Pharmacology & Therapeutics 162 (2016) 35–57

Contents lists available at ScienceDirect

Pharmacology & Therapeutics

journal homepage: www.elsevier.com/locate/pharmthera

Associate Editor: Y. Zhang

Pharmacological modulation of oncogenic Ras by natural products and

their derivatives: Renewed hope in the discovery of novel anti-Ras drugs
Shun Ying Quah a,1, Michelle Siying Tan a,1, Yuan Han Teh a,1, Johnson Stanslas a,b,⁎
a
Pharmacotherapeutics Unit, Department of Medicine, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor, Malaysia
b
Laboratory of Natural Products, Institute of Bioscience, Universiti Putra Malaysia, 43400 UPM, Serdang, Selangor, Malaysia

a r t i c l e i n f o a b s t r a c t

Available online 22 March 2016 Oncogenic rat sarcoma (Ras) is linked to the most fatal cancers such as those of the pancreas, colon, and lung. De-
cades of research to discover an efficacious drug that can block oncogenic Ras signaling have yielded disappoint-
Keywords: ing results; thus, Ras was considered “undruggable” until recently. Inhibitors that directly target Ras by binding to
Oncogenic Ras previously undiscovered pockets have been recently identified. Some of these molecules are either isolated from
Anti-Ras natural products or derived from natural compounds. In this review, we described the potential of these com-
Cancer
pounds and other inhibitors of Ras signaling in drugging Ras. We highlighted the modes of action of these com-
Natural products
Derivatives
pounds in suppressing signaling pathways activated by oncogenic Ras, such as mitogen-activated protein kinase
Strategy (MAPK) signaling and the phosphoinositide-3-kinase (PI3K) pathways. The anti-Ras strategy of these com-
pounds can be categorized into four main types: inhibition of Ras–effector interaction, interference of Ras mem-
brane association, prevention of Ras–guanosine triphosphate (GTP) formation, and downregulation of Ras
proteins. Another promising strategy that must be validated experimentally is enhancement of the intrinsic
Ras–guanosine triphosphatase (GTPase) activity by small chemical entities. Among the inhibitors of Ras signaling
that were reported thus far, salirasib and TLN-4601 have been tested for their clinical efficacy. Although both
compounds passed phase I trials, they failed in their respective phase II trials. Therefore, new compounds of nat-
ural origin with relevant clinical activity against Ras-driven malignancies are urgently needed. Apart from
salirasib and TLN-4601, some other compounds with a proven inhibitory effect on Ras signaling include deriva-
tives of salirasib, sulindac, polyamine, andrographolide, lipstatin, levoglucosenone, rasfonin, and quercetin.
© 2016 Elsevier Inc. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2. Approaches of developing lead compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3. Inhibition of Ras–effector interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4. Interference of Ras membrane association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
41
5. Prevention of Ras–GTP formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
44
6. Downregulation of Ras proteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
48
7. Discussion and conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
51
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
55
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
55
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
55

Abbreviations: AGP, Andrographolide; APT1, Acyl protein thioesterase 1; COX-2, cyclooxygenase-2; DPI, Dual prenylation inhibitor; ERK, Extracellular signal-regulated kinase; FBDD,
Fragment-based drug discovery; FTase, Farnesyltransferase; FTIs, Farnesyltransferase inhibitors; G12, Glycine-12; G13, Glycine-13; Q61, Glutamine-61; GAP, GTPase-activating protein;
GDP, Guanosine diphosphate; GEF, Guanine nucleotide exchange factor; GGTase I, Geranylgeranyltransferase I; GGTI, Geranylgeranyltransferase inhibitor; GTP, Guanosine triphosphate;
GTPase, Guanosine triphosphatase; H-Ras, Harvey-Ras; HTS, High-throughput screening; HVR, Hypervariable region; IC50, Half-maximal inhibitory concentration; ICMT, Isoprenylcysteine
carboxymethyltransferase; K-Ras, Kirsten-Ras; MAPK, Mitogen-activated protein kinase; MEK, Mitogen-activated protein/extracellular signal-regulated kinase kinase; Mg2+, Magnesium
ion; N-Ras, Neuroblastoma-Ras; NMR, Nuclear magnetic resonance; NSCLC, Non-small cell lung cancer; PDAC, Pancreatic ductal adenocarcinoma; PDB, Protein Data Bank; PFTS, Phospho-
farnesylthiosalicylic acid; PI3K, Phosphoinositide-3-kinase; Raf, Rapidly accelerated fibrosarcoma; Ras, Rat sarcoma; SAR, Structure–activity relationship; SBDD, Structure-based drug de-
sign; SRJ23, 3,19-(3-chloro-4-fluorobenzylidene) andrographolide; SOS, Son of Sevenless.
⁎ Corresponding author. Tel.: +60 3 89472310; fax: +60 3 89472759.
E-mail addresses: rcxjs@upm.edu.my, jstanslas@yahoo.co.uk (J. Stanslas).
1
Equal contribution.

http://dx.doi.org/10.1016/j.pharmthera.2016.03.010
0163-7258/© 2016 Elsevier Inc. All rights reserved.
36 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

