Mitochondrially targeted ceramide LCL-30 inhibits colorectal cancer in mice.

Abstract

The sphingolipid ceramide is intimately involved in the growth, differentiation, senescence, and death of normal and cancerous cells. Mitochondria are increasingly appreciated to play a key role in ceramide-induced cell death. Recent work showed the C16-pyridinium ceramide analogue LCL-30 to induce cell death in vitro by mitochondrial targeting. The aim of the current study was to translate these results to an in vivo model. We found that LCL-30 accumulated in mitochondria in the murine colorectal cancer cell line CT-26 and reduced cellular ATP content, leading to dose- and time-dependent cytotoxicity. Although the mitochondrial levels of sphingosine-1-phosphate (S1P) became elevated, transcription levels of ceramide-metabolising enzymes were not affected. In mice, LCL-30 was rapidly absorbed from the peritoneal cavity and cleared from the circulation within 24 h, but local peritoneal toxicity was dose-limiting. In a model of subcutaneous tumour inoculation, LCL-30 significantly reduced the proliferative activity and the growth rate of established tumours. Sphingolipid profiles in tumour tissue also showed increased levels of S1P. In summary, we present the first in vivo application of a long-chain pyridinium ceramide for the treatment of experimental metastatic colorectal cancer, together with its pharmacokinetic parameters. LCL-30 was an efficacious and safe agent. Future studies should identify an improved application route and effective partners for combination treatment.

Full text

Page 1

Mitochondrially targeted ceramide LCL-30 inhibits colorectal
cancer in mice
F Dahm1, A Bielawska2, A Nocito1, P Georgiev1, ZM Szulc2, J Bielawski2, W Jochum3, D Dindo1, YA Hannun2
and P-A Clavien*,1
1Swiss HPB (Hepato-Pancreato-Biliary) Centre, Department of Visceral and Transplantation Surgery, University Hospital Zurich, Ra ¨mistrasse 100, Zurich
CH-8091, Switzerland;2Department of Biochemistry and Molecular Biology, Medical University of South Carolina, 173 Ashley Avenue, Charleston, SC
29425, USA;3Department of Pathology, University Hospital Zurich, Ra ¨mistrasse 100, Zurich CH-8091, Switzerland
The sphingolipid ceramide is intimately involved in the growth, differentiation, senescence, and death of normal and cancerous cells.
Mitochondria are increasingly appreciated to play a key role in ceramide-induced cell death. Recent work showed the C16-pyridinium
ceramide analogue LCL-30 to induce cell death in vitro by mitochondrial targeting. The aim of the current study was to translate these
results to an in vivo model. We found that LCL-30 accumulated in mitochondria in the murine colorectal cancer cell line CT-26 and
reduced cellular ATP content, leading to dose- and time-dependent cytotoxicity. Although the mitochondrial levels of sphingosine-1-
phosphate (S1P) became elevated, transcription levels of ceramide-metabolising enzymes were not affected. In mice, LCL-30 was
rapidly absorbed from the peritoneal cavity and cleared from the circulation within 24h, but local peritoneal toxicity was dose-
limiting. In a model of subcutaneous tumour inoculation, LCL-30 significantly reduced the proliferative activity and the growth rate of
established tumours. Sphingolipid profiles in tumour tissue also showed increased levels of S1P. In summary, we present the first in
vivo application of a long-chain pyridinium ceramide for the treatment of experimental metastatic colorectal cancer, together with its
pharmacokinetic parameters. LCL-30 was an efficacious and safe agent. Future studies should identify an improved application route
and effective partners for combination treatment.
British Journal of Cancer (2008) 98, 98–105. doi:10.1038/sj.bjc.6604099
Published online 20 November 2007
& 2008 Cancer Research UK
www.bjcancer.com
Keywords: colorectal neoplasms; cell death; ceramides; ceramidoids; pharmacokinetics; drug therapy
????????????????????????????????????????????????????
The last two decades have seen an explosive growth in the
understanding of sphingolipid biology. Initially considered inert
structural constituents of cell membranes or precursors thereof,
sphingolipids have emerged as key messenger and bioactive
molecules in a wide range of biological processes (Futerman and
Hannun, 2004). The sphingolipid ceramide can be formed by the
breakdown of sphingomyelin or through de novo synthesis. It is
intimately involved in growth, differentiation, senescence, and
death of normal and cancerous cells. Several inductors of cell death,
for example, TNFa (Obeid et al, 1993), anthracyclines (Bose et al,
1995), or irradiation (El-Assaad et al, 2003) involve ceramide
signalling. Administration of exogenous ceramide also causes cell
death in various cancer cell lines (Oh et al, 1998; Bras et al, 2000). It
is noteworthy that, many cancer cells have a specific ‘sphingolipid–
phenotype’, including lower endogenous ceramide levels (Itoh et al,
2003) and a higher sensitivity to the effects of exogenous ceramide
(Selzner et al, 2001). This offers the opportunity to selectively target
cancer cells with ceramide compounds.
Mitochondria are increasingly appreciated to play a key role in
ceramide-induced cell death. Ceramide treatment of isolated
mitochondria leads to the activation of a mitochondrial protein
phosphatase (PP2A), which dephosphorylates the antiapoptotic
Bcl-2 (Ruvolo et al, 1999) and causes cytochrome c release
(Ghafourifar et al, 1999). Furthermore, ceramide has been shown
to inhibit mitochondrial complex I (Di Paola et al, 2000) and to
induce the formation of reactive oxygen species in mitochondria
(Garcia-Ruiz et al, 1997). Targeted delivery of sphingomyelinase to
mitochondria, but not to other subcellular compartments, results
in bax translocation and the activation of the mitochondrial
pathway of apoptosis (Birbes et al, 2001).
