo-GM1) (Yu et al. 1994
). Virus binding
can be mediated via sphingolipids, including HIV-1 gp120
(galactosylceramide) (Fantini et al. 1997
), Sendai virus
(ganglioside GD1a) (Epand et al. 1995
) and
influenza viruses (gangliosides, sulfatides and polyglycosylceramides)
(Fakih et al. 1997
, Matrosovich et al. 1996 and 1997
, Sato et al. 1996
, Suzuki et al. 1996
).
Synthetic sphingolipids are effective in inhibiting the binding of
bacteria and viruses (Fantini et al. 1997
); therefore,
it is plausible that sphingolipids in food also compete for cellular
binding sites and facilitate the elimination of pathologic organisms
from the intestine. Glycosphingolipids have been hypothesized to be one
of the nonimmunoglobulin compounds in human milk that confer protection
against pathogens (Newburg and Chaturvedi 1992
,
Zopf 1996
). Rueda et al. (1998)
recently
reported that preterm newborn infants given an adapted milk formula
supplemented with gangliosides (1.43 mg/100 kcal) had significantly
fewer E. coli in feces (and higher fecal bifidobacterial
counts) than infants fed the control formula. Interestingly,
sphingolipids help protect plants against necrotic lesions induced by
parasitic fungi (Lhomme et al. 1990
).
Unfortunately, some glycosphingolipids also appear to be participants
in disease induced by microorganisms. A fraction of the persons
infected with Campylobacter jejuni develop
Guillain-Barre or Miller Fisher syndrome, which appears to involve
development of cross-reactive antibodies against gangliosides and
C. jejuni lipopolysaccharides (Jacobs et al. 1997
).
| SUMMARY AND PERSPECTIVES FOR THE FUTURE |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
2 Current address: National Center of
Environmental Health, Centers for Disease Control and Prevention,
Atlanta, GA 30341. ![]()
3 Current address: Department of Physiology,
University of Kentucky, Lexington, KY 40236. ![]()
5 Sphingosine is sometimes used as a generic term
for all sphingoid bases, but most often refers specifically to
D-erythro-1,3-dihydroxy,
2-aminooctadec-4-ene or trans-4-sphingenine (d18:1). ![]()
6 Abbreviations used: Cer, ceramide; DMH,
1,2-dimethylhydrazine; Gal, galactose; GC, gas chromatography; Glc,
glucose; GPI, glycosylphosphatidylinositol; HMG,
ß-hydroxyl-ß-methyl glutarate; Man, manose; MS, mass spectrometry;
PDGF, platelet-derived growth factor; PUFA, polyunsaturated fatty
acids; TNF-
, tumor necrosis factor-
; ZDF rats, Zucker diabetic
fatty rats. ![]()
7 In this regard, some of the estimates in Table 1
are puzzling because plants are generally not thought to contain
substantial amounts of sphingomyelin (Lynch 1993
). ![]()
8 The composition may depend on the source because
we have recently analyzed soy cerebrosides and found one major GlcCer,
with d18:2
4,8 and
-hydroxypalmitic acid (h16:0)
(M. C. Sullards, D. V. Lynch, E. M. Schmelz, E. Wang,
A. H. Merrill. Jr. & J. Adams, unpublished data). ![]()
9 Because sphingolipids are associated with the
globule membrane rather than with the lipid droplet per se, a
substantial portion remains in low fat dairy products, including
"nonfat" dry milk (Jenson 1995
). ![]()
10 This report described sphingomyelin hydrolysis
to ceramide; in a recent collaboration (N. Augé, M.
Nikolova-Karakashian, S. Carpentier, S. Parthasarathy, A.
Nègre-Salvayre, R. Salvayre, A. H. Merrill, Jr. & T. Levade,
J. Biol. Chem., in press), we have also found activation of sphingosine
kinase, which is consistent with sphingosine 1-phosphate mediating the
growth stimulation (ceramide formation may play a role in the toxicity
of oxidized lipoproteins). ![]()
Manuscript received March 1, 1999. Revision accepted April 3, 1999.
| REFERENCES |
|---|
|
|
|---|
1. Abdelali H., Cassand P., Soussotte V., Daubeze M., Bouley C., Narbonne J. F. Effect of dairy products on initiation of precursor lesions of colon cancer in rats. Nutr. Cancer 1995;24:121-132[Medline]
2. Adams J., Ann Q. Structure determination of sphingolipids by mass spectrometry. Mass Spectrom. Rev. 1993;12:51-85
3.
