Download poly-drug cancer therapy based on ceramide

Experimental Oncology 25, 3-10, 2004 (March)
Exp Oncol 2004
26, 1, 3-10
3
REVIEW
POLY-DRUG CANCER THERAPY BASED ON CERAMIDE
Norman S. Radin*
Mental Health Research Institute University of Michigan, Ann Arbor, Michigan, USA
Thousands of research studies have reported that many kinds of cancer cells and tumors can be killed by treatments
that increase the concentration of a simple cellular sphingolipid, ceramide (Cer). While there are many ways to
elevate tumor Cer levels, this approach is complicated by the central, complex role of Cer in cell homeostasis: Cer
is readily metabolized to form other sphingolipids that increase the tumor’s growth rate, metastasis, and resistance
to the patient’s immune system. This review points out the need to prevent this metabolic conversion while
simultaneously stimulating the enzymes that increase the formation of Cer. I describe here many of the enzymes
that need stimulation or inhibition, and drugs or metabolites or dietary components that modify each of the
enzymes. The review also points to the importance of the allylic alcohol group in Cer and in many cancer drugs,
suggesting that the hydroxyl group participates in phosphate transfer to and from proteins by forming a temporary
phosphate ester. The allylic hydroxyl may also reduce the ketone moieties in mitochondrial ubiquinone, with
formation of reactive oxygen species and apoptogenic breakdown. The level of Cer in tumors can be increased by:
(1) direct administration of Cer or a Cer analogue, and (2) stimulation of Cer synthesis from its elementary
precursors, or from (3) sphingomyelin by hydrolysis, or from (4) the glucosphingolipids by hydrolysis, or (5) by
acylation of sphingosine. In addition, Cer concentration can be raised by slowing its conversion to (6) sphingomyelin, (7) glucosylCer, (8) Cer phosphate, and (9) sphingosine + fatty acid by hydrolysis. Therapeutic radiation
stimulates the de novo synthesis of Cer in tumors. Conversion of sphingosine (from Cer) to sphingosine phosphate
probably also ought to be blocked.
Key Words: apoptosis, sphingolipids, allylic alcohols, poly-drug therapy, ceramide concentration.
Poly-drug therapy is based on a simple principle: if
you wish to influence a metabolic process in a patient,
you must remember that more than one process will
be affected. In any given individual, it is important to
control several interacting processes. In the example
described in this paper, ceramide (the fatty acid amide
of sphingosine = Cer) (Figure) is involved in many physiological processes of normal function and disease.
Cer has been found by many researchers to block cell
growth and proliferation in a large number of cell types,
producing apoptosis (suicidal death) in cancer cells if
its concentration is high enough. Thus the process
needed for cancer treatment is one that forces the patient’s tumor to accumulate Cer. Drugs are known which
can 1) accelerate Cer biosynthesis from several precursors, 2) slow its hydrolysis, and 3) slow its conversion to sphingomyelin (SM) which does not seem to
affect cell growth. However there is a big complication:
Cer also forms products that stimulate cell growth and
proliferation, and thus balance growth vs death or differentiation in normal cells. In addition, a group of Cer
metabolites secreted by tumors (sialo-glucosphingoli-
Figure. Structure of ceramide
Received: January 16, 2004.
*Correspondence: E-mail: [email protected]
Abbreviations used: Cer — ceramide; GlcCer — glucosylceramide;
GSH — glutathione; ROS — reactive oxygen species; SM —
sphingomyelin.
pids = gangliosides) inhibits the immune system’s dendritic cells, which are vital for the complete destruction
of cancer cells [1]. These competing, interacting effects make the body’s control of Cer metabolism a vital
factor in normal cell function and stability.
If you treat a patient with a drug that speeds Cer
synthesis, the resultant accumulation of Cer can be
expected to induce more rapid hydrolysis of the Cer by
ceramidase or faster conversion to other metabolites.
Thus the drug’s beneficial effects will be temporary and
the tumor might mutate to a more virulent form in which
the extra Cer is converted rapidly to proliferative sphingolipids (sphingosine-1-phosphate, Cer-1-phosphate, glucosylceramide (GlcCer), or lactosylceramide). It seems evident that all these processes should
be controlled simultaneously, using at least four different drugs in a “Cer cocktail” that act on different Cer
enzymes. Since Cer and its metabolites exert many
different effects in cells, it may be necessary to control
those “side-effects” with additional drugs. The polydrug proposal has been described [2, 3]. The reader
has a wide choice of reviews on sphingolipids [4–9].
Physicians are very familiar with the problem of
side-effects, such as the nausea, or diarrhea, or cell
damage that results from many antineoplastic drugs,
so empirical poly-drug therapy is quite common, using
even five drugs simultaneously. Moreover, more and
more oncologists are using several different kinds of
cancer therapy simultaneously or almost simultaneously, such as radiation together with one or more antiproliferation drugs. (Radiation causes cells to make Cer
faster.) The poly-drug therapy that I am recommending here is the use of drugs to control a single, critical
metabolite that will kill the tumor in a “natural” way.
4
An important factor in this approach is that tumors —
at the time of diagnosis — generally consist of several
different cell lines that differ in their Cer processing.
The dominant cancer cell line might contain a very high
activity of a single Cer enzyme, such as ceramidase
(which lowers tumor Cer), so a single drug that blocks
the enzyme will seem to bring the tumor under control.
However this buys the patient only temporary remission. As tumors progress toward higher virulence and
metastasis, new mutations appear that repress Cer levels and elevate the proliferative Cer metabolites even
more [10]. Thus a wide-acting drug cocktail that controls many aspects of Cer metabolism is essential for
(apparently) all cancer patients. This consideration is
also important for researchers who test drugs with only
one or two cell strains of a given type of cancer instead
of a complex mixture of cells.
Another factor that is rarely recognized by researchers working with cultured cells is the difference in drug
sensitivity between normal cells and cancer cells. Drugs
expected to have antineoplastic activity are typically
tested in available cancer cells, which are easier to
handle. Researchers should recognize the need to
promptly move to animal testing to evaluate the true
potential of their drug. While the evidence is incomplete, all the reported studies show that cancer cells
have an abnormally high rate of sphingolipid synthesis, and that tumors that have progressed to the drug
resistant stage make sphingolipids even faster [11]. It
is thus likely that realistic poly-drug Cer therapy will
work quickly, using below-typical dosages, minimizing
toxicity to the patient.
