Embed Size (px)
Citation preview
Vertical Inhibition of the mTORC1/mTORC2/PI3KPathway Shows Synergistic Effects against MelanomaIn Vitro and In VivoJohannes Werzowa1, Stefan Koehrer1, Sabine Strommer1, Daniel Cejka1, Thorsten Fuereder1, Eva Zebedin2
and Volker Wacheck1,3
The phosphatidyl inositol 3-kinase/mammalian target of rapamycin (PI3K/mTOR) pathway has been shown to beinvolved in the development of melanoma. PI-103 is a kinase inhibitor blocking PI3K class IA and mTOR complex1 and 2. Here, we studied the effect of targeting the PI3K/mTORC1/mTORC2 pathway by PI-103 and rapamycin inmelanoma cells and in a melanoma mouse model. Dual targeting of PI3K and mTOR by PI-103 induced apoptosisand cell-cycle arrest, and inhibited viability of melanoma cells in vitro. Combined treatment with PI-103 and theprototypic mTORC1 inhibitor rapamycin led to the synergistic suppression of AKT and ribosomal S6 proteinphosphorylation and to the induction of apoptosis. In vivo, PI-103 and rapamycin displayed only modest single-agent activity, but the combination significantly reduced the tumor growth compared with both single agents.These data show that blocking the PI3K/mTORC1/mTORC2 pathway using the combination of two distinct small-molecule inhibitors (‘‘vertical inhibition’’) leads to superior efficacy against malignant melanoma in vitro andin vivo.
Journal of Investigative Dermatology (2011) 131, 495–503; doi:10.1038/jid.2010.327; published online 4 November 2010
INTRODUCTIONThe phosphatidyl inositol 3-kinase/mammalian target ofrapamycin (PI3K/mTOR) pathway has emerged as a promis-ing target for anti-cancer therapies. PI3K and its downstreamtargets AKT and mTOR have a central role not only inphysiological processes such as growth, survival, andproliferation, but also in the development of malignantdisease (Bjornsti and Houghton, 2004; Samuels and Ericson,2006). There are many possible mechanisms that can lead toaberrant activation of this pathway, such as mutations ofgrowth factor receptors, PI3K, phosphatase and tensin
homolog deleted on chromosome 10 (PTEN), 3-phosphoino-sitide-dependent kinase 1, RAS, tuberous sclerosis protein 1and 2 (TSC1/2), and AKT (Samuels and Ericson, 2006).
PI3K proteins form a family that is divided into threeclasses. In regulating proliferation and tumorigenesis, themost important PI3K proteins are those that belong to classIA. Class IA PI3Ks have a catalytic subunit consisting either ofp110a, b, or d. p110a is the primary insulin responsive PI3Kvia its close association with insulin receptor substrate-1, andhas been linked to solid cancers (Knight et al., 2006; Denleyet al., 2007). Signaling via p110a leads to phosphorylationand activation of proteins containing a pleckstrin homologydomain. Among these are the serine/threonine kinase AKTand its activating kinase 3-phosphoinositide-dependentkinase 1 (Corvera and Czech, 1998). AKT controls cellsurvival, proliferation, and invasion via its numerous down-stream targets (Stahl et al., 2004; Cheng et al., 2005; Meieret al., 2005). The phosphoinositide phosphatase PTEN is thephysiological antagonist of PI3K signaling, as it catalyzes thedephosphorylation of active phosphatidylinositol 3,4,5-tri-phosphate to inactive phosphatidylinositol 4,5-biphosphate(Stambolic et al., 1998). Besides being the most oftenactivated pathway in cancer in general, the PI3K/mTORpathway has also been shown to be frequently activatedin human melanoma (Meier et al., 2005; Karbowniczeket al., 2007).
The pyridinylfuranopyrimidine PI-103 is a drug candidatethat has been the product of an extensive search for isoform-specific PI3K inhibitors (Knight et al., 2006). Unlike other
& 2011 The Society for Investigative Dermatology www.jidonline.org 495
ORIGINAL ARTICLE
Received 9 December 2009; revised 14 September 2010; accepted 18September 2010; published online 4 November 2010
1Department of Clinical Pharmacology, Section of Experimental Oncology/Molecular Pharmacology, Medical University of Vienna, Vienna, Austria and2Centre for Biomolecular Medicine, Institute of Pharmacology, MedicalUniversity of Vienna, Vienna, Austria
Correspondence: Volker Wacheck, Department of Clinical Pharmacology,Section of Experimental Oncology/Molecular Pharmacology, MedicalUniversity of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria.E-mail: [email protected]
3Current address: Centre for Biomolecular Medicine, Institute ofPharmacology, Medical University of Vienna, Waehringerstrasse 13a, ViennaA-1090, Austria. E-mail: [email protected]
Abbreviations: ATM, ataxia telangiectasia mutated; ATR, ataxia telangiectasiaand Rad3 related protein; DPBS, Dulbecco’s phosphate buffered saline; ERK,extracellular signal-regulated kinase; mTOR, mammalian target of rapamycin;PI3K, phosphatidyl inositol 3-kinase; PTEN, phosphatase and tensin homologdeleted on chromosome 10; siRNA, small interfering RNA; TSC1/2, tuberoussclerosis protein 1 and 2
PI3K inhibitors, it selectively inhibits members of thephosphoinositide-3-kinase-related kinase superfamily, namelyclass I PI3Ks (with the lowest IC50 for p110a), mTORC1,mTORC2, and DNA-PK, whereas leaving other memberssuch as ataxia telangiectasia mutated (ATM) or ataxiatelangiectasia and Rad3 related protein (ATR) unaffected(Bain et al., 2007). Its high affinity for the p110a catalyticsubunit of PI3K is of particular interest, as p110a is theprincipal insulin receptor substrate-1-associated PI3K incancer cell lines (Foukas et al., 2006). PI-103 has beensuccessfully employed in vitro and in animal models fortumors such as glioma, Kaposi’s sarcoma and leukemia (Fanet al., 2006; Chaisuparat et al., 2008; Kharas et al., 2008).