1. Introduction RAS is a proto-oncogene, mutations of which have been implicated in

The rat sarcoma (Ras) protein superfamily comprises small guano-
sine triphosphate (GTP)-binding proteins with inherent guanosine
triphosphatase (GTPase) activity, hydrolyzing GTP to guanosine diphos-
phate (GDP) (Wennerberg et al., 2005). The members of the Ras family
include Harvey-Ras (H-Ras); neuroblastoma-Ras (N-Ras); and the
splice variants of Kirsten-Ras (K-Ras), K-Ras-4A, and K-Ras-4B. These
isoforms share highly conserved GDP/GTP-binding motif elements
(also known as G boxes) that span the N-terminal G domain
(Wennerberg et al., 2005). The C-terminal hypervariable region (HVR)
of Ras is subjected to posttranslational modifications to facilitate the
lipid-mediated membrane localization of Ras (Konstantinopoulos
et al., 2007). The G domain possesses two switches: switch I (residues
25–40) and switch II (residues 59–75) (Vetter & Wittinghofer, 2001;
Kapoor & Travesset, 2015). These switches are essential for the recogni-
tion of the two nucleotide-bound states of Ras by regulatory proteins
and effectors, thus determining the protein–protein interactions be-
tween Ras and these Ras-binding proteins (Vetter & Wittinghofer,
2001; Herrmann, 2003; Wennerberg et al., 2005).
Ras proteins are molecular switches that can be turned “on” and
“off” by effecting conformational changes of switch I and II via the bind-
ing of GTP and GDP, respectively (Vetter & Wittinghofer, 2001). As Ras
exhibits slow intrinsic activities of GDP–GTP exchange and GTP hydro-
lysis, the interconversion between its active and inactive conformations
must be catalyzed by two regulatory proteins: guanine nucleotide ex-
change factor (GEF) and GTPase-activating protein (GAP). The GEFs in-
clude Son of Sevenless (SOS) and Ras guanine nucleotide-releasing
factor 1 (RasGRF1). GEF promotes an active GTP-bound state of RAS
by facilitating GDP–GTP exchange, whereas GAP maintains Ras in the in-
active GDP-bound state by accelerating the intrinsic GTPase activity of
Ras (Schmidt & Hall, 2002; Bernards & Settleman, 2004; Fig. 1).
abnormalities of cell proliferation, survival, migration, and differentia-
tion, which ultimately lead to cancer initiation and progression
(Wennerberg et al., 2005). When extracellular signals, such as epidermal
growth factor (EGF), nerve growth factor (NGF), platelet-derived growth
factor (PDGF), fibroblast growth factor (FGF), activate receptor tyrosine
kinases (RTKs), Ras activates two canonical cancer-driving mitogen-
activated protein kinase (MAPK) and phosphoinositide-3-kinase (PI3K)
pathways (Rodriguez-Viciana et al., 1994; Lodish et al., 2000; Vivanco
& Sawyers, 2002; Sever & Brugge, 2015; Fig. 2). Mutations of Ras have
been found in 25%–30% of human cancers, including 90% of pancreatic
ductal adenocarcinomas (PDACs) (Scheffzek et al., 1997; Prior et al.,
2012). Among the Ras isoforms, K-Ras mutations account for 86% of
RAS-driven cancers, followed by N-Ras (11%) and H-Ras (3%) (Forbes
et al., 2011). Ras mutations most commonly occur at codon glycine-12
(G12), glycine-13 (G13), and glutamine-61 (Q61) (Prior et al., 2012).
G12 mutations cause steric hindrance that prevents Ras–GAP interac-
tion, whereas Q61 mutations attenuate the GAP-activating GTPase activ-
ity of Ras, resulting in prolonged activation of Ras signaling, ultimately
leading to cancer initiation and progression (Gideon et al., 1992;
Scheffzek et al., 1997). The effect of G13 mutations on the biochemical
properties of Ras remains to be elucidated (Gideon et al., 1992). It follows
from the structure of the Ras–GAP complex that G13 mutations involv-
ing structurally larger amino acids would cause steric hindrance in the
Ras–GAP interaction (Prior et al., 2012). Nevertheless, the structural biol-
ogy of G13 mutations must be investigated further.
Decades of research into drugs that can inhibit oncogenic Ras have
been unsuccessful. For this reason, Ras is described as “undruggable.”
The undruggability of Ras is mainly attributed to the lack of a well-
defined binding pocket in the Ras protein structure to accommodate a
drug (Ledford, 2015). As the GTP loading of Ras is the hallmark of Ras ac-
tivation, using a GTP-competitive inhibitor to block oncogenic Ras

Fig. 1. Regulation of Ras functions. Extracellular signaling molecules (e.g., EGF) activate RTKs, leading to dimerization of the receptor and reciprocal phosphorylation between the two
adjacent kinases (also known as autophosphorylation). An adaptor protein, GRB2, binds activated RTK at its phosphotyrosine-binding domain and subsequently recruits GEF to the
inner plasma membrane. Ras latches onto the plasma membrane via a lipid molecule and interacts with GEF for its activation. The GEF facilitates the GTP loading of Ras by removing
GDP, allowing the intracellularly abundant GTP to diffuse into the GDP/GTP-binding pocket and bind to Ras. The active GTP-bound form of Ras interacts with the effector (Eff) to relay
a cascade of signaling. To stop this process, GAP interacts physically with GTP-bound active Ras and activates its GTPase activity. The formation of the Ras–GAP complex favors the
hydrolysis of GTP into GDP. The inactive GDP-bound form of Ras acquires a protein conformation incompatible for Ras–effector binding, thus turning off Ras signaling. *Oncogenic Ras
harbors mutations at codon G12, G13, and Q61. These mutations attenuate or completely disable the GAP-activating GTPase activity of Ras, thus leading us to conclude that oncogenic
Ras is constitutively active. Nevertheless, oncogenic Ras retains its low intrinsic GTPase activity, which slowly hydrolyzes Ras-bound GTP to GDP. Therefore, oncogenic Ras is not
constitutively active per se; rather, its active state is prolonged. Abbreviations: EGF, Epidermal growth factor; GAP, GTPase-activating protein; GDP, Guanosine diphosphate; GEF,
Guanine nucleotide exchange factor; GRB2, Growth factor receptor-bound protein 2; GTP, Guanosine triphosphate; GTPase, Guanosine triphosphatase; Ras, Ras sarcoma; RTK, Receptor
tyrosine kinase.
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57 37

Fig. 2. Ras-mediated signaling pathways. Ras activates several signaling pathways in the mammalian cells, and the two best-studied pathways are MAPK pathway (Raf–MEK–ERK
pathway) and the PI3K pathway. The two direct downstream effectors of active Ras in the MAPK and PI3K pathways are Raf and PI3K, respectively. The interactions of these effectors
with Ras at the RBD initiate these signaling pathways critical for cell proliferation, survival, migration, and differentiation. The Ras–Raf interaction facilitates the activation of Raf via
phosphorylation. Subsequently, a cascade of activating phosphorylation involving downstream kinases (MEK and ERK) relays Ras-mediated intracellular signaling. Active ERK
subsequently initiates gene transcription by activating transcription factors via phosphorylation. In the PI3K signaling pathway, the active membrane-bound Ras recruits PI3K to the
plasma membrane, where PI3K activates AKT via indirect phosphorylation. Active AKT mediates the activation of cellular targets important in cell growth, proliferation, and survival.
Abbreviations: AKT, v-akt murine thymoma viral oncogene; ERK, Extracellular signal-regulated kinase; GDP, Guanosine diphosphate; GEF, Guanine nucleotide exchange factor; GRB2,
Growth factor receptor-bound protein 2; GTP, Guanosine triphosphate; MAPK, Mitogen-activated protein kinase; MEK, Mitogen-activated protein/extracellular signal-regulated kinase
kinase; PI3K, Phosphoinositide-3-kinase; Raf, Rapidly accelerated fibrosarcoma; Ras, Ras sarcoma; RBD, Ras-binding domain; RTK, Receptor tyrosine kinase.

