Immunohistochemistry Antibody Validation Report for Anti-Akt1 Antibody (STJ91542)
1 Akt1 and Akt3 exert opposing roles in the regulation of vascular ...
Transcript of 1 Akt1 and Akt3 exert opposing roles in the regulation of vascular ...
1
Akt1 and Akt3 exert opposing roles in the regulation of vascular tumor growth
Thuy L. Phung1#, Wa Du1, Qi Xue2, Sriram Ayyaswamy1, Damien Gerald2, Zeus Antonello3,
Sokha Nhek2, Carole A. Perruzzi2, Isabel Acevedo3, Rajesh Ramanna-Valmiki1, Paul Rodriguez-
Waitkus1, Ladan Enayati1, Marcelo L. Hochman4, Dina Lev5, Sandaruwan Geeganage6 and
Laura E. Benjamin2,3,6#
1Department of Pathology, Texas Children’s Hospital and Baylor College of Medicine, Houston,
Texas, USA; 2ImClone Systems, New York, NY, USA; 3Department of Pathology, Beth Israel
Deaconess Medical Center, Boston, Massachusetts, USA; 4Hemangioma International
Treatment Center, Charleston, South Carolina, USA; 5Department of Cancer Biology, MD
Anderson Cancer Center, Houston, Texas, USA; and 6Eli Lilly and Company, Indianapolis, IN,
USA
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
2
Running title: Differential regulation of vascular tumors by Akt isoforms
Key words: Akt3, vascular tumors, mTOR, S6-Kinase
Funding: This work was supported in part by NIH grants R01 CA106263 and P01 CA09264401
(LEB); American Heart Association 11BGIA5590018, NIH K08 HL087008, NIH R03 AR063223,
American Cancer Society 122019-RSG-12-054-01-CSM and the Baylor Clinical and
Translational Research Program (TLP).
# Corresponding authors:
Laura E. Benjamin, PhD
ImClone Systems, a wholly-owned subsidiary of Eli Lilly
450 East 29th Street, 12th Floor
New York, NY 10016, USA
Phone: 646-638-6381
Email: [email protected]
Thuy L. Phung, MD, PhD
Texas Children’s Hospital
1102 Bates Avenue, Suite 830
Houston, TX 77030, USA
Phone: 832-824-5202
Email: [email protected]
Potential conflict of interest: QX, SN, DG, CAP, SG and LEB are employees of ImClone
Systems, a wholly-owned subsidiary of Eli Lilly and Company. The other authors report no
conflict of interest.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
3
ABSTRACT
Vascular tumors are endothelial cell neoplasms whose mechanisms of tumorigenesis
are poorly understood. Moreover, current therapies, particularly those for malignant lesions,
have little beneficial effect on clinical outcomes. In this study, we show that endothelial
activation of the Akt1 kinase is sufficient to drive de novo tumor formation. Mechanistic
investigations uncovered opposing functions for different Akt isoforms in this regulation, where
Akt1 promotes and Akt3 inhibits vascular tumor growth. Akt3 exerted negative effects on tumor
endothelial cell growth and migration by inhibiting activation of the translation regulatory kinase
S6K through modulation of Rictor expression. S6K in turn acted through a negative feedback
loop to restrain Akt3 expression. Conversely, S6K signaling was increased in vascular tumor
cells where Akt3 was silenced, and the growth of these tumor cells was inhibited by a novel S6K
inhibitor. Overall, our findings offer a preclinical proof-of-concept for the therapeutic utility of
treating vascular tumors, such as angiosarcomas, with S6K inhibitors.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
4
INTRODUCTION
Vascular tumors are endothelial cell (EC) neoplasms with a wide spectrum of clinical
presentations, ranging from benign infantile hemangiomas in children to low-grade malignant
hemangioendotheliomas and highly aggressive angiosarcomas in adults. To date, the
molecular pathogenesis of vascular tumors is poorly understood and current therapies,
particularly those for malignant vascular tumors, do not significantly improve patient outcome
(1).
Akt is a major signaling pathway activated by vascular endothelial growth factor (VEGF)
that regulates EC survival (2). In infantile hemangioma (IH), hemangioma-derived endothelial
cells (hemeEC) have constitutively active VEGF receptor-2 signaling with high phosphorylation
levels of ERK1/2 and Akt (3). Human angiosarcoma (AS) expresses VEGF-A and the VEGF
receptors (4). We have shown increased phosphorylation of Akt and 4E-BP1 in AS (5).
Hyperactivation of PI3-kinase results in hemangiosarcoma formation in chicken chorioallantoic
membrane (6). Akt1, Akt2 and Akt3 are isoforms that have shared as well as distinct functions
in cancer cells. Both Akt1 and Akt2 promote cancer cell survival and growth. However, in
breast and ovarian cancer, Akt1 decreases cell motility and metastasis and blocks epithelial-to-
mesenchymal phenotype, whereas Akt2 enhances these processes (7-9). Akt3 is preferentially
required for the growth of triple-negative breast cancer (10), and a gene fusion of Akt3 with
MAGI3 leads to constitutive Akt3 activation and is enriched in these tumors (11). Interestingly,
there is some evidence suggesting that Akt3 exerts inhibitory effects in cancer. N-Cadherin
promotes breast cancer metastasis by inhibiting Akt3, and Akt3 has been shown to inhibit lung
tumor growth in mice (12-14). Studies of animal models of breast cancer with simultaneous
deletion or overexpression of Akt1, Akt2 and Akt3 lend further support to Akt isoform-specific
roles in cancer (8, 9).
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
5
Mammalian target of rapamycin (mTOR) complex-1 (mTORC1) and complex-2
(mTORC2) are composed of multiple subunits, including mTOR and Raptor (in mTORC1), and
mTOR and Rictor (in mTORC2) (15, 16). Akt activates mTORC1, which phosphorylates the
translational regulators 4E-BP1 and p70 S6-Kinase (S6K). S6K in turn activates S6 ribosomal
protein (S6) (17, 18). mTORC2 directly activates Akt by phosphorylating it at serine 473,
thereby exerting feedback regulation on the Akt signaling pathway (15). The S6K pathway is
important in protein synthesis and cell growth, and acts as a regulator of actin cytoskeleton
dynamics in cell migration (19, 20).
In this study, we showed that endothelial Akt1 drives vascular tumor growth.
Importantly, we have uncovered the opposing functions of Akt1 and Akt3 in the regulation of
tumor growth, which is mediated through S6K, and found a novel negative feedback regulation
on Akt3 by S6K. We also demonstrated the clinical utility of a novel S6K inhibitor in the
treatment of vascular lesions.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
6
MATERIALS AND METHODS
Animals. Animal studies were conducted in compliance with the Beth Israel Deaconess
Medical Center (BIDMC) Institutional Animal Care and Use Committee guidelines. Double
transgenic myristoylated Akt1 mice in mixed FVB genetic background have been previously
described (21). C57 Bl/6 Akt3-/- mice were from Argiris Efstratiadis and Morris Birnbaum (22).
Cell lines and reagents. The use of human tissues was approved by the Institutional Review
Boards at BIDMC and Baylor College of Medicine. Primary human dermal microvascular
endothelial cells (EC), infantile hemangioma and mouse EC were isolated as described (21, 23).
ASM.5 cells were from Vera Krump-Konvalinkov and EOMA cells were from ATCC as
previously published (23-25). Cell line authentication and validation by short tandem repeats
was performed. HA-tagged-MyrAkt3 and constitutively active S6K (R3A) have been described
(26, 27). LY2584702 (Eli Lilly, Inc.) was prepared in 0.25% Tween-80 and 0.05% antifoam, and
administered orally to mice (12.5 mg/kg twice daily). All antibodies used were from Cell
Signaling Technologies, except for antibodies to smooth muscle actin, β-actin and -tubulin
(Sigma), CD31 (BD Biosciences) and glucose transporter-1 (Dr. Morris Birnbaum).
