mTOR as a Target for Cancer Therapy
`
`P. J. Houghton- S. Huang
`Department of Molecular Pharmacology, St. Jude Children's Research Hospital,
`332 N. Lauderdale, Memphis, TN 38105-2794, USA
`E-mail: peter.houghton@stjude.org
`
`Introduction
`
`.
`
`2
`
`3
`
`Autocrine Growth of Rhabdomyosarcomas . . . . . . .
`
`Selective Tumor Growth Inhibition by Rapamycin .
`
`4 Mechanism of Action of Rapamycin . . . . . . . . .
`
`5
`
`6
`
`Rapamycin Induces Apoptosis in Rhabdomyosarcoma Cells .
`
`The Tumor Suppressor p53 Protects
`from Rapamycin -Induced Apoptosis . .. ..
`
`7 Mechanism(s) of Resistance to Rapamycin ..
`
`8
`
`9
`
`Preclinical Antitumor Activity for Rapamycins .
`
`Clinical Trials for Antitumor Efficacy of Rapamycin s .
`
`10 Conclusions
`
`.
`
`References.
`
`. . . . .
`
`. . . .
`
`340
`
`340
`
`341
`
`343
`
`343
`
`347
`
`347
`
`351
`
`352
`
`353
`
`354
`
`Abstract The target of rapamycin, mTOR, acts as a sensor for mitogenic
`stimuli, such as insulin-like growth factors and cellular nutritional sta(cid:173)
`tus, regulating cellular growth and division. As many tumors are driven
`by autocrine or paracrine growth through the type -I insulin-like growth
`factor receptor, mTOR is potentially an attractive target for molecular(cid:173)
`targeted treatment. Further, a rationale for anticipating tumor-selective
`activity based on transforming events frequently identified in malignant
`disease is becoming established.
`
`G. Thomas et al.(eds.), TOR Target of Rapamycin
`© Springer-Verlag Berlin Heidelberg, 2004
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 001
`
`

`
`P. ]. Houghton and S. Huang
`
`340
`
`1 I
`
`ntroduction
`
`Since the early 1960s the introduction of cytotoxic agents, thence their
`dose intensification (starting in the 1970s), has dramatically improved
`survival of children with hematologic and solid tumors. For example,
`data from 1960-1963 shows that overall survival for children under fif(cid:173)
`teen years old with a diagno sis of neuroblastoma, bone or joint sarco(cid:173)
`mas, or CNS tumors was 25%, 20%, and 35%, respectively. In contrast,
`for the period 1985-1994 the survival for these same groups had in(cid:173)
`creased to 63%-65%. However, for soft tissue sarcomas, the focus of this
`laboratory, increases in sur vival have been less impressive, increasing
`from 60% in 1974 to 71% (1985-1994)1. Although these result s demon(cid:173)
`strate clear progress in treating childhood malignancies, they do not re(cid:173)
`flect the morbidity and long-term sequelae often associated with inten(cid:173)
`sive use of cytotoxic agents. Consequently, almost two decade s ago we
`started to study pediatric soft tissue sarcomas with the ultimate goal of
`developing novel therapeutic approaches based on specific biological
`characteristics of these tumors. In this chapter we will review the pro(cid:173)
`cess that allowed us to stumble onto rapamycin as a potential therapeu(cid:173)
`tic agent, and the progress in understanding why this macrolide antibi(cid:173)
`otic may exert tumor-specific cytotoxicity.
`
`utocrine Growth of Rhabdomyosarcomas
`
`2 A
`
`Our laboratory has focused on rhabdomyosarcoma, a tumor of skeletal
`mu scle origin, and in parti cular a particularly aggressive "alveolar" vari(cid:173)
`ant thereof. Cytogenetic analysis of several of the se tumors from inde(cid:173)
`pend ent patients that were established as xenografts in immune-de(cid:173)
`prived mice showed a consistent chromosomal
`translocation t(2:13)
`(q35;qI4; Hazelton et al. 1987). A more systematic sur vey of patient tu(cid:173)
`mor biopsies demonstr ated consistent
`translocations in greater than
`90% of alveolar rh abdomyosarcomas (Douglass et al. 1987). We now
`know that this translocation results in expression of a chimeric tran-
`
`I Years 1974 to presen t based on SEE R data [Ries LAG, Miller BA, Guemey JG,
`Linet M, Tamra T, Young JL, Bunin GR (cds)]. SEE R Progra m, 1975-1 995; Tables
`and Graphs, Nationa l Cancer Institute. NIH Pub. No. 99-4649. Bethesda, MD, 1999
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 002
`
`