activation by competing with cellular GTP for binding to mutant Ras can
be considered. Drawing from the success of adenosine triphosphate
(ATP)-competitive kinase inhibitors, the use of GTP analogs can be con-
sidered via modifications at the sugar and base moieties of GTP to inhibit
Ras activation (Noonan et al., 1991). Nevertheless, the potency of GTP-
competitive inhibitors was found to be lacking for two reasons: GTP
binds Ras with an extremely high affinity and GTP is abundantly present
intracellularly (John et al., 1993).
As the research on drugs targeting oncogenic Ras directly was un-
successful, the focus was shifted to the modulation of proteins involved
in Ras activation. Elucidating the mechanism of Ras membrane associa-
tion, which explains the spatiotemporal activation of Ras, has
highlighted the possibility of targeting proteins other than Ras to
block oncogenic Ras signaling (Cox et al., 2015). Ras membrane associ-
ation is a prerequisite for Ras activation and signaling, which is initiated
by the addition of a fatty acid or a farnesyl group to the C-terminal HVR
of Ras (Prior & Hancock, 2012; Fig. 4). The farnesyl group localizes Ras to
the plasma membrane, which is then embedded in the inner surface of
the plasma membrane (Fig. 4). The farnesylation of Ras is initiated by
farnesyltransferase (FTase), an enzyme that is relatively more druggable
than Ras (Fig. 4). Indeed, farnesyltransferase inhibitors (FTIs) have
shown promising in vitro and preclinical anticancer activity (Berndt
et al., 2011). With this development, FTIs were considered as solutions
to the blocking of oncogenic Ras signaling. However, the anticancer
38 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

activity of FTIs at bench could not be translated to the expected efficacy
in clinical trials (Berndt et al., 2011). Subsequently, the lipidation of Ras
was found to occur via the attachment of a geranylgeranyl group at the
C-terminal HVR of N-Ras and K-Ras in the presence of an alternative en-
zyme known as geranylgeranyltransferase I (GGTase I) despite the inhi-
bition of farnesylation by FTIs (Whyte et al., 1997; Fig. 4).
Despite the unsuccessful attempts over the decades, researchers
continue to search for inhibitors that block oncogenic Ras signaling. The
recent years have seen a major shift in focus back to targeting oncogenic
Ras itself. With the advancement of computer modeling in the past
5 years, drugging oncogenic Ras via in silico compound screening is an
exciting prospect for this field (Ledford, 2015). Nevertheless, drugging
the regulatory proteins of Ras signaling continues to be the optimal
method of blocking oncogenic Ras signaling, in addition to drugging
Ras directly.
Natural products are organic compounds produced by living
organisms. Natural compounds with no direct role in the growth
and development of host organisms but a biological effect on other
organisms are known as secondary metabolites (Hanson, 2003).
Secondary metabolites represent an abundant resource in drug discov-
ery (Newman & Cragg, 2012). From 1981 to 2010, natural products and
their derivatives made up 79.8% of the anticancer drugs approved glob-
ally (Newman & Cragg, 2012). Most of the small-molecule anticancer
drugs were discovered by isolating phytochemical compounds from
plants, such as alkaloids, flavonoids, phenolic acids, and terpenoids,
followed by structurally modifying active compounds without altering
their pharmacophores to enhance the biological activity of the com-
pounds. Here, we discuss the application of natural products and their
derivatives in blocking oncogenic Ras signaling based on four known
strategies (Fig. 3): inhibition of Ras–effector interactions, interference
of Ras membrane association, prevention of Ras–GTP formation, and
downregulation of Ras proteins. Another promising strategy that has
not been investigated in depth is enhancement of the intrinsic Ras
GTPase activity by small chemical entities, which forms the basis for de-
veloping drugs that target oncogenic Ras signaling in the future (Fig. 3).
In this review, we specifically focused on small chemical entities of nat-
ural products and their derivatives, which emerge as potential drug can-
didates in the treatment of RAS-driven malignancies.

