Biochimica et Biophysica Acta 1585 (2002) 188 – 192
www.bba-direct.com
Review
Fumonisins and fumonisin analogs as inhibitors of
ceramide synthase and inducers of apoptosis
Kena Desai a, M. Cameron Sullards b, Jeremy Allegood a, Elaine Wang a, Eva M. Schmelz c,
Michaela Hartl d, Hans-Ulrich Humpf d, D.C. Liotta e, Qiong Peng a, Alfred H. Merrill Jr. a,*
b
a
School of Biology, Georgia Institute of Technology, Atlanta, GA 30332-0230, USA
Department of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, GA 30332-0230, USA
c
Karmanos Cancer Institute, Wayne State University School of Medicine, Detroit, MI 48201, USA
d
Institut für Lebensmittelchemie, Universität Münster, D-48149 Münster, Germany
e
Department of Chemistry, Emory University, Atlanta, GA 30322, USA
Received 20 March 2002; received in revised form 24 July 2002; accepted 29 October 2002
Abstract
Sphingoid bases are growth inhibitory and pro-apoptotic for many types of cells when added to cells exogenously, and can be elevated to
toxic amounts endogenously when cells are exposed to inhibitors of ceramide synthase. An important category of naturally occurring
inhibitors are the fumonisins, which inhibit ceramide synthase through structural similarities with both the sphingoid base and fatty acyl-CoA
co-substrates. Fumonisins cause a wide spectrum of disease (liver and renal toxicity and carcinogenesis, neurotoxicity, induction of
pulmonary edema, and others), and most—possibly all—of the pathophysiologic effects of fumonisins are attributable to disruption of the
sphingolipid metabolism. The products of alkaline hydrolysis of fumonisins (which occurs during the preparation of masa flour for tortillas)
are aminopentols that also inhibit ceramide synthase, but more weakly. Nonetheless, the aminopentols (and other 1-deoxy analogs of
sphinganine) are acylated to derivatives that inhibit ceramide synthase, perhaps as product analogs, elevate sphinganine, and kill the cells.
Somewhat paradoxically, fumonisins sometimes stimulate growth and inhibit apoptosis, possibly due to elevation of sphinganine 1phosphate, which is known to have these cellular effects. These findings underscore the complexity of sphingolipid metabolism and the
difficulty of identifying the pertinent mediators unless a full profile of the potentially bioactive species is evaluated.
D 2002 Elsevier Science B.V. All rights reserved.
Keywords: Fumonisin; Aminopentol; Ceramide synthase; Sphinganine; Sphinganine 1-phosphate; Toxicity; Apoptosis; Carcinogenesis
Fumonisins are a family of mycotoxins produced by
Fusarium verticillioides (formerly F. moniliforme) [1],
which are common fungal contaminants of corn and some
other grains [2]. As much as 59% of the corn and cornbased products worldwide have been estimated to be contaminated with variable amounts of fumonisin B1 (FB1)
(Fig. 1), the most prevalent of the fumonisin subspecies [2–
4]. Fumonisins cause an at-first-glance puzzling range of
diseases: liver and kidney toxicity and carcinogenicity,
pulmonary edema, immunosuppression (and sometimes
immunostimulation), neurotoxicity, and probably others
[1,5]. Most or all of the toxicities resulting from exposure
to these compounds can be explained by their ability to alter
sphingolipid metabolism by inhibiting ceramide synthase
[5,6].
Fumonisins have also been used extensively to explore
the functions of ceramides and complex sphingolipids in
cell culture. Quite paradoxically, some of these studies use
fumonisins to inhibit apoptosis [7], and in some contexts,
fumonisins are growth stimulatory and anti-apoptotic
rather than cytotoxic [8]. These reasons for these dichotomous effects are not fully known, but this review will
explore the most likely answers as well as discuss other
surprises that have surfaced while studying this category
of compounds.
1. Inhibition of ceramide synthase by fumonisins
* Corresponding author. Tel.: +1-404-385-2842; fax: +1-404-3852917, +1-404-894-0519.
E-mail address: [email protected] (A.H. Merrill).
FB1 is comprised of a long-chain aminopentol backbone
(AP1) (Fig. 1) with two ester-linked tricarballylic acids [1].
1388-1981/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved.
PII: S 1 3 8 8 - 1 9 8 1 ( 0 2 ) 0 0 3 4 0 - 2
K. Desai et al. / Biochimica et Biophysica Acta 1585 (2002) 188–192
189
ition [10]; and (3) upon removing the tricarballylic acids,
AP1 becomes not only an inhibitor but also a substrate for
acylation by ceramide synthase [10]. As will be discussed
later, this acylation may explain why aminopentols remain
toxic to animals despite the lower potency of these compounds as ceramide synthase inhibitors.