`
`mTOR as a Target for Cancer Therapy
`
`341
`
`scription factor (PAX3/FKHR) that appears to block myogenic differen(cid:173)
`tiation (Galili et al. 1993; Shapiro et al. 1993; Epstein et al. 1995) leading
`to tumor formation. These studies lead to further characterization of al(cid:173)
`veolar rhabdomyosarcoma cell lines, and revealed overexpression of
`transcripts for type II insulin-like growth factor (IGF-II) specifically
`from the fetal P3 promoter. In collaboration with Lee Helman at the Pe(cid:173)
`diatric Branch, National Cancer Institute (NCI), Bethesda, we were able
`to show that growth of rhabdomyosarcoma cells was driven by an auto (cid:173)
`crine loop. Specifically, cells secreted IGF-II, and signaled through the
`IGF-I receptor (El-Badry et al. 1990). Inhibition of the IGF-I receptor by
`using a neutralizing antibody inhibited tumor cell growth. This suggest(cid:173)
`ed to us that interference with IGF-I-receptor-mediated signaling may
`be a therapeutic strategy for these tumors. To test this concept, the IGF(cid:173)
`1 receptor was downregulated using a stable expression of antisense con(cid:173)
`structs (Shapiro et al. 1994). These studies showed a high correlation be(cid:173)
`tween downregulation of the receptor, decreased growth in soft agar
`(such growth being a characteristic of malignant cells), and decreased
`formation of tumors when cells were inoculated into immune-deprived
`mice. Further, clones with the lowest expression of IGF-I receptor ex(cid:173)
`pressed the highest levels of the myogenic marker, MyoD, and formed
`multinucleate syncytia, thus recapitulating some characteristics of myo(cid:173)
`genic differentiation. While our studies focus on rhabdomyosarcomas,
`there is increasing evidence that many tumors are "driven" by either au(cid:173)
`tocrine or paracrine signaling through the IGF-I receptor. Clearly dereg(cid:173)
`ulated IGF-I signaling is frequent in many pediatric solid tumors (neu(cid:173)
`roblastoma, Ewing's sarcoma, Wilms' tumor, medulloblastoma, glioblas(cid:173)
`toma), as well as many adult carcinomas (Macauly, 1992; Toretsky and
`Helman 1996). Direct inhibition of the IGF-I receptor with antibody,
`while effective in mouse models (Kalebic et al. 1994), was not at that
`time plausible in humans due to antigenicity, although this is now less of
`a problem and such reagents are being developed for clinical application.
`Alternatively, we sought a small molecule inhibitor ofIGF-I signaling.
`
`3 S
`
`elective Tumor Growth Inhibition by Rapamycin
`
`Our studies with rapamycin started in 1992, and were stimulated by a
`chance conversation with Randall Johnson at Smith Kline Beecham. At
`that time it was known that the immunosuppressive agents FK506 and
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 003
`
`

`
`342
`
`P. J. Houghton and S. Huang
`
`Table 1 Sensitivity of childhood rhabdomyosarcoma cell lines and human colon car(cid:173)
`cinoma celIlines to rapamycin and geldan amycin
`
`Rapamycin ICso (nglml)
`
`Geldanamycin Cso (nM)
`
`Rhabdomyosarcoma cell lines
`Rhl
`Rh18
`Rh28
`Rh30
`Colon carcinoma cell line s
`GC3/cl
`VRCs/cl
`CaCo
`HCT8
`HCT29
`HCn 16
`National Cancer Institute
`screen (60 celIlines)
`
`4,680
`0.1
`8.0
`0.37
`
`9,800
`1,280
`1,570
`8,400
`> 10,000
`> 10,000
`3,160
`
`5.9
`14.3
`17.9
`1.9
`
`3.6
`1.4
`3.4
`NDa
`2.6
`ND
`
`a ND, not determined (from Dilling et aI. 1994).
`
`cyclosporin A blocked T cell activation prior to expression of interleu(cid:173)
`kin-2, whereas rapamycin acted downstream of interleukin-2 expression
`(Flanagan et al. 1991; Schreiber and Crabtree 1992; McCaffrey et al.
`1993). There was also some suggestion that for T cells to progress to S
`phase,
`the IGF-I receptor had to be expressed (Reiss et al. 1992). We
`speculated that perhaps rapamycin acted downstream of the IGF-I re(cid:173)
`ceptor to block cell cycle progression, and if so, may act to inhibit the
`growth of rhabdomyosarcoma cells. The results, shown in Table 1, were
`both surprising and exciting. Three of four rhabdomyosarcoma cell lines
`were exquisitely sensitive to rapamycin whereas one line, (Rh l ), which
`is less dependent on IGF-I mitogenic signaling , was highly resistant. The
`other interesting aspect of these results was the marked selectivity for
`rhabdomyosarcoma cells relative to colon carcinoma cells, or cells used
`in the NCI in vitro screen (Dilling et al. 1994). Intriguing, but not com(cid:173)
`prehended (at that time) was the observation that under serum-free con(cid:173)
`ditions Rh l cells became very sensitive to rapamycin, with the IC50 de(cid:173)
`creasing from 5,800 ng/ml
`to 3.6 ng/ml. Consistent with results from
`other laboratories,
`the effect of rapamycin could be competed using
`FK506, indicating that initial formation of the FKBP-rapamycin complex
`was important.
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 004
`
`