Fig. 3. Strategies for targeting Ras signaling. There are four known strategies by which natural products and their derivatives inhibit Ras function: inhibition of Ras–effector interactions (1),
interference of Ras membrane association (2), abrogation of Ras–GTP formation (3), and downregulation of Ras proteins (4). The efficacy of a prospective strategy that enhances the
intrinsic Ras GTPase activity (?) is yet to be proven, and it can be considered in the future development of small-molecule Ras inhibitors. Further investigations are required to
delineate the exact mechanism of anti-Ras signaling activity in some of the promising compounds (!). Abbreviations: GDP, Guanosine diphosphate; GEF, Guanine nucleotide exchange
factor; GTP, Guanosine triphosphate; GTPase, Guanosine triphosphatase; Ras, Ras sarcoma.
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
2. Approaches of developing lead compounds
2.1. Structure-based drug design
For developing anti-Ras cancer drugs, the focus of research has
shifted to a structure-based drug design (SBDD) to obtain structural in-
formation in terms of the interaction of small compounds with a drug
target. This approach is useful in establishing the structure–activity rela-
tionship (SAR) as a measure of therapeutic efficacy (Greer et al., 1994).
The SBDD approach can be used to identify small-molecule compounds
that can fit into pockets of unique conformation that are found on the
three-dimensional (3D) surface of activated Ras proteins (i.e., GTP
bound). The GTP-bound Ras protein has several active sites. One of
these sites is filled by a molecule that complements the space in terms
of its shape, charge, and other binding components. In fact, SBDD can
be performed even when the target structure is lacking in
pharmacophore modeling, which works on the principle of “molecular
similarity” of small molecules (Chen et al., 2012). In brief, the SBDD
approach provides information on the active conformations or
pharmacophores essential for biological activities and allows the con-
structive modification of small ligands for producing novel structures
with specific activity and potency. Molecular mechanics such as the mo-
lecular mechanics/generalized Born surface area (MM-GBSA) method
can be used to predict the free energy of binding and the strain energy
of a group of small-molecule ligands in the protein–ligand complex
(Prime v3.4, 2013).
To identify the molecular target of anticancer agents, that is,
andrographolide (AGP) and its benzylidene derivatives (Fig. 7), we ap-
plied the concept of SBDD to demonstrate the ability of these com-
pounds to bind Ras (Hocker et al., 2013). A total of 75 K-RasQ61H
conformers derived from molecular dynamic (MD) simulations were
used for ensemble docking to determine the binding properties of
these compounds to multiple protein structures. These conformers
were found to have rare sampled structures with some open binding
sites that were invisible in crystal structures (extracted from the Protein
Data Bank (PDB) managed by Research Collaboratory for Structural Bio-
informatics (RCSB)). These sites could potentially be accessible to small-
molecule inhibitors. Without specifying any particular pocket, Hocker
et al. (2013) blindly docked the compounds against such structures
(Fig. 8) in the presence of guanine nucleotides to exclude nonspecific
hits at the catalytic site. In particular, three distinct binding pockets
were found to be preferentially targeted by the compounds. Blind
docking, which requires the receptor to be rigid while the rotatable
bonds of the ligands are flexible, is a ligand-based scheme that seeks
binding hot spots and identifies binding sites based on the size of both
the ligand and receptor clusters.
2.2. Fragment-based drug discovery
The fragment-based drug discovery (FBDD) approach is generally
used to screen and identify small-molecule fragments or lead molecules
that target the binding sites on protein surfaces or at the interface of
protein–protein interaction sites (Valkov et al., 2012). Some of these
sites normally lack a defined binding pocket or are considered
“undruggable” (Maurer et al., 2012; Chandrashekar & Adams, 2013).
Small-molecule fragments are low-molecular-weight (≥ 120 Da but
b300 Da), moderately lipophilic, highly soluble organic molecules that
usually act as ligands to bind target proteins with low affinity (Rees
et al., 2004). These fragments are usually screened at concentrations
that permit structure characterization with biophysical techniques,
such as nuclear magnetic resonance (NMR) spectroscopy and X-ray
crystallography (Brenke et al., 2009) and biochemical examination
(Maurer et al., 2012). All of these techniques form the underlying prin-
ciple of the hit-to-lead optimization process. FBDD allows the ultrasen-
sitive detection of interactions between proteins and ligands (Maurer
et al., 2012). For instance, researchers such as Maurer et al. (2012).
39
focused on identifying small molecules that can bind to pockets adja-
cent to the Ras–SOS interaction surface by an NMR-based saturation
transfer difference (STD) assay. This assay pooled a group of fragments
for screening and identified ligand segments that are in direct contact
with the protein. When the fragment binds to a Ras protein, the change
in chemical shift of a labeled Ras protein is measured by the changes in
the resonance position of cross-peaks in the H/N two-dimensional
heteronuclear single-quantum coherence (HSQC) spectra (Zartler &
Mo, 2007). Similarly, the free energy of binding for the protein–ligand
complexes can be calculated via molecular mechanics. It is worth noting
that ligand screening is based on the conservation of the binding mode
of fragments on the protein targets with an increase in ligand size, given
that the fragment overlaps well within the main hot spot of the binding
site with its interacting functional groups (Kozakov et al., 2015). In ad-
dition to identifying ligands for a known binding site, FBDD also gener-
ated high-resolution structures of the ligand–protein complexes and
revealed unique binding cavities in proteins.
2.3. High-throughput screening
High-throughput screening (HTS) is used to identify chemical com-
pounds with diverse structures, which show activity against RAS pheno-
types and can inhibit Ras signaling at different stages (Wang et al.,
2013). Synthetic lethal screening is an example of HTS, which is a cell-
based screening method for identifying small molecules that can selec-
tively kill cells harboring a specific mutation (Guo et al., 2008). Synthetic
lethality is dependent on two distinct genes or two related molecular
signaling pathways that influence a common essential biological func-
tion, by which cells will die in the case of mutations in the two genes
or abnormalities in both signaling pathways. Dysfunction of either one
gene or one signaling pathway will not lead to cell death (Guo et al.,
2008). The mutant Ras proteins in tumor cells are highly active, which
prompts cells to eventually rely on other molecular pathways for sur-
vival (de la Cruz et al., 2015). For example, mutant KRAS-driven non-
small cell lung cancer (NSCLC) cells produce antiapoptotic signals es-
sential for NSCLC survival by activating the nuclear factor kappa light-
chain enhancer of activated B cells (NF-κB) pathway (de Castro
Carpeño & Belda-Iniesta, 2013). Genetically defined cell lines are used
in the HTS approach for screening, with the underlying principle that
RAS and other molecular pathways, such as NF-κB pathway in the
above-mentioned case of NSCLC, are blocked and only cancer cells
with the mutant RAS gene are killed, thus sparing the normal counter-
parts (Guo et al., 2008). Despite its potential, the use of HTS is limited
by its ability to only determine compounds that can destroy an onco-
genic protein without identifying the protein target (Guo et al., 2008).
3. Inhibition of Ras–effector interactions
Oncogenic Ras mediates constitutive signal transduction through
protein–protein interaction with effector proteins such as PI3K and rap-
idly accelerated fibrosarcoma (Raf), Raf-1 in particular. To inhibit Ras
signaling, small molecules that can fit and inhibit protein–protein inter-
actions are promising candidates in cancer therapy. Ras–GTP can inter-
convert between two conformational states (states 1 and 2) at a
millisecond time scale; this dynamic equilibrium involves two main re-
gions of the protein: switch I and switch II. State 2 corresponds to the
high-affinity state to effectors whereas state 1 corresponds to a state
of lessened affinity to effectors (Spoerner et al., 2005).
Sulindac sulfide (Fig. 5) is obtained via the reductive metabolism of
the nonsteroidal anti-inflammatory prodrug sulindac (Fig. 5), which is
synthesized from natural safrole (Fig. 5), the main chemical constituent
of the essential oil of a Sassafras plant (Barreiro & Lima, 1992; Brunell
et al., 2011). Sulindac sulfide is a highly specific dual inhibitor that
binds reversibly to Ras (Herrmann et al., 1998). This compound
interferes with the interaction between Ras and Raf-1 kinase protein, ul-
timately inhibiting Ras signaling (Herrmann et al., 1998). At
40 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

concentrations ranging from 10 to 50 μM, this compound clearly signal-regulated kinase kinase (MEK) (Herrmann et al., 1998). In addi-
inhibited Ras-induced Raf-1 kinase activity as indicated by the de- tion, sulindac sulfide impairs the nucleotide exchange of Ras by
creased phosphorylation of mitogen-activated protein/extracellular inhibiting Ras–GEF interaction (Herrmann et al., 1998). Sulindac sulfone

