Endocrine-Related Cancer (2001) 8 249–258
`
`mTOR, a novel target in breast cancer: the
`effect of CCI-779, an mTOR inhibitor, in
`preclinical models of breast cancer
`
`K Yu, L Toral-Barza, C Discafani, W-G Zhang, J Skotnicki, P Frost
`and J J Gibbons
`Wyeth-Ayerst Research, Department of Oncology, 401 North Middletown Road, Pearl River, New York 10965, USA
`(Requests for offprints should be addressed to J J Gibbons Jr; Email: gibbonj@war.wyeth.com)
`
`Abstract
`
`The mammalian target of rapamycin (mTOR) is a central regulator of G1 cell cycle protein synthesis
`that precedes commitment to normal cellular replication. We have studied the effect of cell cycle
`inhibitor-779 (CCI-779), a rapamycin ester that inhibits mTOR function, on the proliferation of a panel
`of breast cancer cell lines. Six of eight lines studied were sensitive (IC50압 50 nM) and two lines were
`resistant (IC50>1.0 µM) to CCI-779. Sensitive lines were estrogen dependent (MCF-7, BT-474,
`T-47D), or lacked expression of
`the tumor suppressor PTEN (MDA-MB-468, BT-549), and/or
`overexpressed the Her-2/neu oncogene (SKBR-3, BT-474). Resistant
`lines (MDA-MB-435,
`MDA-MB-231) shared none of these properties. CCI-779 (50 nM) inhibited mTOR function in both a
`sensitive and a resistant line. In nu/nu mouse xenografts, CCI-779 inhibited growth of MDA-MB-468
`(sensitive) but not MDA-MB-435 resistant tumors. Treatment of sensitive lines with CCI-779 resulted
`in a decrease in D-type cyclin and c-myc levels and an increase in p27kip-1 levels. There was good
`correlation between activation of
`the Akt pathway and sensitivity to CCI-779. Amplification of
`mTOR-regulated p70S6 kinase, which is downstream of Akt, may also have conferred CCI-779
`sensitivity to MCF-7 cells. Taken together, the data suggest that mTOR may be a good target for
`breast cancer therapy, especially in tumors with Akt activation resulting from either growth factor
`dependency or loss of PTEN function.
`Endocrine-Related Cancer (2001) 8 249–258
`
`Introduction
`
`Cell cycle inhibitor-779 (CCI-779) is an ester derivative of
`the natural product
`rapamycin that was developed for
`intravenous use for cancer chemotherapy. Rapamycin is a
`macrolide antibiotic with anti-fungal, immunosuppressive,
`and anti-tumor properties (Sehgal et al. 1994). Genetic
`studies in yeast showed that rapamycin inhibited cell growth
`by blocking the function of the proteins TOR1 and TOR2
`(targets of rapamycin 1 and 2) (Heitman et al. 1991). The
`TOR proteins are members of
`the phosphatidylinositol
`3-kinase (PI3-K)-related family of kinases and regulate
`several cellular functions (Schmelzle & Hall 2000). In order
`to inhibit TOR function, rapamycin initially binds to the
`cytoplasmic immunophilin FKBP-12 and the complex then
`inhibits TOR (Brown et al. 1994).
`A mammalian homolog of the yeast TOR proteins has
`been cloned independently by several groups and will be
`referred to in this report as mTOR (Sabers et al. 1995), but
`
`is also known as FRAP (Brown et al. 1994), RAFT
`it
`(Sabatini et al. 1994), and RAPT (Chiu et al. 1994). The
`mTOR protein regulates cell cycle progression, in part, by
`enhancing translation initiation and/or the stability of cell
`cycle
`regulatory
`proteins
`such
`as D-type
`cyclins
`(Hashemolhosseini et al. 1998, Muise-Helmericks et al.
`1998), c-myc (West et al. 1998), and p27kip-1 (Nourse et al.
`1994) among others. At least two direct targets of mTOR,
`p70 S6 kinase and 4E-BP1/PHAS-1, have been suggested to
`mediate the effect of mTOR on protein translation (Brunn et
`al. 1997, Thomas & Hall 1997, Burnett et al. 1998). 4E-BP1
`(eIF-4E binding protein-1) binds
`to the mRNA cap
`recognition element of the translation initiation complex
`protein eIF-4E (eukaryotic initiation factor 4E) and thereby
`inhibits translation initiation (Beretta et al. 1996). mTOR
`phosphorylation of 4E-BP1 causes it
`to dissociate from
`eIF-4E,
`thus enhancing the translation initiation complex
`interactions with the mRNA 5′ cap. The kinase p70 S6K is
`phosphorylated
`and
`activated
`by mTOR,
`and
`then
`
`Endocrine-Related Cancer (2001) 8 249–258
`1351-0088/01/008–249  2001 Society for Endocrinology Printed in Great Britain
`
`Online version via http://www.endocrinology.org
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 001
`
`

`
`Yu et al.: mTOR, a novel target in breast cancer
`
`phosphorylates the S6 protein of the 40S ribosomal complex.