The central role of mitochondria in ceramide-induced cell death
makes them an alluring target for the specific delivery of ceramide
compounds. Naturally occurring ceramides contain a relatively
long N-linked fatty acyl chain (14–24 carbon atoms), rendering
them practically insoluble in water. Ceramides modified with a
o-pyridinium moiety contain a positive charge delocalised over
the p-electron system (Szulc et al, 2006). These ceramide analogues
exhibit a much higher water solubility and preferentially
accumulate within the mitochondrial matrix driven by the
electrochemical gradient (Novgorodov et al, 2005; Dindo et al,
2006; Senkal et al, 2006). The approach of attacking mitochondria
is further supported by the fact that cancer cells’ mitochondrial
membrane potential tends to be more polarised than that of
normal cells (Chen, 1988).
Revised 22 October 2007; accepted 29 October 2007; published online
20 November 2007
*Correspondence: Dr P-A Clavien; E-mail: clavien@chir.unizh.ch
British Journal of Cancer (2008) 98, 98–105
& 2008 Cancer Research UKAll rights reserved 0007– 0920/08$30.00
www.bjcancer.com
Translational Therapeutics

Page 2

Despite many pleas for ceramide-based treatment regimens
against cancer (Radin, 2003), progression from cell-culture to
in vivo applications has been slow, and no clinical trials have been
reported to date. Previous work in animal models has shown the
ceramidase inhibitor B13 to profoundly suppress the growth of
colorectal liver metastases (Selzner et al, 2001) and to reduce the
progression of metastatic prostate cancer (Samsel et al, 2004).
A short-chain pyridinium ceramide (C6-analogue) was recently
shown to inhibit tumour growth in a model of metastatic head and
neck squamous cell carcinoma (Senkal et al, 2006). These studies
illustrate the large potential of a targeted approach to cancer
therapy by interference with ceramide signalling.
Our recent work showed LCL-30, a C16-pyridinium ceramide
analogue, to induce cell death in vitro by mitochondrial targeting
(Dindo et al, 2006). The aim of the current study was to translate
these results to an in vivo model. This study represents the first
application of a long-chain pyridinium ceramide in vivo as well as
a determination of its tolerability and efficacy in a widely used
animal model of metastatic colorectal cancer.
MATERIAL AND METHODS
Cell culture and biological reagents
D-erythro-2-N-(160-(1’’-Pyridinium)-hexadecanoyl)-sphingosine
bromide (LCL30) was prepared in the Lipidomics Core of the
Medical University of South Carolina (Szulc et al, 2006). CT-26
murine colon carcinoma cells (ATCC, Manassas, VA, USA) were
cultured in RPMI Medium (Invitrogen, Basel, Switzerland)
supplemented with 10% fetal bovine serum (PAA Laboratories,
Austria), 100Uml?1of penicillin, and 100mgml?1of streptomycin
(Invitrogen). The cells were maintained at 371C in a 5% CO2
atmosphere. Actinomycin D and doxorubicin hydrochloride were
from Sigma-Aldrich (Buchs, Switzerland). Recombinant human
TNFa was purchased from R&D Systems Inc. (Minneapolis, MN,
USA). Caspase-3 and -8 substrates (Ac-DEVD-AFC and Ac-LETD-
AFC, respectively) as well as caspase-3- and pan-caspase-inhibitor
(Z-DEVD-CHO and Z-VAD-fmk, respectively) were from Alexis
Biochemicals (Lausen, Switzerland).
Cell viability assay
Cells were seeded into 12-well plates at a density of approximately
50%, corresponding to 5?105cells per well and allowed to adhere
overnight, before the medium was changed to the specified
conditions. The MTT assay was performed as described previously
(Dindo et al, 2006). In parallel, cells were detached using 1%
trypsin (Invitrogen) and centrifuged at 800g. Cell pellets were
resuspended in PBS with trypan blue (Sigma-Aldrich) and both
stained and unstained cells were counted.
Mitochondrial isolation and determination of cytochrome c
All procedures were performed on ice. Cells were scraped and
washed twice in PBS before being resuspended in five volumes of
isolation buffer (250mM sucrose, 20mM HEPES, pH 7.5, 10mM
KCl, 1.5mM MgCl2, 1mM EDTA, 1mM EGTA, 1mM DTT, 0.2mM
PMSF). Cells were broken by repeated aspiration through a pipette.
Centrifugation for 10min at 700g yielded unbroken cells as well as
nuclei. Supernatants were centrifuged for 15min at 10000g to
pellet a crude mitochondrial fraction. Mitochondrial enrichment
was confirmed by Western blotting for cytochrome c oxidase
complex IV (Abcam, 20E8, Cambridge, UK). Mitochondrial
cytochrome c release into the cytosol was assessed quantitatively
with the Quantikine enzyme-linked immunosorbent assay kit
(R&D Systems, Abingdon, UK).
ATP measurement
Cellular ATP content was measured in cellular lysates with the
Enliten Luciferase/Luciferinreagent
Germany) according to the manufacturer’s instructions and
normalised to protein content.
(Promega,Mannheim,
Caspase-3/-8 activities
Cells were scraped and lysed (10mM Tris–HCl, pH 7.4, 2mM
EDTA, 0.1% NP-40) for 10min at 41C. After centrifugation for
10min at 10000g, the lysate corresponding to 25mg of protein was
incubated for 30min at room temperature with or without 1mM
caspase-3 inhibitor Z-DEVD-FMK. Then, caspase-3 substrate
Ac-DEVD-AFC (10mM) or caspase-8 substrate Ac-LETD-AFC
(10mM) and dithiothreitol (10mM final concentration) were added,
and enzyme activity was monitored by measuring fluorescence at
390ex/538emnm (Biolise software and Fluostar microtiter plate
reader, Crailsheim, Germany). Caspase activity was then calculated
by determining the relative fluorescence units generated under
steady state kinetics from which values of caspase-independent
protease activity in the presence of the corresponding inhibitor
was subtracted. Actinomycin D and TNFa were used as positive
controls.