Fig. 4. Targeting Ras intracellular trafficking and membrane association. Intracellular localization of Ras to the plasma membrane is a prerequisite for GEF-mediated Ras activation. In the
cytoplasm, all Ras isoforms are attached with a C15 farnesyl isoprenoid by FTase at the cysteine residue of a tetrapeptide CAAX motif (AA: aliphatic residues; X: variable residue) or a C20
geranylgeranyl isoprenoid (N-Ras and K-Ras) by GGTase I in response to the inhibition of farnesylation by FTIs. Subsequently, the AAX peptide of the farnesylated Ras is removed by RCE1
via catalytic proteolysis, producing the terminal farnesylated cysteine residue that is reversibly carboxylmethylated by ICMT in the ER. In the vesicle-mediated classic secretory pathway,
PAT adds C16 palmitate fatty acid to H-Ras, N-Ras, and K-Ras-4 A at the immediate cysteine residues upstream of the terminal farnesylated cysteine residue on the Golgi membrane (Cox,
2010). The palmitoyl group provides Ras with an additional anchorage for plasma membrane association. Once on the plasma membrane, APT1 removes the palmitoyl group on Ras,
allowing Ras to dissociate from the plasma membrane. Conversely, K-Ras-4B lacks the immediate cysteine residues upstream of the terminal farnesylated cysteine residue, making
palmityolation impossible. Instead, farnesylated K-Ras-4B is transported to the plasma membrane via a non-vesicular pathway with the help of PDEδ, which solubilizes farnesylated K-
Ras-4B. In contrast to other Ras isoforms, K-Ras-4B possesses a positively charged polylysine stretch (KKKKKK) as an additional anchoring feature for membrane association by
interacting with the negatively charged phosphate group of phospholipids (Cox et al., 2014). Upon GEF-mediated activation, GTP-bound H-Ras and K-Ras-4B were found to bind to a
membrane-bound protein (galectin), also known as the docking protein of the oncoprotein, via the farnesyl tail (Elad-Sfadia et al., 2002, 2004; Meder & Simons, 2005). *N-Ras is
associated with the plasma membrane in a different manner, although all Ras isoforms are significantly homologous. **The interaction between K-Ras-4 A and galectin has not been
documented. Several natural compounds and their derivatives (salirasib, palmostatin B, spermatinamine, and aplysamine 6) have been reported to block Ras signaling by interfering
with its membrane association. Abbreviations: APT1, acyl protein thioesterase 1; ER: endoplasmic reticulum; RCE1, Ras-converting enzyme-1; FTase, Farnesyltransferase; FTIs,
Farnesyltransferase inhibitors; GEF, Guanine nucleotide exchange factor; GGTase I, Geranylgeranyltransferase I; GTP, Guanosine triphosphate; HVR, Hypervariable region; ICMT,
Isoprenylcysteine carboxymethyltransferase; PAT, protein acyltransferase; Ras, Ras sarcoma.
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
is another active metabolite obtained by the oxidation of sulindac
(Brunell et al., 2011). The sulfide and sulfone metabolites have been
studied extensively in the case of colorectal cancer with a high frequen-
cy of K-Ras mutations. Despite the structural similarities between
sulindac sulfide and sulindac sulfone, these compounds inhibit the
Ras-dependent cyclooxygenase-2 (COX-2) pathway in colorectal cancer
cells through two distinct mechanisms: the inhibition of COX-2 enzyme
activity and of COX-2 expression, respectively (Williams et al., 1999;
Taylor et al., 2000). Treatment with sulindac sulfone reduced tumorige-
nicity in mice with severe combined immunodeficiency disease (SCID)
bearing human KRAS-driven colorectal and HRAS-driven mammary can-
cer cells (Thompson et al., 1997; Taylor et al., 2000). Nevertheless, the
direct binding of sulindac sulfone to Ras and its inhibitory effect on
the Ras signaling pathway are yet to be elucidated.
4. Interference of Ras membrane association
The posttranslational lipid modifications of Ras proteins are essential
for the localization and membrane association of Ras (Fig. 4). Enzymes
that control the posttranslational modification of Ras have emerged as
important targets in the design of anti-Ras drugs (Zhang & Casey,
1996). Small molecules that inhibit these enzymes can be potentially
developed into anti-Ras drugs. Prenylation of Ras for its localization to
the plasma membrane involves farnesylation and geranylgeranylation.
The carboxyl-terminal residue of the CAAX motif is essential for deter-
mining the isoprenoid added to the protein (Kloog & Cox, 2004). The
initial strategy used to inhibit the localization of Ras led to the develop-
ment of farnesyltransferase inhibitors (Wang et al., 2013). Most FTIs
showed in vivo efficacy against Ras; however, they were ineffective in
late-stage clinical trials, as they could not inhibit the signaling of activat-
ed K-Ras and N-Ras isoforms, which could be alternatively prenylated
via geranylgeranylation (Whyte et al., 1997; Fig. 4). This finding have
led researchers to shift their focus from drugs targeting Ras by interfer-
ing with the Ras membrane association to inhibitors of signaling cas-
cades downstream of Ras.
Alternatively, farnesyl cysteine mimetics, such as S-trans, trans-
farnesylthiosalicylic acid (FTS, also known as salirasib; Marom et al.,
1995; Fig. 6A), have also been developed. This compound is derived
from salicylic acid (Fig. 6A), a phytochemical found in willow tree
(Chen et al., 2009). The farnesylcysteine of salirasib mimics the
carboxyl-terminal farnesylcysteine carboxymethyl ester of Ras crucial
for its attachment to the membrane. This compound competes with
GTP-bound H-Ras and K-Ras-4B proteins for binding to the docking pro-
tein of the oncoprotein (galectin) on the inner leaflet of the plasma
membrane; thus, the Ras membrane association and its localization to
specific cellular localities for signaling are disrupted (Aharonson et al.,
1998; Kloog & Cox, 2004; Haklai et al., 2008; Baker & Der, 2013).
Reduction
Sulindac sulfide
Fig. 5. Chemical structures of compounds that inhibit Ras–effector interactions.
41
Salirasib does not inhibit farnesyltransferase nor does it affect Ras mat-
uration. According to previous reports, salirasib acts independently of
farnesyltransferases and has been found to antagonize Ras activity spe-
cifically by competing with the active GTP-bound form of Ras (Egozi
et al., 1999; Kloog & Cox, 2004). This compound can dislodge GTP-
bound Ras from its membrane anchorage domain (galectin) and accel-
erate the degradation of the GTP-bound Ras in the cytoplasm (Egozi
et al., 1999; Baker & Der, 2013). An in vitro growth inhibitory assay
using a PDAC cell line, PANC-1, revealed that salirasib can decrease the
total amount of activated Ras proteins at concentrations ranging from
25 to 50 μM, with a maximal reduction of approximately 50%; serum-
dependent and EGF-stimulated extracellular signal-regulated kinase
(ERK) activation was also found to be inhibited (Weisz et al., 1999). Fur-
thermore, this compound could prevent the transformation of pancreat-
ic cancer cells harboring oncogenic Ras by inhibiting their anchorage-
dependent and anchorage-independent growth (Weisz et al., 1999).
On testing the compound for any antitumor activity in vivo, no toxic ef-
fect was observed in PDAC-xenografted nude mice (Weisz et al., 1999).
It is worth noting that prolonged salirasib treatment in vitro did not
cause drug resistance in pancreatic and colorectal cancer cells; more im-
portantly, these salirasib-treated cells became highly sensitive to the
chemotherapeutic drug gemcitabine (Gana-Weisz et al., 2002). The sys-
temic exposure of salirasib did not change even when administered in
combination with gemcitabine in the treatment of pancreatic cancer
(Laheru et al., 2012). Although salirasib monotherapy was unsuccessful
in a phase II clinical trial involving patients with lung adenocarci-
nomas harboring K-Ras mutations, a combination therapy of
gemcitabine and salirasib in phase I clinical trials was successful in
patients with advanced PDAC (Riely et al., 2011; Laheru et al.,
2012). A recent study showed that patients with refractory hemato-
logic malignancies were appreciably tolerant to salirasib monother-
apy, which warranted the initiation of phase II clinical trials (Badar
et al., 2015). Although all previous clinical trials involving salirasib
focused on cancers driven by Ras mutations, none of the patients in
the recent clinical trial harbored Ras mutations (Badar et al., 2015).
Salirasib can be considered in the treatment of glioblastoma
multiforme (GBM), which arises from an overactivation of Ras via
an overexpression of EGF and PDGF receptors (Guha et al., 1997;
Goldberg & Kloog, 2006). Salirasib was found to inhibit glioblastoma
cell migration and invasion by the downregulation of PI3K, thus re-
ducing Rac-1 activity and increasing RhoA activation; both of these
activities play a role in the arrangement of the actin cytoskeleton
(Goldberg & Kloog, 2006). Salirasib treatment immobilized glioblas-
toma cells by strengthening the cell attachment to the extracellular
matrix via actin rearrangement (Goldberg & Kloog, 2006). In addi-
tion, salirasib induced apoptotic cell death by decreasing the survivin
gene transcripts (Blum et al., 2006).
Oxidation
Sulindac Sulindac sulfone
Safrole
42 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

