Dihydroceramide accumulation mediates cytotoxic autophagy of cancer cells via autolysosome destabilization.

TL;DR: It is shown that THC (but not nutrient deprivation) increases the dihydroceramide:ceramide ratio in the endoplasmic reticulum of glioma cells, and this alteration is directed to autophagosomes and autolysosomes to promote lysosomal membrane permeabilization, cathepsin release and the subsequent activation of apoptotic cell death.

Abstract: Autophagy is considered primarily a cell survival process, although it can also lead to cell death. However, the factors that dictate the shift between these 2 opposite outcomes remain largely unknown. In this work, we used Δ9-tetrahydrocannabinol (THC, the main active component of marijuana, a compound that triggers autophagy-mediated cancer cell death) and nutrient deprivation (an autophagic stimulus that triggers cytoprotective autophagy) to investigate the precise molecular mechanisms responsible for the activation of cytotoxic autophagy in cancer cells. By using a wide array of experimental approaches we show that THC (but not nutrient deprivation) increases the dihydroceramide:ceramide ratio in the endoplasmic reticulum of glioma cells, and this alteration is directed to autophagosomes and autolysosomes to promote lysosomal membrane permeabilization, cathepsin release and the subsequent activation of apoptotic cell death. These findings pave the way to clarify the regulatory mechanisms that determine the selective activation of autophagy-mediated cancer cell death.

Full text

Autophagy
ISSN: 1554-8627 (Print) 1554-8635 (Online) Journal homepage: http://www.tandfonline.com/loi/kaup20
Dihydroceramide accumulation mediates cytotoxic
autophagy of cancer cells via autolysosome
destabilization
Sonia Hernández-Tiedra, Gemma Fabriàs, David Dávila, Íñigo J. Salanueva,
Josefina Casas, L. Ruth Montes, Zuriñe Antón, Elena García-Taboada, María
Salazar-Roa, Mar Lorente, Jesper Nylandsted, Jane Armstrong, Israel López-
Valero, Christopher S. McKee, Ana Serrano-Puebla, Roberto García-López,
José González-Martínez, José L. Abad, Kentaro Hanada, Patricia Boya,
Félix Goñi, Manuel Guzmán, Penny Lovat, Marja Jäättelä, Alicia Alonso &
Guillermo Velasco
To cite this article: Sonia Hernández-Tiedra, Gemma Fabriàs, David Dávila, Íñigo J. Salanueva,
Josefina Casas, L. Ruth Montes, Zuriñe Antón, Elena García-Taboada, María Salazar-Roa,
Mar Lorente, Jesper Nylandsted, Jane Armstrong, Israel López-Valero, Christopher S. McKee,
Ana Serrano-Puebla, Roberto García-López, José González-Martínez, José L. Abad, Kentaro
Hanada, Patricia Boya, Félix Goñi, Manuel Guzmán, Penny Lovat, Marja Jäättelä, Alicia
Alonso & Guillermo Velasco (2016) Dihydroceramide accumulation mediates cytotoxic
autophagy of cancer cells via autolysosome destabilization, Autophagy, 12:11, 2213-2229, DOI:
10.1080/15548627.2016.1213927
To link to this article: http://dx.doi.org/10.1080/15548627.2016.1213927
© 2016 The Author(s). Published with
license by Taylor & Francis.© Sonia
Hernández-Tiedra, Gemma Fabrias, David
Dávila, Íñigo J. Salanueva, Josefina Casas, L.
Ruth Montes, Zuriñe Antón, Elena Garcıa-
Taboada, María Salazar-Roa, Mar Lorente,
Jesper Nylandsted, Jane Armstrong, Israel
López-Valero, Christopher S. McKee, Ana
Serrano-Puebla, Roberto García-López, José
González-Martínez, José L. Abad, Kentaro
Hanada, Patricia Boya, Felix Goñi, Manuel
Guzman, Penny Lovat, Marja Jäättelä, Alicia
Alonso, and Guillermo Velasco.
View supplementary material
Published online: 16 Sep 2016.
Submit your article to this journal
Article views: 2834
View related articles

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=kaup20
Download by: [CSIC Centro de Investigaciones Biologicas] Date: 02 February 2017, At: 05:19
View Crossmark data
Citing articles: 1 View citing articles