In the current study, we hypothesized that verticalinhibition of the PI3K/AKT/mTOR signaling pathway by thePI3K/mTORC1/mTOR2 triple inhibitor PI-103 alone and incombination with the mTORC1 inhibitor rapamycin results inpotent anti-tumor activity against melanoma in vitro andin vivo.
RESULTSBlockade of PI3Ka and mTOR with PI-103 inhibits melanomacell viabilityCell viability following combined PI3Ka/mTORC1/mTORC2inhibition by PI-103 was tested in a panel of differentmelanoma cell lines. Cells were treated with PI-103 in thedose range from 10 nmol l�1 to 10 mmol l�1. Incubation withPI-103 led to a dose-dependent decrease in cell viability in all
cell lines tested (i.e., Mel-Juso and 518A2 (Figure 1a), 607B,A375, and SK-MEL-28 (Supplementary Figure S1 online)). TheIC50 values ranged from 0.05 (SK-MEL-28) to 0.8mmol l�1 (607B).
To investigate the effects of PI-103 on PI3K/mTOR signa-ling, we performed phospho-specific western blot analyses ofAKT and ribosomal S6 protein (S6; Figure 1b)—two well-described targets downstream of PI3K and mTOR (Soulardand Hall, 2007). Concentrations below 0.05 and 0.5 mmol l�1
in Mel-Juso and 518A2 cells, respectively, led to a slightincrease in AKT phosphorylation. Above these concentra-tions, AKT phosphorylation was reduced and this effectwas stronger in Mel-Juso cells than in 518A2 cells, whichappeared to be relatively resistant to inhibition of AKTphosphorylation. Phosphorylation of ribosomal S6 proteinwas markedly reduced at 500 nmol l�1 PI-103 in bothcell lines.
Rapamycin and PI-103 show synergistic effects on melanomacells in vitro
As relatively high concentrations of PI-103 were needed toaffect melanoma cell viability and phosphorylation of ribosomalS6 protein and AKT, we next asked whether ‘‘vertical inhi-bition’’ of the PI3K/mTORC1/mTOR2 by an additional inhibitormight improve the activity of PI-103 against melanoma. Thus,we treated Mel-Juso and 518A2 melanoma cells with combina-tions of PI-103 and rapamycin, as rapamycin and its analogs arethe most potent mTORC1 inhibitors available. The combinationof rapamycin and PI-103 led to a more pronounced decrease in
Via
bilit
y (%
)
100
90
80
70
60
50
40
30
20
10
0
PI-103 (μmol l–1)
PI-103 (μmol l–1)
Mel-Juso 518A2p-AKT (S473)
p-AKT (T308)
p-S6 (S240/244)
p-ERK (T202/Y204)
ERK
Actin
S6
AKT
0 0.001 0.01 0.05 0.1 0.5 1 0 0.001 0.01 0.05 0.1 0.5 1
0 0.01 0.1 1 10
Figure 1. Effects of PI-103 on melanoma cells in vitro. (a) PI-103 inhibits viability of melanoma cells in vitro. Mel-Juso (black circles) and 518A2 (open squares)
melanoma cells were treated with PI-103 from 0.01–10mmol l�1 for 72 hours and viability assays were performed. Control treatment was with DMSO. Data show
mean values ± SD of four independent experiments. (b) PI-103 suppresses phosphorylation of phosphatidyl inositol 3-kinase downstream targets (phospho-specific
western blot). Mel-Juso and 518A2 cells were incubated with PI-103 (0.001–1mmol l�1) for 24 hours. ERK, extracellular signal-regulated kinase.
496 Journal of Investigative Dermatology (2011), Volume 131
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
cell viability in all the melanoma cell lines used compared withthe respective single-agent treatments, and resulted in astatistically significant reduction in viability relative to therespective single-agent treatments (P-values o0.05; Figure 2aand Supplementary Figure S2 online). Combination indicesfrom isobolograms were calculated for all combinations. Acombination index below 1 indicates synergy between the twodrugs. The highest rates of synergy were observed for Mel-Juso,518A2, and 607B cells (see Supplementary Figure S2 online).