A.

Salicylic acid

S-trans-geranyl thiosalicylic acid (GTS) S-geranylgeranyl thiosalicylic acid (GGTS)

Farnesyl group

R1 R2 R3 R4 Compound

OH H H H FTS, also known

as salirasib

OCH3 (methyl ester) H H H FTS-ME

OCH2OCH3 H H H FTS-MOME

(methoxymethylester)

OCH(CH3)2 H H H FTS-IE

(isopropylester)

OCH2C6H5 (bezylester) H H H FTS-BE

NH2 (amide) H H H FTS-A

NHCH3 (methylamide) H H H FTS-MA

N(CH3)2 (dimethylamide) H H H FTS-DMA

OH F (5-fluoro) H H 5-F-FTS

OH Cl (5-chloro) H H 5-Cl-FTS

OH H F (4-fluoro) H 4-F-FTS

OH H Cl (4-chloro) H 4-Cl-FTS

OH H H Cl (3-chloro) 3-Cl-FTS

Phospho-farnesylthiosalicylic acid (PFTS)

 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57 43

B. C.

TLN-4601

D.

Aplysamine 6 Spermatinamine

E.

Palmostatin B

Lipstatin
Fig. 6. Chemical structures of compounds that interfere with Ras membrane association. The 15-carbon farnesyl group of salirasib (solid box) and the presence of free carboxylic acid are
crucial for its Ras inhibitory activity. However, the addition of amide or ester and butane–diethyl phosphate (dashed box) to the parent compound could enhance the efficacy of the
derivatives in inhibiting pancreatic cancer cell growth.