Mouse hemangioma skin graft model. Ten-mm circular pieces of flank skin from donor
myrAkt1 mice were grafted onto the back of recipient mice with absorbable Vicryl sutures
(Ethicon). Recipient animals were maintained on 1.5 mg/ml tetracycline in the drinking water as
described (21) to turn off myrAkt1 while the grafts were left to heal for 2 weeks. After this time,
half of the recipients continued to receive tetracycline for 4 weeks, while the other half was
taken off tetracycline to turn on myrAkt1 expression.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
7
Immunoprecipitation and western blots. For Akt isoform immunoprecipitations, cells were
lysed in RIPA buffer with protease/phosphatase inhibitors. Cell lysates were precleared with
Protein A/G Agarose, 1 hour, 40C, followed with addition of primary antibodies to 0.8 mg protein
lysates and rotated overnight, 40C. Protein A/G Agarose was then added to the lysates and
rotated for 2 hours, 40C. Immunoprecipitated proteins were denatured in Laemmli buffer and
analyzed by western blotting as described (23).
Tumor growth. EOMA cells (0.3 x 106) were injected subcutaneously in 6-8 week old nu/nu
female mice (2 sites/mouse, 4-5 mice/group). Tumor size was measured daily. For drug
treatment, when tumors reached 0.01 cm3 in size, the animals were treated with vehicle control
or LY2584702 (12.5 mg/kg twice daily, oral dosing). Tumor size was measured every 3-4 days.
Lentiviral shRNA and quantitative real-time PCR. shRNA clones used are listed in
Supplementary Table S2. qPCR primers are provided in Supplementary Materials and
Methods. Lentivirus packaging and qPCR were performed as described (23).
Statistical analysis. Data were presented as mean ± SD. The difference between multiple
experimental groups was assessed by one-way ANOVA, and the difference between multiple
experimental groups at different time points (two independent variables) was assessed by two-
way ANOVA using GraphPad Prism.
Other standard reagents and methods are provided in Supplementary Materials and
Methods.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
8
RESULTS
Akt is activated in human vascular tumors. Clinical specimens of 17 benign infantile
hemangioma (IH), 6 kaposiform hemangioendothelioma (KHE), 16 Kaposi’s sarcoma (KS) and
9 angiosarcoma (AS) were stained for phosphorylated Akt Serine 473 (pAkt S473). All of the
tumors expressed increased pAkt levels compared to adjacent normal blood vessels in the
same tissue sections (Figure 1A). We also compared the levels of pAkt in these tumors with
blood vessels in 20 normal skin specimens as the normal counterpart of neoplastic vessels.
The percent positively stained tumor cells (stain reactivity) and the stain intensity per group
were calculated. Higher pAkt stain reactivity was seen in vascular tumors than in normal skin
(73.3±2.1% in normal skin vs. 92.3±1.3%* in IH, 95.5±1.2%* in KHE, 92.3±1.3%* in KS and
89.7±1.9%* in AS; *P<0.0001) (Figure 1B). To evaluate the stain intensity, the staining was
scored using a 3-tier system (1, low; 2, moderate; and 3, high) by two pathologists and the
average score was graphed. Representative pictures of low and high stain intensity are shown
(Supplementary Figure S1). Significantly higher pAkt stain intensity was seen in tumors (1.6 ±
0.1 in normal skin vs. 2.1 ± 0.2* in IH, 2.2 ± 0.3* in KHE, 2.5 ± 0.2* in KS and 3.0 ± 0.03* in AS;
*P<0.01) (Figure 1C and Supplementary Figure S1). These results showed increased Akt
activation across different types of human vascular tumors.
To determine whether Akt is hyperactivated in neoplastic EC, we focused on IH, which is
a common soft tissue tumor of infancy and fresh tissues are available for studies. We purified
hemangioma-derived endothelial cells (hemeEC) from IH using CD31-magnetic bead isolation
and stained for endothelial markers (Supplementary Figure S2A). HemeEC showed a 1.9-fold
increase in pAkt as compared with normal human dermal microvascular endothelial cells
(HDMEC) by western blot (Figure 1D). Consistent with a previous report of constitutive VEGFR-
2 activation in IH (3), we observed increased phosphorylated VEGFR-2 in a subset of hemeEC
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
9
samples. Interestingly, PTEN levels were also reduced in some hemangiomas. These findings
showed increased Akt activation in IH, which is associated with decreased PTEN and increased
VEGFR-2 activation.
We next examined Akt activation in malignant vascular tumors. We utilized ASM.5 cells
isolated from a spontaneous human angiosarcoma (24), and EOMA cells derived from a
spontaneous mouse hemangioendothelioma (25). ASM.5 and EOMA cells had increased Akt
activation as compared with normal EC (Figures 1E-F) (1.0 ± 0.2 in HDMEC vs. 2.8 ± 1.1* in
ASM.5, *P<0.05, N=3), (1.0 ± 0.5 in mouse EC vs. 6.7 ± 2.0* in EOMA, *P<0.05, N=4).
Furthermore, lower levels of PTEN were seen in both cell types. These findings indicate
common aberrations in the PTEN/Akt pathway in both benign and malignant vascular tumors.
Akt1 promotes the growth and migration of vascular tumor cells. HemeEC and ASM.5
cells were transduced with lentiviral short-hairpin RNA to Akt1 (shAkt1). Significant Akt1
knockdown was achieved using two independent shAkt1 clones as compared with pLKO
scramble shRNA control (Figures 2A-B). Akt1 knockdown significantly reduced the growth of
hemeEC and ASM.5 cells as assessed by cellular DNA content (Figures 2C-D). Since Akt1 is
known to affect cell survival (28), we determined the effects of shAkt1 on apoptosis. Akt1
knockdown resulted in increased tumor cell apoptosis in response to serum-starvation as
determined by the apoptotic marker Annexin V (Supplementary Figures S2B-C). Akt1
knockdown decreased basal and VEGF-stimulated migration of hemeEC and ASM.5 cells in
Boyden chamber migration assay (Figures 2E-F). Akt1 knockdown also significantly inhibited
basal and VEGF-induced cord formation in hemeEC (Figures 2G-H). These findings showed
that Akt1 plays a key role in promoting the survival and migration of vascular tumor cells.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
10
Endothelial Akt1 activation induces hemangioma formation in mice. To determine whether
Akt1 activation in EC is sufficient to drive de novo vascular tumor formation, we utilized a double
transgenic mouse model that expresses tetracycline-inducible and endothelial cell-specific
activated myristoylated Akt1 (myrAkt1) (21). myrAkt1 transgene induction increased Akt1
expression by 3.9-fold, but did not affect Akt2 and Akt3 levels in EC isolated from these mice
(Supplementary Figures S3A-B). myrAkt1 mice have systemic pathological angiogenesis, but
they do not develop vascular tumors due to shortened lifespan from systemic edema (21). To
study the long-term effects of endothelial Akt1 activation, we developed a skin graft model in
which 10-mm circular pieces of skin from myrAkt1 donors (FVB background) were transplanted
onto the back skin of immunodeficient nu/nu mice (Figure 3A). Recipient animals were
maintained on tetracycline to turn off myrAkt1 while the grafts were left to heal for 2 weeks.