`
`mTOR as a Target for Cancer Therapy
`
`343
`
`echanism of Action of Rapamycin
`
`4 M
`
`The action of rapamycin will be covered in detail elsewhere. Briefly, ra(cid:173)
`pamycin first binds the immunophilin FKBP-12 (the 12-kDa FK506
`binding protein) and this complex is now known to be a specific inhibi(cid:173)
`tor of a serine/threonine kinase mTOR (the mammalian target of ra(cid:173)
`pamycin, also called FRAP/RAPT/RAFT; Brown et al. 1994; Sabatini et
`al. 1994; Chiu et al. 1994; Sabers et al. 1995). Kinase mTOR is a member
`of the PIKK superfamily that includes ATR, ATM, Mecl, and Tell pro (cid:173)
`teins having homology to phosphatidylinositol lipid kinases. Evidence
`increasingly implicates mTOR as a central controller of cell growth and
`proliferation, and it controls initiation of translation of ribosomal pro(cid:173)
`teins and several proteins that regulate cell cycle. Activation of ribosom(cid:173)
`al S6Kl after mitogen stimulation is dependent on mTOR (Chung et al.
`1992; Kuo et al. 1992; Terada et al. 1992). Cap-dependent translation is
`facilitated by mTOR's phosphorylation and inactivation of 4E-BPs, sup (cid:173)
`pressors of eukaryotic initiation factor 4E (eIF4E; Lin et al. 1994, Pause
`et al. 1994, Beretta et al. 1996). More recent findings have shown that
`mTOR may directly or indirectly control transcription, ribosomal bio(cid:173)
`genesis, actin cytoskeleton organization, and protein kinase C (reviewed
`in Schmelzle and Hall, 2000). Recently, our studies with rhabdomyosar(cid:173)
`coma cells showed that activation of MAP kinases (p44/42) by growth
`factors was mTOR-dependent (Houghton et al. 2001; Harwood et al.,
`manuscript submitted),
`implicating mTOR in cross-talk between the
`PI3K and MAPK pathways in some cell lines. Thus, the emerging picture
`places mTOR in a central role in which it senses mitogenic stimuli and
`amino acid (Iiboshi et al. 1999), ATP (Dennis et al. 2001), or nutrient
`(Rohde et al. 2001) conditions; mTOR coordinates many cellular pro(cid:173)
`cesses related to growth and proliferation. The signaling pathways from
`IGF-IR to mTOR and downstream targets are depicted in Fig. 1.
`
`apamycin Induces Apoptosis in Rhabdomyosarcoma Cells
`
`5 R
`
`The basis for the differential sensitivity of Rhl cells under serum-con(cid:173)
`taining or serum-free conditions was of interest. The results suggested
`that some component of serum was able to rescue Rh l cells but not
`Rh30 cells. Consequently, we attempted to identify components of serum
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 005
`
`

`
`344
`
`P. J. Houghton and S. Huang
`
`Ribosomal
`proteins
`IGF-II
`
`Cyelin DI
`ODe.
`c- M YC
`, I IIIF I
`
`,=-
`
`-
`
`Fig. 1 Simplified schema of signaling pathways from the IGF-I receptor to mTOR
`and Ras in mammalian cells
`
`that protected from rapamycin (Hosoi et al. 1999). As shown in Fig. 2,
`under serum-free conditions IGF-I completely protected Rh l cells, but
`had no apparent effect on the action of rapamycin against Rh30 cells.
`Further, IGF-I rescue was not a consequence of reversing rapamycin in(cid:173)
`hibition of ribo somal S6Kl activation, as S6Kl activity remained sup(cid:173)
`pressed. So far the only growth factors shown to protect Rh1 cells are
`IGF-I, IGF-II, and to a lesser extent insulin (Thimmaiah et al. 2003). As
`IGF-I has been reported to prevent apoptosis in cells undergoing differ(cid:173)
`ent forms of stress (Butt et al. 1999; Fujio et al. 2000; Kulik et al. 1997;
`Sell et al. 1995), we examined the fate of Rh l and Rh30 cells treated un(cid:173)
`der serum-free conditions with or without IGF-I supplementation. As
`shown in Fig. 3, under serum-free conditions rapamycin induces quite
`dramatic apoptosis of both Rhl and Rh30 cells, and IGF-I essentially
`completely protects against the effect of rapamycin. Thus , under serum(cid:173)
`free conditions the response to rapamycin is apoptosis, whereas under
`seru m-containing conditions Rh30 cells are growth-arrested, but Rhl
`cells continue to proliferate. Interestingly, expression of a rapamycin re(cid:173)
`sistant mTOR mutant (Ser2035-> Ile) conferred resist ance to both the
`growth-inhibiting and apoptosis-inducing activities of rap amycin. Thus,
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 006
`
`