1. Therapy by stimulating Cer synthesis de novo
1a. Cer is made from palmitic acid and serine, which
form 3-keto sphinganine and CO2 (the latter is from
the COOH of serine). The amino group of the lipid is
then acylated by a fatty acid, typically ranging in size
from C16 to C26, forming 3-keto dihydroceramide. The
ketone moiety is reduced to an OH, forming dihydroceramide. This lipid seems to have little or no physiological activity, but a dehydrogenase acting between
carbons 4 and 5 produces a trans-∆4 double bond that
makes true Cer. The vital difference between the two
amides is the double bond, which — together with the
C3 hydroxyl — is part of an allylic alcohol group.
Several researchers have shown that elevating tissue precursor levels (palmitic acid and serine) will speed
the de novo synthesis of Cer and aid the appearance
of apoptosis. Conversely, it is helpful to use an inhibitor
(etomoxir) to slow the transport of palmitic acid to mitochondria by carnitine, which leads to the “burning” of
the fatty acid. Administration of extra (dietary) carnitine
results in slower Cer formation and reduction in apoptosis, evidently because it slows de novo synthesis of
Cer [12]. For some tumors, increasing the patient’s intake of pyridoxine might speed Cer synthesis, since
the first step in the synthesis is catalyzed by pyridoxal
phosphate.
1b. Radiation stimulates the serine-palmitic condensation by inducing faster formation of keto-sphinganine
Experimental Oncology 25, 3-10, 2004 (March)
synthase, thus elevates the Cer in the tumor. In addition, the high-energy photons generate reactive oxygen
species (ROS), which speed the hydrolysis of sphingomyelin (SM = Cer-phosphocholine) to form additional
Cer (see section 3a). Apparently any energetic form of
radiation works: UVA, UVB, gamma-rays, and (probably) X-rays. Recent publications have supported the use
of simultaneous or near-simultaneous radiation, combined with chemotherapy that elevates tumor Cer. A merit
of using chemotherapy first, to shrink the tumor, is that
one can reduce the volume of normal tissue exposed to
radiation. However the advantage of simultaneous use
is that the tumor is killed promptly, before mutated, more
resistant cancer cells can appear.
1c. Several drugs with antineoplastic activity have
been shown to stimulate the serine-palmitic condensation: 1) tetrahydrocannabinol (the active component
of marijuana); 2) PSC 833 (valspodar, which neutralizes the multi-drug insensitivity of advanced tumors);
3) retinoic acid (a metabolite of Vitamin A); 4) camptothecin and irinotecan (popular antineoplastic drugs);
5) etoposide (an anticancer drug that inhibits topoisomerase II and breaks DNA strands); and 6) paclitaxel (Taxol, a well-known anticancer drug).
Modification of these structures has yielded superior anticancer drugs, e.g., fenretinide (N-(4-hydroxyphenyl) retinamide) and homocamptothecin. Fenretinide produces Cer and apoptosis and fludarabine, an
antineoplastic drug that causes Cer accumulation,
probably also acts by stimulating keto-sphinganine
synthesis. Modifications of tetrahydrocannabinol that
produce less change in behavior appear to be more
effective for chemotherapy.
2. Therapy by treatment with ceramide
2a. Cer is very insoluble in water but it can be dissolved in organic solvents, such as ethanol and dimethylsulfoxide, or incorporated into a water/oil cream. Thus it
can readily be applied to skin tumors. For injection, Cer
can be incorporated into an emulsion with a suitable
detergent or incorporated into liposomes. Use of shortchain Cer (acetyl sphingosine or a little longer) greatly
enhances the dispersibility. The short-chain ceramides
are readily absorbed by cells and metabolized (mainly
by hydrolysis and reacylation of the released sphingosine), so they may be excellent therapeutic agents.
2b. Cer can also be converted to a more dispersible prodrug by attaching a water-soluble polymer —
poly(ethylene glycol) — with a connecting spacer, such
as succinic acid. Presumably the ester linkages would
be hydrolyzed inside the tumor cells, liberating Cer. It
might also be useful to make Cer with sphingosine and
a water-soluble fatty acid, such as the ω-poly(ethylene
glycol) ether of decanoic acid. This might act directly to
produce apoptosis.
2c. Cer has been made more dispersible by attachment to glucuronic acid as a glycoside in the C-1 position. Added to the diet, it was found to undergo hydrolysis and reduce carcinogenesis in the colon [13].
This approach deserves more study, especially with
polysaccharide substituents, which would allow the use
Experimental Oncology 25, 3-10, 2004 (March)
of higher doses. Natural sphingolipids in food also undergo some hydrolysis to Cer in the intestine and are
easy to prepare from animal spinal cords or brains.
While nervous system products might contain pathogenic prions, treatment with alkali would not harm the
lipids but would surely denature the prions.
2d. The recent growth of interest in the potential use
of Cer in chemotherapy has encouraged chemists to synthesize variations on Cer structure [3, 14, 15]. Since dihydroceramide shows little activity, it would appear that
the allylic alcohol group of Cer must play an important
role in producing tumor death. This concept finds support
in the observation that many antineoplastic drugs also
contain an allylic alcohol or generate an allylic alcohol group
under metabolic activation [14]. Evidence has been cited
to support the possibility that the C-3 OH of Cer is oxidized to a ketone group in mitochondria, by interaction
with coenzyme Q or a factor closely related to CoQ [16].
The allylic ketone presumably then condenses in a Michael
reaction with glutathione or cysteine thiol groups or other
reactive substances, possibly an enzyme acting on CoQ.
The apoptogenic consequences of such a condensation
are described in section 3a. It is probably significant that
an allylic ketone group is seen in many antineoplastic
drugs. Readers of this journal may be especially interested to note that landomycin, an allylic ketone (quinone),
has potent antitumor and antibiotic activity [17].
Another suggestion [18] is based on the known ability
of Cer to stimulate some protein kinases and protein
phosphatases [19, 20]. Since the activity of many proteins depends on the presence or absence of phosphate
esters at certain sites, Cer catalyzes a wide variety of
biological phenomena. An appropriate complex of phosphate changes might result in apoptosis. This catalytic
or cofactor activity also requires the high reactivity of an
allylic OH, so the hypothesis was that the phosphate
anion forms a transient 1,3-cyclic ester with Cer which
is then transferred to a protein or — in a phosphatase —
is simply hydrolyzed. Conceivably the transient phosphate group is simply on the C-3 hydroxyl.