`Phosphorylation of S6 results in enhanced translation of
`in the 5′
`proteins that contain a polypyrimidine tract
`untranslated region (Jeffries et al. 1997, Volarevic & Thomas
`2000). In addition to regulating protein translation, mTOR
`can also regulate the stability of some cell cycle regulatory
`proteins such as D-type cyclins and p27kip-1. Activation of
`mTOR
`appears
`to
`stabilize
`D-type
`cyclins
`(Hashemolhosseini et al. 1998) and to destabilize the
`(Nourse et al.
`cyclin-dependent kinase inhibitor p27kip-1
`1994).
`Several of the cell cycle targets that are regulated by
`mTOR have been reported to be dysregulated in human
`breast cancer,
`including eIF-4E (Kerekatte et al. 1995),
`D-type cyclins
`(Weinstat-Saslow et al. 1995), p27kip-1
`(Fredersdorf et al. 1997), and c-myc (Liao & Dickson 2000).
`Therefore, we have begun to study the effect of the mTOR
`inhibitor CCI-779 in models of human breast cancer. Cell
`growth in culture revealed that 6 of 8 breast cancer lines
`studied were inhibited by CCI-779 with IC50s in the low nM
`range. Two lines, however, were found to be markedly
`resistant (IC50>1 µM). Cell lines sensitive to CCI-779 were
`estrogen receptor positive, or overexpressed Her-2/neu, or
`had lost
`the
`tumor
`suppressor gene product PTEN
`(phosphatase related to tensin and deleted on chromosome
`10) (Li et al. 1997). Sensitive lines contained higher levels
`of the activated form of Akt, suggesting that this PI3-K
`downstream target may be a common link between growth
`factor-dependent
`lines
`(estrogen,
`Her-2/neu)
`and
`PTEN-deleted lines that are sensitive to mTOR inhibition.
`The potential for therapeutic use of an mTOR inhibitor is
`discussed in terms of the rationale provided by known
`genetic alterations in human breast cancer cells.
`
`Materials and methods
`
`Chemicals and cell culture methods
`
`All chemicals were obtained from Sigma-Aldrich (St Louis,
`MO, USA). CCI-779 was synthesized at Wyeth-Ayerst
`Research. Cell
`lines
`of MDA-MB-468
`(MDA-468),
`MDA-MB-435 (MDA-435), MDA-MB-231 (MDA-231),
`MCF-7, T-47D, SKBR-3 and BT-474 were obtained from
`the American Type Culture Collection (ATCC) (Rockville,
`MD, USA). BT-474G is a sub-clone derived from BT-474.
`All cell lines were cultured in Minimum Essential Medium
`(MEM) containing 10% fetal bovine serum (FBS) and 1 mM
`MEM sodium pyruvate in a 37°C incubator containing 5%
`CO2. All cell culture reagents were purchased from
`Gibco-BRL (Grand Island, NY, USA).
`
`Proliferation assay
`Cells were plated in 96-well culture plates at about 3000 cells
`per well. One day following plating, drugs were added to
`
`cells. Three days after drug treatment, viable cell densities
`were determined by measuring metabolic conversion (by
`viable cells) of the dye MTS, a previously established cell
`proliferation assay. Stock solutions of MTS and PMS were
`purchased from Promega Corp. (Madison, WI, USA). For
`each assay, MTS and PMS stocks were freshly thawed and
`mixed (MTS/PMS, 20:1). The MTS/PMS mixture was then
`added to 96-well cell plates at 20 µl/well, and plates were
`incubated for 1–2 h in cell culture incubator. MTS assay
`results were read in a 96-well
`format plate reader by
`measuring absorbance at 490 nm. The effect of each drug
`treatment was calculated as a percentage of control cell
`growth obtained from vehicle-treated cells grown in the same
`culture plate.
`
`In vivo tumor inhibition
`Xenograft model athymic nu/nu female mice, 5–6 weeks of
`age, were obtained from Charles River Laboratories,
`Wilmington, MA, USA and maintained in a barrier facility
`in accordance with Institutional Animal Care and Use
`Committee (IACUC) regulations. Animals were injected s.c.