Animal experiments
All animal experiments were in accordance with Swiss federal
animal regulations and approved by the cantonal veterinary office
of Zurich. Specific pathogen-free Balb/c mice 10–12 weeks of age
(Harlan, Netherlands), syngeneic with the CT-26 colon carcinoma
cell line, were kept on a 12h day/night cycle with free access to
food and water. Animal health, weight, and food intake were
monitored daily, and animals were killed according to predefined
criteria (signs of pain, reduction of food intake 450%, weight loss
420%). For subcutaneous tumour cell inoculations, CT-26 cells,
cultured in the exponential growth phase, were treated with trypsin
and washed in PBS. Cells were then suspended in serum-free
medium, and 200ml (corresponding to 5?105cells) were injected
subcutaneously. Therapy commenced after a subcutaneous tumour
became detectable, which occurred after 4–5 days. Tumour growth
was monitored daily with a sliding calliper, and tumour volume
was calculated according to the ellipsoid formula 4/3 p ? l/2 ? w/
2 ? w/2, where l is the length and w is the width. Vehicles for
intraperitoneal injection of LCL-30 and doxorubicin were 30%
Cremophore (Sigma-Aldrich) and NaCl respectively. All animals
received an equal number of intraperitoneal injections with active
compound or appropriate vehicle controls (daily injections of
LCL-30 and weekly injections of doxorubicin). Blood cell counts
were determined with a Coulter AcT Diff counter (Beckman
Coulter, Nyon, Switzerland). Plasma aspartate aminotransferase
(AST), alkaline phosphatase, and creatinine were determined with
the serum multiple analyzer (Ektachem DTSCII, Johnson &
Johnson Inc., Rochester, USA).
Histology and immunohistochemistry
Formalin-fixed tissue was paraffin-embedded, sectioned, and
stained with H&E using standard techniques. For immunohisto-
chemistry, tissue sections were incubated with anti-Ki-67 antibody
(NeoMarkers). Pretreatment of sections, antibody incubation, and
detection of primary antibody (Ventana DAB iView Kit) were
performed on a Nexes immunohistochemistry staining system
(Ventana Medical Systems, Tucson, AZ, USA). For CD31 staining,
detection of primary antibody was performed with a Histofine
staining kit (Nichirei Corporation, Tokyo, Japan) and diamino-
benzidine (DAB) as a chromogen. All immunostains were counter-
stained with hematoxylin. TUNEL staining was performed with the
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
99
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 3

in situ cell death detection kit (Roche Applied Science, Rotkreuz,
Switzerland) according to the manufacturer’s instructions. Ki-67
staining was quantified on 10 images with the analySIS^D imaging
software using a semi-automatic thresholding algorithm (Olympus,
Volketswil, Switzerland). Microvascular density was counted on 10
high-power fields of CD31-immunostains.
Enzyme-linked immunosorbent assay (ELISA) for TNFa
TNF-a levels in plasma were determined by ELISA (Quantikine
mouse TNF-a, R&D systems, Minneapolis, USA) following the
manufacturer’s instructions. The lower detection limit of this assay
is 5.1pgml?1.
Quantitative real-time polymerase chain reaction
Total RNA was extracted from cells or tissue using TRIzol reagent
(Invitrogen) following the manufacturer’s instructions. Five micro-
grams of RNA were reverse transcribed to cDNA using the
ThermoScript RT-PCR System (Invitrogen) kit. Quantitative real-
time PCR amplification and data analysis were performed using an
ABI Prism 7000 Sequence Detector System (PE Applied Biosystems,
Rotkreuz, Switzerland). TaqMan gene expression assays (PE
Applied Biosystems) were used to quantify mRNA expression of
the respective genes. Results were quantified as fold induction in
comparison to baseline after normalisation to 18S RNA (TaqMan
ribosomal RNA control reagents, PE Applied Biosystems).
Detection of LCL-30 and of endogenous sphingolipids
Liquid chromatography and mass spectrometry (LC-MS) analysis
was performed in the Lipidomics Core of the Medical University of
South Carolina on a Thermo Finnigan TSQ 7000 triple quadrupole
mass spectrometer operating in a Multiple Reaction Monitoring
(MRM) positive ionisation mode. Briefly, cell lysates or tumour
homogenate fractions were fortified with internal standards for
quantification, and lipids were extracted and analysed as
previously described (Bielawski et al, 2006).
Statistical analysis
Data represent mean±s.d. of n independent experiments. Mann–
Whitney U-test or one-way ANOVA with Student–Newman–
Keuls post hoc testing was used as appropriate to compare groups,
using SPSS 12.0 (SPSS Inc., Chicago, USA). A P-value below 0.05
was considered to indicate statistical significance.
RESULTS
LCL-30 elicits cytotoxicity in vitro
We previously tested the cytotoxicity of LCL-30 on a range of
human and murine cancer cell lines (Dindo et al, 2006). For the
present study, we focused on the colon cancer cell line CT-26,
which can be used as a syngeneic in vivo model of colorectal
cancer in Balb/c mice. LCL-30 treatment of CT-26 cells was able to
effectively induce cell death in vitro in a dose-dependent manner
(Figure 1). A 50% inhibition of cell viability (IC50) was achieved at
10.6mM. Time course experiments with different concentrations
showed a time-dependent reduction of cell viability with a steady
slope (not shown). These results were confirmed by trypan blue
exclusion (not shown).