`
`mTOR as a Targe t for Cancer Therapy
`
`345
`
`A
`
`1 20
`
`Rh1
`
`___ t
`
`12 0
`
`100
`
`Rh30
`
`100
`
`.._-.t- .-~ 80
`
`e~
`
`8 0
`
`o"
`
`60
`
`40
`
`20
`
`O L..J...---'-
`
`-'--
`
`---'--
`
`--'---
`
`-'----'
`
`100 1000
`1 0
`1
`0 .1
`0.01
`Rapamycin (ng/ml)
`
`0
`
`2 , '
`
`L'
`
`.,.
`
`Q'!.
`:;;6 0
`
`~ C
`
`)40
`
`OJo
`
`20
`
`o L..-!._'--_ -'-_ -'-_ -'-_ -'-_ .....
`o
`
`10
`1
`0 .1
`0 .0 1
`Aapamycin (ng/mll
`
`1 0 0 10 0 0
`
`2 ,'
`
`B
`
`m
`:;
`
`cE
`
`~1
`.2:'t:l
`u~
`c( ~ 1 .5
`m(cid:173)
`
`.. "'"'cc'"~::
`
`1
`
`., .!
`..
`IIIc.c
`,..0.
`o.~-a.,.
`
`a:
`
`:E
`0.
`
`~
`
`.
`..
`
`l:
`.t
`..
`E
`•
`;;
`
`e
`:i
`...
`!2
`
`e
`:i
`...
`~
`!2
`
`..
`c
`:i
`...
`...
`!2
`
`s:
`:i
`
`..
`0.
`...
`a:
`s
`
`e
`:i
`~
`0.
`
`..
`...
`0:
`S!
`
`e
`:i
`...
`..
`0
`...
`...
`S!
`
`0:
`
`.
`..
`
`E
`;;
`
`e
`:i
`
`. ...
`...
`~
`..
`E
`;;
`
`!2
`
`e
`2
`:e
`...
`!2
`
`e
`2
`II
`!2
`
`..
`
`e
`:i
`...
`..
`0.
`...
`0:
`!2
`
`e
`:i
`=
`..
`
`0.
`~
`
`!2
`
`e
`:i
`..
`~
`...
`..
`
`0:
`!2
`
`Fig.2A, B Inhibitory effect of rapamycin and effect of IGF-I on Rh i (left panels) and
`Rh30 (right panels) rhabdomyosarcoma cells. A Rapamycin sensitivity. Cells were seeded
`and allowed to attach overnight, washed, and cultured in serum-free N2E in medium con(cid:173)
`taining serial concentr ations of rapamycin with (closed square) or without (open square)
`addition of IGF-I (10 ng/ml) for 7 days, Growth was assessed by lysing cells and counting
`nuclei. Results are presen ted as percent control. Cell nu mber for control Rhl cells was
`I.74xl05, and for Rh30 was s.zxio', respectively. B Activation of p70S6 K by IGF-I is
`blocked in Rhl (left panel) and Rh30 (right panel) by rap amycin, Cells were serum(cid:173)
`starved overn ight, then stimulated with IGF-I (IO nglml) without or with prein cubation
`for 15 min with rapamycin (IOOng/ml). Ribosomal p70S6 K assays were performed on im(cid:173)
`munoprecipitates derived from 3x106 cells. Each value repre sents the mean±SD for 4 de(cid:173)
`terminations, and shows a representative experi ment. (From Hosoi et al. 1999)
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 007
`
`

`
`346
`
`P. J. Houghton and S. Huang
`
`Rh30
`
`~tit=j B. Serum -free
`
`C. Rapamycin 100ng/ml
`
`D. Rapamy cin 100n g/m l
`IGF- 1 10ng /ml
`
`,a,
`
`'0'
`
`10'
`
`Annexin V-FITC Fluorescence
`
`Fig.3 Rapamycin induces apoptosis that is blocked by IGF-I. Rhl (left panels), and
`Rh30 cells (right panels) were grown und er seru m-containing (FBS) or serum-free
`(N2E) culture conditions supplemented with rapamycin (lOO ng/m!) or rapamycin
`plus IGF-I (lO nglml). Cells were har vested after 6 (RhI) or 4 days (Rh30), and ap(cid:173)
`optosis measured by the ApoAlert method. Cells were analyzed by FACS for annex(cid:173)
`in-V-FITC and propidium iodide fluorescence. (From Hosoi et al. 1999)
`
`apoptosis is a consequence of inhibiting mTOR, and not through a sec(cid:173)
`ond mechanism of action. Of note is that both RhI and Rh30 cells are
`mutant for the p53 tumor suppressor gene. This raised the possibility
`that p53 may be involved in sensing mTOR inhibition and cooperating
`in enforcing a G, arrest.
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 008
`
`