TRANSLATIONAL RESEARCH PAPER
Dihydroceramide accumulation mediates cytotoxic autophagy of cancer cells via
autolysosome destabilization
Sonia Hern

andez-Tiedra
a
,
b
, Gemma Fabri

as
c
, David D

avila
a
,
b
,

I
~
nigo J. Salanueva
a
, Josefina Casas
c
, L. Ruth Montes
d
,
Zuri
~
ne Ant

on
d
, Elena Garc

ıa-Taboada
a
, Mar

ıa Salazar-Roa
a
, Mar Lorente
a
,
b
, Jesper Nylandsted
e
, Jane Armstrong
f
,
g
,
Israel L

opez-Valero
a
,
b
, Christopher S. McKee
f
, Ana Serrano-Puebla
a
,
h
, Roberto Garc

ıa-L

opez
a
, Jos

e Gonz

alez-Mart

ınez
a
,
b
,
Jos

e L. Abad
c
, Kentaro Hanada
i
, Patricia Boya
h
,F

elix Go
~
ni
d
, Manuel Guzm

an
a
,
j
, Penny Lovat
f
, Marja J
€
a
€
attel
€
a
e
,
Alicia Alonso
d
, and Guillermo Velasco
a
,
b
a
Department of Biochemistry and Molecular Biology I, School of Biology, Complutense University, Madrid, Spain;
b
Instituto de Investigaciones
Sanitarias San Carlos (IdISSC), Madrid, Spain;
c
Research Unit on BioActive Molecules (RUBAM), Departments of Biomedicinal Chemistry, Institute for
Advanced Chemistry of Catalonia (IQAC-CSIC), Barcelona, Spain;
d
Biofisika Institute (UPV/EHU, CSIC), and Departamento de Bioqu

ımica, Universidad del
Pa

ıs Vasco, Barrio Sarriena s/n, Leioa, Spain;
e
Unit of Cell Death and Metabolism, Center for Autophagy, Recycling and Disease, Danish Cancer Society
Research Center (DCRC), Copenhagen, Denmark;
f
Dermatological Sciences, Institute of Cellular Medicine, Newcastle University, Newcastle-upon-Tyne,
UK;
g
Faculty of Applied Sciences, University of Sunderland, Sunderland, UK;
h
Departament of Cellular and Molecular Biology, Centro de Investigaciones
Biol

ogicas, CSIC, Madrid, Spain;
i
Department of Biochemistry and Cell Biology, National Institute of Infectious Diseases, Shinjuku-ku, Tokyo, Japan;
j
Centro de Investigaci

on Biom

edica en Red Sobre Enfermedades Neurodegenerativas, Instituto Ram

on y Cajal de Investigaci

on Sanitaria, Madrid, Spain,
Instituto Universitario de Investigaci

on Neuroqu

ımica, Complutense University, Madrid, Spain
ARTICLE HISTORY
Received 21 October 2015
Revised 21 June 2016
Accepted 13 July 2016
ABSTRACT
Autophagy is considered primarily a cell survival process, although it can also lead to cell death. However,
the factors that dictate the shift between these 2 opposite outcomes remain largely unknown. In this
work, we used D
9
-tetrahydrocannabinol (THC, the main active component of marijuana, a compound that
triggers autophagy-mediated cancer cell death) and nutrient deprivation (an autophagic stimulus that
triggers cytoprotective autophagy) to investigate the precise molecular mechanisms responsible for the
activation of cytotoxic autophagy in cancer cells. By using a wide array of experimental approaches we
show that THC (but not nutrient deprivation) increases the dihydroceramide:ceramide ratio in the
endoplasmic reticulum of glioma cells, and this alteration is directed to autophagosomes and
autolysosomes to promote lysosomal membrane permeabilization, cathepsin release and the subsequent
activation of apoptotic cell death. These findings pave the way to clarify the regulatory mechanisms that
determine the selective activation of autophagy-mediated cancer cell death.
KEYWORDS
autophagy; cancer;
cannabinoids; cell death;
sphingolipids
Introduction
Macroautophagy, hereafter named autophagy, is a highly con-
served cellular process in which cytoplasmic materials, including
organelles, are sequestered into double-membrane compart-
ments, phagophores, that mature into autophagosomes; the
cargo is subsequently delivered to lysosomes for degradation
and recycling.
1-3
In many cellular settings, triggering of auto-
phagy relies on the inhibition of MTORC1 (mechanistic target
of rapamycin [serine/threonine kinase] complex 1), an event
that promotes the activation (de-inhibition) of several ATG
(autophagy-related) proteins involved in the initial phase of
phagophore formation.
1-3
The membrane source from which
autophagosomes are derived is still debatable, as it has been pro-
posed that it could be derived either from de novo synthesized
lipids or generated by vesicle budding from the endoplasmic
reticulum (ER), Golgi apparatus or endosomes,
4,5
or the plasma
membrane.
6
In particular, an ER-derived structure termed the
omegasome has been proposed as an origin of the phagophore
membrane.
5,7
Enlargement of this compartment to form the
autophagosome requires the participation of 2 ubiquitin-like
conjugation systems, one involving the conjugation of ATG12
(autophagy-related 12) to ATG5 (autophagy-related 5), and the
other of phosphatidylethanolamine to MAP1LC3/LC3 (microtu-
bule-associated protein 1 light chain 3).
2
The final outcome of
the activation of the autophagy program is highly dependent on
the cellular context and the strength and duration of the stress-
inducing signals. Thus, autophagy plays an important role in
cellular homeostasis and is considered primarily a cell-survival
mechanism, for example in situations of nutrient deprivation.
8-11
However, stimulation of autophagy can also have a cytotoxic
effect. For example, several anticancer agents activate auto-
phagy-associated cell death.
8-10,12
However, the molecular
CONTACT Guillermo Velasco gvelasco@ucm.es Department of Biochemistry and Molecular Biology I, School of Biology, Complutense University Calle Jos