We next examined whether the combination of PI-103 andrapamycin affects the phosphorylation status of AKT and S6protein in Mel-Juso and 518A2 cells. As shown in Figure 2b,the combined treatment led to a synergistic reduction inAKT phosphorylation. Rapamycin at 50 nmol l�1 plus PI-103at 0.5 mmol l�1 completely inhibited AKT phosphorylation.Treatment with rapamycin alone caused a clear increase inAKT phosphorylation, as previously described in melanomaand other cell lines (O’Reilly et al., 2006). Phosphorylation
518A2
518A2
Mel-Jusoa
b
c
Mel-Juso
*
*
* **
**
*
0 0 0.1 0.5 1 0 0 0.1 0.5 1
PI-103 (μmol l–1)
PI-103 (μmol l–1)
Rapamycin Rapamycin
Rapamycin
*
*
*
*
Cel
ls in
G0/
G1
(%)
60
70
80
90
50
400
20
40
60
Apo
ptot
ic c
ells
(%
)
Rapamycin Rapamycin +PI-103
Control PI-103 Rapamycin Rapamycin +PI-103
Control PI-103
0 0 0.05 0.05 0.1 0.1 0.5 0.5 1 1+++++ –– – – –
+ + + + + + + +– –
PI-103 (μmol l–1)
p-AKT (S473)p-AKT (T308)
p-S6 (S240/244)
p-ERK (T202/Y204)ERKActin
S6
AKT
0 0 0.050.05 0.1 0.1 0.5 0.5 1 1+++++ –– – – –
Via
bilit
y (%
)100
80
60
40
20
0
Via
bilit
y (%
)
100
80
60
40
20
0
Figure 2. Effects of combined treatment with PI-103 and rapamycin on melanoma cells in vitro. (a) PI-103 and rapamycin synergistically reduce melanoma
cell viability. Mel-Juso and 518A2 cells were treated with rapamycin, PI-103, and combinations of the two inhibitors for 24 hours. Rapamycin was used at a
concentration of 50 nmol l�1. Results of at least three independent experiments (viability assays) are depicted as box plots. Asterisks indicate significant
differences between the combined treatment and respective single-agent treatments (Po0.05). (b) Combination of rapamycin and PI-103 leads to suppression of
AKT phosphorylation at S473 and T308 (phospho-specific western blot). Mel-Juso and 518A2 cells were treated with rapamycin (50 nmol l�1), PI-103
(0.1–1 mmol l�1), and combinations of the two for 24 hours. Bands that were not originally run next to each other are separated by a dividing line. Control
treatment was with DMSO. (c) Combination of rapamycin and PI-103 leads to induction of cell-cycle arrest and apoptosis: Mel-Juso (white boxes) and 518A2
(gray boxes) cells were treated with rapamycin (50 nmol l�1), PI-103 (1 mmol l�1), and combination of the two for 48 hours. Cells were stained with propidium
iodide, and cell-cycle analysis was performed using flow cytometry. The fraction of apoptotic cells after 24 hours treatment was determined
using flow cytometry. Asterisks indicate that differences between combined treatments, and both single-agent therapies are significant. ERK, extracellular
signal-regulated kinase.
www.jidonline.org 497
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
of extracellular signal-regulated kinase was increased byboth agents alone and in combination. An increase inextracellular signal-regulated kinase phosphorylation afterrapamycin treatment has also been observed by others(Chen et al., 2010).
Combination of PI-103 and rapamycin led to a signi-ficant increase of cells in the G0/G1 phase of the cell cycle(Figure 2c). Treatment with rapamycin and PI-103 alone ledto a reduction in G2/M phase, whereas the combinedtreatment also strongly suppressed the S-phase of the cellcycle (see Supplementary Figure S4 online). With respect toapoptosis induction, the effect of the combination treatmentwith PI-103 and rapamycin was even more pronounced.Whereas rapamycin treatment did not induce apoptosis andPI-103 treatment led to modest increases in apoptosis, thecombination of PI-103 and rapamycin more than doubled thenumber of apoptotic cells in both cell lines (Figure 2c).Therefore, combination of PI-103 and rapamycin is stronglysynergistic in the induction of apoptosis in melanoma cells.
p110a inhibition is essential for the effects of PI-103
To gain more mechanistic insights into the observed effects, wetested GDC-0941, which is a highly specific small-moleculeinhibitor of p110a, with reduced activity against mTOR and noactivity against DNA-PK (Raynaud et al., 2009). Mel-Juso and518A2 were sensitive toward GDC-0941 treatment, with IC50
values of B0.75mmol l�1 (Figure 3a). In combination withrapamycin, synergistic reduction in cell viability was observedto a similar extent as for PI-103 (Figure 3a and SupplementaryFigure S5 online). In addition, GDC-0941 treatment blockedthe phosphorylation of AKT protein (S473 and T308) at evenlower concentrations than PI-103, and S6 protein phosphor-ylation was blocked at similar concentrations (Figure 3b).Combination of GDC-0941 with rapamycin, analogous toPI-103 plus rapamycin, completely blocked AKT and S6protein phosphorylation at GDC-0941 concentrations as lowas 0.05mmol l�1 (Figure 3b). Extracellular signal-regulatedkinase phosphorylation was not significantly influenced.
To elucidate the role of other p110 isoforms, we employedIC87114, a highly specific p110d inhibitor, which—at higherconcentrations—also inhibits p110g (Ali et al., 2008). Asshown in Figure 3c, IC87114 had no effects on the viability ofMel-Juso and 518A2 cells at concentrations that inhibitp110d (around 0.1 mmol l�1) and p110g (above 1 mmol l�1;only concentrations from 2.5–10 mmol l�1 are shown), indi-cating that other PI3K isoforms than p110a do not have a rolein the cell lines tested.
We next inhibited p110a using a small interfering RNA(siRNA) directed against PIK3CA, the gene encoding the p110acatalytic subunit of PI3K. PIK3CA siRNA treatment led to areduction in p110a protein levels by B50%, as determined bydensitometric analysis of western blot bands of p110a (Figure3e). This p110a downregulation resulted in about 20%reduction in cell viability in Mel-Juso and 518A2 cells, aneffect comparable to the treatment with rapamycin at50 nmol l�1 alone (Figure 3d). Combination of PIK3CA siRNAand rapamycin demonstrated a mild additive effect on cellviability. At the protein level, PIK3CA siRNA further promoted
the reduction of pS6 by rapamycin, whereas it slightly reducedthe p-AKT induction caused by rapamycin (Figure 3e).