Variations on the structure of Cer should then include features that enhance the reactivity of the allylic
group. Moving the C-3 OH to the C-5 position, and
the ∆4 double bond to the ∆3 position (i.e., inverting the
position of the allylic alcohol) yielded an iso-Cer that
was much more active in slowing cell proliferation [21].
The OH in this isomer is further from the amide group
and thus more exposed to the environment than the
C-3 OH. Another approach is to add an OH to position
C-6, which would yield Cer with two allylic alcohols instead of one [22]. Adding one or more conjugated double bonds to natural Cer (i.e., at ∆6, ∆8, etc.) may enhance the reactivity of the C-3 OH.
Modifying the fatty acid part of Cer might also be
productive. For example, retinoic acid and conjugated
linoleic acid have anticancer activity, probably as the
result of oxygenase action that inserts an allylic OH into
the chain. They could be readily incorporated into Cer.
Ceramides containing a 2-OH fatty acid are common
in mammals and addition of a ∆3 double bond to the
5
fatty acid (yielding an extra allylic alcohol group) would
probably enhance the Cer activity. Many other Cer analogs with promising properties have recently been described [15]. It seems evident that the synthesis of new
Cer analogs will yield powerful antineoplastic drugs.
It should be remembered that a test with a new Cer
analog might show good interference with cancer cell proliferation simply because it protects endogenous Cer
against conversion to a proliferative metabolite. For example, it might simply inhibit ceramidase rather than mimic
Cer’s ability to produce apoptosis. Conversely, a new Cer
analog might have little effect on cell proliferation simply
because it inhibits a Cer enzyme that occurs at a very
high level of activity in the cell type chosen for the evaluation. Cer containing a pyrrolidine ring at C-1 instead of
an OH failed to inhibit growth of MDCK cells, yet showed
high inhibition of GlcCer synthase, thus is potentially very
useful in a cancer cocktail (see section 4).
3. Therapy by stimulating sphingomyelin hydrolysis
3a. Several kinds of SMase hydrolyze SM to Cer and
phosphocholine. Neutral, Mg2+-stimulated SMase has
received the most attention, probably because its activity
depends on the oxido/reduction status of the tumor and
because the substrate (SM) occurs at a relatively high
concentration in cells. Thus SM acts as a reservoir for
the rapid formation of Cer. Glutathione (GSH), the major
thiol in tissues, inhibits the enzyme, so any factor that
lowers its concentration will speed Cer synthesis [23]. It
is significant that many tumors contain high levels of GSH
and γ-glutamylcysteine synthase, thus tend to keep Cer
levels low and avoid apoptosis.
GSH levels are lowered by oxidative stress. While
oxygen is relatively scarce in solid tumors, some parts
of the tumors are oxygenated enough to attain the benefits of oxidative stress. The sulfur atom in GSH can be
oxidized to a disulfide (as a dimer or reaction with protein-bound cysteine), or an S-O or S=O or S-NO derivative. Agents that produce ROS oxidize GSH, elevate
Cer, and kill cancer cells. An important generator of ROS
is Cer itself, which produces ROS in mitochondria. This
ROS oxidizes GSH and stimulates SM hydrolysis, generating additional Cer. Thus one sees a gradual rise in
cellular Cer by what amounts to an autocatalytic growing spiral. GSH can also be oxidized to a mixed disulfide
by various disulfide drugs, such as disulfiram. This antialcoholism drug was shown to induce apoptosis of melanoma cells. Betathine (β-alanyl cysteamine disulfide)
has shown some promise in cancer patients, possibly
by forming a mixed disulfide with GSH.
GSH can also be depleted by inhibiting its biosynthesis with buthionine sulfoximine, which also induces
apoptosis. Various detoxification reactions consume
GSH, generally forming thio ethers. The common antiinflammatory drug, acetaminophen, is detoxified in part
via this reaction. Glutamic acid, one of the GSH precursors, can be removed from the body in relatively large
amounts by letting it react (after conversion to glutamine)
with phenylacetic acid. The resultant amide, phenylacetylglutamine, is excreted in urine. 4-Phenylbutyrate
6
reacts similarly. These simple molecules decreased
plasma glutamine levels in patients, as well as GSH [24].
The two arylalkanoic acids have shown some anticancer activity and ability to prevent carcinogenesis, particularly when used with other Cer inducers, such as radiation and doxorubicin. (Incidentally, they are useful for
treating patients whose urea cycle is inadequate.)
Because of the many normal functions of GSH, reducing its concentration in tumors should not be complete and the method used should be part of other Cerelevating treatments. Many antineoplastic drugs react
with GSH by a Michael condensation or thiol condensation, an effect which must enhance the effectiveness
of the drugs.
3b. SMase is stimulated by arachidonic acid, one of
the major unsaturated fatty acids of phospholipids [25].
Arachidonate can be elevated by ingestion or by stimulating phospholipase A2, which liberates arachidonate,
increases Cer, and produces apoptosis. Hydrogen peroxide stimulates the phosphatase, so Cer itself (which
produces peroxide in cells) may increase its concentration in an expanding spiral like the one described
above. Short-chain Cer — and probably ordinary Cer —
can act as an arachidonate acceptor from phospholipids, especially plasmalogen, forming the 1-arachidonoyl-Cer ester. The ester is hydrolyzed readily, offering another way to stimulate SM hydrolysis. In other
pathways, arachidonate is attacked by oxygenases,
forming several allylic alcohols, which can be expected
to produce apoptosis [26]. Arachidonate metabolism is
complex and conversion of the acid to prostaglandins
should be minimized since they stimulate tumor growth.
3c. Oxidized low density lipoprotein stimulates neutral SMase, alkaline and acid ceramidase, sphingosine
kinase, and the formation of lactosylceramide from
GlcCer [27]. Cer levels are raised by the SMase. but
the other enzymes lower Cer and form proliferative
sphingolipids, so the lipoprotein must be used together
with inhibitors of the other enzymes. In a study with
human macrophages, the Cer increase was also accompanied by increased neutral and acidic SMase.