After this time, animals were either kept on tetracycline (myrAkt1 expression off) or taken off
tetracycline (myrAkt1 on) for 4 weeks. Without myrAkt1 expression, a dermal scar was present
but no tumor was seen (Figure 3B). However, induction of myrAkt1 resulted in the development
of red tumor masses at the graft sites. The tumors measured 0.39 ± 0.14 cm3 in size and had
histologic features consistent with benign hemangiomas (Figure 3C). The tumor vessels co-
expressed the endothelial marker CD31 and HA-tagged myrAkt1 (Figure 3D). Abundant pAkt
was seen in tumor vessels, some of which contained a covering of smooth muscle actin (SMA)
while many were SMA-deficient. Interestingly, the expression of glucose transporter-1 (Glut-1),
a biomarker found uniquely in human IH (29), was seen in CD31+ vessels. We also performed
myrAkt1 skin grafts in syngeneic immunocompetent FVB recipients and observed similar tumor
development in these mice with tumors measuring 0.49 ± 0.21 cm3 in size (Supplementary
Figures S3C-D).
Sustained endothelial Akt1 activation is necessary to maintain hemangioma growth. Skin
graft transplantation was performed and myrAkt1 expression was turned on to allow
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
11
hemangioma to develop. After 4 weeks, myrAkt1 expression was turned off in half of the
animals while myrAkt1 expression continued to be on in the other half. Tumors in animals with
sustained myrAkt1 expression continued to grow, but those in animals with myrAkt1 turned off
regressed dramatically over the next 3.5 weeks and had few vessels and more fibrofatty tissue
(Figures 3E-F). Thus, sustained endothelial Akt1 activation is required to maintain the integrity
of blood vessels in this vascular tumor model, and supports our previous finding of the
“plasticity” of the microvasculature in response to Akt1 signaling (21).
Akt3 expression is reduced in vascular tumors. Immunostains for Akt1, Akt2 and Akt3 in 15
IH and 10 AS samples showed that the levels of Akt1 and Akt2 in these tumors were similar to
those present in adjacent normal blood vessels in the same tissue sections (Figure 4A). By
contrast, Akt3 levels were reduced in tumor tissues. We next compared the levels of Akt
isoforms in vascular tumors with the blood vessels in 6 normal human skin specimens (Figure
4B). The stain reactivity and stain intensity were calculated as described for pAkt stains in
Figure 1. Representative pictures of low and high Akt isoform stain intensity are shown in
Supplementary Figure S4 at different magnifications. The stain reactivity (%) and stain intensity
for endogenous Akt1 were similar in normal skin and vascular tumors (for Akt1 stain reactivity,
85.6 ± 4.4% in normal skin vs. 77.0 ± 5.4% in IH and 87.7 ± 1.5% in AS; P-values = not
significant. For Akt1 stain intensity, 2.3 ± 0.4 in normal skin vs. 1.5 ± 0.2 in IH and 2.1 ± 0.3 in
AS; P-values = not significant). Likewise, endogenous Akt2 levels were similar in normal skin
and vascular tumors (for Akt2 stain reactivity, 78.0 ± 13.5% in normal skin vs. 89.1 ± 1.7% in IH
and 86.0 ± 3.2% in AS; P-values = not significant. For Akt2 stain intensity, 1.8 ± 0.5 in normal
skin vs. 2.3 ± 0.2 in IH and 2.3 ± 0.2 in AS; P-values = not significant). In contrast, endogenous
Akt3 levels were reduced in vascular tumors as compared to normal skin (for Akt3 stain
intensity, 2.4 ± 0.3 in normal skin vs. 1.1 ± 0.1* in IH and 1.4 ± 0.2* in AS; *P<0.05). The stain
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
12
reactivity for Akt3 was lower in IH (88.7 ± 2.3% in normal skin vs. 69.5 ± 8.3%* in IH; *P<0.05),
but not in AS (88.7 ± 2.3% in normal skin vs. 87.8 ± 2.0% in AS; P-value = not significant).
Immunoblots for Akt3 in ASM.5 and EOMA cells showed that they had significantly lower
Akt3 levels than normal EC, which is consistent with the findings in patient tumor tissues
(Supplementary Figure S5). We did not observe a significant change in Akt3 in hemeEC. To
examine the phosphorylation status of Akt isoforms, ASM.5 cells were immunoprecipitated with
antibodies specific for Akt1, Akt2 and Akt3, and immunoblotted for each Akt isoform and
phospho-Akt. Akt1 and Akt3 were phosphorylated at both threonine 308 (T308) and S473
residues, sites that are required for full Akt activation (Figure 4C). Interestingly, Akt2 did not
appear to be phosphorylated in these studies, indicating that only Akt1 and Akt3 are the two
active isoforms in these tumor cells.
Akt1 promotes, whereas Akt3 inhibits vascular tumor growth. We assessed the effects of
loss of each Akt isoform on sprouting angiogenesis in IH. Since Akt2 does not appear to be
activated in vascular tumor cells (Figure 4C), we chose to focus on Akt1 and Akt3. Western blot
analysis and quantitation by densitometry showed effective knockdown of each Akt isoform in
hemeEC by shRNA (Supplementary Figure S6A). Of note, the levels of Akt3 appeared higher
than Akt1 and Akt2 in the blots. However, because the antibodies to detect Akt1, Akt2 and Akt3
were different antibodies, it would not be possible to cross compare the levels of one Akt
isoform to another by western blots.
In spheroid sprouting assay, loss of Akt1 reduced basal and VEGF-A-stimulated sprout
formation as compared with pLKO (Figures 5A-B). However, loss of Akt3 significantly increased
sprout formation under basal conditions and with VEGF-A using independent shRNAs. Similar
findings were observed in ASM.5 and EOMA cells. Effective knockdown of each Akt isoform
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
13
was achieved as shown by densitometric quantitation of western blots (Supplementary Figures
S6B-C). Compared with pLKO, loss of Akt1 reduced tumor cell migration in scratch wound
assays (Figures 5C-D). In contrast, loss of Akt3 significantly increased cell migration. Akt1 and
Akt3 also exert opposing effects on cell growth – loss of Akt1 reduced, whereas loss of Akt3
increased EOMA growth in vitro (Figure 5E).
To evaluate the functions of Akt isoforms in vivo, Akt1, Akt2 and Akt3 were knocked-
down in EOMA cells using independent shRNA clones for each isoform (Supplementary Figure
S6D). Cells were then injected subcutaneously in nu/nu mice and tumor size was monitored for
12 days. Loss of Akt1 reduced tumor growth; loss of Akt2 had no effect, whereas loss of Akt3
enhanced tumor growth (Figure 5F and Supplementary Figure S6E). These findings
demonstrate the opposing roles of Akt1 and Akt3 in vascular tumor growth.
The distinct functions of Akt1 and Akt3 are mediated through p70 S6-Kinase. To evaluate
downstream effectors of Akt isoforms, Akt1 and Akt3 were knocked-down in hemeEC and
EOMA cells and analyzed for Akt, S6K and S6. pAkt levels were normalized to total Akt levels
and calculated relative to that in pLKO. Knockdown of either Akt1 or Akt3 decreased pAkt in
hemeEC (Figure 6A) (1.0 ± 0.0 in pLKO vs. 0.6 ± 0.1* in shAkt1 and 0.7 ± 0.04* in shAkt3;
*P<0.05, N=4). Similar reduction in pAkt was found in EOMA cells (1.0 ± 0.0 in pLKO vs. 0.4 ±
0.02* in shAkt1 and 0.7 ± 0.1* in shAkt3; *P<0.05, N=3). These results indicate that both Akt1
and Akt3 contribute to the total pool of phosphorylated Akt in these cells.