Rapamycin and PI-103 show synergistic effects in a humanmelanoma xenograft model
To investigate the in vivo activity of combined rapamycinand PI-103 treatment, we treated athymic nude micebearing 518A2 xenograft melanomas with rapamycin (inthe form of rapamune), PI-103, and the combination of both.As shown in Figure 4a, both rapamycin and PI-103monotherapy showed only modest effects on tumor growth.In contrast, the combination treatment resulted in a majorinhibition of tumor growth by about 50% relative to anyother treatment group. At the end of the study, mice treatedwith rapamycin plus PI-103 displayed a statistically signifi-cant reduction in tumor volume compared with all othertreatment groups (Po0.05). None of the treatment regimensresulted in observable toxicity for the animals. There were nosignificant differences in body weight between the animals inall the four groups over the whole course of the experiment(Figure 4b).
To study the mechanism of action for the prominent anti-tumor activity of rapamycin plus PI-103 in vivo, cell lysateswere prepared from tumors at the end of study and wereexamined for PI3K/mTOR target expression. We observed aclear effect of the combined treatment on AKT phosphoryla-tion (Figure 4c). Similar to the in vitro results, rapamycintreatment alone induced AKT phosphorylation in vivo, andcombination of PI-103 with rapamycin reversed this activa-tion. After normalization for actin loading, densitometricanalysis revealed for the combination treatment a reductionin AKT phosphorylation by B85 and 50% relative torapamycin and PI-103 monotherapies, respectively (Figure4d). These findings were confirmed by immunohistochemicalanalysis of tumor tissue sections. p-AKT and p-S6 proteinlevels were clearly reduced following combined treatmentwith PI-103 and rapamycin (Figure 4e).
DISCUSSIONTargeting the mTOR pathway appears to be a promisingstrategy against numerous types of cancer and has alreadybeen approved for the treatment of kidney cancer (Hudeset al., 2007). Unfortunately, rapamycin and its analogs(rapalogs) have not been successful in the treatment ofmelanoma (Margolin et al., 2005). The reason for this lack ofresponse is not clear although it has been argued that theability of rapalogs to induce phosphorylation of the oncogeneAKT in certain cancer cells including melanoma mighthamper their efficacy (Hay, 2005). Targeting both, the PI3Kand mTOR, pathways with a set of newly identified ATP-competitive PI3K/mTOR inhibitors inhibited the tumorgrowth in a syngeneic B16 mouse melanoma tumor model(Marone et al., 2009). Furthermore, combination of PI3Kinhibitors and rapamycin has shown cooperative effects inother tumor entities in vitro (Takeuchi et al., 2005). Incontrast, horizontal inhibition using PI-103 and the BRAFinhibitor sorafenib did not lead to synergistic effects againstmelanoma in vivo (Lopez-Fauqued et al., 2009).
498 Journal of Investigative Dermatology (2011), Volume 131
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
In the present study, we hypothesized that ‘‘verticalinhibition’’ of the PI3K/AKT/mTOR signaling pathway byPI-103 and rapamycin results in potent anti-tumor activityagainst human malignant melanoma in vitro and in vivo.‘‘Vertical inhibition’’ refers to the optimal blockade of a given
pathway by combination of compounds that inhibit differentcomponents within this pathway.
The combination of PI-103 with the highly specificmTORC1 inhibitor rapamycin led to synergistic effectsin vitro and in vivo. Thus, the main finding of our work is
0.1++
0.1 0.5 0.5
518A2
518A2p-AKT (S473)p-AKT (T308)
p-ERK (T202/Y204)ERKActin
AKTp-S6 (S240/244)S6
Mel-Juso
Mel-Jusoa
b
c d
e
1 1 GDC-0941 (μmol l–1)
* * **
*
**
**
Rapamycin0.050.05
140
120
100
80
60
40
20
0
100
80
60
40
20
00 2.5 5 7.5 10 IC87114 (μmol l–1)
Mel-Juso 518A2p110αp-AKT (S473)p-AKT (T308)
p-S6 (S240/244)
p-ERK (T202/Y204)
+ + + + ––––+ + + +––––
+ + + –––– +
ERKActin
siRNA PIK3CAsiRNA controlRapamycin
S6
AKT
siRNA
cont
rol
siRNA
cont
rol+r
apa
siRNA
PIK3CA
siRNA
PIK3CA
+rap
a
Via
bilit
y (%
)
Via
bilit
y (%
)
Via
bilit
y (%
)
40
20
60
80
100
0
Via
bilit
y (%
)
40
20
60
80
100
00
– +– +–+0– –
0.1++
0.1 0.5 0.5 1 1 GDC-0941 (μmol l–1)Rapamycin
GDC-0941 (μmol l–1)Rapamycin
0.050.05
0 0 0.1 0.1 0.5 0.5 1 10.05 0.05
0– +– +–
+– +– +– +– +–0 0 0.1 0.1 0.5 0.5 1 10.05 0.05
+– +– +– +– +–
+0– –
Figure 3. Effects of inhibition of different p110 isoforms. (a) GDC-0941 and rapamycin show synergistic effects on cell viability. Mel-Juso and 518A2 cells
were treated with GDC-0941 (0.05–1 mmol l�1) alone and in combination with rapamycin (50 nmol l�1) for 72 hours, and cell viability assays of three
independent experiments were performed. Asterisks indicate significant differences between combination and respective single-agent treatments (Po0.05).