Oxidized LDL may contain an oxidized form of cholesterol, 4-cholesten-3-one (an allylic ketone), as well
as allylic alcohol oxidation products of polyunsaturated
fatty acids, which should help its apoptogenic activity.
3d. Exogenous SM, as a component of the Cer cocktail, can be considered a prodrug, since it is converted to
Cer in the tumor. Some tumors contain little SM, probably because they convert most of their Cer to the proliferative sphingolipids. Use of a short-chain SM might be
better than natural SM, since it should penetrate tumors
more rapidly. Dietary SM was reported to slow the appearance of colonic tumors [28], and the effectiveness
of 5-fluorouracil and irinotecan for treating tumor xenografts in mice was enhanced by the inclusion of SM.
(Irinotecan, like camptothecin, is an allylic alcohol that
induces faster Cer biosynthesis.) SM resembles lecithin
structurally, so it can be incorporated into a liposomal
dispersion of anticancer drugs (such as vincristin). It is
readily prepared from natural sources.
Experimental Oncology 25, 3-10, 2004 (March)
3e. Cholesterol binds strongly to SM, apparently reducing its sensitivity to SMase. Thus any steps available for reducing tumor levels of cholesterol may prove
helpful in cancer therapy. Dietary control of fat (a known
stimulator of cancer growth) and cholesterol intake
seems like a simple way to enhance the Cer cocktail.
Statins, which slow cholesterol biosynthesis, have
shown promising anticancer activity and apoptosis in
cultured cells, but human studies have not been encouraging, possibly because statins also reduce the
cellular level of ubiquinone. One mechanism by which
Cer induces apoptosis is by interference with ubiquinone
function (section 2d) so a deficiency of ubiquinone in a
tumor might be undesirable. Perhaps statins would be
effective if used together with supplementary ubiquinone and the Cer cocktail.
3f. Dihydroxy-vitamin D3, the active form of vitamin D,
stimulates SM hydrolysis [29], produces apoptosis, and
has been found to be helpful in cancer chemotherapy. It
is significant that it is an allylic alcohol. High doses upset
body Ca2+ metabolism and efforts are currently underway to synthesize improved analogues.
4. Forcing ceramide accumulation by blocking
its glucosylation
4a. Glucosylceramide and lactosylCer stimulate cell
proliferation and appear to be involved in tumor multidrug resistance, so the enzymes that produce these
glucosphingolipids should be blocked. Cer analogs in
which the terminal OH is replaced by an amine are
potent inhibitors of Cer glucosylation. Similar analogs
in which the last 15 carbon atoms of sphingosine are
replaced by a phenyl ring (the “P-drugs”) are also highly
effective, in vitro and in vivo [30, 31]. Injection of Dthreo-1-(ethylenedioxy)phenyl-2-palmitoylamino-3pyrrolidino-propanol-1 into mice with Fabry disease
for 3 days at 1 mg every 12 h resulted in 50% depletion
of renal GlcCer [32]. During the period of treatment,
GlcCer synthesis is strongly slowed and the complex
glucosphingolipids and GlcCer gradually disappear by
hydrolytic turnover. This sort of compound, now under
commercialization, has been recommended for use in
treating sphingolipidoses [33], cancer, and other diseases of proliferation. In mice bearing Ehrlich ascites
cancer cells, administration of a P-drug for only 10 days
succeeded in permanently curing 30-40% of the mice
and greatly prolonged life in the others [30]. The cured
mice were unaffected by reinoculation with more cancer cells, suggesting that the drug also enabled the
development of immune resistance.
N-butyldeoxynojirimycin has also been reported to be
useful for slowing the conversion of Cer to GlcCer, but the
data on its effectiveness are not entirely satisfying.
4b. Tamoxifen is an anti-estrogen with valuable
anticancer activity. It also inhibits Cer glucosylation,
especially in multi-drug resistant cancer cells [34].
Tamoxifen seems to help prevent the appearance of
cancer, a feature in agreement with other data suggesting that chronic elevation of tissue Cer slows or
prevents cancer. Raloxifene, a related antineoplastic
drug, may also slow GlcCer synthesis.
Experimental Oncology 25, 3-10, 2004 (March)
4c. RU486 (mifepristone), the abortion-inducing
drug, is an inhibitor of Cer glucosylation, producing Cer
accumulation and apoptosis [35]. It has also shown
promise in human chemotherapy. It is a conjugated allylic ketone which inactivates a cytochrome c, probably
by Michael condensation with a CySH moiety.
4d. Since testosterone stimulates Cer glucosylation and inhibits GlcCer hydrolysis in mouse kidneys, it
elevates the synthesis of the proliferative glucosphingolipids [36]. This effect is seen in growing mice, in
which the male kidneys grow much faster than female
kidneys. Anti-androgen chemotherapy can be expected
to slow Cer glucosylation, especially in androgen sensitive tumors (prostate, kidney, lung, brain). The effects on other sphingolipids need study.
4e. Cellular glucose levels affect the rate of UDP-glucose biosynthesis, thereby affecting Cer glucosylation.
High blood glucose levels, as in diabetics, speed the diversion of Cer into glucosphingolipids, and thus should
be avoided. Studies have shown the proliferative effects
of high glucose availability, as well as high fat availability
and the consequent stimulation of tumor growth.
5. Elevating ceramide in tumors by blocking ceramidase
Acid and neutral/alkaline ceramidases, particularly
in mitochondria, tend to prevent Cer accumulation and
apoptosis. Some tumors have a high ceramidase activity and thus protect themselves from apoptosis [37].
Blocking ceramidase is especially important for cancer
therapy since its product, sphingosine, is converted in
part to the mitogenic lipid, sphingosine-1-phosphate.
5a. N-oleoylethanolamine, a very simple amide that
resembles the first two carbons of Cer, inhibits the acid
type of ceramidase and tends to produce the typical
changes of Cer-induced apoptosis [38, 39]. Some
studies have shown that the amide also stimulates neutral SMase and inhibits Cer glucosylation, thus it elevates tumor Cer by modulating three Cer enzymes.
More effort should be invested in searching for a more
stable and potent substitute.
5b. An aromatic Cer analog, D-erythro-2-tetradecanoylamino-1-phenyl-1-propanol, inhibits neutral/
alkaline ceramidase, elevating Cer and slowing cell
growth. Like natural Cer, it has the D-erythro configuration. This analog, as well as a related one (2-Nmyristoylamino-1-(4-nitrophenyl)-1,3-propandiol)
was active against acid ceramidase, producing Cer
accumulation and apoptosis in normal and melanoma
cells [40]. As the authors suggest, this approach to cancer chemotherapy looks very promising.