e
Antonio Nov

ais 12, 28040-Madrid, Spain.
Supplemental data for this article can be accessed on the publisher’s website.
© 2016 Sonia Hern

andez-Tiedra, Gemma Fabrias, David D

avila,

I
~
nigo J. Salanueva, Josefina Casas, L. Ruth Montes, Zuri
~
ne Ant

on, Elena Garcıa-Taboada, Mar

ıa Salazar-Roa, Mar Lorente, Jesper
Nylandsted, Jane Armstrong, Israel L

opez-Valero, Christopher S. McKee, Ana Serrano-Puebla, Roberto Garc

ıa-L

opez, Jos

e Gonz

alez-Mart

ınez, Jos

e L. Abad, Kentaro Hanada, Patricia Boya, Felix
Go
~
ni, Manuel Guzman, Penny Lovat, Marja J
€
a
€
attel
€
a, Alicia Alonso, and Guillermo Velasco. Published with license by Taylor & Francis.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (
http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribu-
tion, and reproduction in any medium, provided the original work is properly cited.
AUTOPHAGY
2016, VOL. 12, NO. 11, 2213 –2229
http://dx.doi.org/10.1080/15548627.2016.1213927

mechanisms that determine the outcome of autophagy activa-
tion for the survival or death of cancer cells remain to be
clarified.
D
9
-Tetrahydrocannabinol (THC), the main active compo-
nent of Cannabis sativa,
13,14
exerts a wide variety of biological
effects by mimicking endogeno us substances—the endocanna-
binoids anandamide
15
and 2-arachidonoylglycerol (2-AG)
16,17
that engage specific cell-surface G protein-coupled cannabinoid
receptors.
14
So far, 2 major cannabinoid-specific receptors,
CNR1/CB
1
(cannabinoid receptor 1 [brain]) and CNR2/CB
2
(cannabinoid receptor 2 [macrop hage]), have been cloned and
characterized from mammalian tissues.
18,19
Cannabinoid
administration curbs the growth of several genetic and xeno-
graft models of cancer in rats and mice, and therefore these
compounds are considered a novel family of potential antican-
cer agents.
20
The mechanism of cannabinoid anticancer action
relies, at least largely, on the ability of these agent s to stimulate
autophagy-mediated cancer cell death.
20
Thus, THC binds can-
nabinoid receptors, which leads to the stimulation of de novo
sphingolipid synthesis and the subsequent activation of an
endoplasmic reticulum (ER) stress-related signaling route that
involves the upregulation of the transcriptional co-activator
NUPR1/p8 (nuclear protein 1, transcriptional regulator) and
its effector TRIB3 (tribbles pseudokinase 3).
20-23
The stimula-
tion of this pathway promotes in turn autophagy via
TRIB3-mediated inhibition of the AKT (thymoma viral proto-
oncogene)-MTORC1 axis, which is indispensable for the pro-
apoptotic and antitumoral action of cannabinoids.
24,25
In this study, we have investigated the molecular mechanism
underlying the activation of autophagy-mediated cancer cell death
by comparing the effects of THC treatment and nutrient depriva-
tion, 2 autophagic stimuli that produce opposite effects on the reg-
ulation of cancer cell survival/death. Using this experimental
model, we found that treatment with THC—but not exposure to
nutrient deprivation—leads to an alteration of the balance
between different molecular species of ceramides and dihydrocer-
amides in the microsomal (endoplasmic reticulum-enriched) frac-
tion of cancer cells. Moreover, our findings support the hypothesis
that such modification can be transmitted to autophagosomes and
autolysosomes, where it can promote the permeabilization of the
organellar membrane, the release of cathepsins to the cytoplasm
and the subsequent activation of apoptotic cell death.
Results
THC-induced, but not nutrient deprivation-induced,
autophagy relies on the stimulation of sphingolipid
biosynthesis
As a first approach to investigate the molecular mechanisms
responsible for the activation of autophagy-mediated cancer
cell death we analyzed the effect of 2 different stimuli, namely
nutrient deprivation and THC treatment, that trigger cytopro-