`
`mTOR as a Target for Cancer Therapy
`
`347
`
`6 T
`
`he Tumor Suppressor p53 Protects
`from Rapamycin-Induced Apoptosis
`
`Cellular response to mTOR inhibition by rapamycin is generally charac(cid:173)
`terized by p53-independent (Metcalfe et al. 1997) cytostasis without cell
`death. In contrast, our studies have shown that
`the response to ra(cid:173)
`pamycin in rhabdomyosarcoma cells lacking functional p53 is apoptosis
`(Hosoi et al. 1999; Huang et al. 2001). Apoptotic cells appeared to pro(cid:173)
`gress from Gl to S phase, as >90% had incorporated BrdUrd in the pres(cid:173)
`ence of rapamycin. To determine whether cellular response to rapamycin
`was determined by p53 function, Rh30 cells were infected with aden(cid:173)
`oviruses expressing wild-type p53 or the cyclin-dependent kinase inhib(cid:173)
`itor p21Cipl (Huang et al. 2001). Expression of p53 or p21cipl protected
`cells. We speculated that the mechanism was probably an enforcement
`of Gl arrest and prevention of cells from initiating DNA replication.
`More recent data (Huang et al. 2003) suggest that cell cycle arrest is not
`the mechanism, as expression of a truncated p21Cipl allele lacking the
`nuclear localization signal does not cause G1 accumulation, but protects
`against rapamycin-induced apoptosis.
`To further investigate the role of p53 and p21Cipl
`in a non-trans(cid:173)
`formed background we used murine embryo fibroblasts (MEFs) with
`disrupted p53 or p21c ipl genes. Specifically, MEFs with disruption of ei(cid:173)
`ther p53 or p21cipl undergo apoptosis whereas wild-type cells arrest in
`Gl without loss of viability when treated with rapamycin (Fig. 4). Re-ex(cid:173)
`pression of wild-type p53 confers resistance to rapamycin-induced ap(cid:173)
`optosis (Huang et al. 2001). Exactly how mTOR signaling feeds into p53
`remains to be determined. However, of importance for potential thera(cid:173)
`peutic application to cancer treatment is that rapamycin is selectively
`toxic to cells with attenuated p53-mediated Gl checkpoint responses, at
`least in the absence ofIGF-I.
`
`echanism(s) of Resistance to Rapamycin
`
`7 M
`
`Rapamycin resistance was first found in yeast (Heitman et al. 1991).
`Mechanisms of rapamycin resistance are multiple and complicated;
`some of them have been identified and some remain to be elucidated.
`Cells may acquire resistance with or without mutagenesis. For instance,
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 009
`
`

`
`348
`A.
`.,.r-- .,....-- ...,
`
`P. J. Houghton and S. Huang
`
`B.
`,,' y---
`
`.,--
`
`...,
`
`g
`8
`"j
`...
`'"
`'2
`~
`E
`.2
`X
`0
`0.:.
`
`\.,.lo'_..
`
`p53 ·/·
`C<>ntrol
`
`p21 ·' ·
`C ontrof
`
`8
`c:
`~
`'"~
`ii:
`~
`:B
`§
`:2a.
`£
`
`,,'
`
`..I
`
`\
`
`I
`
`.
`.".
`:' V
`
`j
`
`!\.. '\
`I
`\
`
`p53 . '+
`Rap 100
`
`p53 -I.
`Control
`
`p53 .f·
`Rap 100
`.h.
`\
`
`"-
`,
`
`p53 ./.
`Ad-p53
`
`\.A .~
`
`p53 .f·
`Ad-p53
`• Rapl00
`
`\,
`
`w ,IS'
`,OJ' wI
`10' 1iT .0'
`trI'
`Anne llin Y·FITC Fluonlscence
`
`to'
`
`td 10' ttl
`,,'
`to'
`1ft'
`10'
`Anne xin Y· FITC Fluorescence
`
`\G'
`
`Fig. 4 The cellular response to rapamycin is dependent on functional p53 or p21cip l
`in murine embryo fibrobl asts. Left: Wild-type, p53-/', and p21-/- MEFs were grown
`und er serum-free condition s with or without rapamycin (IOO ng/ml). After 5 days
`cells were harvested and apopt osis determined by ApoAlert assay. Result s show a
`representative experiment. Percent distribution of cells in each qu adrant is present(cid:173)
`ed. Right: Wild-type, and p53-/- MEFs, and p53-1- MEFs infected with Ad-p53 (MOl of
`100) were grown without or with rapamycin (100 ng/ml). Cells were harvested after
`5 days and apoptosis determined by quantitative FACS analy sis (ApoAlert) , as de(cid:173)
`scribed in Fig. 3. The percent distribution of cells in each quad rant is presented. Re(cid:173)
`sults show a representative experiment. (From Huang et al. 2001)
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 010
`
`