2. Induction of apoptosis by fumonisins
Fig. 1. Structures of FB1, AP1, and PAP1.
In the current model for how fumonisins inhibit ceramide
synthase, the aminopentol backbone competes for binding
of the sphingoid base substrate, whereas the anionic tricarballylic acids interfere with binding of the fatty acyl-CoA
[5]. Three lines of evidence support this model: (1) the
potency with which FB1 inhibits ceramide synthase is
sensitive to the concentrations of both sphingoid bases
and fatty acyl-CoA [9]; (2) removal of the tricarballylic
acids diminishes the potency of ceramide synthase inhib-
When fumonisins are added to cells in culture, they
cause rapid and dramatic elevations (i.e., one to two orders
of magnitude) in sphinganine due to inhibition of de novo
sphingolipid biosynthesis (Fig. 2). This most often results
in growth arrest [11,12] and apoptosis [13]. The growth
arrest has been shown to be due to up-regulation of
inhibitors of cyclin-dependent protein kinases (Cip1, Kip1
and Kip2) at both transcriptional and post-transcriptional
levels [11,12]. The mechanism(s) for the induction of
apoptosis are less clear, but probably involve disruption
of many cell regulatory pathways, including inhibition of
protein kinase C [14], key hallmarks of apoptotic signaling
[15], disruption of the endothelial barrier [16] and others.
With so many changes, it is possible that fumonisins are
acting via more than one initial target, and it appears that
they may affect MAPK directly [17], although this is
complicated by the involvement of sphingolipids in many
MAPK pathways [18]. All in all, the most conclusive
evidence to date that elevations in sphinganine mediate at
least the earliest toxicity of fumonisins is that an inhibitor
Fig. 2. Disruption of sphingolipid metabolism by fumonisins. Shown are the inhibition of the acylation of sphinganine and sphingosine by FB1, resulting in
elevations in these sphingoid bases (and sometimes the sphingoid base 1-phosphates) and reductions in complex sphingolipids. Free sphinganine (and
sometimes sphingosine) efflux from the cell and appear in serum and urine.
190
K. Desai et al. / Biochimica et Biophysica Acta 1585 (2002) 188–192
of upstream enzyme of sphingolipid biosynthesis (inhibition of serine palmitoyltransferase by ISP1) suppresses the
toxicity [13].
The results with fumonisins in cell culture are paralleled
with findings in vivo. When animals consume fumonisins,
the earliest biochemical alteration that is detected is a timeand dose-dependent elevation in tissue, serum and urinary
sphinganine, and usually at later times or higher doses
(when major tissue damage has occurred) an elevation in
sphingosine [19]. On this basis, increases in the ratio of
sphinganine to sphingosine has proven to be a useful
biomarker for exposure to fumonisins [20]. Hepatotoxicity
occurs in response to FB1 consumption in most exposed
animals, and is characterized by cell death in the form of
both apoptosis and necrosis; additionally, there are indications of spontaneous cell regeneration (mitogenesis), a
condition favorable for carcinogenesis [21 – 24]. According
to the recently published study by the National Center for
Toxicologic Research [24], there is clear evidence of carcinogenic activity of FB1 in male, but not female, F344/N rats
based on the increased incidences of renal tubule neoplasms,
and there was clear evidence of carcinogenic activity in
female (but not male) B6C3F1 mice based on the increased
incidences of hepatocellular neoplasms.
Considering the complexity of sphingolipid metabolism
(with both pro- and anti-apoptotic aspects, as will be
discussed below), it is probably difficult to select the precise
timepoint to conduct in vivo analyses of the effects of
fumonisins. For example, in the earliest phase, the tissues
that may be most affected are those with high rates of
sphingolipid synthesis de novo (such as liver and kidney),
whereas other tissues with low rates of sphingolipid biosynthesis may be damaged more by uptake of sphingoid bases
from circulation or neighboring cells. In vivo studies are
additionally complicated by the complexities of its uptake,
distribution and excretion. This has been studied mostly in
rodents, where it has been shown that FB1 is largely
excreted in fecal matter (80%) and to a lesser degree in
urine (3%) [25]; however, en route, it undergoes biliary
secretion and enterhepatic circulation [26]. The portion that
is retained is primarily found in liver, and to a lesser extent,
the kidneys [23,25].