`with either 6 × 106 MDA-MB468 cells or 6 × 106
`MDA-MB435 cells. When tumors reached a weight of
`between 80 and 120 mg, animals were randomized into
`treatment groups (5 mice/group). Animals were treated
`intraperitoneally (i.p.) for 5 consecutive days with 40, 20, or
`10 mg/kg CCI-779 prepared in 2% ethanol, 8% cremophor
`el, (Sigma, St Louis, MO, USA), or vehicle alone. Tumor
`mass ([length × width2]/2) was determined on days 7, 14, 21
`and 28 post staging. The data were analyzed via Student’s
`t-test. A P-value <0.05 indicates a statistically significant
`reduction in relative tumor growth of the treated group
`compared with that of the vehicle control group.
`
`Protein lysates and immunoblotting
`For immunoblotting experiments, cells were plated in 10-cm
`dishes or 6-well plates. Depending on the study, after the
`cells had completely attached, they were either serum-starved
`or incubated in growth medium overnight. Treatment with
`various inhibitors ranged from 2 to 16 h. After drug
`pretreatment,
`the cells were rinsed once with cold PBS
`(phosphate buffered saline without Mg++ and Ca++) and then
`lysed in cold gentle lysis buffer (25 mM Hepes, pH 7.55,
`100 mM NaCl, 20 mM β-glycerophosphate, 1.5 mM MgCl2,
`0.5 mM EGTA, 0.25 mM EDTA, 1% NP-40, 10 mM
`Na3VO4, 10 µg/ml aprotinin, 10 µg/ml leupeptin, 1 mM
`phenylmethylsulfonylfluoride, 1 µM microcystin LR and
`0.1% 2–mercaptoethanol). In some experiments, cells from
`6-well plates were lysed in NuPAGE-LDS sample buffer
`(Invitrogen, Carlsbad, CA, USA). The crude lysates were
`briefly sonicated and then clarified by centrifugation for 15
`min at 14 000 r.p.m. Cleared lysates (20–50 µg) were
`subjected to SDS-PAGE electrophoresis using the NuPAGE
`system (Invitrogen, Carlsbad, CA, USA) and transferred to
`
`250
`
`www.endocrinology.org
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 002
`
`

`
`a nitrocellulose membrane. The sources of various primary
`antibodies are as follows. Phospho-AKT (Ser473), AKT,
`phospho-p70 S6
`kinase
`(Thr389),
`p70 S6
`kinase,
`phospho-4E-BP1
`(Thr-37, Thr-46
`and Ser-65) were
`purchased from Cell Signaling Technology (Beverly, MA,
`USA). Antibodies against cyclin D3, c-myc and 4E-BP1
`were from Santa Cruz Biotechnology (Santa Cruz, CA,
`USA). Anti-p27kip1 was from Transduction Laboratories (San
`Diego, CA, USA) and phospho-FKHRL-1 (Thr32) was from
`Upstate Biotechnology
`(Lake
`Placid, NY, USA).
`Immunoblots were blocked for 1 h with blocking buffer
`TBST (20 mM Tris, pH 7.5, 500 mM NaCl2, 0.1%-
`Tween-20) and 5% non-fat milk. After washing, they were
`at 4°C
`incubated with primary antibodies overnight
`according to the manufacturer’s suggestions, washed and
`treated
`with
`appropriate
`secondary
`antibodies.
`Immunoreactive proteins were detected using enhanced
`chemiluminescence (ECL) (Amersham Pharmacia Biotech,
`Piscataway, NJ, USA).
`
`Results
`
`The effect of the mTOR inhibitor CCI-779
`on the growth of human breast cancer lines
`in vitro
`
`The growth inhibitory properties of CCI-779 were studied in
`vitro on a panel of 8 human breast cancer cell lines (Table 1).
`The MCF-7, BT-474, and T-47D cell lines are all estradiol
`responsive (Yarden et al. 1996) and all were strongly growth
`inhibited by CCI-779 (IC50 low nM). Similarly, growth of lines
`BT-549 and MDA-MB-468 which contain deletions of the
`PTEN tumor suppressor gene (Lu et al. 1999) was highly
`sensitive
`to treatment with CCI-779. The Her-2/neu
`overexpressing (Chen et al. 2000) SKBR-3 line and ER
`positive, Her-2 overexpressing BT-474 cells were also
`inhibited at low nM concentrations of CCI-779. Two lines,
`MDA-MB-435 and MDA-MB-231, were resistant to treatment
`with CCI-779 (IC50s at low µM concentrations). These lines do
`
`Table 1 Effect of CCI-779 on the growth of human breast
`cancer lines in vitro.