5c. Mitochondrial alkaline Cerase can be efficiently
inhibited with a secondary amine version of Cer
(stearoyl sphingosine that has been reduced to stearyl
sphingosine) [41].
6. Blocking the conversion of ceramide to
sphingomyelin
A portion of freshly synthesized Cer is diverted by
reaction with a phospholipid, particularly phosphatidylcholine (lecithin), forming SM. The Cer is thus unavailable for producing apoptosis (although the hydrolysis
7
of SM can be stimulated — section 3). The reaction is
a simple, reversible transesterification, in which the
phosphocholine moiety is attached to the Cer C-1 OH,
leaving diacyl glycerol. A similar exchange between
phosphatidylethanolamine and Cer yields Cer-phosphoethanol, which then gets methylated to form more
SM. An elevated lecithin content in cells, seen in many
tumors, minimizes the apoptogenic action of several
Cer elevating agents. Conversely, subnormal lecithin
content tends to produce apoptosis. The P-drug inhibitors of Cer glucosylation divert the extra Cer into
SM synthesis, so such drugs should be used together
with agents that lower tumor phospholipid content.
Emulsification of anticancer drugs with the aid of lecithin should also be avoided.
Consistent with these observations are the findings
that Cer inhibits lecithin synthesis and GlcCer stimulates it. Moreover Cer destroys lecithin and phosphatidylethanolamine by transferring the phospholipid
arachidonate to their C-1 position (section 3b). Tumors,
by keeping their Cer content low and GlcCer high thus
stimulate their own growth.
6a. Phospholipases lower phospholipid levels, helping to keep Cer levels high. Several agents stimulate
these enzymes. Cer-1-phosphate is a direct activator
of cytosolic phospholipase A2 [42], a factor that needs
study. Hexadecylphosphocholine probably acts this
way, inducing an increase in ceramide level, lowering
lecithin and SM, and producing apoptosis. It is considered too toxic to use in patients as a solitary drug but
might well be useful at low dosages in the Cer cocktail.
Other lecithin-like drugs might be better.
6b. Tamoxifen (section 4b) stimulates phospholipases C and D, lowering lecithin levels. This effect augments its inhibition of Cer glucosylation.
7. Slowing synthesis of sphingosine phosphate
Sphingosine-1-phosphate, like Cer-1-phosphate
and deacylated SM (sphingosyl phosphocholine) is a
highly potent stimulator of cell growth and proliferation,
also producing a wide variety of physiological effects.
Since it is formed by phosphorylation of sphingosine,
the product of ceramidase action, it is important to slow
Cer hydrolysis (section 5). It is possible to inhibit the
kinase that attacks the free base with N,N-dimethyl
sphingosine, N,N,N-trimethyl sphingosine, or L-threosphinganine. The latter, also called safingol, showed
some promise in cancer patients when used alone. One
of these inhibitors would be useful if the ceramidases
are not fully inhibited. No stimulator is yet available for
the phosphatases acting on the above phosphate esters or on the lyase that cleaves sphingosine phosphate, but it might be useful to try synthesizing one.
The same possibility applies to the kinase that converts Cer to Cer-1-phosphate.
Unfortunately, sphingosine phosphate and Cer
phosphate stimulate mitochondrial ceramidase, an additional reason to prevent their formation [42].
8. Stimulating glucosylceramide hydrolysis
8a. GlcCer stimulates cell growth and proliferation
and generates the gangliosides, which prevent the pa-
8
tient’s immune system from attacking the abnormal components of tumors. Thus its synthesis should not only be
slowed (section 4) but its hydrolysis (to yield Cer) should
be stimulated. The glucosidase acting on GlcCer is normally activated by acidic phospholipids, and ingesting
them could be useful. However, the natural acidic phospholipids are readily attacked by enzymes and might
not last long enough to be useful. A search for a stable
synthetic phospholipid should be made.
8b. The glucosidase also needs a small, heat-stable protein called saposin C, which stimulates the enzyme and protects it against proteolytic breakdown.
Studies of saposin C showed that only a small portion
of the peptide is needed for activity, and efforts have
been made to synthesize an unnatural but stable mimic
of this peptide. A 14-mer peptide analog, Prosaptide
TX14(A), looks promising and has shown promise as a
neurological deficit drug in human trials. While its ability to stimulate the glucosidase does not seem to have
been measured, it produces the expected increase in
gangliosides when added to neuroblastoma cells [43].
8c. GlcCer glucosidase catalyzes transglycosidation as well as hydrolysis, allowing the degradation of
GlcCer to proceed faster than hydrolysis. Retinol and
pentanol accept the glucose moiety, liberating Cer [44].
Individuals whose tumors possess good β-glucosidase
activity (a glucosidase unable to attack GlcCer) may
be able to destroy the tumor’s GlcCer rapidly. Retinol
(Vitamin A) is a known antioxidant, helping to prevent
the appearance of cancer. It undergoes oxidation to
4-keto retinol, retinoic acid, and 4-keto retinoic acid.
Retinoic acid and the keto metabolites (conjugated allylic ketones) very likely exert anticancer activity [14].
The transglycosidation approach deserves more study.
8d. The human glucosidase that hydrolyzes GlcCer
is commercially available for Gaucher disease patients
(whose enzyme is too slow). Supplementing a cancer
patient’s glucosidase with the medicinal enzyme may
increase the tumor’s rate of GlcCer destruction and help
force the tumor to accumulate Cer. While the enzyme is
very expensive, it would not be needed for a long time.
9. Other Cer-controlling factors
The mechanisms by which Cer exerts its signals for
cell function are under intense investigation, and these
undoubtedly raise additional potentially useful possibilities for therapy [45].