Since it appears that an imbalance between cell death
and regeneration primarily contributes to the FB1-induced
carcinogenesis [6]; the variable expression or activity of
enzymes that mediate sphingolipid metabolism in cells
must be considered when one seeks to understand species
and tissue differences in the responses to these mycotoxins. It is possible that a relatively simple relationship may
exist: toxicity occurring in cells that accumulate sphinganine (and sphingosine), and mitogenesis in cells that
accumulate sphingoid-1-phoshates [8]. One argument
against this interpretation is that addition of sphingosine
1-phosphate to cells in culture sometimes induces apoptosis, however, this may not be caused by sphingosine 1phosphate per se, but to the sphingosine (or its metabolite
ceramide) which is released upon removal of the phosphate group, as has been recently shown with mesangial
cells [28].
Because the elevation in sphinganine is so rapid and
large, it usually dominates over depletion of complex
sphingolipids in the effects of fumonisins on cells. Nonetheless, loss of glycosphingolipids from A431 epithelial
cells causes a significant increase in diffusible lipids (i.e.,
the mobile fraction) and this correlates with a loss of
epithelial cell morphology, a reduced rate of cell growth
and the inhibition of cell –substrate adhesion [29]. Exogenous administration of G M3 gangliosides to cells,
whereby restoring cellular glycosphingolipids, blocks elevations in diffusible lipids and reverses the effects of
glycosphingolipid depletion on cell morphology and substrate adhesion [29]. Depletion of complex sphingolipids
also alters the function of glycophosphatidylinositolanchor proteins, such as the folate receptor [30]. The
treatment of Caco-2 cells with FB1, at concentrations that
cause 40% depletion of complex sphingolipids, results in a
complete halt in the cellular uptake of folate by its
receptor; potentially compromising the cellular processes
dependent on this vitamin [30]. Most recently, post-transcriptional modifications, occurring at membrane micro-
3. Other consequences of inhibition of ceramide synthase
As sphinganine accumulates, there is increased synthesis
of sphinganine 1-phosphate [26] as well as formation of the
downstream degradation product(s) (for example, in fumonisin-treated J774 cells, as much as one-third of the ethanolamine in phosphatidylethanolamine was derived from
sphingoid bases) [27]. The elevation in sphinganine 1phosphate is likely to be a factor in cells where fumonisins
are growth stimulatory instead of toxic, and may play an
important role in the inhibition of apoptosis in studies where
fumonisin has that effect.
Fig. 3. Schematic representation of the inhibition of the acylation of
sphingosine by AP1, which is acylated to PAP1, a more potent inhibitor of
ceramide synthase.
K. Desai et al. / Biochimica et Biophysica Acta 1585 (2002) 188–192
191
domains which transform prions to their infectious form
have been linked to complex sphingolipids; the depletion
of sphingomyelin in cells, by treatment with FB1, has been
found to increase the formation of infectious prions by 3 –
4-fold [31].
4. Metabolism and cytotoxicity of fumonisin
aminopentols
The aminopentols are formed during the treatment of
corn with lye, a common practice in regions of Central
and South America; furthermore, the intestinal microflora
of primates has been reported to hydrolyze FB1 to AP1
[5,32]. As noted above, AP1 is a weak inhibitor of ceramide synthase, but in the presence of palmitoyl-CoA is
also acylated to form N-palmitoyl-AP1 (PAP1), which is a
more potent inhibitor of ceramide synthase [10] (Fig. 3).
Interestingly, PAP1, which is half as effective as FB1 in
inhibiting ceramide synthase in vitro, causes significantly
greater accumulation of sphinganine and toxicity in HT-29
cells [10]. In recent studies, we have found that other
fatty acids (e.g., nervonic acid) are even more rapidly
added to AP1, hence, all together, such acylation may
explain why AP1 causes liver and kidney lesions and
carcinogenicity in rodents that resembles those caused by
FB1 [33,34].
5. Studies with 1-deoxy-sphinganine analogs
Structural differences between the fumonisins and
sphinganine include the absence of a hydroxyl group at
the 1-position, the presence of an additional hydroxyl at
position 5, and a threo- versus erythro-stereochemistry at
positions 2 and 3. To investigate the contributions of these
features to the behavior of AP1, the analogous 1-deoxy 5hydroxy-sphinganine was prepared by highly stereoselective synthesis [10]. Comparison of this analog with other
homologs (i.e., 1-deoxysphinganines) and stereoisomers
revealed that all were acylated, but the D-erthyro-stereoisomers were the best substrates and inhibitors for ceramide
synthase [10].