(b) Combination of rapamycin and PI-103 leads to the suppression of AKT phosphorylation (phospho-specific western blot). Mel-Juso and 518A2 cells were treated
with GDC-0941 from 0.05–1mmol l�1 alone and in combination with rapamycin (50 nmol l�1) for 24 hours. Control treatment was with DMSO. (c) IC87114 has no
effect on cell viability. Mel-Juso (black circles) and 518A2 (open squares) cells were treated with IC87114 from 2.5–10mmol l�1 for 72 hours, and cell viability
assays of three independent experiments were performed. (d) p110a suppression and rapamycin show additive effects in inhibition of melanoma cell viability.
Mel-Juso (white bars) and 518A2 (black bars) cells were transfected with a small interfering RNA (siRNA) against p110a or control siRNA and treated with
rapamycin (all 50 nmol l�1) for 24 hours, and cell viability assays of three independent experiments were performed. Asterisks indicate significant differences
between combination and respective single-agent treatments (Po0.05). (e) Combination of PIK3CA siRNA and rapamycin suppresses phosphorylation of AKT
(phosphorylation-specific western blot). Mel-Juso and 518A2 cells were treated wit PIK3CA siRNA, control siRNA, and rapamycin (all 50 nmol l�1) for 24 hours.
Bands that were not originally run next to each other are separated by a dividing line. ERK, extracellular signal-regulated kinase.
www.jidonline.org 499
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
that vertical inhibition of the PI3K/mTORC1/mTORC2signaling pathway with two distinct small molecule inhibitorsshowed superior efficacy compared with the effects of eitheragent alone.
In our melanoma model, relatively high PI-103 concen-trations (between 0.5 and 1 mmol l�1) were needed tosuppress AKT phosphorylation compared with other in vitro
models such as Kaposi’s sarcoma (Chaisuparat et al., 2008).This relative resistance of the melanoma cells in our studywas overcome when we combined PI-103 with rapamycin.Interestingly, low concentrations of PI-103 even led toincreases in p-AKT phosphorylation in both cell lines(Figure 1b). This effect has also been observed with theprototypic PI3K/mTOR inhibitor LY294002 (Werzowa et al.,
**
Mou
se w
eigh
t (g)
Days
30
25
20
15
10
5
05 9 13 16 19 23 26 30 331 3
*
Tum
or s
ize
(mm
3 )
4,000a
c
d
e
b
3,000
2,000
1,000
0
p-AKT (S473)
AKT
p-S6 (S240/244)
S6
Actin
PI-103+ + ++++– – –– – –++++++ – – –– – – Rapamycin
Nor
mal
ized
ban
d in
tens
ity, p
-AK
T
p-AKT pS6
Rapa RapaCon Con
Comb CombPI-103 PI-103
Control Rapamycin PI-103 Combination
3.5
2.5
1.5
0.5
0
3
2
1
Contro
l
Rapam
ycin
Rapam
ycin
+ PI-1
03PI-103
Figure 4. Combination of rapamycin and PI-103 leads to synergistic effects in vivo. (a) Tumor volume (in mm3) of all the four treatment groups at the end of the
study. Each treatment group consisted of six animals. 518A2 cells were injected subcutaneously into athymic nude mice. Mice were treated five times weekly
with PI-103 (20 mg kg�1), rapamycin (1 mg kg�1), PI-103 (20 mg kg�1) plus rapamycin (1 mg kg�1), and solvent control until abortion criteria in any group were
reached (day 33). Asterisks indicate statistically significant differences (Po0.05). (b) Mouse weight (in grams) of all the four treatment groups from day 1
(treatment start) to the end of the study. (c) Cell lysates from excised tumors were analyzed using phospho-specific western blot analysis. (d) p-AKT band
intensities were measured using densitometry and were normalized with the respective actin bands. (e) Immunohistochemical staining of xenograft tumors
for p-AKT (S473) and pS6 (S240þ244). Bar¼0.1 mm.
500 Journal of Investigative Dermatology (2011), Volume 131
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
2009). A possible explanation for this may be that at lowerconcentrations, the PI3K and mTORC2 inhibition is notsufficient to block the activating effect on AKT of mTORC1inhibition.
When PI-103 and rapamycin were combined in vitro, thephosphorylation of AKT was suppressed even at low PI-103concentrations (Figure 2b). Synergistic suppression of cellviability occurred even at much lower PI-103 concentrations(Figure 2a), indicating that suppression of AKT phosphory-lation alone cannot explain the synergistic effects of thecombination. 518A2 cells showed less response to PI-103 interms of reduction in AKT phosphorylation, but still theeffects on cell viability were greater. In this context, it isnoteworthy that the cell lines used differ in their PTEN status:Mel-Juso cells are PTEN wild-type whereas 518A2 cells carrya mutation (Wu et al., 2003; Werzowa et al., 2009). 607Band A375 cells are PTEN wild-type and these cells were alsoless sensitive to PI-103 compared with SK-MEL-28 cells,which carry mutated PTEN (Tsao et al., 2000; Thallingeret al., 2007). Overall, suppression of AKT phosphorylationalone may not be an ideal biomarker for effectiveness of thisclass of inhibitors, at least not in vitro.