`
`Cell
`line
`
`MCF-7
`BT-474
`T-47D*
`BT-549*
`MDA-MB-468
`SKBR-3
`MDA-MB-435
`MDA-MB-231
`
`IC50
`(nM)
`
`10–50
`0.6
`<10.0
`<10.0
`0.7
`0.7
`1600
`5900
`
`Estrogen
`receptor α Her-2/Neu PTEN−/−
`+
`−
`−
`+
`+
`−
`+
`−
`−
`+
`+
`−
`−
`−
`
`−
`−
`−
`
`+
`−
`−
`
`Endocrine-Related Cancer (2001) 8 249–258
`
`not respond to estradiol, do not overexpress Her-2/neu, and are
`wild-type for the tumor suppressor, PTEN.
`
`The mTOR pathway is activated in CCI-779
`sensitive MDA-MB-468 cells and minimally
`activated in CCI-779 resistant MDA-MB-435
`cells
`
`There was a marked difference in the ability of CCI-779 to
`inhibit MDA-468 (PTEN−/−) cells compared with MDA-435
`(PTEN+/+) cells (Fig. 1A). The PTEN−/−, CCI-779 sensitive
`MDA-468 line showed evidence of Akt activation, as has
`been reported by others (Lu et al. 1999, Weng et al. 1999).
`A comparison of Western blots using a phospho-specific
`antibody for
`the activated form of Akt shows marked
`activation relative to total Akt protein in the MDA-468 cells
`compared with the MDA-435 cells (Fig. 1B). The forkhead
`transcription factor (FKHRL−1), a downstream target of Akt
`(Biggs et al. 1999) is also highly phosphorylated, confirming
`that the Akt signaling is activated in these cells. p70 S6K, a
`direct target of mTOR, is also highly phosphorylated in the
`MDA-468 cells, suggesting that
`the mTOR pathway is
`activated in the PTEN−/− cells. Both the Akt and mTOR
`pathways were only minimally activated in the MDA-435
`PTEN wild-type cells.
`
`Inhibition of mTOR function inhibits growth in
`xenografts of MDA-468 (PTEN−/−) cells but not
`MDA-435 (PTEN+/+) cells
`
`CCI-779 produced a similar differential effect on the growth
`of MDA-468 cells compared with MDA-435 cells in vivo
`(Fig. 2). The PTEN mutant MDA-468 cells were sensitive
`whereas the PTEN wild-type MDA-435 cells were not. In
`these experiments, nu/nu mice were injected in the flank with
`tumor cells and tumors were allowed to grow to a size of
`100 mg. Staged mice were then randomized into treatment
`groups and treated with CCI-779 by i.p. injection at 10, 20,
`or 40 mg/kg or with vehicle for 5 consecutive days. Although
`the MDA-468 cells did not grow as well as the MDA-435
`cells in nude mice, there was a clear regression of tumor size
`in MDA-468–treated tumors at doses of 40 and 20 mg/kg.
`Even at the low dose of 10 mg/kg, the delay in growth of
`MDA-468 tumors extended 10 days beyond the last dose of
`CCI-779 on day 5. The growth of MDA-435 tumors was not
`affected by CCI-779 at any dose. The effect of CCI-779 in
`vivo has not been limited to slow growing tumors as we have
`seen similar growth inhibitory effects in the fast growing
`U87 MG glioblastoma (data not shown).
`
`Treatment with CCI-779 inhibits mTOR
`function in both sensitive (MDA-468) and
`resistant (MDA-435) cells
`
`*Data from National Cancer Institute 60 cell panel screen
`(Monks et al. 1991).
`
`MDA-468 cells were sensitive to growth inhibition by the
`mTOR inhibitor CCI-779 in vitro and in vivo while
`
`www.endocrinology.org
`
`251
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 003
`
`

`
`Yu et al.: mTOR, a novel target in breast cancer
`
`Figure 1 Loss of PTEN tumor suppressor in MDA-468 cells results in an elevated sensitivity to growth inhibition by CCI-779.
`(A) Growth inhibition curves. Cells were plated in 96-well cell culture plates at 3000 cells per well overnight before treatment
`with CCI-779 for 3 days. Cell growth assays were performed by standard MTS assay as described in Materials and methods.
`(B) Elevated AKT and mTOR signaling in MDA-468 cells. Cells were plated in 10-cm culture plates overnight and were then
`serum-starved for 24 h with culture medium containing 0.1% serum. Total cellular lysates were prepared using the gentle lysis
`buffer described in Materials and methods. Equal amounts (50 µg) of total proteins were analyzed by immunoblotting with
`antibodies of AKT, phospho (P)-AKT (S473), phospho (P)-FKHRL-1 (T32) and phospho (P)-p70 S6K (T389) as described in
`Materials and methods.
`
`Figure 2 Nu/nu mice (5 mice/group) were injected in the flank with either MDA-MB-468 or MDA-MB-435 cells (6×106/mouse).