In the late 20th century, some researchers made modifications to the
S-prenyl backbone of salirasib to generate new anti-Ras inhibitors
(Fig. 6A) for a more effective inhibitory effect on the growth of tumor
cells with overactivated Ras (Aharonson et al., 1998). However, studies
into targeting Ras using the S-prenyl derivatives of salirasib were unsuc-
cessful, as suggested by the growth inhibitory activity of the S-prenyl
derivatives (IC50 = 10–N50 μM) compared with salirasib (IC50 = 5–
15 μM) in H-ras-transformed EJ cells, an endometrial adenocarcinoma
cell line (Aharonson et al., 1998). Similarly, Goldberg et al. synthesized
a series of compounds by modifying the carboxyl group of salirasib
(Fig. 6A). Both studies clearly showed that the salirasib derivatives in-
hibit Ras activation by dislodging the Ras proteins from the membrane,
as proven by the marked reduction in Ras–GTP levels and subsequent
suppression of the downstream Ras–MAPK pathway (Aharonson et al.,
44 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
1998; Goldberg et al., 2009). Some of the derivatives, such as halogenic-
substituted derivatives at the C5 position of the benzene ring and ester
or amide derivatives of thiosalicylic acid, showed greater potency than
their parent compound by selectively inhibiting the growth of cells
with activated Ras. Mackenzie et al. (2013) reported a derivative of
salirasib with improved biological activity, phospho-
farnesylthiosalicylic acid (PFTS, also known as MDC-1016; Fig. 6A), syn-
thesized by adding butane–diethyl phosphate to the carboxylic group of
salirasib. Indeed, the in vitro anticancer activity of PFTS (IC50 = 44.1 ±
2.8 μM) against MIA PaCa-2 pancreatic cancer cells was found to be su-
perior to the activity of its parental compound (IC50 N 80 μM). This com-
pound is nontoxic with significant antitumor activity reported at
tolerable doses in both an MIA PaCa-2 tumor xenograft model in nude
mice (50 mg/kg per day, 5 days/week, orally for 25 days) and a KRAS-
driven pancreatic carcinogenesis murine model (100 mg/kg per day,
5 days/week, orally for 9 days). Interestingly, only 1 μM PFTS was left
unreacted and the rest were metabolized immediately into two metab-
olites, namely salirasib (29.7 μM) and salirasib glucuronide (10.3 μM), in
the murine plasma 1 h after oral administration at a dose of 200 mg/kg.
An equimolar dose of its parental compound, salirasib, remains partially
intact (19.8 μM) whereas the rest is metabolized into salirasib glucuro-
nide (37.4 μM) (Mackenzie et al., 2013). PFTS treatment in vitro and
in vivo led to the inhibition of Ras activation and subsequent reduction
of Ras–MAPK and Ras–PI3K signaling (Mackenzie et al., 2013). To deter-
mine the preventive effect of PFTS on cell proliferation, the compound
was tested in combination with another compound, phospho-valproic
acid (P-V), a novel signal transducer and activator of transcription 3
(STAT3) inhibitor. Interestingly, Mackenzie et al. (2013) showed that
this combination exhibited a synergistic and additive inhibitory effect
on the growth of cells harboring mutant K-Ras (MIA PaCa-2) and
wild-type K-Ras (BxPC-3), respectively. It is interesting to note that
Aharonson et al. (1998) reported that the presence of at least 15 carbon
chains (as in the farnesyl group) and free carboxylic acid were required
for Ras inhibition. Conversely, Goldberg et al. and Mackenzie et al.
showed that modifications to the carboxylic group of salirasib can en-
hance the compound's anti-Ras and growth inhibitory activities.
Another natural compound called TLN-4601 (Fig. 6B), a novel
farnesylated dibenzodiazepinone, was isolated as a secondary metabo-
lite from the marine actinomycete bacterium Micromonospora sp.
(Charan et al., 2004). Due to its farnesylated moiety, this compound
was initially considered to inhibit Ras prenylation, but Boufaied et al.
subsequently showed that this compound induces proteasomal-
dependent Raf-1 degradation instead. This compound was also found
to inhibit the formation of active Ras–GTP, which leads to the downreg-
ulation of Ras–MAPK signaling (Campbell et al., 2010). A phase I clinical
study involving patients with GBM, showed the safety and excellent tol-
erance of this compound at drug plasma concentrations in the micro-
molar range (Mason et al., 2010). In addition, the in vitro and in vivo
anticancer activity of K-Ras-driven PDAC was also reported, which war-
ranted clinical trials in patients with advanced pancreatic cancer
(Campbell et al., 2010). As GBM is caused by dysregulated Ras–MAPK
signaling, the anti-GBM activity of the compound can be attributed to
the inhibition of Ras–MAPK signaling. Despite the promising results of
the phase I clinical trials, this compound was ineffective in phase II trials
involving patients with progressive GBM (Mason et al., 2012). The exact
mechanism of Raf-1 protein degradation and Ras–GTP depletion in-
duced by TLN-4601 remains to be elucidated.
In addition to FTIs, inhibitors of isoprenylcysteine carboxy-
methyltransferase (ICMT) can prevent the localization of Ras to the
plasma membrane (Fig. 4) (Baines et al., 2011). ICMT, an integral mem-
brane protein found on the endoplasmic reticulum (ER), catalyzes the
methylation of the C-terminal prenylcysteine on prenylated proteins;
thus, it is crucial for the correct localization of prenylated proteins to cel-
lular membranes (Wright et al., 2009). While conducting an HTS of nat-
ural products for inhibitors of ICMT, a novel polyamine alkaloid,
spermatinamine, was identified (Fig. 6C), which was purified from the
extract of an Australian Verongida marine sponge, Pseudoceratina sp.
This compound was found to inhibit ICMT at an IC50 of 1.9 μM
(Buchanan et al., 2007). Although it is chemically unsuitable as a drug
lead, this compound exhibits good potency, which can be further opti-
mized via chemical modifications to target ICMT (Buchanan et al.,
2007). Subsequently, this compound was proven to be highly potent
in inhibiting the growth of cancer cell lines such as leukemia (K562)
and colon cancer (HT29) cell lines with a mean 50% growth inhibition
(GI50) value of 0.23 μM (Kottakota et al., 2012). Aplysamine 6 is also a
natural product that inhibits ICMT, with bromotyrosine as the scaffold
(Fig. 6D). This is also an alkaloid derived from the same species as
spermatinamine, with an IC50 value of 14 μM for ICMT (Buchanan
et al., 2008).
As illustrated in Fig. 4, it is critical that certain Ras isoforms such as
N-Ras, H-Ras, and K-Ras-4A are palmitoylated and depalmitoylated
after methylation so as to control the steady-state transportation of
Ras between the plasma membrane and the Golgi complex as well as
to prevent nonspecific localization on endomembranes other than the
Golgi complex (Rocks et al., 2005). After the CAAX motif is processed
at ER, the cysteine residues that are palmitoylated at the Golgi complex
serve as second signals necessary for vesicular transport to the plasma
membrane for association (Plowman & Hancock, 2005). In contrast to
the other Ras isoforms, K-Ras-4B does not undergo palmitoylation and
its localization to the plasma membrane is facilitated by a prenyl
binding protein, known as PDEδ (Chandra et al., 2012). The polylysine
region of K-Ras-4B participates in an ionic interaction with the negative-
ly charged plasma membrane for membrane association (Plowman
& Hancock, 2005). Depalmitoylation catalyzed by acyl protein
thioesterase 1 (APT1) is crucial for the redistribution of farnesylated
Ras from all other endomembranes to the Golgi complex, the site at
which farnesylated Ras is repalmitoylated and relocalized to the plasma
membrane via a classical secretory pathway (Rocks et al., 2005; Cox,
2010). Palmostatin B (Fig. 6E), derived from the natural product
lipstatin, has been shown to inhibit APT1 (Dekker et al., 2010;
Hernandez et al., 2013). Lipstatin, a natural product isolated from the
actinobacterium Streptomyces toxytricini (Zhu et al., 2014), has been
found to inhibit porcine pancreatic lipase (Weibel et al., 1987; Dekker
et al., 2010). Based on protein structure similarity clustering (PSSC),
the active site of the lipase, which is also the binding site of lipstatin,
is highly similar to APT1 (Dekker et al., 2010). Lipstatin contains a
β-lactone, and palmostatin B is derived by chemical modification of
the constituents surrounding the β-lactone core (Dekker et al., 2010).
Palmostatin B inhibits APT1 via reversible competitive inhibition
(IC50 = 670 nM) in vitro (Dekker et al., 2010). This β-lactone inhibitor
disturbs the palmitoylation and depalmitoylation cycle of Ras, thus lo-
calizing the palmitoylated Ras to the endomembranes, instead of the
plasma membrane (Cox, 2010). This phenomenon subsequently leads
to partial phenotypic reversion in oncogenic H-RasG12V-transformed fi-
broblasts, particularly the loss of spindle-like phenotype and reestab-
lishment of cell–cell contact (Dekker et al., 2010).
5. Prevention of Ras–GTP formation
Like other small G proteins, Ras contains a natural binding pocket
that can accommodate either GDP or GTP in its inactive or active form,
respectively. In Ras activation, the GDP–GTP nucleotide exchange is a
crucial step regulated by the GEF (Fig. 1). The GEF proteins, particularly
SOS, contain a catalytic subunit composed of a Ras exchanger motif
(REM) and a cell division cycle 25 (Cdc25) homology core domain. Ac-
cording to Boriack-Sjodin et al. (1998), SOS facilitates GTP loading in
two ways. First, SOS inserts a helical hairpin motif of the Cdc25 domain
to displace the switch I region. Second, insertion of SOS distorts the con-
formation of the switch II region and alters the chemical environment of
the nucleotide-binding site for the associated magnesium ion (Mg2+)
and the phosphate groups of the nucleotide. The resulting steric hin-
drance on switch I and the remodeling of switch II cause the
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57 45

AGP SRJ09

SRJ10 SRJ23
Fig. 7. Chemical structures of AGP and its benzylidene derivatives. The three-membered ring scaffold (dashed box shown in SRJ09) of all AGP derivatives is an important pharmacophore
that allows the binding of these compounds to p1 of the Ras protein.

Switch I

p1

p2

Switch II

Fig. 8. Structural basis for inhibition of Ras–SOS interaction by andrographolide-based compounds. Docking pose of SRJ23 in p1 and p2 of K-RasQ61H protein. SRJ23 bound at p1, which is
located behind switch I (amino acid residues 25–40), by anchoring the lactone ring into the pocket and interacting with the nucleotide. SRJ23 also bound at the groove between switches I
and II (amino acid residues 59–75), with the lactone ring directed up towards the nucleotide binding site. Notably, p2 was found to be critical for accommodating small-molecule inhibitors
in mutant K-Ras.
46 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

A.
Phenylhydroxyl-
amine group

Naphthyl group

Sulfonamide
group

R= H

SCH-53239 SCH-54292 SCH-53870

B.