Knockdown of Akt1 in hemeEC decreased levels of phosphorylated S6K (pS6K) and its
downstream effector phosphorylated S6 (pS6). By contrast, loss of Akt3 increased pS6K and
pS6 (Figure 6A). Quantitative analysis of western blots showed that pS6K levels in hemeEC
were 1.0 ± 0.0 in pLKO vs. 0.5 ± 0.2* in shAkt1 and 3.3 ± 0.5* in shAkt3; *P<0.05, N=3.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
14
Similarly, the levels of pS6 were 1.0 ± 0.0 in pLKO vs. 0.5 ± 0.1* in shAkt1 and 1.9 ± 0.3* in
shAkt3; P<0.05. We also evaluated EOMA cells and found similar opposing effects of Akt1 and
Akt3 on S6K and S6 activation (Figure 6A). pS6K levels in EOMA cells were 1.0 ± 0.0 in pLKO
vs. 0.3 ± 0.1* in shAkt1 and 3.2 ± 0.3* in shAkt3; *P<0.05, N=3. Similarly, levels of pS6 were
1.0 ± 0.0 in pLKO vs. 0.3 ± 0.2* in shAkt1 and 2.2 ± 0.7* in shAkt3; *P<0.05. Endothelial cells
from Akt3-/- mice similarly showed increased levels of pS6K and pS6 (Figure 6A). pAkt levels
were 1.0 ± 0.0 in wild type (WT) cells vs. 0.8 ± 0.03* in Akt3-/- cells; *P<0.05, N=3. pS6K levels
were 1.0 ± 0.0 in WT cells vs. 4.8 ± 0.8* in Akt3-/- cells; *P<0.05. Levels of pS6 were 1.0 ± 0.0
in WT cells vs. 1.8 ± 0.3* in Akt3-/- cells; *P<0.05. Thus, we have observed in vascular tumor
cells and confirmed in Akt3-/- cells that Akt1 and Akt3 exert opposite effects on S6K signaling
pathway, in which Akt1 promotes whereas Akt3 inhibits S6K activation.
To determine whether S6K mediates the biological effects observed in knockdown
studies, we rescued Akt1 knockdown cells with over-expression of constitutively activated S6K
R3A, and rescued Akt3 knockdown cells with concurrent knockdown of S6K. S6K rescued
shAkt1 and shAkt3 effects on the migration and proliferation of EOMA and ASM.5 cells (Figures
6B-E and Supplementary Figure S7). These findings showed that S6K is a mediator of the
inhibitory effects of Akt3 in vascular tumor cells.
We also utilized a small molecule inhibitor of S6K, LY2584702. LY2584702 is highly
selective for S6K1 when tested against 83 different kinases and 45 cell surface receptors. In
S6K1 enzyme assay, the drug’s IC50 = 2 nM. For pS6 inhibition in cells, the IC50 = 100 nM
(Supplementary Table S1). The drug has some activity against the S6K-related kinases MSK2
and RSK at high concentrations (enzyme assay IC50 = 58-176 nM). LY2584702 inhibits S6K
activity in EOMA cells, as determined by the phosphorylation of its downstream effector S6, in a
dose dependent manner (Figure 6F). To examine the role of S6K in vivo, EOMA cells
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
15
expressing shAkt3 were implanted in nu/nu mice, then treated for 14 days with LY2584702 or
rapamycin, which is a potent inhibitor of mTORC1 and S6K activation (30). Analysis of tumors
removed after 14 days showed that LY2584702 inhibited S6 phosphorylation almost as
effectively as rapamycin (Figure 6G). Loss of Akt3 increased tumor growth as compared with
pLKO (Figure 6H). LY2584702 treatment alone did not significantly affect the growth of pLKO
tumors. However, it significantly reduced the growth of tumors with shAkt3. These findings
showed that down-regulation of Akt3 increased S6K activation in vascular tumors, and
enhanced the anti-tumor efficacy of S6K inhibition with LY2584702.
Akt3 modulates Rictor levels. We evaluated the effects of Akt3 knockdown on the
protein components of mTOR complexes in EOMA and ASM.5 cells. The levels of mTOR and
Raptor (a key component of mTORC1) were not affected by loss of Akt1 or Akt3 (Figures 7A-B).
However, the levels of Rictor (a key component of mTORC2) were significantly reduced in cells
with loss of Akt3, but not Akt1. Rictor immunoblots were quantified by densitometry, normalized
to β-actin and calculated relative to pLKO. In EOMA cells, Rictor levels were 1.0 ± 0.0 in pLKO
vs. 1.4 ± 0.2 in shAkt1 and 0.4 ± 0.2* in shAkt3; *P<0.05, N=9. Similarly in ASM.5 cells, Rictor
levels were 1.0 ± 0.0 in pLKO vs. 1.1 ± 0.2 in shAkt1 and 0.5 ± 0.03* in shAkt3; *P<0.05, N=3.
We observed that knockdown of Akt3, but not Akt1, significantly reduced Rictor mRNA (Figure
7C). These findings indicate that Akt3 positively regulates Rictor levels, at least in part by
modulating Rictor mRNA expression.
To determine whether Rictor affects S6K phosphorylation, we knocked down Rictor and
immunoblotted for pS6K and pS6 (mTORC1 activity) and pAkt S473 (mTORC2 activity). Rictor
knockdown decreased pAkt levels in both EOMA and ASM.5 cells (Figure 7D and
Supplementary Figure S8A). Importantly, loss of Rictor resulted in increased pS6K in both cell
lines. Densitometric analysis showed that in EOMA cells, pS6K levels were 1.0 ± 0.0 in pLKO
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
16
vs. 1.6 ± 0.2* in shRictor; *P<0.05, N=4. pS6 downstream of S6K also increased with Rictor
knockdown. pS6 Levels were 1.0 ± 0.0 in pLKO vs. 1.7 ± 0.3* in shRictor; *P<0.05, N=4. Rictor
knockdown appeared to recapitulate the effects of Akt3 knockdown on S6K and S6, suggesting
that Akt3 regulates S6K activation by modulating Rictor levels.
To more definitely demonstrate that the inhibitory effects of Akt3 on S6K is mediated by
Rictor, we performed a rescue experiment in which constitutively activated myristoylated Akt3
(myrAkt3) was overexpressed in EOMA cells. Expression of HA-tagged myrAkt3 was confirmed
by immunoblotting (Figure 7E). Overexpression of myrAkt3 increased pAkt (T308 and S473)
and Rictor, but decreased pS6K. Concurrent knockdown of Rictor in cells with myrAkt3
overexpression rescued the inhibitory effects of myrAkt3 on S6K phosphorylation.
Densitometric analysis showed that pS6K levels were 1.0±0.0 in vector vs. 0.3±0.04* in HA-
myrAkt3 and 1.5±0.5* in HA-myrAkt3 + shRictor; *P<0.05, N=3. These findings further support
our hypothesis that Akt3 is a negative regulator of S6K signaling pathway and Rictor is a
potential mediator of the effects of Akt3.
S6K exerts negative feedback regulation on Akt3. It has been shown that S6K exerts
negative feedback regulation on Akt signaling via down-regulating IRS-1 and receptor tyrosine
kinase signaling (31). Given the observed differential effects of Akt1 and Akt3 on S6K pathway,
we investigated whether S6K exerts differential feedback regulation on Akt isoforms.