`Mice were randomized into treatment groups after tumors reached a size of about 100 mg. Treatment with CCI-779 or vehicle
`was for days 1–5 after staging and mice were not treated thereafter.
`
`252
`
`www.endocrinology.org
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 004
`
`

`
`MDA-435 cells were not. We studied by Western blot the
`effect of CCI-779 on mTOR function in these cells by
`looking at the phosphorylation status of the mTOR targets
`p70 S6K and 4E-BP1 (Fig. 3). Cells were treated in vitro for
`16 h with CCI-779 at a concentration (50 nM) that inhibited
`growth of MDA-468 cells by about 50% and had no effect
`on growth of MDA-435 cells. The level of phosphorylated
`p70 S6K was higher in the MDA-468 cells, presumably due
`to loss of PTEN and activation of Akt (Fig. 1B), and
`CCI-779 treatment resulted in complete dephosphorylation of
`p70 S6K without affecting the total protein levels of p70
`S6K (not shown). There was considerably less but detectable
`phosphorylated p70 S6K in the PTEN+/+ MDA-435 cells;
`however, treatment with CCI-779 also resulted in complete
`dephosphorylation of p70 S6K in these cells (Fig. 3). Similar
`results have been reported for
`rhabdomyosarcoma lines
`sensitive or resistant to rapamycin (Hosoi et al. 1998). Using
`an antibody specific for threonine 46 (T46) on 4E-BP1, a site
`directly phosphorylated by mTOR (Gingras et al. 1999), we
`
`Figure 3 Differential effects of CCI-779 on cellular mTOR
`targets in CCI-sensitive and CCI–resistant cells. MDA-435
`and MDA-468 cells were plated in complete growth medium
`in 6-well culture plates and treated with CCI-779 for 16 h.
`Total cellular proteins were prepared by NuPAGE-LDS
`sample buffer. Immunoblotting assays with antibodies of
`phospho (P)-p70 S6K (T389), phospho (P)-4E-BP1 (T46),
`4E-BP1, c-Myc, cyclin D3 and p27kip1 were performed using
`the NuPAGE system. Cont, control.
`
`Endocrine-Related Cancer (2001) 8 249–258
`
`observed that CCI-779 treatment in MDA-468 cells resulted
`in a shift to a faster migrating species, suggesting inhibition
`of phosphorylation of other residues due to mTOR inhibition
`(Fig. 3). We were surprised to find that
`the resistant
`MDA-435 cells markedly overexpressed 4E-BP1 relative to
`MDA-468 cells. Nevertheless, there was also a shift to faster
`migrating species of 4E-BP1 after CCI-779 treatment of
`MDA-435 cells. This shift was more easily observed with an
`antibody that recognized total 4E-BP1 levels where a more
`condensed faster migrating band was seen after CCI-779
`treatment. Thus treatment with CCI-779 resulted in mTOR
`inhibition in both a sensitive and a resistant line as evidenced
`by a decrease in phosphorylation of the direct mTOR targets
`p70 S6K and 4E-BP1.
`We also looked at downstream cell cycle regulatory
`proteins that are reported to be modulated by mTOR (Fig.
`3). In the sensitive MDA-468 cells, we observed decreases
`in total c-myc and cyclin D3 protein levels after treatment
`with CCI-779 but not
`in the resistant MDA-435 cells.
`Similarly, we observed an increase in p27kip-1 levels in the
`sensitive MDA-468 cells, but not in the resistant cells. These
`data suggest
`that mTOR function is inhibited in both
`sensitive
`and
`resistant
`lines
`but
`the
`downstream
`in PTEN−/− MDA-468 cells,
`consequences are greater
`suggesting they are more dependent on mTOR function.
`Alternatively, there may be other targets of mTOR inhibition
`besides p70 S6K and 4E-BP-1 that are operative in sensitive
`lines but not in resistant lines.
`
`The effect of mTOR inhibition on the growth
`of estrogen-dependent MCF-7 cells
`
`MCF-7 cells do not contain a PTEN mutation or deletion but
`are growth inhibited by about 50% by treatment with 50 nM
`CCI-779. We looked by Western blot at
`the effect of
`CCI-779 treatment on proximal (p70 S6K, 4E-BP1) and
`downstream (c-myc, cyclin D3, p27kip-1) targets in these cells
`(Fig.
`4A). We
`found
`that MCF-7
`cells markedly
`overexpressed the mTOR target p70 S6K (data not shown).