`
`mTOR as a Target for Cancer Therapy
`
`349
`
`specific mutations in FKBP12 that prevent the formation of FKBP-ra(cid:173)
`pamycin complex (Fruman et al. 1995), or certain mutations in FKBP-ra(cid:173)
`pamycin-binding domain of mTOR (Ser2035~Ile) that block binding of
`FKBP-rapamycin complex to mTOR (Chen et al. 1995; Hosoi et al. 1998),
`conferred rapamycin resistance. Mutation of Thr389 ~Glu (Dennis et al.
`1996) or Thr229~Ala or Glu (Sugiyama et al. 1996) of S6K1, a down(cid:173)
`stream effector of mTOR, also rendered S6K1 insensitivity to rapamycin.
`As discussed in other chapters, the signaling pathway directly down(cid:173)
`stream of mTOR bifurcates (von Manteuffel et al. 1997). In mammalian
`cells, mTOR signals to both S6K1 and 4E-BP, and probably to other
`unidentified effectors. We were interested in understanding whether cel(cid:173)
`lular responses to rapamycin (growth inhibition and apoptosis) were
`dictated by inhibition of signaling to either S6K1 or 4E-BP, or both. One
`approach to understanding which of these pathways is critical to cellular
`proliferation and survival was to select for rapamycin resistance and de(cid:173)
`termine which pathway(s) altered to accommodate rapamycin inhibition
`of mTOR. Rapamycin-resistant clones (Rh30/RapalOK) were obtained
`by continuously growing Rh30 cells in the presence of increasing con(cid:173)
`centrations of rapamycin, without prior mutagenesis (Hosoi et al. 1998).
`These initial studies showed that in resistant clones, rapamycin still in(cid:173)
`hibited IGF-I-stimulated mTOR-dependent activation of S6Kl. However,
`of note, resistance was characterized by elevated c-MYC. Further inde(cid:173)
`pendent clones were selected, either with or without mutagenesis (Dil(cid:173)
`ling et al. 2002). Interestingly, resistance was unstable in each of the
`clones characterized. When the selecting pressure (i.e., rapamycin) was
`withdrawn, cells reverted to rapamycin sensitivity within 6-10 weeks. In
`resistant cells, as compared to parental cells, approximately tenfold less
`4E-BP was bound to eIF4E, and total cellular 4E-BP was markedly re(cid:173)
`duced . In contrast, levels of eIF4E were unchanged. Steady-state levels of
`4E-BP transcripts remained unaltered, but the rate of 4E-BP synthesis
`was reduced in resistant cells. Of importance, in cells that reverted to ra(cid:173)
`pamycin sensitivity, levels of total 4E-BP returned to those of parental
`cells. Compared to parental cells, resistant clones had either similar or
`lower levels and activity of ribosomal S6K1, but c-MYC levels were ele(cid:173)
`vated in both resistant and revertant clones. Further, anchorage-inde(cid:173)
`pendent growth was enhanced in c-MYC-overexpressing cells, suggesting
`that the eIF4E pathway controls some aspects of the malignant pheno(cid:173)
`type (Lazaris-Karatzas et al. 1990; DeBenedetti et al. 1994). These results
`indicate that accommodation in the eukaryotic-initiation-factor 4E
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 011
`
`

`
`350
`
`P. J. Houghton and S. Huang
`
`_
`
`4E·BPI
`
`_
`
`fH ubo lin
`
`_
`
`_
`
`eIF4E
`
`13 ·tubu lon
`
`co
`l(cid:173)o
`J:
`
`HCT8·4E-BPI clon es
`
`2
`
`3
`
`4
`
`5
`
`I
`
`B.
`
`4E-BP1 . .
`
`elF4E . .
`
`c.
`
`l00 '-lli!= _ _
`
`ec
`8 60
`
`40
`
`C2
`
`lCii
`a.
`
`o i~~_......
`,--__"",
`"","--
`lit
`1000
`100
`10
`1
`Rapamyc in (ng/ml)
`
`!
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 012
`
`