Since the stereochemistry of the 3- and 5-position
hydroxyl groups on AP1 (and FB1) is the same as that of
L-threo, 1-deoxy-, 5-hydroxysphinganine, this analog was of
particular interest. This compound is significantly more
toxic for HT-29 cells than either of the naturally occurring
sphingoid bases (sphingosine and sphinganine) or a shortchain (C2-) ceramide (Fig. 4). In preliminary studies, we
have observed that this compound is also an inhibitor of
sphingosine kinase, which may contribute to its greater
toxicity (i.e., by both elevating sphingoid bases—which
are cytotoxic—and suppressing sphingoid base 1-phosphates, which are often mitogenic and inhibitors of apoptosis). Consistent with this interpretation, another sphingoid
Fig. 4. Relative cytotoxicities of N-acetyl-sphingosine (C2Cer), sphingosine
(So), sphinganine (Sa), 1-deoxy, 5-hydroxysphinganine (DHSa) and N,N,
dimethylsphingosine (DMSo) for HT-29 cells. The cells were incubated
with 10 AM of the shown compounds for 24 h as described in Ref. [10],
then viable cells were analyzed by the MTT assay.
base that inhibits sphingosine kinase (N,N-dimethylsphingosine) is also highly toxic (Fig. 4).
6. Perspectives for future research
Much remains to be determined about the signaling
pathways that mediate the cellular effects of fumonisins
and how they are interwoven with sphingolipid metabolism.
Particularly intriguing in this regard are the findings about
fumonisins and tumor necrosis factor (TNF) receptors and
signaling [15,35,36] since this is one of the most studied
pathways involving sphingolipids, and involves both mitogenesis and apoptosis, depending on the types of cells
studied and the context. Also somewhat surprising has been
the recent identification of UOG1 as a modulator of ceramide synthesis (either as an isoform of ceramide synthase or a
regulator of ceramide synthase) from stearoyl-CoA in a
FB1-independent manner [37].
It should be evident from the many fumonisin subspecies
that may be involved plus the various aspects of sphingolipid metabolism/signaling that are involved, that when these
are finally interwoven, it will give a very complex tapestry.
Acknowledgements
We thank the many collaborators who have assisted in
our research on fumonisins, especially Dr. Ron Riley, Dr.
Ken Voss and others at the U.S. Department of Agriculture
in Athens, GA. Work in our lab has been supported
primarily by NIH grant GM46368.
References
[1] W.F.O. Marasas, Environ. Health Perspect. 109 (Suppl. 2) (2001)
239 – 243.
[2] E.W. Sydenham, G.S. Shepard, P.G. Theil, W.F.O. Marasas, J. Agric.
Food Chem. 40 (1992) 1352 – 1354.
192
K. Desai et al. / Biochimica et Biophysica Acta 1585 (2002) 188–192
[3] A. Pittet, V. Parisod, M. Schellenberg, J. Agric. Food Chem. 40
(1992) 1352 – 1354.
[4] W.M. Haschek, L.A. Gumprecht, G. Smith, M.E. Tumbleson, P.D.
Constable, Environ. Health Perspect. 109 (Suppl. 2) (2001) 251 – 257.
[5] A.H. Merrill Jr., M.C. Sullards, E. Wang, K.A. Voss, R.T. Riley,
Environ. Health Perspect. 109 (Suppl. 2) (2001) 283 – 289.
[6] R.T. Riley, E. Enongene, K.A. Voss, W.P. Norred, F.I. Meredith, R.P.
Sharme, J. Spitsbergen, D.E. Williams, D.B. Carlson, A.H. Merrill Jr.,
Environ. Health Perspect. 109 (Suppl. 2) (2001) 301 – 308.
[7] R. Bose, M. Verheij, A. Haimovitz-Friedman, K. Scotto, Z. Fuks,
R. Kolesnick, Cell 82 (1995) 405 – 414.
[8] J.J. Schroeder, H.M. Crane, J. Xia, D.C. Liotta, A.H. Merrill Jr.,
J. Biol. Chem. 269 (1994) 3475 – 3481.
[9] A.H. Merrill Jr., G. van Echten, E. Wang, K. Sandhoff, J. Biol. Chem.
268 (1993) 27299 – 27306.
[10] H.U. Humpf, E.M. Schmelz, F.I. Meredith, H. Vesper, T.R. Vales,
D.S. Menaldino, D.C. Liotta, A.H. Merrill Jr., J. Biol. Chem. 273
(1998) 19060 – 19064.
[11] J.R. Ciacci-Zanella Jr., A.H. Merrill Jr., E. Wang, C. Jones, Food
Chem. Toxicol. 36 (1998) 791 – 804.
[12] Y. Zhang, M.B. Dickman, C. Jones, J. Biol. Chem. 274 (1999)
12367 – 12371.
[13] E.M. Schmelz, M.A. Dombrink-Kurtzman, P.C. Roberts, Y. Kozutsumi, T. Kawasaki, A.H. Merrill Jr., Toxicol. Appl. Pharmacol. 148
(1998) 252 – 260.