The synergistic effects of PI-103 and rapamycin in vitroand in vivo might be explained by the fact that these twoinhibitors exert their inhibitory actions via distinct molecularmechanisms. Whereas PI-103 blocks the ATP-bindingdomain of its targets, rapamycin exerts its action via a uniqueand highly specific allosteric mechanism. First, it binds itsintracellular binding partner FK506 binding protein and thenthe rapamycin-binding domain of mTORC1 probably result-ing in an inhibition of substrate binding (Banaszynski et al.,2005; Bain et al., 2007). It is therefore tempting to speculatethat the combination of these two distinct modes of action ledto optimal suppression of mTORC1 and PI3K class 1A, withsynergistic effects in vitro and in vivo. It also appears unlikelythat other targets of PI-103 such as DNA-PK are responsiblefor the observed effects, as GDC-0941, a highly specificinhibitor of p110a, had similar effects on cell viability andprotein phosphorylation (Figure 3a and b) at even lowerconcentrations than PI-103. Importantly, GDC-0941 has littleactivity against mTOR underscoring the fact that p110a butnot mTOR inhibition is mainly responsible for the anti-tumoreffects of PI-103. In general, mTOR inhibition seems lessimportant than p110a inhibition at least in cells with anoncogenic mutation in PI3K signaling (Sutherlin et al., 2010;Workman et al., 2010). GDC-0941 is currently undergoingphase I clinical trials as anti-cancer drug (Workman et al.,2010). In our study, the combination of p110a inhibition withthe allosteric mTORC1 inhibitor rapamycin yielded synergis-tic effects on target protein phosphorylation and cell viability.These data suggest that combination of rapamycin with GDC-0941 seems to be a highly promising strategy for further pre-clinical and clinical evaluation.
The results obtained with the siRNA directed againstPIK3CA support the finding that p110a is a relevant targetin melanoma cells and that combination with rapamycincan augment the effects on viability and protein phosphory-lation. Still, the effects of the siRNA were weaker than
the effects of PI-103 and GDC-0941, which might reflect theincomplete downregulation of p110a protein levels byPIK3CA siRNA.
Taken together, our study underscores the importance ofthe PI3K/mTORC1/mTORC2 pathway in melanoma anddemonstrates that rational combination of compounds thatlead to an optimal blockade of a critical pathway (‘‘verticalinhibition’’) may provide an effective strategy for futuretreatment of melanoma.
MATERIALS AND METHODSReagents
PI-103 was kindly provided by Genentech (San Francisco, CA). For
experiments, fresh PI-103 solutions in DMSO were prepared before
use. We obtained rapamycin from Alexis Biochemicals (now Enzo
Life Sciences AG, Lausen, Switzerland). A 100mmol l�1 stock
solution in DMSO was prepared and stored at �20 1C. Sirolimus
solution (Rapamune) was purchased from Wyeth Pharma (Madison,
NJ) and diluted to the desired concentration in PEG200. IC87114
and GDC-0941 were obtained from Selleck Chemicals LLC
(Houston, TX), dissolved in DMSO, and diluted to the desired final
concentration using cell culture media.
Cell lines
The human melanoma cell lines 518A2 and 607B were obtained
from P. Schrier (University of Leiden, The Netherlands). The cell line
Mel-Juso was kindly provided by Dr Judith Johnson (University of
Munich, Germany). A375 and SK-MEL-28 cells were obtained from
the American Type Culture Collection (ATCC; Manassas, VA). All
cells were cultured in Dulbecco’s modified eagle medium (Gibco
Invitrogen, Paisley, UK) with 4500 mg l�1 glucose, L-glutamine, and
pyruvate supplemented with 10% fetal calf serum, and an antibiotic
mixture containing penicillin, streptomycin, and amphotericin B
(Gibco Invitrogen) in a fully humidified 5% CO2, 95% ambient air
atmosphere at 37 1C.
Cell viability assay
The Cell Titer Blue cell viability assay (Promega, Madison, WI) was
used to quantify the fraction of viable cells in our samples. Tumor
cells in exponential growth were harvested and seeded at 2� 103
cells per well (0.1 ml) in 96-well plates and were incubated
overnight at 37 1C. The cells were then incubated for 24, 48, and
72 hours with the respective inhibitor. DMSO at the appropriate
concentrations was used as control. After treatment, 20 ml of cell titer
blue reagent was added and the reaction mixtures were incubated
for 3 hours at 37 1C. Color reactions were measured using a
fluorescence detection system (Victor 1620 Multilabel Detector,
Wallac/Perkin Elmer, Wellesley, MA).
siRNA transfection
In all, 7.5� 104 cells were seeded 1 day before transfection in six-
well plates in cell culture medium without antibiotics. Cells were
transfected with 50 nmol l�1 Silencer Select siRNA against PIK3CA
(p110a catalytic subunit of PI3K; Ambion, Austin, TX) in the
presence of Saint-Red siRNA delivery system according to the
manufacturer’s instructions (Synvolux, Groningen, The Netherlands).
As control, a Silencer Select Negative Control at 50 nmol l�1 was
employed (Ambion/Applied Biosystems, Austin, TX).
www.jidonline.org 501
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
Western blottingCell extracts were prepared in lysis buffer containing 100 mmol l�1
NaCl, 0.1% SDS, 1% Nonidet P40, 50 mM Tris, pH 7.4, 10 mmol l�1
EDTA, supplemented with 10 mmol l�1 p-nitrophenolphosphate,
40 mmol l�1 b-glycerophosphate, and Complete protease inhibitor
cocktail tablets (Roche, Mannheim, Germany). The amount of
soluble proteins was quantified by means of modified Bradford
analysis (Bio-Rad, Richmond, CA). Total lysates (10 mg per lane)
were separated by electrophoresis through a 7.5, 12, or 15% SDS
polyacrylamide gel and were blotted onto nitrocellulose membranes
(Whatman, Dassel, Germany). Membranes were blocked with 0.2%
I-Block (Tropix, Bedford, MA) in 1� Tris-buffered saline plus 0.1%
Tween 20 and incubated with the primary antibodies at 4 1C
overnight. The primary antibody against actin was purchased from
Sigma (St Louis, MI). The antibodies against total and phospho-
(Ser473) and phospho-(T308) AKT, total and phospho-(Ser240þ 244) S6,
total and phospho-(Thr202/Thyr204) p44/42 mitogen-activated protein
kinase, and PI3 kinase p110a were obtained from Cell Signaling
Technology (Beverly, MA). The membranes were washed and
incubated again for 30 minutes at room temperature with alkaline
phosphatase-conjugated goat anti-mouse and goat anti-rabbit
immunoglobulins (Tropix). The bound antibody was detected using
an enhanced chemiluminescence reagent (Tropix). Band intensities
were quantified using TotalLab software (Nonlinear Dynamics,
Newcastle upon Tyne, UK).