`There were high levels of mTOR-dependent phosphorylated
`p70 S6K in these cells that were completely inhibited by
`CCI-779. Similarly, there was a shift to faster migrating
`species of 4E-BP1 in CCI-779-treated MCF-7 cells. These
`results suggest that phosphorylation of two specific targets of
`mTOR (p70 S6K and 4E-BP1) is inhibited by the drug. We
`also observed a slight decrease in c-myc and cyclin D3 levels
`in CCI-779-treated MCF-7 cells. The levels of p27kip-1 appear
`unchanged after CCI-779 treatment, although very high
`levels of p27kip-1 were seen in untreated MCF-7 cells making
`small changes difficult to detect. We have also looked at the
`status of other targets regulated by Akt signaling (Brunet et
`al. 1999). Phosphorylation of the forkhead transcription
`factor FKHRL-1 is highly elevated (Fig. 4B) compared with
`MDA-435 and MDA-231 cells. Similarly, we also observed
`
`www.endocrinology.org
`
`253
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 005
`
`

`
`Yu et al.: mTOR, a novel target in breast cancer
`
`Figure 4 Biochemical characterizations of cellular mTOR and AKT targets in MCF-7 cells. (A) Effects of CCI-779 on cellular
`mTOR targets. Cells were plated, treated and analyzed in a similar manner as described in Fig. 3. (B) Cells of the indicated cell
`lines were plated in complete growth medium in 10-cm culture plates. Total cell lysates were made from exponentially growing
`cells in the gentle lysis buffer. Fifty micrograms total proteins per lane were immunoblotted for phospho (P)-AKT (S473) and
`phospho(P)-FKHRL-1 (T32) in a similar manner as described in Fig. 1B.
`
`an increased phosphorylation of the Akt target GSK-3β (data
`not shown). Since MCF-7 cells do not have a high level of
`active Akt-1, the mechanism for elevated phosphorylation of
`FKHRL-1 and GSK-3β remains to be identified. It is also
`possible that deregulation of these targets may contribute to
`its sensitive response to inhibition of mTOR function.
`
`Cells resistant to mTOR inhibition contain
`lower levels of activated Akt than sensitive
`lines
`
`The phosphorylation status of 4E-BP1 is also affected by Akt
`(Gingras et al. 1999) and several laboratories have suggested
`that mTOR is either downstream of PI3-K씮Akt activation
`or activated in parallel with PI3-K/Akt to collaborate on the
`regulation of 4E-BP1 and p70 S6K. Therefore, we compared
`the levels of Akt phosphorylation in a panel of breast cancer
`lines containing cells sensitive or resistant to the mTOR
`
`inhibitor CCI-779 (Fig. 5). Cells were grown in 10% serum
`and harvested prior
`to achieving confluency. The two
`resistant lines, MDA-231 and MDA-435, showed the least
`phosphorylation of Akt relative to total Akt
`levels. The
`highest levels of phospho-Akt, as expected, were seen in the
`PTEN−/− MDA-468 cells. In the Her-2 overexpressing and
`ER positive BT-474
`cells, Akt was
`also
`highly
`phosphorylated and only slightly less so in SKBR-3 cells
`which also overexpress Her-2/neu. The estrogen responsive
`cells MCF-7 and T-47D were intermediate in the level of
`Akt phosphorylation.
`
`Discussion
`
`We have studied the effect of the mTOR inhibitor CCI-779
`on cell growth and cell signal transduction in a panel of
`human breast cancer cell lines. We found that most breast
`cancer
`lines were responsive to CCI-779. While the
`
`254
`
`www.endocrinology.org
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 006
`
`

`
`Endocrine-Related Cancer (2001) 8 249–258
`
`Figure 5 Survey of the phosphorylated AKT proteins in breast tumor lines. Indicated cell lines were plated in 10-cm dishes in
`complete growth medium. Lysates were prepared in the gentle lysis buffer and immunoblotted for phospho (P)-AKT (S473) and
`AKT in a similar manner as described in Fig. 1B.
`
`molecular basis of responsiveness to mTOR inhibition is not
`known, some observations from the literature regarding the
`responsive cells may be pertinent. Cells with known
`dependence on the hormonal growth factor estradiol or with
`aberrant expression of a growth factor receptor (Her-2/neu)
`were sensitive to in vitro growth inhibition by the mTOR
`inhibitor. Similarly, cells that had lost PTEN, a negative
`regulator of growth factor signaling through PI3-K, were also
`sensitive to CCI-779. These observations are consistent with
`the known effect of mTOR regulation
`on growth
`factor-induced proliferation (Sehgal et al. 1994, Wiederrecht
`et al. 1995). Cell lines in which growth was not inhibited
`by the mTOR inhibitor did not respond to estrogen, did not
`overexpress Her-2/neu, and were wild-type for PTEN.