LCL-30 targets mitochondria
Ceramide has been detected in mitochondria (Dindo et al, 2006) as
well as some ceramide-metabolising enzymes such as ceramide
synthase and ceramidase. LCL-30 represents a cationic lipid
designed to be enriched in the positively charged mitochondria.
To investigate whether this mitochondrial accumulation occurs in
CT-26, cells were treated with the IC50 concentration of LCL-30
(10mM) for up to 8h. Whole cells and mitochondrially enriched
fractions were isolated at different time points and subjected to
mass spectrometry, allowing a detailed detection of endogenous
sphingolipids as well as LCL-30. As illustrated in Figure 2A, LCL-30
was progressively taken up into cells with levels achieving
approximately 0.95pmolemg?1protein after 4h of incubation.
Notably, LCL-30 was significantly enriched in the mitochondrial
fraction such that its levels in isolated mitochondria reached about
6pmolemg?1protein by 4h. Cellular and mitochondrial uptake
appeared to level off after 4h. These results demonstrate significant
enrichment of LCL-30 in the mitochondria of CT-26 cells.
LCL-30 induces an endogenous sphingolipid response
Next, the effects of LCL-30 on endogenous sphingolipids were
examined. Cellular levels of total ceramide decreased gradually in
response to LCL-30 treatment (Figure 2B), whereas mitochondrial
levels peaked after 2h before returning to normal. Levels of
sphingosine and dihydrosphingosine reacted in an analogous
fashion (not shown). Interestingly, cellular and mitochondrial levels
of sphingosine-1-phosphate (S1P) rose continuously from almost
undetectable levels (Figure 2C), with a more marked rise in the
mitochondrially enriched fraction. These experiments demonstrate
that LCL-30 enters CT-26 cells, is enriched in the mitochondrial
fraction, and leads to a transient rise of endogenous mitochondrial
ceramides as well as a marked rise of mitochondrial S1P.
LCL-30 does not regulate ceramide-metabolising enzymes
at the transcriptional level
The changes in cellular sphingolipid levels in response to
treatment, especially the rapid rise of S1P levels, prompted us to
analyse the expression levels of key enzymes in the metabolic
conversion of ceramide to sphingosine and further to sphingosine-
1-phosphate. Quantitative real-time PCR was performed for acid
ceramidase (Asah1), neutral ceramidase (Asah2), alkaline cerami-
dase (Asah3), sphingosine kinase 1 (SphK1), and sphingosine
kinase 2 (SphK2). Neutral ceramidase was never detectable. None
of the expressed enzymes showed any changes in their transcript
levels during 8h of treatment with LCL-30 (data not shown).
LCL-30 decreases cellular ATP levels
As the above experiments showed LCL-30 to be enriched in
mitochondria, we assessed whether mitochondrial function was
0510 1520
0
20
40
60
80
100
Concentration of LCL-30 (?M)
Viability (% of control)
Figure 1
with increasing concentrations of LCL-30, and viability was assessed by the
MTT assay (IC50¼10.6mM).
Cytotoxicity of LCL-30. CT-26 cells were incubated for 24h
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
100
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 4

affected. Cellular ATP levels showed a continuous time- and dose-
dependent decrease, with different kinetics than the mitochondrial
uncoupler FCCP (Figure 3A). Surprisingly, we could not detect
cytosolic cytochrome c release by ELISA after 8 or 24h of
incubation with LCL-30 at a concentration of 10mM (Figure 3B). In
line with this observation, we could not detect any caspase 3
(Figure 3C) or caspase 8 activity (data not shown). Actinomycin D
and TNFa (ActD/TNF) were used as positive controls for these
assays. Taken together, LCL-30 accumulates in mitochondria of
CT-26 cells and decreases ATP production, ultimately causing cell
012345678
0
1
2
3
4
5
6
7
8
Mitochondria
Whole cells
Time of exposure (h)
LCL-30 (pmole per ?g protein)
012345678
0
1
2
3
Whole cells
Mitochondria
Time of exposure (h)
Ceramide (pmole per ?g protein)
012345678
0.00
0.01
0.02
0.03
0.04
0.05
0.06
Whole cells
Mitochondria
Time of exposure (h)
S1P (pmole per ?g protein)
Figure 2
for the indicated time to 10mM of LCL-30 and then separately analyzed by mass spectrometry for whole cells (squares) and mitochondrially enriched
fractions (circles). Results are mean±s.d. of n¼3 and normalised to protein content.
Cellular uptake of LCL-30 (A) and the resulting changes in endogenous ceramide (B) and sphingosine-1-phosphate (C) CT-26 were exposed
0123456
0
20
40
60
80
100
120
LCL-30 (10 ?M)
LCL-30 (20 ?M)
FCCP (20 ?M)
Time of exposure (h)
ATP (% of control)
0 h4 h8 h 12 h24 h ActD/TNF
0.05
0.10
0.15
0.20
Cytochrome c ratio
(cytosol/mitochondria)
ControlLCL-30ActD/TNF
0
50
100
150
200
250
Caspase-3 activity
(RFU ?g–1 min–1)
Figure 3
mean±s.d. of n¼4 and normalised to protein content. (B) Ratio of cytosolic to mitochondrial cytochrome c at different durations of incubation with 10mM
LCL-30, as assessed by ELISA. (C) Caspase-3-activity after 6h of exposure to 10mM LCL-30.
(A) Cellular ATP content during exposure to 20mM FCCP (circles), 10mM LCL-30 (squares) and 20mM LCL-30 (triangles). Results are
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
101
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 5

death, without detectable cytochrome c release or caspase
activation.