Apoptosis assay
For quantification of apoptosis, a PE-conjugated monoclonal
Apo2.7 antibody (Beckman-Coulter, Fullerton, CA) was used.
After treatment, 0.5–1.0� 106 cells were detached from six-well
plates using accutase (PAA, Linz, Austria) and then washed twice
with ice-cold Dulbecco’s phosphate-buffered saline (DPBS; Lonza,
Vervier, Belgium). Cells were permeabilized with cold (4 1C)
digitonin (Sigma) at 100mg ml�1 in DPBSF (DPBS with 2.5% fetal
calf serum (v/v) and 0.01% NaN3 (w/v)) for 20 minutes. After
washing, 20 ml of Apo2.7-PE and 80 ml of DPBSF were added and
the mixture was incubated for 15 minutes at room temperature in
the dark. After a washing step, the cells were suspended in DPBSF
and analyzed using the FACScalibur device (BD biosciences,
San Jose, CA).
Cell-cycle distribution
For cell-cycle measurements, cells were seeded in six-well plates
(80,000 per well), treated with therapeutics of interest, harvested in
log-phase, and detached with Trypsin (Gibco Invitrogen). The
suspension was centrifuged at 2500 r.p.m. for 5 minutes at room
temperature, pellets were washed once with DPBS, resuspended in
200ml DPBS, and fixed by adding 2 ml of ice-cold ethanol-DPBS mix
(70:30). After 30 minutes on ice, fixed cells were centrifuged at
1200 r.p.m. for 5 minutes at room temperature, supernatant was
discarded, cells were resuspended in 800ml DPBS, 10 ml of RNase A
was added, cells were stained with 10 ml of propidium iodide and
incubated for at least 30 minutes at 37 1C. Following this procedure,
samples were protected from light and analyzed directly with the
FACScalibur device.
FACS data were further analyzed using ModFIT software (Verity
software house, Topsham, ME) to determine the relative amounts of
cell-cycle phases.
Xenografts518A2 human melanoma cells (107) were resuspended in DPBS and
injected subcutaneously into lower flanks on both sides of 8- to 10-
week-old female athymic nude mice (Harlan Winkelmann, Borchen,
Germany). Mice with established tumors (50–100 mm3) were
randomly allocated to treatment with PI-103 (20 mg kg�1 per day
in DMSO, administered intraperitoneally), sirolimus (1 mg kg�1 per
day in PEG200, administered per os by oral gavage), combination of
the two or solvent control (PEG200 plus DMSO). Animals were
treated five times a week and tumors were measured with callipers
twice a week. Volumes were measured from six mice on both sides
for each data point (mm3¼ 4/3� 3.142� (width2 � length/2).
When tumor size in control group reached abortion criteria (day 33
after initiation of treatment), all animals were killed and tumors were
snap frozen for extraction of protein lysates. The study protocol was
approved by the Medical University of Vienna Animal Welfare
Committee. All animal studies were performed according to the
Austrian laws and guidelines for animal care and protection.
Immunohistochemistry
Paraffin-embedded mouse tissue sections were stained for p-AKT
(Ser473) and p-S6 (Ser240þ 244; Cell Signaling Technology) and a
DAKO LSAB 2 system.
Statistical analysis
Data were expressed as the mean ± SD. The statistical analysis was
done using Student’s t-test. For multiple comparisons, analysis of
variance was followed by the least significant difference post-hoc
test. P-values o0.05 were considered statistically significant.
Statistical analysis was performed using SPSS 13 software (SPSS,
Chicago, IL). For evaluation of synergistic effects and calculation of
combination indices, CalcuSyn Version 1.1 Software (Biosoft,
Cambridge, UK) was used.
CONFLICT OF INTERESTThe authors state no conflict of interest.