Augé N., Andrieu N., Negre-Salvayre A., Thiery J. C., Levade T., Salvayre R. The sphingomyelin-ceramide pathway is involved in oxidized low density lipoprotein-induced cell proliferation. J. Biol. Chem. 1996;271:19251-19255
4. Barenholz Y., Gatt S. Phospholipids. Hawthorne J. N. Ansell G. B. eds. Phospholipids, ch. 4 1982 Elsevier Biomedical Press Amsterdam, The Netherlands.
5. Barenholz Y., Thompson T. E. Sphingomyelins in bilayers and biological membranes. Biochim. Biophys. Acta 1980;604:129-158[Medline]
6. Bast D. J., Brunton J. L., Karmali M. A., Richardson S. E. Toxicity and immunogenicity of a verotoxin 1 mutant with reduced globotriaosylceramide receptor binding in rabbits. Infect. Immun. 1997;65:2019-2028[Abstract]
7.
Bennun F. R., Roth G. A., Monferan C. G., Cumar F. A. Binding of cholera toxin to pig intestinal mucosa glycosphingolipids: relationship with ABO blood group system. Infect. Immun. 1989;57:969-974
8. Bettger W. J., Blackadar C. B., McCorquodale M. L. The effect of dietary fat type on the fatty acid composition of sphingomyelin in rat liver and heart. Nutr. Res. 1996;16:1761-1765
9.
Bielawska A., Crane H. M., Liotta D. C., Obeid L. M., Hannun Y. A. Selectivity of ceramide-mediated biology. Lack of activity of erythro-dihydroceramide. J. Biol. Chem. 1993;268:26226-26232
10.
Bilderback T. R., Grigsby R. J., Dobrowsky R. T. Association of p75NTR with caveolin and localization of neurotrophin-induced sphingomyelin hydrolysis to caveolae. J. Biol. Chem. 1997;272:10922-10927
11. Birt D. F., Merrill A. H., Jr, Barnett T., Enkvetchakul B., Pour P. M., Liotta D. C., Geisler V., Menaldino D. S., Schwartzbauer J. Inhibition of skin papillomas by sphingosine, N-methyl sphingosine, and N-acetyl sphingosine. Nutr. Cancer 1998;31:119-126[Medline]
12. Blank M. L., Cress E. A., Smith Z. L., Snyder F. Meats and fish consumed in the American diet contain substantial amounts of ether-linked phospholipids. J. Nutr. 1992;122:1656-1661
13.
Blomberg L., Krivan H. C., Cohen P. S., Conway P. L. Piglet ileal mucus contains protein and glycolipid (galactosylceramide) receptors specific for Escherichia coli K88 fimbriae. Infect. Immun. 1993;61:2526-2531
14.
Boldin D. J., Jonas A. Sphingomyelin inhibits the lecithin-cholesterol acyltransferase reaction with reconstituted high density lipoproteins by decreasing enzyme binding. J. Biol. Chem. 1996;271:19152-19158
15. Borek C., Merrill A. H., Jr Sphingolipids inhibit multistage carcinogenesis and protein kinase C. Bronzetti G. Hayatsu M. DeFlora S. Waters M. D. Shankel D. M. eds. Antimutagenesis and Anticarcinogenesis Mechanisms III 1993:367-371 Plenum Press New York, NY.
16.
Borek C., Ong A., Stevens V. L., Wang E., Merrill A. H., Jr Long-chain (sphingoid) bases inhibit multistage carcinogenesis in mouse C3H/10T1/2 cells treated with radiation and phorbol 12-myristate 13-acetate. Proc. Natl. Acad. Sci. U.S.A. 1991;88:1953-1957
17.
Cahoon E. B., Lynch D. V. Analysis of glucocerebrosides of rye (Secale cereale L. Cv. Puma) leaf and plasma membrane. Plant Physiol 1991;95:58-68
18. Candiloros H., Zeghari N., Ziegler O., Donner M., Drouin P. Hyperinsulinemia is related to erythrocyte phospholipid composition and membrane fluidity changes in obese nondiabetic women. J. Clin. Endocrinol. Metab. 1996;81:2912-2918[Abstract]
19.
Chatterjee S. Neutral sphingomyelinase increases the binding, internalization, and degradation of low density lipoproteins and synthesis of cholesteryl ester in cultured human fibroblasts. J. Biol. Chem. 1993;268:3401-3406
20.
Chatterjee S. Neutral sphingomyelinase action stimulates signal transduction of tumor necrosis factor in the synthesis of cholesterol esters. J. Biol. Chem. 1994;269:879-889
21. Chatterjee S. Sphingolipids in atherosclerosis and vascular biology. Arterioscler. Thromb. Vasc. Biol. 1998;18:1523-1533