Dose-finding and toxicity in mice
A major aim of this study was to investigate the effects of the long
chain pyridinium ceramide LCL-30 in an animal model. To define
the toxicity and the tolerable dose, escalating doses of LCL-30
dissolved in 30% cremophore were administered to mice
intraperitoneally. A single dose of 100mgkg?1was lethal, while
50mgkg?1was lethal after several doses. Toxicity manifested itself
by an eosinophilic inflammatory reaction of the peritoneum,
followed by fibrinous exudation and fibrous organisation, result-
ing in severe intraperitoneal adhesions and ileus. No other toxic
effects were identified either by histological analysis of internal
organs, including brain and bone marrow, or by blood counts and
plasma tests (aspartate aminotransferase, alkaline phosphatase,
and creatinine). Daily injections of 20mgkg?1for 1 week were
tolerated. This dose was then combined with different doses of
doxorubicin. The addition of 6mgkg?1doxorubicin to 20mgkg?1
LCL-30 was also established as safe. Average weight loss in animals
undergoing treatment was 6.8% (control), 9.1% (doxorubicin),
12.7% (LCL-30), and 13.9% (LCL-30þdoxorubicin). None of these
animals had to be killed according to the pre-defined criteria.
Thus, the in vivo application of LCL-30 caused localised
inflammation and did not result in systemic toxicity.
Pharmacokinetics of LCL-30
After defining the tolerable dose of LCL-30 in vivo, we sought to
identify the pharmacokinetic behaviour of this compound. The
levels of LCL-30 as well as endogenous sphingolipids were
determined by mass spectrometry after a single intraperitoneal
dose of 20mgkg?1of LCL-30 (Figure 4). LCL-30 was rapidly
absorbed from the peritoneal cavity, leading to a peak blood
concentration after 2h. Elimination from blood was almost
complete after 24h. Concentrations were determined separately
for plasma and the cellular components of blood: at its peak level,
LCL-30 partitioned equally into the aqueous as well as the cellular
phase of blood, with slower elimination from the cellular
compartment.
Therapeutic efficacy of LCL-30 in subcutaneous tumour
grafts
The next step was the assessment of the therapeutic efficacy of
LCL-30 in the treatment of established subcutaneous tumour
grafts. Pharmacokinetic data led us to choose a dosing regimen of
once per day for LCL-30. After the establishment of solid tumours,
animals were randomised to one of four treatment regimens and
received an equal number of injections over the course of 1 week.
At the beginning of treatment, there were no differences between
the animal groups. The growth curves under treatment are shown
in Figure 5A; whereas LCL-30 significantly reduced tumour
growth, doxorubicin did not cause a significant growth reduction.
0510152025
0
1
2
3
4
Hours after i.p. injection
LCL-30 in blood (?M)
Figure 4
injection of 20mgkg?1. Results are mean±s.d. of n¼4 and normalised to
protein content.
Concentration of LCL-30 in blood after a single intraperitoneal
Control DOX LCL-30 LCL-30 + DOX
0
20
40
60
80
100
*
*
*
Ki67-positive nuclei (%)
123456789
0
200
400
600
800
1000
**
Days of treatment
Tumour volume (mm3)
Tumour volume (mm3)
ControlDOXLCL-30 LCL-30 + DOX
0
100
200
300
400
500
600
**
**
*
Figure 5
(triangles), or LCL-30þdoxorubicin (circles), *P¼0.027 vs control (ANOVA); (B) Volume of explanted tumours after the end of treatment, *Po0.05,
**Po0.01 vs control (ANOVA with Student–Newman-Keuls post hoc test). (C) Proliferative activity of treated tumours assessed by Ki67
immunohistochemistry, *Po0.001 vs control (ANOVA with Student–Newman-Keuls post hoc test). Results are mean±s.d. of n¼8.
(A) Progression of subcutaneous tumours in animals receiving daily injections with vehicle (open squares), doxorubicin (closed squares), LCL-30
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
102
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 6

The combination of LCL-30 and doxorubicin did not add to the
efficacy of LCL-30 alone. Figure 5B shows the volumes of explanted
tumours measured ex-vivo, where doxorubicin-treated tumours
were also significantly smaller than controls. Again, LCL-30 was
more efficacious than doxorubicin, and the combined treatment
did not produce any additive effects. It should be noted that
the absolute values of tumour volumes in this protocol are lower
due to the systematic overestimation of tumours measured
through the skin of live animals. Importantly, the anti-tumour
effect could not be explained by an unspecific systemic inflamma-
tion, as plasma levels of TNFa remained undetectable in all
animals, while positive controls (LPS injection) yielded values of
1098±261pgml?1(not shown). Therefore, LCL-30 demonstrates
significant efficacy against colon cancer in this in vivo model.
Histological assessment of subcutaneous tumours
Histological assessment of H&E-stained tumour sections revealed
no differences between the groups with respect to tumour
morphology or necrotic area. To analyse whether apoptotic cell
death plays an important role in tumours, we examined TUNEL
stains. There was a low number of single-cell apoptosis in all
groups with no differences between groups (not shown). However,
the mitotic count was reduced from a baseline of six per high
power field (HPF) to five in the doxorubicin group and to four in
the LCL-30 and LCL-30þdoxorubicin groups. This led us to use
Ki67-immunostaining to further analyse the proliferating tumour
cell fraction. The results also showed a reduced proliferative
activity of treated tumours (Figure 5C). It is noteworthy that
microvascular density (as assessed by CD31-immunostaining) was
not different between groups (not shown).