Chlorpromazine, the anti-psychotic drug, increased
the Cer concentration in mouse liver by 25% within 6 h
after injection [46] and produced apoptosis in cultured CHO
cells. Administered to mice bearing sarcoma 37 cells,
chlorpromazine (at a high dosage) almost completely
halted the growth of the tumors [47]. The cancer cells in
the treated mice revealed formation of fatty granules,
stained with Oil Red O, consistent with accumulation of
Cer. However, the slow growth could be attributed to the
heavy sedation of the mice. A more recent study with rats
also showed that chlorpromazine produced apoptosis; in
this study, the cells were in the thymus and revealed an
increase of 15-fold in the degree of apoptosis [48]. The
authors attributed the effect to interference with Ca2+ me-
Experimental Oncology 25, 3-10, 2004 (March)
tabolism. Many reports of Cer effects have shown that
marked changes in Ca2+ distribution can occur.
Cytokines affect sphingolipid enzymes and might
be useful to control but also may exert too many effects. The Cer content of cultured cells seems to rise
rapidly as the cells approach confluency, which can
make in vitro experiments difficult to interpret. Age of
the patient may be an important variable, since GlcCer
synthase decreases rapidly in late life.
Since so many established or proposed anticancer
drugs contain an allylic alcohol moiety in their structure, they may work like ceramide, interfering with
Coenzyme Q function in mitochondria or by activating
protein phosphorylation and dephosphorylation. They
differ from each other in the tissue of action or type of
tumor that is attacked, so it seems worth using the most
suitable one for each patient together with the Cer cocktail. In addition, the many antineoplastic drugs that contain an allylic ketone moiety may mimic my proposed
keto ceramide and could then also be used together
with the Cer-inducing cocktail. It is possible that they
undergo reduction to allylic alcohols in the tumor, and
mimic Cer more closely. Perhaps their main function is
to decrease the tumor’s GSH content (section 3a).
The impact of dietary allylic alcohols and allylic ketones should not be overlooked, since so many foods
contain them and so many are available in purified form.
Curcumin (in Indian spice), indole-3-carbinol (in broccoli), genistein (in soy tofu), and conjugated linoleic acid
(in milk fat) are commercially available and show usefulness in cancer studies. While this last item does not
contain an allylic group, it is undoubtedly oxygenated
in the body to yield an allylic alcohol. This same is true
of vitamin A, which is metabolized to allylic ketones
(section 8c). Vitamin D3, normally formed in the skin
by sunlight and activated by metabolic hydroxylation to
form the allylic alcohol, seems to protect against the
appearance of cancer by stimulating SM hydrolysis
(section 3f). People living in northern latitudes or who
work indoors have been shown to resist carcinogenesis by eating extra vitamine D. Another kind of dietary
factor to consider is the need to eliminate the contamination of corn by a parasitic mold, which forms an inhibitor of Cer synthesis, fumonisin B1. This widely prevalent contaminant has been found at dangerous levels
that chronically slow the acylation of sphingosine and
keto sphinganine. The inhibitor, a sphingosine analog,
tends to produce cancer.
Recent work with nitric oxide has shown that the
chronic application of NO sources to cultured cells results in gradual Cer elevation, then apoptosis [49]. Other
work has suggested there is a mutual stimulation relationship that should be considered.
Readers who plan to try the above recommendations
in animals or patients should keep in mind the complexity
of Cer’s interactions with normal tissues and tumors, as
opposed to cultured cancer cells. It is essential to make
many measurements of the changes induced by their polydrug Cer-inducing cocktail. This review definitely does not
list all the known (and unknown) relevant interactions.
Experimental Oncology 25, 3-10, 2004 (March)
REFERENCES
1. Caldwell S, Heitger A, Shen WP, Liu YH, Taylor B,
Ladisch S. Mechanisms of ganglioside inhibition of APC
function. J Immunol 2003; 171: 1676–83.
2. Radin NS. Killing cancer cells by poly-drug elevation of ceramide levels: a hypothesis whose time has come?
Eur J Biochem 2001; 268: 193–204.
3. Radin NS. Killing tumors by ceramide-induced apoptosis: critique of available drugs. Biochem J 2003; 371: 243–56.
4. Cox TM. Future perspectives for glycolipid research
in medicine. Philosoph Trans Royal Soc London — Series
B: Biol Sci 2003; 358: 967–73.
5. Kolter T, Sandhoff K. Sphingolipids — their metabolic pathways and the pathobiochemistry of neurodegenerative diseases. Angew Chem (Engl) 1999; 38: 1533–68.
6. Kester M, Kolesnick R. Sphingolipids as therapeutics. Pharmacol Res 2003; 47: 365–71.
7. Radin NS, Inokuchi J. Glucosphingolipids as sites
of action in the chemotherapy of cancer. Biochem Pharmacol 1988; 37: 2879–86.
8. Pettus BJ, Chalfant CE, Hannun YA. Ceramide in
apoptosis: an overview and current perspectives. Biochim
Biophys Acta 2002; 1585: 114–25.
9. Shayman JS, Radin NS. Structure and function of renal glycosphingolipids. Am J Physiol 1991; 260: F291–F302.
10. Radin NS. The development of aggressive cancer —
a possible role for sphingolipids. Cancer Invest 2002;
20: 779–86.
11. Lavie Y, Cao H, Volner A, Lucci A, Han T-Y, Geffen V, Giuliano AE, Cabot MC. Agents that reverse multidrug resistance, tamoxifen, verapamil, and cyclosporin A, block
glycosphingolipid metabolism by inhibiting ceramide glycosylation in human cancer cells. J Biol Chem 1997; 272: 1682–7.
12. Vescovo G, Ravara B, Gobbo V, Sandri M, Angelini A, Della Barbera M, Dona M, Peluso G, Calvani M,
Mosconi L, Dalla Libera L. L-Carnitine: a potential treatment for blocking apoptosis and preventing skeletal muscle myopathy in heart failure. Am J Physiol Cell Physiol
2002; 283: C802–10.
13. Schmelz EM, Bushnev AS, Dillehay DL, Sullards MC,
Liotta DC, Merrill AHJr. Ceramide-β-D-glucuronide: Synthesis, digestion, and suppression of early markers of colon
carcinogenesis. Cancer Res 1999; 59: 5768–72.
14. Radin NS. Designing anticancer drugs via the Achilles heel: ceramide, allylic ketones, and mitochondria. Bioorg
Med Chem 2003; 11: 2123–42.
15. Brodesser S, Sawatzki P, Kolter T. Bioorganic
chemistry of ceramide. Eur J Org Chem 2003; 2021–34.
16. Radin NS. Apoptotic death by ceramide: will the real
killer please stand up? Med Hypotheses 2001; 57: 96–100.