Knockdown of S6K increased pAkt (T308 and S473) in EOMA cells, consistent with the relief of
feedback inhibition of Akt signaling by S6K (Figure 7F). S6K knockdown increased Akt3, but
not Akt1 levels, in EOMA and ASM.5 cells (Figure 7F and Supplementary Figure S8B). Akt3
levels in EOMA cells were 1.0±0.0 in pLKO vs. 1.9±0.5* in shS6K; *P<0.05, N=3. Inhibition of
S6K activation with rapamycin showed a significant increase in Akt3, but not Akt1
(Supplementary Figure S8C). In immunoprecipitation experiments, more Akt3 was pulled down
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
17
in cells with S6K knockdown as compared with pLKO (1.0 in pLKO vs. 1.8 in shS6K, N=2)
(Figure 7G). No changes were seen with Akt1 (1.0 in pLKO vs. 0.9 in shS6K). Conversely,
overexpression of constitutively active S6K reduced levels of Akt3 (Figure 7H). These findings
show that Akt3 inhibits S6K activation, which in turn exerts negative feedback regulation on Akt3
itself.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
18
DISCUSSION
Our studies showed increased Akt activation and decreased PTEN levels in both benign
and malignant vascular tumors. These tumors are composed of multiple cell types, including
neoplastic endothelial cells, inflammatory and stromal cells. It is not known whether the
endothelial cellular component alone is sufficient for vascular tumor development. We showed
in myrAkt1 animal model that de novo hemangioma formation is driven by endothelial Akt1
activation, and is “endothelial-cell autonomous.” Tumor regression was observed upon loss of
endothelial myrAkt1 in our animal model, indicating that sustained Akt signaling is required for
tumor maintenance. Spontaneous regression is a distinctive characteristic of infantile
hemangioma that is biologically programmed in the natural progression of the tumor. It is
conceivable that a pre-programmed network that switches off downstream VEGF signaling
pathways, such as Akt, may be a mechanism of hemangioma regression.
Little is known about Akt3 function. Limited studies on Akt3 showed that it is required for
VEGF stimulation of mitochondrial biogenesis and autophagy in EC, and is important for growth-
factor induced angiogenic responses (32). Akt3 is the dominant isoform in melanoma and
ovarian cancer and regulates cellular senescence, VEGF secretion and angiogenesis in these
tumors (33-35). Our studies of vascular tumors and the emerging literature on Akt3 provide a
new perspective on Akt signaling: there is a “check-and-balance” by different Akt isoforms to
modulate overall Akt signaling output and limit unchecked growth signals downstream of Akt.
Thus, one Akt isoform may regulate growth-promoting biological output, while another isoform
may regulate growth-inhibitory output in order to ensure homeostatic regulation of Akt signaling.
We have found that Akt1 and Akt3 have unique opposing roles, in which Akt1 promotes,
whereas Akt3 inhibits tumor endothelial cell growth. It is possible that the net balance of Akt
signaling output drives vascular tumors. Akt3 is expressed at lower levels in vascular tumors
than in normal blood vessels. Such a scenario has also been observed in malignant glioma, in
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
19
which the expression of Akt3 is lower in the tumor than in normal brain tissue; however, the
remaining Akt3 has kinase activity (36). This suggests that although the level of endogenous
Akt3 is reduced, it still retains the functional capacity as an active kinase.
Different Akt isoforms can signal via distinct downstream pathways that may vary
depending on the cellular context and subcellular localization (37). Akt1 and Akt3, but not Akt2
ablation in mice predominantly influences GSK-3 α/β signaling in a mouse lung tumor model
(14). Akt3, but not Akt1, is crucial for the activation of the mTORC1/S6K signaling pathway in
the brain, thus reinforcing the differential effects of Akt isoforms on the S6K pathway (22). We
have found that Akt1 and Akt3 exert distinct effects on S6K signaling: Akt1 stimulates, whereas
Akt3 inhibits S6K activation. To determine how Akt3 regulates S6K activation, we showed that
knockdown of Akt3 leads to reduced Rictor protein levels, which is due at least in part to a
reduction of Rictor mRNA expression.
Overexpression of myrAkt3 increases Rictor levels and reduces S6K activation.
Importantly, knockdown of Rictor in cells with myrAkt3 overexpression rescues the effects of
myrAkt3 on S6K. Taken together, these findings suggest a potential mechanism by which Akt3
regulates S6K activation through Rictor. We postulate that by regulating Rictor levels, Akt3 can
affect the formation of mTOR complexes and the balance of mTORC1 and mTORC2 activities
in the cell. Published studies lend support to this hypothesis. The formation of TOR complexes
in Caenorhabditis elegans is controlled by semaphorin-plexin signaling, in which semaphorin
promotes the formation of TORC1 and inhibits TORC2 by promoting a shift of TOR from Rictor
towards Raptor, thereby altering the ratio of TORC1 and TORC2 (38). Another mechanism of
regulation of mTOR complex assembly involves the mTOR inhibitor rapamycin. Rapamycin
mainly inhibits mTORC1, but long-term drug treatment can lead to the inhibition of mTORC2, in
which rapamycin may sequester mTOR and interfere with the assembly of mTORC2 (39).
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
20
Emerging studies have highlighted the importance of negative feedback loops that
normally operate to dampen various types of signaling, and therefore ensure homeostatic
regulation of signals in the cell. We have found that S6K exerts negative feedback regulation on
Akt3. This is consistent with the known S6K-mediated negative feedback on the PI3 Kinase-Akt
pathway (31). The selective feedback effects of S6K on Akt3 may reflect the collective pressure
in tumor cells to lower Akt3 levels as a way to counteract the inhibitory effects of active Akt3.
Improved understanding of the unique roles of Akt isoforms has led to the recent
development of Akt inhibitors that preferentially block Akt1 and Akt2, such as Akti-1/2 (40). We
have found that inhibition of S6K activity with LY2584702 was more effective in reducing the
growth of vascular tumors with loss of Akt3 than tumors with normal levels of Akt3. These
findings highlight the potential clinical utility of treating vascular tumors, such as angiosarcoma,
with agents that block S6K signaling. While our studies are selective for vascular tumors at this
time, the findings provide an impetus for further investigation of Akt isoforms in other tumor
types, which could potentially improve our ability to integrate molecular data with therapeutic
treatment regimens.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
21
ACKNOWLEDGEMENTS
We wish to thank the following people for providing reagents and technical support:
Argiris Efstratiadis, Morris Birnbaum, Vera Krump-Konvalinkov, Rebecca Chin, Alex Toker,
William Hahn, Andrew Aplin, Milton Finegold, Cecilia Rosales, Bhuvaneswari Krishnan,
Ningning Zheng, Benjamin Hopkins, Rafael Rojano and Tareq Qdaisat.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
22
REFERENCES
1. Young RJ, Brown NJ, Reed MW, Hughes D, Woll PJ. Angiosarcoma. Lancet Oncol
2010;11: 983-91.
2. Shiojima I, Walsh K. Role of Akt signaling in vascular homeostasis and angiogenesis.
Circ Res 2002;90: 1243-50.
3. Jinnin M, Medici D, Park L, Limaye N., Liu Y, Boscolo E, et al. Suppressed NFAT-
dependent VEGFR1 expression and constitutive VEGFR2 signaling in infantile hemangioma.
Nat Med 2008;14: 1236-46.
4. Itakura E, Yamamoto H, Oda Y, Tsuneyoshi M. Detection and characterization of
vascular endothelial growth factors and their receptors in a series of angiosarcomas. J Surg
Oncol 2008;97: 74-81.
5. Lahat G, Dhuka AR, Hallevi H, Xiao L, Zou G, Smith KD, et al. Angiosarcoma: clinical
and molecular insights. Ann Surg 2010;251: 1098-106.
6. Bader AG, Kang S, Vogt PK. Cancer-specific mutations in PIK3CA are oncogenic in
vivo. Proc Natl Acad Sci U S A 2006;103: 1475-9.
7. Yoeli-Lerner M, Yiu GK, Rabinovitz I, Erhardt P, Jauliac S, Toker A. Akt blocks breast
cancer cell motility and invasion through the transcription factor NFAT. Mol Cell 2005;20: 539-
50.
8. Hutchinson JN, Jin J, Cardiff RD, Woodgett JR, Muller WJ. Activation of Akt-1 (PKB-
alpha) can accelerate ErbB-2-mediated mammary tumorigenesis but suppresses tumor
invasion. Cancer Res 2004;64: 3171-8.
9. Maroulakou IG, Oemler W, Naber SP, Tsichlis PN. Akt1 ablation inhibits, whereas Akt2
ablation accelerates, the development of mammary adenocarcinomas in mouse mammary
tumor virus (MMTV)-ErbB2/neu and MMTV-polyoma middle T transgenic mice. Cancer Res
2007;67: 167-77.