[14] E.R. Smith, P.L. Jones, J.M. Boss, A.H. Merrill Jr., J. Biol. Chem.
272 (1997) 5640 – 5646.
[15] C. Jones, J.R. Ciacci-Zanella, Y. Zhang, G. Henderson, M. Dickman,
Environ. Health Perspect. 109 (Suppl. 2) (2001) 315 – 320.
[16] S. Ramasamy, E. Wang, B. Hennig, A.H. Merrill Jr., Toxicol. Appl.
Pharmacol. 133 (1995) 343 – 348.
[17] E.V. Wattenberg, F.A. Badria, W.T. Shier, Biochem. Biophys. Res.
Commun. 227 (1996) 622 – 627.
[18] E. Pinelli, N. Poux, L. Garren, B. Pipy, M. Castegnaro, D.J. Miller, A.
Pfohl-Leszkowicz, Carcinogenesis 20 (1999) 1683 – 1688.
[19] E. Wang, P.F. Ross, T.M. Wilson, R.T. Riley, A.H. Merrill Jr., J. Nutr.
122 (1992) 1706 – 1716.
[20] R.T. Riley, N.H. An, J.L. Showker, H.S. Yoo, W.P. Norred, W.J.
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
Chamberlain, E. Wang, A.H. Merrill Jr., G. Motelin, V.R. Beasley,
W.M. Hascheck, Toxicol. Appl. Pharmacol. 118 (1993) 105 – 112.
W.C.A. Gelderblom, S. Abel, C.M. Smuts, J. Marnewick, W.F.O.
Marasas, Environ. Health Perspect. 109 (Suppl. 2) (2001) 291 – 300.
P.C. Howard, A. Warbritton, K.A. Voss, R.J. Lorentzen, J.D. Thurman,
R.M. Kovach, T.J. Bucci, Environ. Health Perspect. 109 (Suppl. 2)
(2001) 309 – 314.
W.C.A. Gelderblom, N.P.J. Kriek, W.F.O. Marasas, P.G. Thiel, Carcinogen 12 (1991) 1247 – 1251.
Natl. Toxicol. Program Tech. Rep. Ser. (496) (2001) 1 – 352.
W.P. Norred, R.D. Plattner, W.J. Chamberlain, Nat. Toxins 1 (1993)
246 – 341.
M.C. Sullards, A.H. Merrill Jr., Sci. Signal Transduct. Knowl. Environ. (January 30, 2001) (67) PL1. Available at: http://www.stke.
sciencemag.org.
E.R. Smith, A.H. Merrill Jr., J. Biol. Chem. 270 (1995) 18749 – 18758.
I. Gennero, J. Fauvel, M. Nie’to, C. Cariven, F. Gaits, F. BriandMe’sange, H. Chap, J.P. Salles, J. Biol. Chem. 277 (2002)
12724 – 12734.
S. Barbour, M. Edidin, B. Felding-Habermann, J. Taylor-Norton, N.S.
Radin, B.A. Fenderson, J. Cell. Physiol. 150 (1992) 610 – 619.
V.L. Stevens, J. Tang, J. Biol. Chem. 272 (1997) 18020 – 18025.
N. Naslavsky, H. Shmeeda, G. Friedlander, A. Yanai, A.H. Futerman,
Y. Barenholz, A. Taraboulos, J. Biol. Chem. 274 (1999) 20763 – 20771.
A.H. Merrill Jr., E.M. Schmelz, D.L. Dillehay, S. Spiegel, J.A. Shayman, J.J. Schroeder, R.T. Riley, K.A. Voss, E. Wang, Toxicol. Appl.
Pharmacol. 142 (1997) 208 – 225.
K.A. Voss, C.W. Bacon, F.I. Meredith, W.P. Norred, Food, Chem.
Toxicol. 34 (1996) 623 – 632.
S. Hendrich, K.A. Miller, T.M. Wilson, P.A. Murphy, J. Agric. Food
Chem. 41 (1993) 1649 – 1654.
R.P. Sharma, N. Bhandari, R.T. Riley, K.A. Voss, F.I. Meredith, Toxicology 143 (2000) 183 – 194.
R.P. Sharma, N. Bhandari, Q. He, R.T. Riley, K.A. Voss, Toxicology
159 (2001) 69 – 79.
K. Venkataraman, C. Riebeling, J. Bodennec, H. Riezman, J.C. Allegood, M.C. Sullards, A.H. Merrill Jr., A.H. Futerman, J. Biol.
Chem. 277 (2002) 35642 – 35649.