`
`mTOR as a Target for Cancer Therapy
`
`351
`
`(eIF4E) pathway occurs in rapamycin resistance through suppression of
`translation of 4E-BP. This suggests that
`inhibition of cap-dependent
`translation mediated by eIF4E is critical for rapamycin-induced growth
`arrest and apoptosis. Further, intrinsic or acquired resistance may be a
`consequence of low levels of 4E-BP suppressor proteins. To examine
`whether intrinsic resistance was associated with low 4E-BP, we examined
`levels in colon carcinoma cell lines previously characterized as ra(cid:173)
`pamycin resistant (Dilling et al. 1994). Several of these cell lines with in(cid:173)
`trinsic rapamycin resistance were found to have low 4E-BP relative to
`eIF4E (Fig. 5A). To explore the role of 4E-BP in rapamycin resistance,
`stable clones of HCT8 colon carcinoma were engineered to overexpress
`4E-BP. As shown in Fig. 5, rapamycin sensitivity increased markedly
`(> l,OOO-fold) as 4E-BP expression increased. These results suggest that
`the 4E-BP:eIF4E ratio is an important determinant of rapamycin resis(cid:173)
`tance in some cell lines. However,
`this mechanism clearly does not ac(cid:173)
`count for resistance in other cell lines.
`
`reclinical Antitumor Activity for Rapamycins
`
`8 P
`
`Initial in vivo testing of rapamycin as a potential antitumor agent was
`undertaken through the Division of Cancer Treatment at the National
`Cancer Institute (NCI; Douros and Suffness 1981; Houchens et al. 1983;
`
`Fig.5A-C Overexpression of 4E-BP abrogates resistance to rapamycin. A Western
`blot analysis of 4E-BP,elF4E, and tubulin (loading control) in cell lines that have dif(cid:173)
`ferent intrinsic sensitivities to rapamycin. Colon carcinoma cell lines CaCo2, GC3/cl,
`HCTS, HCT29, Hcn I6, and VRC5/cI are intrinsically resistant to rapamycin with
`IC50 concentrations >1,200 ng/ml. Pediatric solid tumor lines SJ-G2 (glioblastoma)
`and RhlS and Rh30 (rhabdomyosarcoma) are sensit ive to rapamycin (lC50 < I ng/
`ml). B Expression of 4E-BP and elF4E in HCTS clones stably transfected with a 4E(cid:173)
`BP expression plasmid (pcDNA3-PHAS-I). Expression of 4E-BP was greater in clones
`C2, C4, and C5 than in parental HCT8 cells, but expression of was similar in parental
`and CI and C3 transfected clones. C Sensitivity to rapamycin. Cells were plated at
`low density in increasing concentrations of raparnycin, and colonies were counted
`after 7 days of exposure to rapamycin. Symbols: Parental HCT8 (closed circle) and
`clones Cl (open circle), C2 (closed square), C3 (open square), C4 (closed triangle),
`and C5 (open triangle) . Each point is the mean±SD of three determinations. (From
`Dilling et al. 2002)
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 013
`
`

`
`352
`
`P. J. Houghton and S. Huang
`
`Eng et al. 1984). These preliminary results revealed that rapamycin sig(cid:173)
`nificantly inhibited the growth of syngeneic B16 melanoma, colon carci(cid:173)
`noma models 26 and 38 tumor, CD8Fl mammary tumor, and EM epen(cid:173)
`dymoblastoma. Rapamycin, in the active dose range, was less toxic than
`conventional antitumor drugs, such as 5-fluorouracil, cyclophospha(cid:173)
`mide, and adriamycin (Eng et al. 1984). However, at that time Ayerst Re(cid:173)
`search Laboratories (Montreal, Canada), where rapamycin was first iso(cid:173)
`lated and characterized, abandoned rapamycin as an antitumor agent
`because they failed to develop a satisfactory intravenous formulation for
`clinical
`trials. Recently, two rapamycin ester analogues, CCI-779 [ra(cid:173)
`pamycin-42, 2, 2-bis(hydroxymethy1)-propionic acid; Wyeth-Ayerst] and
`RADOOI [everolimus, 40-0-(2-hydroxyethyl)-rapamycin; Novartis, Basel,
`Switzerland], with improved pharmaceutical properties have been syn(cid:173)
`thesized and evaluated. CCI-779 is designed for intravenous injection,
`whereas RADOOI for oral administration. Both have antitumor effects
`similar to rapamycin, are currently being developed as antitumor agents,
`and are undergoing phase I-III clinical trials. In culture, rapamycin and
`CCI-779 potently inhibit growth of numerous malignant cell lines, in(cid:173)
`cluding those derived from rhabdomyosarcoma, neuroblastoma, glio(cid:173)
`blastoma, medulloblastoma, small cell lung cancer (Dilling et al. 1994;
`Shi et al. 1995; Seufferlein and Rozengurt, 1996; Hosoi et al. 1998; Geo(cid:173)
`erger et al. 2001; Dudkin et al. 2001), osteoscarcoma (Ogawa et al. 1998),
`pancreatic carcinoma (Grewe et al. 1999; Shah et al. 2001), breast and
`prostate carcinoma (Gibbons et al. 2000; Yu et al. 2001), murine melano(cid:173)
`ma, T-cell
`leukemia, and B-cell lymphoma (Houchens et al. 1983;
`Hultsch et al. 1992; Gottschalk et al. 1994; Muthukkumar et al. 1995).
`CCI-779 in vivo also inhibited the growth of human U251 malignant gli(cid:173)
`oma cells that were resistant
`to rapamycin in vitro (Geoerger et al.
`2001). CCI-779 induced tumor growth inhibition correlated with de(cid:173)
`creased phosphorylation of 4E-BP (Dudkin et al. 2001).
`
`9 C
`
`linical Trials for Antitumor Efficacy of Rapamycins
`
`As alluded to, RADOOI and CCI-779 are undergoing phase I and phases
`II/III clinical trials, respectively. However, only preliminary results from
`phase I trials of CCI-779 are currently available. Two groups conducted
`phase I trials for CCI-779 using different schedules (Hidalgo et al. 2000;
`Raymond et al. 2000). In the USA, CCI-779 was administered as a 30-
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 014
`
`