10. Chin YR, Yoshida T, Marusyk A, Beck AH, Polyak K, Toker A. Targeting Akt3 signaling
in triple-negative breast cancer. Cancer Res 2014;74: 964-73.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
23
11. Banerji S, Cibulskis K, Rangel-Escareno C, Brown KK, Carter SL, Frederick AM, et al.
Sequence analysis of mutations and translocations across breast cancer subtypes. Nature
2012;486: 405-9.
12. Chung S, Yao J, Suyama K, Bajaj S, Qian X, Loudig OD, et al. N-cadherin regulates
mammary tumor cell migration through Akt3 suppression. Oncogene 2012: 1-9.
13. Hollander MC, Maier CR, Hobbs EA, Ashmore AR, Linnoila RI, Dennis PA. Akt1 deletion
prevents lung tumorigenesis by mutant K-ras. Oncogene 2011;30: 1812-21.
14. Linnerth-Petrik NM, Santry LA, Petrik JJ, Wootton SK. Opposing functions of akt
isoforms in lung tumor initiation and progression. PLoS ONE 2014;9: e94595.
15. Sarbassov dos D, Guertin DA, Ali SM, Sabatini DM. Phosphorylation and regulation of
Akt/PKB by the rictor-mTOR complex. Science 2005;307: 1098-101.
16. Hara K, Maruki Y, Long X, Yoshino K, Oshiro N, Hidayat S, et al. Raptor, a binding
partner of target of rapamycin (TOR), mediates TOR action. Cell 2002;110: 177-89.
17. Burnett PE, Barrow RK, Cohen NA, Snyder SH, Sabatini DM. RAFT1 phosphorylation of
the translational regulators p70 S6 kinase and 4E-BP1. Proc Natl Acad Sci U S A 1998;95:
1432-7.
18. Dufner A, Thomas G. Ribosomal S6 kinase signaling and the control of translation. Exp
Cell Res 1999;253: 100-9.
19. Ip CK, Cheung AN, Ngan HY, Wong AS. p70 S6 kinase in the control of actin
cytoskeleton dynamics and directed migration of ovarian cancer cells. Oncogene 2011;30:
2420-32.
20. Proud CG. mTORC1 signalling and mRNA translation. Biochem Soc Trans 2009;37:
227-31.
21. Phung TL, Ziv K, Dabydeen D, Eyiah-Mensah G, Riveros M, Perruzzi C, et al.
Pathological angiogenesis is induced by sustained Akt signaling and inhibited by rapamycin.
Cancer Cell 2006;10: 159-70.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
24
22. Easton RM, Cho H, Roovers K, Shineman DW, Mizrahi M, Forman, MS, et al. Role for
Akt3/protein kinase Bgamma in attainment of normal brain size. Mol Cell Biol 2005;25: 1869-78.
23. Du W, Gerald D, Perruzzi CA, Rodriguez-Waitkus P, Enayati, L, Krishnan B, et al.
Vascular tumors have increased p70 S6-kinase activation and are inhibited by topical
rapamycin. Lab Invest 2013;93: 1115-27.
24. Krump-Konvalinkova V, Bittinger F, Olert J, Brauninger W, Brunner J, Kirkpatrick CJ.
Establishment and characterization of an angiosarcoma-derived cell line, AS-M. Endothelium
2003;10: 319-28.
25. Obeso J, Weber J, Auerbach R. A hemangioendothelioma-derived cell line: its use as a
model for the study of endothelial cell biology. Lab Invest 1990;63: 259-69.
26. Xue Q, Nagy JA, Manseau EJ, Phung TL, Dvorak HF, Benjamin LE. Rapamycin
inhibition of the Akt/mTOR pathway blocks select stages of VEGF-A164-driven angiogenesis, in
part by blocking S6Kinase. Arterioscler Thromb Vasc Biol 2009;29: 1172-8.
27. Shao Y, Aplin AE. Akt3-mediated resistance to apoptosis in B-RAF-targeted melanoma
cells. Cancer Res 2010;70: 6670-81.
28. Franke TF, Hornik CP, Segev L, Shostak GA, Sugimoto C. PI3K/Akt and apoptosis: size
matters. Oncogene 2003;22: 8983-98.
29. North PE, Waner M, Mizeracki A, Mihm MC, Jr. GLUT1: a newly discovered
immunohistochemical marker for juvenile hemangiomas. Hum Pathol 2000;31: 11-22.
30. Kuo CJ, Chung J, Fiorentino DF, Flanagan WM, Blenis J, Crabtree GR. Rapamycin
selectively inhibits interleukin-2 activation of p70 S6 kinase. Nature 1992;358: 70-3.
31. O'Reilly KE, Rojo F, She QB, Solit D, Mills GB, Smith D, et al. mTOR inhibition induces
upstream receptor tyrosine kinase signaling and activates Akt. Cancer Res 2006;66: 1500-8.
32. Corum DG, Tsichlis PN, Muise-Helmericks RC. AKT3 controls mitochondrial biogenesis
and autophagy via regulation of the major nuclear export protein CRM-1. Faseb J 2013;28: 395-
407.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
25
33. Cheung M, Sharma A, Madhunapantula SV, Robertson GP. Akt3 and mutant V600E B-
Raf cooperate to promote early melanoma development. Cancer Res 2008;68: 3429-39.
34. Cristiano BE, Chan JC, Hannan KM, Lundie, NA, Marmy-Conus NJ, Campbell IG, et al.
A specific role for AKT3 in the genesis of ovarian cancer through modulation of G(2)-M phase
transition. Cancer Res 2006;66: 11718-25.
35. Liby TA, Spyropoulos P, Buff Lindner H, Eldridge J, Beeson C, Hsu T, et al. Akt3
controls vascular endothelial growth factor secretion and angiogenesis in ovarian cancer cells.
Int J Cancer 2012;130: 532-43.
36. Mure H, Matsuzaki K, Kitazato KT, Mizobuchi Y, Kuwayama K, Kageji T, et al. Akt2 and
Akt3 play a pivotal role in malignant gliomas. Neuro Oncol 2010;12: 221-32.
37. Gonzalez E, McGraw TE. Insulin-modulated Akt subcellular localization determines Akt
isoform-specific signaling. Proc Natl Acad Sci U S A 2009;106: 7004-9.
38. Nukazuka A, Tamaki S, Matsumoto K, Oda Y, Fujisawa H, Takagi S. A shift of the TOR
adaptor from Rictor towards Raptor by semaphorin in C. elegans. Nat Commun 2011;2: 484.
39. Sarbassov dos D, Ali SM, Sengupta S, Sheen JH, Hsu PP, Bagley AF, et al. Prolonged
rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell 2006;22: 159-68.
40. Cherrin C, Haskell K, Howell B, Jones B, Leander K, Robinson R, et al. An allosteric Akt
inhibitor effectively blocks Akt signaling and tumor growth with only transient effects on glucose
and insulin levels in vivo. Cancer Biol Ther 2010;9: 493-503.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
26
FIGURE LEGENDS
Figure 1. Akt phosphorylation in benign and malignant human vascular tumors. (A)
Immunohistochemical stains of vascular tumors for pAkt (S473). Scale bar, 100 μm. Insets
show staining in normal blood vessels adjacent to tumor. (B) pAkt reactivity in normal human
skin and vascular tumors. *P<0.01 vs. NL skin. (C) The stain intensity was scored using a 3-tier
system (see main text) with median values (red lines). *P<0.01 vs. NL skin. NL Skin, normal
skin; IH, infantile hemangioma; KHE, kaposiform hemangioendothelioma; KS, Kaposi’s
sarcoma; AS, angiosarcoma. (D) Western blots of human dermal microvascular endothelial
cells (HDMEC) and primary infantile hemangioma EC (HemeEC). Graph shows densitometric
quantitation of pAkt / total Akt ratios, normalized to HDMEC as control. *P<0.05, N=6. (E-F)
Western blots of (E) HDMEC and ASM.5 cells, and (F) normal mouse EC and EOMA cells.