SUPPLEMENTARY MATERIAL
Supplementary material is linked to the online version of the paper at http://www.nature.com/jid
REFERENCES
Ali K, Camps M, Pearce WP et al. (2008) Isoform-specific functions ofphosphoinositide 3-kinases: p110 delta but not p110 gamma promotesoptimal allergic responses in vivo. J Immunol 180:2538–44
Bain J, Plater L, Elliott M et al. (2007) The selectivity of protein kinaseinhibitors: a further update. Biochem J 408:297–315
Banaszynski LA, Liu CW, Wandless TJ (2005) Characterization of theFKBP.rapamycin.FRB ternary complex. J Am Chem Soc 127:4715–21
Bjornsti MA, Houghton PJ (2004) The TOR pathway: a target for cancertherapy. Nat Rev Cancer 4:335–48
Chaisuparat R, Hu J, Jham BC et al. (2008) Dual inhibition of PI3Kalpha andmTOR as an alternative treatment for Kaposi’s sarcoma. Cancer Res68:8361–8
Chen XG, Liu F, Song XF et al. (2010) Rapamycin regulates Akt and ERKphosphorylation through mTORC1 and mTORC2 signaling pathways.Mol Carcinog 49:603–10
Cheng JQ, Lindsley CW, Cheng GZ et al. (2005) The Akt/PKB pathway:molecular target for cancer drug discovery. Oncogene 24:7482–92
502 Journal of Investigative Dermatology (2011), Volume 131
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
Corvera S, Czech MP (1998) Direct targets of phosphoinositide 3-kinase productsin membrane traffic and signal transduction. Trends Cell Biol 8:442–6
Denley A, Kang S, Karst U et al. (2007) Oncogenic signaling of class I PI3Kisoforms. Oncogene 27:2561–74
Fan QW, Knight ZA, Goldenberg DD et al. (2006) A dual PI3 kinase/mTORinhibitor reveals emergent efficacy in glioma. Cancer Cell 9:341–9
Foukas LC, Claret M, Pearce W et al. (2006) Critical role for the p110alphaphosphoinositide-3-OH kinase in growth and metabolic regulation.Nature 441:366–70
Hay N (2005) The Akt-mTOR tango and its relevance to cancer. Cancer Cell8:179–83
Hudes G, Carducci M, Tomczak P et al. (2007) Temsirolimus, interferon alfa,or both for advanced renal-cell carcinoma. N Engl J Med 356:2271–81
Karbowniczek M, Spittle CS, Morrison T et al. (2007) mTOR is activated in themajority of malignant melanomas. J Invest Dermatol 128:980–7
Kharas MG, Janes MR, Scarfone VM et al. (2008) Ablation of PI3K blocksBCR-ABL leukemogenesis in mice, and a dual PI3K/mTOR inhibitorprevents expansion of human BCR-ABL+ leukemia cells. J Clin Invest118:3038–50
Knight ZA, Gonzalez B, Feldman ME et al. (2006) A Pharmacological mapof the PI3-K family defines a role for p110alpha in insulin signaling.Cell 125:733–47
Lopez-Fauqued M, Gil R, Grueso J et al. (2009) The dual PI3K/mTOR inhibitor(PI-103) promotes immunosupression, in vivo tumor growth and increasessurvival of sorafenib treated melanoma cells. Int J Cancer 126:1549–61
Margolin K, Longmate J, Baratta T et al. (2005) CCI-779 in metastaticmelanoma. Cancer 104:1045–8
Marone R, Erhart D, Mertz AC et al. (2009) Targeting melanoma with dualphosphoinositide 3-kinase/mammalian target of rapamycin inhibitors.Mol Cancer Res 7:601–13
Meier F, Schittek B, Busch S et al. (2005) The RAS/RAF/MEK/ERK and PI3K/AKT signaling pathways present molecular targets for the effectivetreatment of advanced melanoma. Front Biosci 10:2986–3001
O’Reilly KE, Rojo F, She QB et al. (2006) mTOR inhibition induces upstreamreceptor tyrosine kinase signaling and activates Akt. Cancer Res 66:1500–8
Raynaud FI, Eccles SA, Patel S et al. (2009) Biological properties of potentinhibitors of class I phosphatidylinositide 3-kinases: from PI-103 throughPI-540, PI-620 to the oral agent GDC-0941. Mol Cancer Ther 8:1725–38
Samuels Y, Ericson K (2006) Oncogenic PI3K and its role in cancer. Curr OpinOncol 18:77–82
Soulard A, Hall MN (2007) SnapShot: mTOR signaling. Cell 129:434
Stahl JM, Sharma A, Cheung M et al. (2004) Deregulated Akt3 activitypromotes development of malignant melanoma. Cancer Res 64:7002–10
Stambolic V, Suzuki A, de la Pompa JL et al. (1998) Negative regulationof PKB/Akt-dependent cell survival by the tumor suppressor PTEN.Cell 95:29–39
Sutherlin DP, Sampath D, Berry M et al. (2010) Discovery of (thienopyr-imidin-2-yl)aminopyrimidines as potent, selective, and orally availablepan-PI3-kinase and dual pan-PI3-kinase/mTOR inhibitors for the treat-ment of cancer. J Med Chem 53:1086–97
Takeuchi H, Kondo Y, Fujiwara K et al. (2005) Synergistic augmentation ofrapamycin-induced autophagy in malignant glioma cells by phosphati-dylinositol 3-kinase/protein kinase B inhibitors. Cancer Res 65:3336–46
Thallinger C, Werzowa J, Poeppl W et al. (2007) Comparison of a treatmentstrategy combining CCI-779 plus DTIC versus DTIC monotreatment inhuman melanoma in SCID mice. J Invest Dermatol 127:2411–7
Tsao H, Zhang X, Fowlkes K et al. (2000) Relative reciprocity of NRAS andPTEN/MMAC1 alterations in cutaneous melanoma cell lines. Cancer Res60:1800–4
Werzowa J, Cejka D, Fuereder T et al. (2009) Suppression of mTORC2dependent AKT phosphorylation in melanoma cells by combinedtreatment with rapamycin and LY294002. Br J Dermatol 160:955–64
Workman P, Clarke PA, Raynaud FI et al. (2010) Drugging the PI3 kinome:from chemical tools to drugs in the clinic. Cancer Res 70:2146–57
Wu H, Goel V, Haluska FG (2003) PTEN signaling pathways in melaoma.Oncogene 22:3113–22
www.jidonline.org 503
J Werzowa et al.Vertical mTORC1/mTORC2/PI3K Inhibition in Melanoma