Sphingolipid profiles of subcutaneous tumours
After completion of treatment, tumour samples were also subjected
to mass spectrometry analysis of sphingolipid content. Impor-
tantly, 24h after the last injection of a 1-week treatment course,
LCL-30 was detected at 0.22 (±0.09) and 0.23 (±0.08) pmolemg?1
protein in the LCL-30 and LCL-30þdoxorubicin-treated groups,
respectively. The content of endogenous ceramides decreased in
the LCL-30-treated tumours, an effect that was less pronounced
after doxorubicin co-treatment (Figure 6A). Similar to the in vitro
effects, LCL-30 caused an increase in S1P levels (Figure 6B), while
sphingosine levels were lowered by LCL-30 treatment (not shown).
Thus, LCL-30 appears to concentrate on tumours even when it has
been cleared from the blood with persistent effects on tumour
sphingolipids that recapitulate its effects in tissue culture.
DISCUSSION
This study is the in vivo continuation of our previous experiments
with LCL-30, the cationic water-soluble analogue of C16-ceramide,
in which we could demonstrate that LCL-30 accumulates in the
mitochondria of SW403 colorectal cancer cells and induces
mitochondrial swelling, cytochrome c release, caspase activation,
and eventually cell death (Dindo et al, 2006). Here, we expand our
analysis to the murine colon carcinoma cell line CT-26 cells and
use CT-26 cells in syngeneic Balb/c mice as an in vivo model of
colorectal cancer.
LCL-30 was cytotoxic for CT-26 cells in a dose- and time-
dependent fashion, in analogy to other cell lines previously tested
(Dindo et al, 2006). Cellular fractionation and mass spectrometric
analyses showed LCL-30 to be enriched in the mitochondrial
fraction, in line with published data on cationic o-pyridinium
analogues of C6and C16ceramide (Novgorodov et al, 2005; Dindo
et al, 2006; Senkal et al, 2006). Incubation with LCL-30 led to a dose-
and time-dependent decrease of cellular ATP-levels, pointing to a
breakdown of mitochondrial respiration, as already described for
the o-pyridinium C6analogue LCL-29 (Novgorodov et al, 2005). Yet
to our surprise, hallmarks of mitochondrially mediated apoptotic
cell death, such as cytochrome c release or caspase activation, could
not be detected. The mechanism of LCL-30-mediated cell death in
CT-26 remains unclear, although ceramide has been implicated as
an endogenous mediator of caspase-independent programmed
cell death (Thon et al, 2005). Delineating the differences between
cell lines that show caspase activation (SW403) and those without
(CT-26) might help define the different mechanisms involved.
Exposure to LCL-30 led to a transient depression of whole-cell
ceramide levels, whereas mitochondrial ceramide levels showed a
transient rise. This ceramide response is somewhat different from
SW403, which show a mitochondrial decrease of ceramide levels
(Dindo et al, 2006). Importantly, the rapid and pronounced rise of
mitochondrial S1P levels is comparable between both cell lines,
raising the possibility of the presence of a mitochondrial
sphingosine kinase (SphK). Additional experiments with isomers
of LCL-30 have revealed S1P to be derived from endogenous
sources and not from the breakdown of LCL-30 (Bielawska A,
unpublished). S1P has been primarily regarded as a counter player
of ceramide activity, although the intracellular compartmentalisa-
tion and the biological context (Ikeda et al, 2003) are important for
its biological effects. Future experiments should take intracellular
distribution of sphingosine kinase proteins into account. While the
activation of SphK1 leads to reduced apoptosis and improved
proliferation, activation of SphK2 has been associated with
enhanced cell death (Liu et al, 2003), which has been attributed
to differential localisation in ER vs cytosol (Wattenberg et al,
2006). Neither SphK1 nor SphK2 has been detected in mitochon-
dria. Nevertheless, there is evidence for additional sphingosine
kinase activity (Fukuda et al, 2003), which might be responsible for
ControlDOX LCL-30LCL-30+DOX
0
250
500
750
*
Ceramide (pmole per ?g protein)
Control DOX LCL-30LCL-30+DOX
0
1
2
*
S1P (pmole per ?g protein)
Figure 6
treatment with the indicated regimen, *P¼0.098 vs control. (B) Levels of
sphingosine-1-phosphate in tumours after 1 week of treatment with the
indicated regimen, *P¼0.045. Results are mean±s.d. of n¼4 and
normalised to protein content.
(A) Levels of endogenous ceramide in tumours after 1 week of
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
103
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 7

the rise in mitochondrial levels of S1P in response to LCL-30.
Alternatively, enhanced levels of S1P might be explained by an
inhibition of S1P-degrading enzymes S1P Lyase or S1P phospha-
tase by LCL-30 (Oskouian et al, 2006). At present, the exact role of
S1P produced in response to exogenous treatment with a ceramide
analogue remains elusive: it could be an antiapoptotic escape
mechanism, a cytotoxic signal, or an epiphenomenon.
Another focus of this study was to assess the safety and efficacy
of LCL-30 in an in vivo mouse tumour model. The maximum
tolerable dose could be established in dose-escalation studies.
Interestingly, dose-limiting toxicity manifested itself as a local
peritoneal reaction. The lack of organ-specific toxic effects is
encouraging as it carries two important implications. First, it hints
at a certain degree of tumour-selectivity of LCL-30, and second,
the locally toxic effects of LCL-30 might be circumvented by
alternative modes of application.
After a single intraperitoneal injection, LCL-30 reached a peak
concentration in blood within 2h and was cleared within 24h. Peak
concentrations were lower than LC50 in vitro, although a higher
peak between the first two pharmacokinetic sampling points
(30min and 2h) cannot be excluded. Clearance was somewhat
slower than for C6pyridinium, which was already cleared from
the circulation after 4h by renal excretion (Senkal et al, 2006),
suggesting that ceramides with longer acyl chains might be cleared
from the circulation more slowly. Such pharmacokinetic behaviour
might be beneficial for therapeutic purposes.