Figure 2. Akt1 promotes vascular tumor cell growth and migration. (A) HemeEC and (B)
ASM.5 expressing pLKO or shAkt1 (independent clones #1 and #2) were analyzed by western
blot. (C-D) HemeEC and ASM.5 DNA content, expressed as fluorescence units relative to
“Day=0”. *P<0.05, N=3. (E-F) Transwell migration assays of (E) hemeEC and (F) ASM.5 cells
expressing shAkt1 ± VEGF-A (50 ng/ml) for 5 hours normalized to “pLKO-VEGF” control.
*P<0.05 vs. pLKO-VEGF; **P<0.05 vs. pLKO+VEGF, N=3. (G) Representative bright field
images of HemeEC cord formation on Collagen I matrix ± VEGF (50 ng/ml) for 14 hours. (H)
Quantitation of cord length. *P<0.01 vs. pLKO-VEGF; **P<0.01 vs. pLKO+VEGF, N=3.
Figure 3. Endothelial myrAkt1 activation drives to hemangioma formation in vivo. (A)
Schematic of myrAkt1 skin graft model of hemangioma (see text for details). (B) Vascular
tumors developed in the grafts 4 weeks following myrAkt1 induction. (C) Microscopic features
of the tumor (scale bar, 100 μm). Arrows indicate tumor boundary. Graph of tumor volume in
myrAkt1 mice is shown. (D) Immunofluorescence stains of myrAkt1 tumor for CD31 (red) and
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
27
HA-tagged myrAkt1 (green); smooth muscle actin (SMA, red) and phospho-Akt (green); glucose
transporter-1 (Glut-1, red) and CD31 (green); nuclei (blue). (E-F) myrAkt1-induced
hemangiomas regressed when myrAkt1 was turn off for 3.5 weeks. Images are representative
of 6 tumors per group. Red arrows, blood vessels; black arrows, fibrofatty tissue. Scale bar,
0.5 cm.
Figure 4. Akt1, Akt2 and Akt3 expression in human vascular tumors. (A) Immunostains of
infantile hemangioma and angiosarcoma tissues for Akt isoforms. Insets show staining in
normal blood vessels adjacent to tumor. (B) Plots of semi-quantitative analysis of Akt1, Akt2
and Akt3 immunostains in normal human skin and vascular tumors. The stain reactivity is the
percentage of tissues with positive staining. The stain intensity was scored using a 3-tier
system (see main text) with median values (red lines). NL Skin, normal skin; IH, infantile
hemangioma; AS, angiosarcoma. *P<0.05 vs. NL Skin. (C) Akt isoforms were
immunoprecipitated from ASM.5 cell lysates and immunoblotted.
Figure 5. Differential effects of Akt isoforms in vascular tumor cells. (A) Spheroids of hemeEC
expressing pLKO, shAkt1 or shAkt3 were cultured in Matrigel/Collagen I matrix ± VEGF (50
ng/ml) for 24 hours. Arrowheads indicate endothelial sprouts. (B) Total sprout length per
spheroid was quantified and normalized to “pLKO -VEGF” control. *P<0.01 vs. pLKO -VEGF;
**P<0.01 vs. pLKO+VEGF; N = 3, independent shAkt clones #1 and #2 for each Akt isoform.
(C) ASM.5 and (D) EOMA cells with Akt1 or Akt3 knockdown were assessed for cell migration in
scratch wound assays. The area of wound closure was quantified and shown as percent
closure relative to t=0 hour. *P<0.05, N= 9. (E) EOMA cells expressing pLKO, shAkt1 or shAkt3
were assessed for cell growth (fluorescence units of DNA content). *P<0.01, N=6. (F) EOMA
cells with stable Akt1, Akt2 or Akt3 knockdown were implanted in mice and monitored for tumor
growth. *P<0.05, N = 8.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
28
Figure 6. Akt1 and Akt3 exert opposing effects on S6K pathway. (A) HemeEC and EOMA cells
expressing pLKO, shAkt1 or shAkt3 were immunoblotted. Lung microvascular EC from wild
type (WT) and Akt3-/- mice were similarly analyzed. (B) EOMA cells expressing pLKO, shAkt1,
shAkt1+S6K R3A, shAkt3, or shAkt3+shS6K were assessed for cell migration by scratch wound
assay. Bright-field images of scratch wounds are shown. (C) EOMA cell migration was
assessed in Boyden chamber transwell assay. *P<0.01 vs. pLKO; **P<0.01 vs. shAkt1;
***p<0.01 vs. shAkt3, N=4. (D-E) ASM.5 cells expressing shAkt3 or shAkt3+shS6K were
assessed for (D) cell migration in transwell assay, and (E) cell growth (relative to “Day=0”).
*P<0.05 vs. pLKO; **P<0.05 vs. shAkt3, N=3. (F) Chemical structure of LY2584702. EOMA
cells were treated ± LY2584702 for 4 hours and then immunoblotted. Cells in 0.1% FBS and
10% FBS served as controls. (G) EOMA tumors grown in mice treated ± Rapamycin (2
mg/kg/day, i.p. injections) or LY2584702 (12.5 mg/kg twice daily, oral dosing) for 14 days and
immunoblotted. Two tumor samples (#1 and #2) per treatment condition were analyzed. (H)
Mice with EOMA tumors expressing pLKO or shAkt3 were treated ± LY2584702. *P<0.05 vs.
pLKO; **P<0.05 vs. shAkt3+Vehicle, N=6.
Figure 7. Akt3 regulates Rictor levels and is under negative feedback regulation by S6K. (A-B)
EOMA and ASM.5 cells expressing pLKO, shAkt1 or shAkt3 were immunoblotted. (C) qPCR for
Rictor mRNA in EOMA cells with Akt1 or Akt3 knockdown, calculated relative to pLKO.
*P<0.01, N=4. (D) EOMA cells expressing pLKO or shRictor were immunoblotted to assess for
Akt/S6K pathway activation. (E) EOMA cells expressing HA-myrAkt3 or HA-myrAkt3 + shRictor
were immunoblotted. (F) Immunoblots of EOMA cells expressing pLKO or shS6K. (G) EOMA
cells were immunoprecipitated for Akt1 and Akt3, and blotted for these isoforms. (H)
Immunoblots of cells expressing pLKO or constitutively active S6K.