17. Korynevska AV, Matselyukh BP, Stoika RS.
In vitro study of landomycin E antitumor activity. Exp
Oncol 2003; 25: 96–104.
18. Radin NS. Killing tumors by ceramide-induced apoptosis: critique of available drugs. Biochem J 2003; 371: 243–56.
19. Kowluru A, Metz SA. Ceramide-activated protein
phosphatase-2A activity in insulin-secreting cells. FEBS Lett
1997; 418: 179–82.
20. Westwick JK, Bielawska AE, Dbaibo G, Hannun YA, Brenner DA. Ceramide activates the stress-activated protein kinases. J Biol Chem 1995; 270: 22689-92.
21. Chun J, Byun HS, Arthur G, Bittman R. Synthesis and growth inhibitory activity of chiral 5-hydroxy-2N-acyl-(3E)-sphingenines: ceramides with an unusual sphingoid backbone. J Org Chem 2003; 68: 355–9.
22. Chun J, Byun HS, Bittman R. First asymmetric
synthesis of 6-hydroxy-4-sphingenine-containing ceramides.
Use of chiral propargylic alcohols to prepare a lipid found
in human skin. J Org Chem 2003; 68: 348–54.
9
23. Liu B, Andrieu-Abadie N, Levade T, Zhang P,
Obeid LM, Hannun YA. Glutathione regulation of neutral sphingomyelinase in tumor necrosis factor-α-induced
cell death. J Biol Chem 1998; 273: 11313–20.
24. Miller AC, Whittaker T, Thibault A, Samid D.
Modulation of radiation response of human tumor cells by
the differentiation inducers, phenylacetate and phenylbutyrate. Int J Radiat Biol 1997; 72: 211–8.
25. Jayadev S, Linardic CM, Hannun YA. Identification of arachidonic acid as a mediator of sphingomyelin
hydrolysis in response to tumor necrosis factor alpha. J Biol
Chem 1994; 269: 5757–63.
26. Taketo MM, Sonoshita M. Phospolipase A2 and
apoptosis. Biochim Biophys Acta 2002; 1585: 72–6.
27. Auge N, Garcia V, Maupas-Schwalm F, Levade T,
Salvayre R, Negre-Salvayre A. Oxidized LDL-induced
smooth muscle cell proliferation involves the EGF receptor/PI-3 kinase/Akt and the sphingolipid signaling pathways. Arterioscler Thromb Vasc Biol 2002; 22: 1990–5.
28. Schmelz EM, Sullards MC, Dillehay DL, Merrill AH.
Colonic cell proliferation and aberrant crypt foci formation
are inhibited by dairy glycosphingolipids in 1,2-dimethylhydrazine-treated CF1 mice. J Nutr 2000; 130: 522–7.
29. Okazaki T, Bell RM, Hannun YA. Sphingomyelin
turnover induced by vitamin D3 in HL-60 cells. Role in
cell differentiation. J Biol Chem 1989; 264: 19076–80.
30. Inokuchi J, Mason I, Radin NS. Antitumor activity in mice of an inhibitor of glycosphingolipid biosynthesis. Cancer Lett 1987; 38: 23–30.
31. Abe A, Radin NS, Shayman JA, Wotring LL,
Zipkin RE, Sivakumar R, Ruggieri JM, Carson KG,
Ganem B. Structural and stereochemical studies of potent
inhibitors of glucosylceramide synthase and tumor cell
growth. J Lipid Res 1995; 36: 611–21.
32. Abe A, Gregory S, Lee L, Killen PD, Brady RO,
Kulkarni A, Shayman JA. Reduction of globotriaosylceramide in Fabry disease mice by substrate deprivation.
J Clin Invest 2000; 105: 1563–71.
33. Radin NS. Treatment of Gaucher disease with an
enzyme inhibitor. Glycoconjugate J 1996; 13: 153–7.
34. Lavie Y, Cao H, Volner A, Lucci A, Han T-Y, Geffen V, Giuliano AE, Cabot MC. Agents that reverse multidrug resistance, tamoxifen, verapamil, and cyclosporin A, block
glycosphingolipid metabolism by inhibiting ceramide glycosylation in human cancer cells. J Biol Chem 1997; 272: 1682–7.
35. Lucci A, Giuliano AE, Han TY, Dinur T, Liu YY,
Senchenkov A, Cabot MC. Ceramide toxicity and metabolism differ in wild-type and multidrug-resistant cancer cells. Int J Oncol 1999; 15: 535–40.
36. Shukla A, Shukla GS, Radin NS. Control of kidney size by sex hormones; possible involvement of glucosylceramide. Amer J Physiol 1992; 262: F24–9.
37. Selzner M, Bielawska A, Morse MA, Rudiger HA,
Sindram D, Hannun YA, Clavien PA. Induction of apoptotic cell death and prevention of tumor growth by ceramide analogues in metastatic human colon cancer. Cancer
Res 2001; 61: 1233–40.
38. Sugita M, Willians M, Dulaney JT, Moser HW.
Ceramidase and ceramide synthesis in human kidney and
cerebellum. Description of a new alkaline ceramidase. Biochim Biophys Acta 1975; 398: 125–31.
39. Scurlock B, Dawson G. Differential responses of
oligodendrocytes to tumor necrosis factor and other proapoptotic agents: Role of ceramide in apoptosis. J Neurosci
Res 1999; 55: 514–22.
40. Raisova M, Goltz G, Bektas M, Bielawska A, Riebeling C, Hossini AM, Eberle J, Hannun YA, Orfanos CE,
Geilen CC. Bcl-2 overexpression prevents apoptosis induced
by ceramidase inhibitors in malignant melanoma and HaCaT
keratinocytes. FEBS Lett 2002; 516: 47–52.
10
Experimental Oncology 25, 3-10, 2004 (March)
41. Usta J, El Bawab S, Roddy P, Szulc ZM, Yusuf,
Hannun A, Bielawska A. Structural requirements of ceramide and sphingosine based inhibitors of mitochondrial
ceramidase. Biochem 2001; 40: 9657–68.
42. Pettus BJ, Bielawska A, Subramanian P, Wijesinghe DS, Maceyka M, Leslie CC, Evans JH, Freiberg J,
Roddy P, Hannun YA, Chalfant CE. Ceramide-1-phosphate is a direct activator of cytosolic phospholipase A2.