`
`mTOR as a Target for Cancer Therapy
`
`353
`
`min intravenous infusion daily for 5 days every 2 weeks at doses of
`15-24 rng/rrr'per day. Of 45 patients with various types of cancer, nine of
`them showed some evidence of tumor response (Hidalgo et al. 2000). In
`Europe, CCI-779 was administered as a 30-min intravenous infusion
`weekly at doses of 7.5-220 mg/rrr' per week (Raymond et al. 2000). After
`2:8 weekly doses, of 18 patients, significant tumor regressions were ob(cid:173)
`served in two patients with lung metastasis of renal cell carcinomas
`(both treated with 15 mg/rrr' per week) and in one patient with a neuro(cid:173)
`endocrine tumor of the lung treated with 22.5 mg/rn! per week. Two pa(cid:173)
`tients experienced tumor stabilization. Consistently, these two groups
`showed that CCI-779 was well tolerated in patients with only mild side
`effects, such as acneform rash, mild mucositis, some thrombocytopenia,
`and elevated triglyceride and cholesterol levels (Hidalgo and Rowinsky
`2000). It is anticipated that these agents will be evaluated against addi(cid:173)
`tional histiotypes in phase II trials.
`
`10
`Conclusions
`
`Rapamycin and its derivatives represent novel agents for therapy of hu(cid:173)
`man cancer. In this review we have detailed some of the studies from this
`and other laboratories that have assisted in development of this class of
`drug specifically as cancer chemotherapeutic agents. Other studies,
`many reported in this book, have elucidated the mechanism of action of
`rapamycin, and have identified pathways proximal and distal to mTOR
`that have lead to a greater understanding of how this class of drug exerts
`cytostatic activity. Our studies have suggested that cellular response to
`rapamycin is converted from G1 cytostasis to apoptosis when p53 is mu(cid:173)
`tated in tumor cells, or disrupted in embryo fibroblasts. These results
`suggest that combining rapamycin, or its derivatives, with an inhibitor
`of the IGF-I receptor may induce apoptosis selectively in tumor cells
`with mutated p53. As both humanized antibodies against this receptor
`and small molecule inhibitors are in development, the validity of this ap(cid:173)
`proach will be amenable to testing relatively quickly. Recently, Neshat et
`al. (2001) have shown that PTEN (phosphatase and tensin homolog
`deleted on chromosome ten) -mutated or PTEN-deficient
`tumor cells
`were more sensitive to CCI-779 (Podsypanina et al. 2001; Neshat et al.
`2001). The PTEN-mutated tumor cells demonstrated elevated levels of
`phosphorylated Akt and activated S6Kl (Podsypanina et al. 2001). Loss
`
`Roxane Labs., Inc.
`Exhibit 1034
`Page 015
`
`

`
`354
`
`P. J. Houghton and S. Huang
`
`of PTEN by deletion or mutation has been reported in approximately
`50% of all solid human tumors (Simpson and Parsons 2001). Thus, there
`are at least two mechanisms by which to anticipate tumor-selective ac(cid:173)
`tivity of rapamycin analogues as cancer chemotherapeutic agents.
`
`Acknowledgment. We wish to dedicate this review to the memory of Charles B. Pratt,
`M.D., who dedicated his life to helping children with cancer at St. Jude and through(cid:173)
`out
`the world . Studies from this laboratory were supported by USPHS awards
`CA23099, CA77776, CA96696 and CA21765 .
`
`References
`
`Beretta L, Gingras AC, Svitkin YV, Hall MN, Sonenberg N (1996) Rapamycin blocks
`the phosphorylation of 4E-BPI and inhibits cap-dependent initiation of tra nsla(cid:173)
`tion . EMBO J 15:658-664
`Brown EJ, Albers MW, Shin TB, Ichikawa K, Keith CT, Lane WS, Schreiber SL (1994)
`A mammalian protein targeted by Gl-arresting rapamycin-receptor complex. Na(cid:173)
`ture 369:756-758
`Butt A.J., Firth S.M. and Baxter R.C. (1999) The IGFaxis and programmed cell death .
`Immunol and Cell BioI. 77:256-262
`Chen J, Zheng XF, Brown EJ, Schreiber SL (1995) Identification of an ll-kDa
`FKBPI2-rapamycin-binding domain within the 289-kDa FKBPI2-rapamycin-as(cid:173)
`sociated protein and characterization of a critical serine residue . Proc Nat! Acad
`Sci USA 92:4947-4951
`Chiu MI, Katz H, Berlin V (1994) RAPTl, a mammalian homolog of yeast Tor, inter(cid:173)
`acts with the FKBPl2/rapamycin complex . Proc Nat! Acad Sci USA 91:12574(cid:173)
`12578
`Chung J, Kuo CJ, Crabtree GR. Blenis J.(1992) Rapamycin -FKBP specifically blocks
`growth-dependent activation of and signaling by the 70 kd S6 protein kinases.
`Cell 69:1227-1236
`Damiens E (2000) Molecular events that regulate cell proliferation : an