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
Figure 1
Infantile
Hemangioma Kaposiform
Hemangioendothelioma Kaposi’s Sarcoma Angiosarcoma A
Phosphorylated Akt
* * *
pAkt % Reactivity
*
Perc
en
tag
e
B * * * *
Sta
in S
co
re
C pAkt Intensity
pAkt
Total Akt
pVEGFR-2
Total VEGFR-2
PTEN
β-Actin
HDMEC HemeEC D
0.0
0.5
1.0
1.5
2.0
2.5
pA
kt / T
ota
l A
kt
(N
orm
aliz
ed to
HD
ME
C)
*
β-Actin
F Mouse
EC EOMA
pAkt
Total Akt
PTEN
pAkt
Total Akt
PTEN
β-Actin
HDMEC ASM.5 E
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
Figure 2
A
Akt1
Akt2
Akt3
β-Actin
HemeEC
Akt1
Akt2
Akt3
β-Actin
ASM.5 B
ASM.5
0.0
1.0
2.0
3.0
4.0
0 1 2 3 4
pLKO
shAkt1 #1
shAkt1 #2
Flu
ore
scence U
nits
(Re
lative
to
Da
y 0
)
* * *
*
Days
*
HemeEC C D
0.0
1.0
2.0
3.0
4.0
0 1 2 3 4
pLKO
shAkt1
Flu
ore
scence U
nits
(Re
lative
to
Da
y 0
)
* *
*
Days
G HemeEC Cord Formation
+V
EG
F
pLKO shAkt1 #1
-VE
GF
shAkt1 #2 E
F H
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
Figure 3
A B C
myrAkt1OFF 4 wks
TET:myrAkt1
VE-C dh i TTA
X
A
0.6
0.8
(cm
3)
B CmyrAkt1 OFF myrAkt1 ON
TUMORmyrAkt1
OFF2 k
Nu/Nu Recipients
OFF 4 wks
myrAkt1 ON 4 wks
Double Transgenic Donor
Cadherin:TTA
0.0
0.2
0.4
Tum
or V
olum
e 2 wks
myrAkt1OFF2 wks
RecipientsTransgenic Donor
myrAkt1 ONmyrAkt1 ON for another 3 5 wks Tumor H&E
ED
myrAkt1OFF
myrAkt1ON
myrAkt1 ON another 3.5 wks Tumor H&EDCD31 HA-Tag Nuclei SMA pAkt Nuclei Glut-1 CD31 Nuclei
myrAkt1 ONmyrAkt1 OFF for another 3.5 wks Tumor H&EF
on April 6, 2018. ©
2014 Am
erican Association for C
ancer Research.
cancerres.aacrjournals.org D
ownloaded from
Author m
anuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author M
anuscript Published O
nlineFirst on N
ovember 11, 2014; D
OI: 10.1158/0008-5472.C
AN
-13-2961
Figure 4
Infantile Hemangioma Angiosarcoma
Akt1A Akt1 % Reactivity
age
ore
Akt1 IntensityB
Infantile Hemangioma Angiosarcoma
NL Skin IH AS NL Skin IH AS
Akt2 % Reactivity Akt2 I t it
Perc
enta
Stai
n Sc
o
Akt2 Stai
n Sc
ore
Akt2 % Reactivity Akt2 Intensity
Perc
enta
ge
NL Skin IH AS
P
Akt3 Intensity
core
* *
Akt3 % Reactivity
age
*
NL Skin IH AS
Akt3 NL Skin IH AS NL Skin IH AS
Stai
n S * *
Perc
enta
pAkt (S473)
IP:
pAkt (T308)
C
Akt1
Akt2
Akt3
on April 6, 2018. ©
2014 Am
erican Association for C
ancer Research.
cancerres.aacrjournals.org D
ownloaded from
Author m
anuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author M
anuscript Published O
nlineFirst on N
ovember 11, 2014; D
OI: 10.1158/0008-5472.C
AN
-13-2961
Figure 5
1200 pLKO
E*
EOMA
6pLKOshAkt1 #1shAkt1 #2 **
BpLKO shAkt1 shAkt3
A HemeEC
600
800
1000
pLKOshAkt1shAkt3
cenc
e U
nits
*
*3
4
5shAkt3 #1shAkt3 #2
out L
engt
ho
pLK
O-V
EG
F) **
*
*
pLKO shAkt1 shAkt3
-VEGF
0
200
400
Fluo
resc *
***0
1
2
VEGF +VEGF
Spr
o(R
elat
ive
to
*
* **+ VEGF
0 2 4 6 8Days
-VEGF +VEGF
CpLKO shAkt1 shAkt3
t = 0h
ASM.5
0.5LKO
F EOMA Tumor Growth
DpLKO shAkt1 shAkt3
EOMA
t 0h
t = 16h0.3
0.4
0.5pLKOshAkt1 #1shAkt2 #1shAkt3 #1
me
(cm
3 )
* *
*
60%
80%
100%
nt C
losu
re
*
*
0.1
0.2
* *
Tum
or V
olum
**
60%
80%
100%
nt C
losu
re
*
*
0%
20%
40%
pLKO shAkt1 shAkt3
Per
cen
0.00 2 4 6 8 10 12
* *
Days0%
20%
40%
pLKO shAkt1 shAkt3
Per
cen
on April 6, 2018. ©
2014 Am
erican Association for C
ancer Research.
cancerres.aacrjournals.org D
ownloaded from
Author m
anuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author M
anuscript Published O
nlineFirst on N
ovember 11, 2014; D
OI: 10.1158/0008-5472.C
AN
-13-2961
Figure 6
**CB pLKO shAkt1
shAkt1 + S6K R3A shAkt3
shAkt3 + shS6K
pAkt
HemeECA
20406080
100
*
****
Mig
rate
d C
ellst = 0 h
pS6
pS6K
Total S6K
Total Akt
F
t = 16 h0#
M
120D
Total S6
EOMA
F
020406080
100120
***
Mig
rate
d C
ells
EOMA Tumor GrowthHpLKO + VehiclepLKO + LY2584702
pAkt
Total Akt
pS6K
Total S6K
10%
FB
S
0.1%
FB
S
1600
800
400
200
100
β Actin
pS6
LY2584702 (nM)0
pLKO shAkt3 shAkt3 + shS6K
# M
0 06
0.08
0.1pLKO + LY2584702shAkt3 + VehicleshAkt3 + LY2584702
e (c
m3)
*E4.5 pLKO *
Total S6KpS6
Total S6
Mouse ECβ-Actin
G#1 #2 #1 #2 #1 #2
Vehicle RAPA LY2584702Tumor: 0.02
0.04
0.06
Tum
or V
olum
e
*
**1.52.02.53.03.54.0
pLKOshAkt3shAkt3 + shS6K
ores
cenc
e U
nits
elat
ive
to D
ay 0
)
* **
***
pS6K
pAktTotal Akt
pS6
α-Tubulin0
3 7 11 14
DaysDays
0.00.51.0
0 1 2 3 4 5
Fluo
(Re
Total S6KpS6
Total S6
on April 6, 2018. ©
2014 Am
erican Association for C
ancer Research.
cancerres.aacrjournals.org D
ownloaded from
Author m
anuscripts have been peer reviewed and accepted for publication but have not yet been edited.
Author M
anuscript Published O
nlineFirst on N
ovember 11, 2014; D
OI: 10.1158/0008-5472.C
AN
-13-2961
Figure 7
A EOMA
Akt1
Akt2
β-Actin
Raptor
Rictor
Akt3
mTOR
IP: Akt1 IP: Akt3
Akt1
Akt3
G
Akt3
Akt1
Total Akt
pS6
Total S6
H
Akt1
Akt2
β-Actin
Rictor
Akt3
Raptor
ASM.5 B C
0.0
0.4
0.8
1.2
Ric
tor
mR
NA
Fold
Chan
ge
(Rela
tive t
o p
LK
O)
*
D
E F
Total S6K
Akt3
Akt1
β-Actin
pAkt (T308)
pAkt (S473)
Total Akt
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961
Published OnlineFirst November 11, 2014.Cancer Res Thuy L Phung, Wa Du, Qi Xue, et al. tumor growthAkt1 and Akt3 exert opposing roles in the regulation of vascular
Updated version
10.1158/0008-5472.CAN-13-2961doi:
Access the most recent version of this article at:
Material
Supplementary
http://cancerres.aacrjournals.org/content/suppl/2014/11/11/0008-5472.CAN-13-2961.DC1
Access the most recent supplemental material at:
Manuscript
Authoredited. Author manuscripts have been peer reviewed and accepted for publication but have not yet been
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cancerres.aacrjournals.org/content/early/2014/11/22/0008-5472.CAN-13-2961To request permission to re-use all or part of this article, use this link
on April 6, 2018. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on November 11, 2014; DOI: 10.1158/0008-5472.CAN-13-2961