tective and cytotoxic autophagy, respectively. We found that
genetic inhibition of the autophagy essential gene ATG5 in
both U87MG cells and oncogene-transformed mouse embry-
onic fibroblasts (MEFs) prevented THC-induced cell death
while it further diminished the nutrient deprivation-induced
decrease in cell viability (
Fig. 1A and Fig. S1A), thus supporting
the notion that stimulation of autophagy may play a dual role
in the regulation of cancer cell survival.
After confirming that incubation with EBSS and treatment
with THC led to an increase in the accumulation of
MAP1LC3B-positive dots in U87MG cells (Fig. S1B) we ana-
lyzed the ability of these 2 stimuli to enhance the autophagic flux
in U87MG cells. To this aim, we performed the treatments in the
presence or the absence of the lysosomal proteases inhibitors
E64d and pepstatin A (C inh); upon stimulation of dynamic
autophagy and in the presence of these inhibitors there is a
blockade of the autophagic flux and therefore an enhanced accu-
mulation of proteins present in the autophagosomes, and specifi-
cally of the lipidated and autophagosome-associated form of
MAP1LC3, MAP1LC3-II. Of note, incubation with EBSS
induced only an early and transient increase in the autophagic
flux (EBSS led to MAP1LC3B-II accumulation, an event that was
enhanced in the presence of E64d C pepstatin A;
Fig. 1A lower
panel
, Fig. 1B and Fig. S1C) whereas stimulation of the autopha-
gic flux by THC occurred at longer times and was sustained for
several hours (
Fig. 1A lower panel, Fig. 1B and Fig. S1C).
Previous reports by our group show that the stimulation of
sphingolipid biosynthesis by THC is involved in the induction
of autophagy-mediated cancer cell death.
20,21,24,26,27
In agre e-
ment with these observations, we found here that THC upregu-
lates mRNA levels of different enzymes involved in
sphingolipid synthesis de novo, an effect that was not observed
when cells were exposed to EBSS (
Fig. 1C). Likewise, pharma-
cological blockade (by using ISP-1) of SPT (serine palmitoyl-
transferase), the enzyme that catalyzes the first step of
sphingolipid biosynthesis, prevented THC-, but not nutrient
deprivation-induce d autophagy (
Fig. 1D). In addition, we con-
firmed that, in agreement with previous observations,
22,24,27
incubation w ith ISP-1 inhibited THC-evoked cell death
(Fig. S1D). Collectively, these results suggest that a general
increase in de novo-synthesized sphingolipids might be a differ-
ential factor in the activation of cytotoxic autophagy by THC.
THC, but not nutrient deprivation, enhances sphingolipid
synthesis de novo and dihydroceramide accumulation
The initial steps of sphingolipid biosynthesis occur at the ER,
28
where ceramides are synthesized (Fig. 2A). Therefore, as a first
approach to investigate the effect of THC and nutrient depriva-
tion on sphingolipid metabolism, we analyzed the sphin golipid
composition of the microsomal fraction of U87MG cells sub-
jected to ei ther stimulus. As shown in
Fig. 2B, THC —but not
incubation with EBSS—increased cerami de levels in the micro-
somal fraction of U87MG cells. We also found that THC but
not EBSS enhanced the levels of dihydr oceramides to a higher
extent than those of ceramides (
Fig. 2B and Fig. 2C). DEGS1/
dihydroceramide desaturase (delta[4]-desaturase, sphingolipid
1) catalyzes the insertion of a 4,5-trans double bond in the
sphingoid backbone of dihydroceramides to generate ceramides
(see
Fig. 2A).
28
Specifically, treatment with THC produced a
2.8-, 2.9- and 4.5-fold increase in the levels of C16, C24 and
C24:1 dihydroceramides, respectively, and a 1.3- and 1.2-fold
increase in the levels of C24 and C24:1 ceramides, respectively
(Fig. S2). It should be noted that ceramide levels were 6- to 10-
fold higher than those of dihydroceramides in vehicle-treated
2214
S. HERN