`We chose one sensitive (MDA-MB-468) and one
`resistant (MDA-MB-435) line to study in vivo in nu/nu
`mouse
`xenografts
`and
`for
`further
`biochemical
`characterization of mTOR inhibition. In vivo, MDA-468
`tumors were inhibited by treatment with CCI-779 (10, 20, or
`40 mg/kg). At the higher doses there was regression of the
`staged tumors and even at the low dose of 10 mg/kg, the
`delay in growth extended 10 days beyond the last dose given.
`The MDA-435 line was not inhibited in the xenograft model
`at any of the doses tested, similar to the refractory phenotype
`observed in vitro.
`By Western blot we showed that the serine threonine
`kinase Akt was highly activated in exponentially growing
`MDA-468 cells but only minimally in MDA-435 cells.
`Activation of Akt in these cells results from loss of function of
`the PTEN tumor suppressor gene (Lu et al. 1999, Weng et al.
`1999) which negatively regulates PI3-K activation of Akt. In
`addition to Akt activation, p70 S6K (a downstream target of
`mTOR) was also highly phosphorylated on Thr-389 in
`MDA-468 cells but only marginally in MDA-435 cells,
`
`suggesting activation of the mTOR pathway in the sensitive
`but not in the resistant line. Treatment of both cell lines with
`CCI-779 resulted in complete dephosphorylation of p70 S6K
`at the mTOR-dependent site Thr-389. With respect to the other
`mTOR target 4E-BP1, treatment with CCI-779 resulted in a
`shift to faster migrating species in both sensitive and resistant
`lines, suggesting that mTOR-dependent phosphorylation of
`4E-BP1 was inhibited in both lines. The level of expression of
`4E-BP1 was markedly higher in the resistant MDA-435 cells
`and although there was a shift to the more dephosphorylated
`form after CCI-779 treatment, constitutive levels of the
`unphosphorylated form were high, suggesting a possible
`mechanism for MDA-435 cell resistance to mTOR regulation.
`This could occur if mTOR was able to phosphorylate only a
`portion of the 4E-BP1,
`leaving high residual
`levels of
`unphosphorylated 4E-BP1 bound to eIF-4E. Additional
`studies to determine if the unphosphorylated 4E-BP1 caused a
`difference in free 4E levels in growing MDA-435 versus
`MDA-468 cells will be necessary to address this possibility.
`Nevertheless, we have shown that
`the difference in
`responsiveness to CCI-779 in these two breast cancer lines is
`not due to the failure of CCI-779 to inhibit mTOR function in
`the resistant cells. Hosoi et al. (1998) have reported the same
`observation
`in
`rapamycin
`sensitive
`and
`resistant
`rhabdomyosarcoma cell lines.
`Although CCI-779 inhibited mTOR function in both
`sensitive and resistant cells, downstream of mTOR the
`response to CCI-779 was different in MDA-468 compared
`with MDA-435 cells (Fig. 3).
`Inhibition of mTOR by
`CCI-779 in MDA-468 cells resulted in decreased cyclin D3
`and c-myc levels and an increase in p27kip-1 levels. These
`effects were not seen in the resistant MDA-435 cells. Lu et
`(1999)
`reported that
`transfection of PTEN into the
`al.
`PTEN−/− MDA-468 cells decreased phosphorylation of p70
`
`www.endocrinology.org
`
`255
`
`Roxane Labs., Inc.
`Exhibit 1035
`Page 007
`
`

`
`Yu et al.: mTOR, a novel target in breast cancer
`
`S6K on the same residues as the mTOR inhibitor rapamycin.
`PTEN transfection also resulted in increased p27kip-1 levels in
`the MDA-468 cells. Thus, PTEN, like the mTOR inhibitor
`CCI-779, inhibited p70 S6K and increased p27kip-1. These
`data suggest that the growth advantage of cells that have lost
`PTEN may, at least in part, be mediated by mTOR. Our data
`showing an increase in p27kip-1 after CCI-779 treatment in
`MDA-468 cells support
`this contention and suggest
`that
`is regulated by PTEN in an mTOR-dependent
`p27kip-1
`manner. The effect of mTOR inhibition on p27kip-1 levels in
`normal T cells has long been known (Nourse et al. 1994).
`Taken together, these data suggest that loss of PTEN results
`in activation of mTOR and that mTOR inhibition may be
`therapeutically effective in breast tumors lacking PTEN.
`D-type cyclins have been reported to be overexpressed
`in human breast cancer (Weinstat-Saslow et al. 1995). The
`mTOR pathway can regulate both the stability and translation
`of D-type
`cyclins
`(Hashemolhosseini
`1998,
`et
`al.
`Muise-Helmericks et al. 1998). Reduction of cyclin D3 in
`the MDA-468 cells by CCI-779, coupled with the increase
`of p27kip-1 may result in redistribution of p27kip-1 towards
`cyclin E–CDK-2 complexes, thereby preventing entry into
`the S phase of the cell cycle.