Treatment of established subcutaneous tumours over the course
of 1 week showed LCL-30 to be an efficacious compound in vivo.
Cytotoxic effects on tumours in vivo were less than expected from
in vitro experiments, possibly due to insufficient peak concentra-
tions being reached in vivo. Inhibition of tumour proliferation
might have been caused by an unspecific inflammatory response to
peritoneal injection of a peritoneal irritant. As TNFa could not be
detected in the plasma of any animal, this is highly unlikely.
Relative ceramide levels were much higher in solid tumours than
in cell culture, which might be caused by a different sphingolipid
composition of tumour cells growing in vivo. Solid tumours
also contain additional cell types, such as stromal or infiltrating
blood-derived cells, with a high ceramide content (Dahm et al,
2006). On the basis of in vitro data showing synergistic cytotoxicity
of doxorubicin with LCL-30 (Dindo et al, 2006), doxorubicin was
also tested alone and in combination treatment. In contrast to the
in vitro observation, doxorubicin conveyed no additive effect
compared to LCL-30 alone. This might be related to the dosing
schedule where LCL-30 was administered daily and doxorubicin,
once per week. Weekly administration of doxorubicin was based
on established dosing regimens (Yoneda et al, 1999; Dubois et al,
2002).
In summary, we present the first in vivo application of a long-
chain cationic ceramide for the treatment of experimental
metastatic colorectal cancer, together with its pharmacokinetic
parameters. Although cytotoxic for CT-26, the mechanism of cell
death was different from the previously studied SW403, and
doxorubicin did not convey additive effects in vivo. Nevertheless,
LCL-30 was an efficacious and safe agent. Future studies should
further elaborate on the mechanism of cell death and aim to
identify an alternative application route as well as more effective
partners for combination treatment.
ACKNOWLEDGEMENTS
We thank Stefan Heinrich, Gerd Kullak, Daniel Fetz, Peter Gehrig,
Riem Ha (University and University Hospital Zurich), Ju ¨rgen
Schiller, Jan Hengstler (University Leipzig), Besim Ogretmen,
Tarek Taha, and Sergei Novgorodov (Medical University South
Carolina) for helpful discussions, and Valentin Rousson (Department
of Biostatistics, Institute for Social and Preventive Medicine,
University of Zurich) for statistical advice. Furthermore, we are
grateful to Udo Ungethu ¨m, Marion Bawohl, Claas Bo ¨rger
(University Hospital Zurich), and Barbara Rembiesa (Medical
University of South Carolina) for technical assistance. This work
was supported by Sassella Foundation, Zurich, Switzerland (to FD)
and the National Cancer Institute, Bethesda, MD (Grant IPO1-
CA097132 to AB and YAH)
REFERENCES
Bielawski J, Szulc ZM, Hannun YA, Bielawska A (2006) Simultaneous
quantitative analysis of bioactive sphingolipids by high-performance
liquid chromatography-tandem mass spectrometry. Methods 39: 82–91
Birbes H, El Bawab S, Hannun YA, Obeid LM (2001) Selective hydrolysis of
a mitochondrial pool of sphingomyelin induces apoptosis. FASEB J 15:
2669–2679
Bose R, Verheij M, Haimovitz-Friedman A, Scotto K, Fuks Z, Kolesnick R
(1995) Ceramide synthase mediates daunorubicin-induced apoptosis: an
alternative mechanism for generating death signals. Cell 82: 405–414
Bras A, Albar JP, Leonardo E, de Buitrago GG, Martinez AC (2000)
Ceramide-induced cell death is independent of the Fas/Fas ligand
pathway and is prevented by Nur77 overexpression in A20 B cells. Cell
Death Differ 7: 262–271
Chen LB (1988) Mitochondrial membrane potential in living cells. Annu
Rev Cell Biol 4: 155–181
Dahm F, Nocito A, Bielawska A, Lang KS, Georgiev P, Asmis LM, Bielawski J,
Madon J, Hannun YA, Clavien PA (2006) Distribution and dynamic
changes of sphingolipids in blood in response to platelet activation.
J Thromb Haemost 4: 2704–2709
Di Paola M, Cocco T, Lorusso M (2000) Ceramide interaction with the
respiratory chain of heart mitochondria. Biochemistry 39: 6660–6668
Dindo D, Dahm F, Szulc Z, Bielawska A, Obeid LM, Hannun YA, Graf R,
Clavien P-A (2006) Cationic long-chain ceramide LCL-30 induces cell
death by mitochondrial targeting in SW403 cells. Mol Cancer Ther 5:
1520–1529
Dubois V, Dasnois L, Lebtahi K, Collot F, Heylen N, Havaux N, Fernandez AM,
Lobl TJ, Oliyai C, Nieder M, Shochat D, Yarranton GT, Trouet A (2002)
CPI-0004Na, a new extracellularly tumor-activated prodrug of doxorubicin:
in vivo toxicity, activity, and tissue distribution confirm tumor cell
selectivity. Cancer Res 62: 2327–2331
El-Assaad W, Kozhaya L, Araysi S, Panjarian S, Bitar FF, Baz E, El-Sabban ME,
Dbaibo GS (2003) Ceramide and glutathione define two independently
regulated pathways of cell death initiated by p53 in Molt-4 leukaemia cells.
Biochem J 376: 725–732
Fukuda Y, Kihara A, Igarashi Y (2003) Distribution of sphingosine kinase
activity in mouse tissues: contribution of SPHK1. Biochem Biophys Res
Commun 309: 155–160
Futerman AH, Hannun YA (2004) The complex life of simple sphingolipids.