J Biol Chem (in press).
43. Misasi R, Sorice M, Carson GS, Griggi T, Lenti L,
Pontieri GM, O’Brien JS. Prosaposin and prosaptide, a
peptide from prosaposin, induce an increase in ganglioside
content on NS20Y neuroblastoma cells. Glycoconj J 1996;
13: 195–202.
44. Vanderjagt DJ, Fry DE, Glew RH. Human glucocerebrosidase catalyses transglucosylation between GlcCer
and retinol. Biochem J 1994; 300: 309–15.
45. Ruvolo PP. Intracellular signal transduction pathways activated by ceramide and its metabolites. Pharmacol Res 2003; 47: 383–92.
46. Hospattankar AV, Vunnam RR, Radin NS. Changes in liver lipids after administration of 2-decanoylamino3-morpholinopropiophenone and chlorpromazine. Lipids
1982; 17: 538–43.
47. Belkin M, Hardy WG. Effect of reserpine and chlorpromazine on sarcoma 37. Science 1957; 123: 233–4.
48. Balakumaran A, Campbell GA, Moslen MT. Calcium channel blockers induce thymic apoptosis in vivo in
rats. Toxicol Appl Pharmacol 1996; 139: 122–7.
49. Huwiler A, Dorsch S, Briner VA, Van den Bosch H,
Pfeilschifter J. Nitric oxide stimulates chronic ceramide
formation in glomerular endothelial cells. Biochem Biophys
Res Commun 1999; 258: 60–5.
ÖÅÐÀÌÈÄ ÊÀÊ ÎÑÍÎÂÍÎÉ ÊÎÌÏÎÍÅÍÒ
ÏÐÎÒÈÂÎÎÏÓÕÎËÅÂÎÉ ÏÎËÈÕÈÌÈÎÒÅÐÀÏÈÈ
Òûñÿ÷è ïóáëèêàöèé ñîîáùàþò, ÷òî ìíîãèå òèïû îïóõîëåâûõ êëåòîê è íîâîîáðàçîâàíèé ìîãóò áûòü óíè÷òîæåíû
âîçäåéñòâèÿìè, ïðèâîäÿùèìè ê ïîâûøåíèþ êîíöåíòðàöèè ñôèíãîëèïèäà öåðàìèäà (Cer). Ñóùåñòâóåò ìíîãî
ñïîñîáîâ óâåëè÷èòü êîëè÷åñòâî Cer â îïóõîëè, íî èõ ïðèìåíåíèå îñëîæíÿåòñÿ òåì, ÷òî Cer âûïîëíÿåò
öåíòðàëüíóþ ðîëü â ãîìåîñòàçå êëåòêè: ëåãêî ìåòàáîëèçèðóåòñÿ ñ îáðàçîâàíèåì äðóãèõ ñôèíãîëèïèäîâ,
ñïîñîáñòâóþùèõ ðîñòó îïóõîëè, ìåòàñòàçèðîâàíèþ è ïðîòèâîäåéñòâèþ èììóííîé ñèñòåìå ïàöèåíòà. Îòìå÷åíà
íåîáõîäèìîñòü ïðåäîòâðàùåíèÿ òàêîé ìåòàáîëè÷åñêîé êîíâåðñèè íà ôîíå îäíîâðåìåííîé àêòèâàöèè ôåðìåíòîâ,
ó÷àñòâóþùèõ â ñèíòåçå Cer, îïèñàíû ôåðìåíòû, êîòîðûå ñëåäóåò àêòèâèðîâàòü èëè èíãèáèðîâàòü, à òàêæå
ëåêàðñòâà, ìåòàáîëèòû è êîìïîíåíòû ðàöèîíà, ìîäèôèöèðóþùèå êàæäûé ôåðìåíò. Îñâåùåíà âàæíîñòü
àëëèëüíîé ñïèðòîâîé ãðóïïû â ìîëåêóëå Cer è ðÿäå ïðîòèâîîïóõîëåâûõ àãåíòîâ, óêàçàíî, ÷òî ãèäðîêñèëüíàÿ
ãðóïïà ó÷àñòâóåò â ïåðåíîñå ôîñôàòà îò áåëêà ê áåëêó ïóòåì îáðàçîâàíèÿ ýôèðà ôîñôàòà. Àëëèëüíàÿ
ãèäðîêñèëüíàÿ ãðóïïà ìîæåò òàêæå ñîêðàùàòü ÷èñëî êåòîíîâ â ìèòîõîíäðèàëüíûõ óáèõèíîíàõ ñ îáðàçîâàíèåì
ðåàêòèâíûõ ôîðì êèñëîðîäà. Óðîâåíü Cer â îïóõîëÿõ ìîæåò áûòü ïîâûøåí çà ñ÷åò 1) ïðÿìîãî ââåäåíèÿ Cer
èëè åãî àíàëîãîâ; 2) ñòèìóëÿöèè îáðàçîâàíèÿ Cer èç åãî ïðåäøåñòâåííèêîâ; 3) ïóòåì ãèäðîëèçà ñôèíãîìèåëèíà
èëè (4) ãèäðîëèçà ãëèêîñôèíãîëèïèäîâ; 5) àöèëèðîâàíèÿ ñôèíãîçèíà. Êðîìå òîãî, áîëåå âûñîêàÿ êîíöåíòðàöèÿ
Cer ìîæåò áûòü îáóñëîâëåíà çàìåäëåíèåì åãî êîíâåðñèè â (6) ñôèíãîìèåëèí, (7) ãëèêîçèëCer, (8) Cer-ôîñôàò,
(9) ñôèíãîçèí + æèðíûå êèñëîòû ïóòåì ãèäðîëèçà. Ëó÷åâàÿ òåðàïèÿ ñòèìóëèðóåò ñèíòåç Cer de novo â îïóõîëÿõ.
Âîçìîæíî, òàêæå íåîáõîäèìî áëîêèðîâàòü êîíâåðñèþ ñôèíãîçèíà (èç Cer) â ñôèíãîçèíôîñôàò.
Êëþ÷åâûå ñëîâà: àïîïòîç, ñôèíãîëèïèäû, àëëèëüíûå ñïèðòîâûå ãðóïïû, ïîëèõèìèîòåðàïèÿ, êîíöåíòðàöèÿ öåðàìèäà.
Copyright © Experimental Oncology, 2004