ANDEZ-TIEDRA ET AL.

cells (Fig. 2C and Table S1). Thus, the observed increase in
dihydroceramides levels triggered by THC led to a striking
modification of the ceramide:dihydroceramide ratio in the
microsomal fraction of U87MG cells (
Fig. 2D).
THC, but not nutrient deprivation, inhibits sphingolipid
transport from the ER to the Golgi
Once synthesized in the ER, ceramides can be delivered via
vesicular transport or through the ceramide transporter protein
COL4A3BP/CERT (collagen, type IV, a 3 [Goodpasture anti-
gen] binding protein)
29
to the Golgi apparatus, where the
synthesis of sphingomyelin and complex glycosphingolipids
takes place.
28
One way to approach the analysis of this process is
to follow the subcellular distribution of fluorescent dye-conju-
gated ceramides, for example BODIPY C5 ceramide. Thus,
when added to cells, BODIPY C5 ceramide is endocytosed and
rapidly transported to the Golgi. We therefore monitored BOD-
IPY C5 ceramide distribution to analyze the effect of THC on
ceramide transport from the ER to the Golgi. As shown in
Fig. 3A, BODIPY C5 ceramide was located in perinuclear struc-
tures resembling the Golgi apparatus in U87MG cells treated
with vehicle or subjected to nutrient deprivation. In contrast,
THC treatment induced a particulate distribution of BODIPY
Figure 1. THC, but not nutrient deprivation, -induced autophagy relies on the stimulation of sphingolipid biosynthesis. (A) Upper panel: Effect of THC (4 m M, 18 h) and
incubation with EBSS (18 h) on the number of U87MG cells stably transfected with control (shC) or ATG5-selective (shATG5) shRNAs as estimated by the MTT test (n D 4;
mean § s.d;

, P < 0.01 from THC-treated or EBSS-incubated U87 shC cells). Lower panel: Effect of THC (4 mM) and incubation with EBSS on the induction of autophagy
(as determined by MAP1LC3B-II lipidation in the presence of E64d, 10 mM; and pepstatin A, 10 mg/ml [Cinh]) of U87 cells stably transfected with control (U87 shC) or
ATG5-selective (shATG5) shRNAs (n D 3, a representative experiment is shown). ATG5 mRNA levels (as determined by real-time quantitative PCR) were reduced by
85 § 3% on U87shATG5 cells when compared with U87shC cells; (n D 4). Values in the bottom of the western blots correspond to the fold change in the MAP1LC3B-II to
TUBA1A ratio relative to shC U87MG cells at the initial time point of the treatments. Nd, nondetectable. (B) Effect of THC (4 mM, 1 h, 3 h and 6 h) and incubation with EBSS
(i.e., nutrient deprivation, 1, 3 and 6 h) on the induction of autophagy (as determined by MAP1LC3B-II lipidation in the presence of E64d, 10 mM; and pepstatin A, 10 mg/
ml [Cinh]) of U87MG cells (n D 3, a representative experiment is shown). (C) Effect of THC (4 mM; 3 h) on the mRNA levels (as determined by quantitative real-time PCR)
of different enzymes involved in sphingolipid biosynthesis (CERS2; CERS5; CERS6 (ceramide synthase 2, 5 and 6), DEGS1/dihydroceramide desaturase (delta[4]-desaturase,
sphingolipid 1) and SPTLC1 (serine palmitoyltransferase long chain base subunit 1) of U87MG cells (n D 5;

, P < 0.05;

, P < 0.01 from Veh-treated cells). (D) Effect of
THC (4 mM), ISP-1 (5 mM) and incubation with EBSS on autophagy (18 h) (as determined by MAP1LC3B immunostaining). Note that incubation with ISP-1 prevents THC
but not starvation-induced autophagy of U87MG cells. Values correspond to the percentage of cells with MAP1LC3B dots relative to the total cell number of cells § s.d;
n D 3.

, P < 0.05;

, P < 0.01 from Veh-treated cells and
#
, P < 0.05 from THC- and EBSS-treated cells. Bar: 20 mm.
AUTOPHAGY 2215