EMBO Rep 5: 777–782
Garcia-Ruiz C, Colell A, Mari M, Morales A, Fernandez-Checa JC (1997)
Direct effect of ceramide on the mitochondrial electron transport chain
leads to generation of reactive oxygen species. Role of mitochondrial
glutathione. J Biol Chem 272: 11369–11377
Ghafourifar P, Klein SD, Schucht O, Schenk U, Pruschy M, Rocha S, Richter C
(1999) Ceramide induces cytochrome c release from isolated mitochondria.
Importance of mitochondrial redox state. J Biol Chem 274: 6080–6084
Ikeda H, Satoh H, Yanase M, Inoue Y, Tomiya T, Arai M, Tejima K,
Nagashima K, Maekawa H, Yahagi N, Yatomi Y, Sakurada S, Takuwa Y,
Ogata I, Kimura S, Fujiwara K (2003) Antiproliferative property of
sphingosine 1-phosphate in rat hepatocytes involves activation of Rho
via Edg-5. Gastroenterology 124: 459–469
Itoh M, Kitano T, Watanabe M, Kondo T, Yabu T, Taguchi Y, Iwai K,
Tashima M, Uchiyama T, Okazaki T (2003) Possible role of ceramide as
an indicator of chemoresistance: decrease of the ceramide content via
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
104
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics

Page 8

activation of glucosylceramide synthase and sphingomyelin synthase in
chemoresistant leukemia. Clin Cancer Res 9: 415–423
Liu H, Toman RE, Goparaju SK, Maceyka M, Nava VE, Sankala H, Payne SG,
Bektas M, Ishii I, Chun J, Milstien S, Spiegel S (2003) Sphingosine kinase
type 2 is a putative BH3-only protein that induces apoptosis. J Biol Chem
278: 40330–40336
Novgorodov SA, Szulc ZM, Luberto C, Jones JA, Bielawski J, Bielawska A,
Hannun YA, Obeid LM (2005) Positively charged ceramide is a potent
inducer of mitochondrial permeabilization. J Biol Chem 280: 16096–16105
Obeid LM, Linardic CM, Karolak LA, Hannun YA (1993) Programmed cell
death induced by ceramide. Science 259: 1769–1771
Oh WJ, Kim WH, Kang KH, Kim TY, Kim MY, Choi KH (1998) Induction of
p21 during ceramide-mediated apoptosis in human hepatocarcinoma
cells. Cancer Lett 129: 215–222
Oskouian B, Sooriyakumaran P, Borowsky AD, Crans A, Dillard-Telm L,
Tam YY, Bandhuvula P, Saba JD (2006) Sphingosine-1-phosphate lyase
potentiates apoptosis via p53- and p38-dependent pathways and is down-
regulated in colon cancer. Proc Natl Acad Sci USA 103: 17384–17389
Radin NS (2003) Designing anticancer drugs via the achilles heel: ceramide,
allylic ketones, and mitochondria. Bioorg Med Chem 11: 2123–2142
Ruvolo PP, Deng X, Ito T, Carr BK, May WS (1999) Ceramide induces Bcl2
dephosphorylation via a mechanism involving mitochondrial PP2A.
J Biol Chem 274: 20296–20300
Samsel L, Zaidel G, Drumgoole HM, Jelovac D, Drachenberg C, Rhee JG,
Brodie AM, Bielawska A, Smyth MJ (2004) The ceramide analog, B13,
induces apoptosis in prostate cancer cell lines and inhibits tumor growth
in prostate cancer xenografts. Prostate 58: 382–393
Selzner M, Bielawska A, Morse MA, Rudiger HA, Sindram D, Hannun YA,
Clavien PA (2001) Induction of apoptotic cell death and prevention of
tumor growth by ceramide analogues in metastatic human colon cancer.
Cancer Res 61: 1233–1240
Senkal C, Ponnusamy S, Rossi M, Sundararaj K, Szulc Z, Bielawski J,
Bielawska A, Meyer M, Cobanoglu B, Koybasi S, Sinha D, Day T, Obeid L,
Hannun Y, Ogretmen B (2006) Potent Anti-Tumor Activity of a Novel
Cationic Pyridinium-Ceramide Alone or In Combination with Gemcita-
bine Against Human Head and Neck Squamous Cell Carcinomas In vitro
and In vivo. J Pharmacol Exp Ther 317: 1188–1199
Szulc ZM, Bielawski J, Gracz H, Gustilo M, Mayroo N, Hannun YA, Obeid LM,
Bielawska A (2006) Tailoring structure-function and targeting pro-
perties of ceramides by site-specific cationization. Bioorg Med Chem 14:
7083–7104
Thon L, Mohlig H, Mathieu S, Lange A, Bulanova E, Winoto-Morbach S,
Schutze S, Bulfone-Paus S, Adam D (2005) Ceramide mediates caspase-
independent programmed cell death. FASEB J 19: 1945–1956
Wattenberg BW, Pitson SM, Raben DM (2006) The sphingosine and
diacylglycerol kinase super family of signaling kinases: localization as a
key to signaling function. J Lipid Res 47: 1128–1139
Yoneda J, Killion JJ, Bucana CD, Fidler IJ (1999) Angiogenesis and growth
of murine colon carcinoma are dependent on infiltrating leukocytes.
Cancer Biother Radiopharm 14: 221–230
Mitochondrially targeted ceramide in colorectal cancer
F Dahm et al
105
British Journal of Cancer (2008) 98(1), 98–105
& 2008 Cancer Research UK
Translational Therapeutics