`Breast lines that are wild-type for PTEN but dependent
`on estrogen were also sensitive to CCI-779. Western analysis
`of MCF-7 cells
`showed that
`these
`cells markedly
`overexpressed p70 S6K and had high levels of
`the
`mTOR-dependent phosphorylated form of p70 S6K when
`grown in 10% serum. The increased p70 S6K expression is
`a function of gene amplification (Barlund et al. 2000, Wu et
`al. 2000) and has been reported to occur in as many as 9%
`of primary breast cancers. Treatment of MCF-7 cells with
`CCI-779 completely inhibited phosphorylation of p70 S6K.
`Similarly, CCI-779 caused a nearly complete shift of 4E-BP1
`to the under-phosphorylated fast migrating species. Cyclin
`D3 levels in MCF-7 cells were reduced by CCI-779 by about
`50% and similar results were seen for cyclin D1 (data not
`shown). It has recently been shown that as little as 30–40%
`reduction in cyclin D1 levels by anti-estrogens in MCF-7
`cells is sufficient to induce a shift in p21cip1 molecules to
`cyclin E–Cdk2 complexes, causing inhibition of progression
`from G1씮S phase of the cell cycle (Carroll et al. 2000).
`Inasmuch as anti-estrogens inhibit D-type cyclin production
`at the transcriptional level and mTOR inhibition decreases
`D-type cyclins at the translational and/or protein stability
`level, there is a strong rationale for combining CCI-779 with
`anti-estrogen therapy.
`In experiments
`to be
`reported
`elsewhere, one of us (P Frost) has found that CCI-779 acted
`synergistically in combination with an anti-estrogen to inhibit
`proliferation of MCF-7 cells in vitro and also potentiated the
`effect of anti-estrogens in vivo in a MCF-7 mouse-xenograft
`model.
`mTOR inhibition also effectively inhibited proliferation
`of Her-2/neu-expressing cells BT-474 and SKBR-3 (Table
`
`1). Lee et al. (2000) recently showed that neu-dependent
`transformation requires cyclin D1 and is induced through an
`E2F-dependent signaling pathway. Although we did not
`directly study the effects of CCI-779 on D-type cyclins in
`the Her-2/neu overexpressing lines, the amply demonstrated
`effect of mTOR inhibition on D-type cyclin levels suggests
`a plausible mechanism for sensitivity of Her-2/neu positive
`tumors to mTOR inhibition.
`The upstream activator(s) of mTOR is not well
`characterized. Studies in yeast (Cardenas et al. 1999) and
`more recently in mammalian cells suggest that mTOR may
`act as a sensor to ensure appropriate nutritional status before
`the cells commit
`to division (Schmelzle & Hall 2000).
`Activation of Akt appears to be upstream of mTOR
`activation in that Akt has been shown to phosphorylate
`mTOR (Sekulic et al. 2000). However, mutation of the site
`on mTOR phosphorylated by Akt did not inhibit downstream
`signaling to p70 S6K or 4E-BP1 (Sekulic et al. 2000). This
`has led to the hypothesis that Akt and mTOR are activated
`by parallel pathways and converge to activate downstream
`targets (Gingras et al. 1999). Nevertheless, it appears that
`Akt and mTOR are activated by growth factors in a
`coordinated if not
`linear fashion and suggests that Akt
`activation may be a marker for enhanced mTOR dependency
`in tumor cells. We found good correlation between Akt
`activation and responsiveness to CCI-779 in the breast cancer
`lines studied. The two resistant lines, MDA-MB-231 and
`MDA-MB-435, showed the least activation of Akt as
`evidenced by phosphorylation of Akt and its downstream
`targets FKHRL-1 (Fig. 4B) and GSK-3β (data not shown).
`One exception was the MCF-7 line, which was sensitive to
`CCI-779 but did not show evidence of strong activation of
`Akt. MCF-7 cells did overexpress highly activated
`mTOR-dependent p70 S6K, suggesting that mTOR may be
`activated in an Akt-independent manner in these cells. The
`resistant line MDA-MB-231 has been shown to overexpress
`Akt-3 (Nakatani et al. 1999). There are three Akt isozymes,
`Akt 1–3, that are reported to be regulated similarly. Our data
`for MDA-MB-231 show no evidence of phosphorylation of
`downstream targets of Akt such as FKHRL-1 and GSK-3β,
`suggesting that even though Akt-3 is overexpressed, it does
`not appear to be constitutively active in these cells. This
`differs from PTEN-deficient cells where it has been shown
`the Akt
`is constitutively active,
`suggesting a greater
`dependency of these cells on the Akt and mTOR pathwa