Compound R1 R2 R3 C-2 bond stereo

(solid arrow)

5 OH

6 OH

7 OH

8 OH

9 OH OH

10 OH OH
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57 47

nucleotide-binding site to open and GDP to be released subsequently
(Cherfils & Zeghouf, 2013). The GEF activity is promoted by picomolar
nucleotide affinities of Ras combined with millimolar cytosolic concen-
trations of GTP (Spiegel et al., 2014).
The GTPase cycle is complete only when active Ras proteins are
deactivated, whereby GTP is cleaved into GDP. It is worth noting that
even when GTPase activity is not initiated by GAP, the Ras protein can
be inactivated slowly due to the intrinsic GTPase activity (Bos et al.,
2007). Thus, GTP hydrolysis is efficient when regulated by the GAPs.
However, single-point mutations in tumorigenic Ras, especially at
Q61, lead to impaired intrinsic GTP hydrolysis and insensitivity to
GAPs (Scheffzek et al., 1997). Impaired intrinsic GTPase activity in-
creases the dependence of the nucleotide state of Ras on the relative
concentration and nucleotide affinity, thus preferring GTP to GDP and
increasing the proportion of active GTP-bound Ras (Scherer et al.,
1989). This limits the development and use of GTP-competitive inhibi-
tors; thus, research efforts have shifted from strategies impairing pro-
tein activation to those preventing the initial Ras–GTP formation. With
the aim of developing novel anti-Ras antitumor drugs, recent research fo-
cused on regulatory proteins, as well as Ras itself, as the targets for
inhibiting nucleotide exchange and the subsequent formation of activated
protein. Some of these agents were derived from natural products and
have led to the discovery of agents that can inhibit Ras–GTP formation.
AGP, a bicyclic diterpenoid lactone extracted from the well-
known Asian herb Andrographis paniculata (“King of Bitters”), and
its benzylidene derivatives, namely 3,19-(2-bromobenzylidene)
andrographolide (SRJ09), 3,19-(3-bromobenzylidene) andrographolide
(SRJ10), and 3,19-(3-chloro-4-fluorobenzylidene) andrographolide
(SRJ23) (Fig. 7), have shown promising anticancer activity in vitro, as re-
ported by Jada et al. (2008). All benzylidene derivatives showed more
potency than the parent compound against breast (MCF-7) and colon
cancer (HCT-116) cells, with the GI50 value ranging from 3.8 to
5.3 μM. Ras was found to be a potential target when determining the
exact molecular target of these compounds by the SBDD approach and
ensemble docking (Hocker et al., 2013). Three binding pockets (p1,
p2, and p3) preferentially targeted by AGP and its benzylidene deriva-
tives were found on the activated protein surface (PDB ID: 3GFT,
Fig. 8). p1 is a transient pore located behind switch I, with which
these compounds interacted the most and bound with good affinity
(micro- to nanomolar). MD simulations showed that SRJ23 bound to
p1 and was stabilized at the binding site by the formation of hydrogen
bonds between the hydroxyl group on the lactone ring and the α-
phosphate of GTP, as well as electrostatic contact with Mg2+ (Hocker
et al., 2013, Fig. 8). The AGP and SRJ derivatives were also found to target
p2, a groove between switches I and II (Fig. 8). When the ligands bound
to these pockets, the key residues on the switch regions were reoriented
and Ras–GEF interaction was subsequently impaired. As it cannot bind
GAP, oncogenic Ras, which is constitutively GTP loaded, is dependent
on intrinsic hydrolysis to return to its GDP-bound ground state. Then,
GEF interacts with oncogenic Ras to facilitate GTP loading. This process
was blocked, as evidenced by the reduced fraction of GTP-bound
K-RasG12V in cell-based assays and a concomitant reduction in associ-
ated MAPK signaling after prolonged treatment with AGP derivatives
(Hocker et al., 2013).
A group of researchers at the Schering-Plough Research Institute de-
veloped a guanine nucleotide analog termed SCH-53239 and two non-
nucleotidic derivatives, SCH-54292 (a derivative of SCH-53239 with a
glucose moiety) and SCH-53870 (Taveras et al., 1997; Fig. 9A), to act
as Ras binders for inhibiting GEF-catalyzed nucleotide exchange. These
compounds were initially designed to compete with GDP for the
nucleotide-binding site. They share a common scaffold but are dissimi-
lar in the substituent group on the sulfonamidic nitrogen (Fig. 9A). As
the guanine and sugar moieties of SCH-53239 and SCH-54292, respec-
tively, were directed away from the binding site, these derivatives
showed insignificant ligand binding. These molecules showed a high
potency in inhibiting the GEF-dependent nucleotide exchange with
considerably low IC50 values of 0.5–0.7 μM (Palmioli, 2009). Without
displacing the bound GDP, these SCH compounds bound to a previously
unidentified hydrophobic pocket adjacent to switch II and close to the
nucleotide-binding site on the mutant Ras–GDP protein (Taveras et al.,
1997). Drawing from these studies, Peri et al. (2005) designed and syn-
thesized a new class of Ras inhibitors by retaining the putative
pharmacophore groups such as phenylhydroxylamine and sulfonimide
groups and attaching them to a bicyclic core derived from the natural
sugar D-arabinose (Fig. 9B). These arabinose-derived bicyclic com-
pounds bound stronger, as shown in in silico modeling. However,
these compounds are limited by the toxicity and poor metabolic stabil-
ity of the hydroxylamine group, which is crucial for inhibiting the nucle-
otide exchange. Few other Ras inhibitors were synthesized based on the
backbone of SCH compounds, in which the sugar moiety was conjugat-
ed to increase water solubility and the hydroxylamine group was re-
moved to eliminate the nonspecific toxicity (Fig. 10) (Palmioli et al.,
2009; Sacco et al., 2011). When the hydroxylamine group was removed,
the nucleotide-exchange inhibitory activity of these newly synthesized
compounds significantly reduced. Nevertheless, the sugar moiety has
been shown to partially compensate the nucleotide-exchange inhibito-
ry activity of these derivatives in vitro (IC50 values = 35.5–100 μM), by
enhancing its binding affinity to Ras. Interestingly, these sugar-based
compounds were found to bind p2, which is also the binding site of
other Ras inhibitors as reported by Grant et al. (2011), Maurer et al.
(2012), Sun et al. (2012) and Hocker et al. (2013). All sugar-derived
compounds were found to inhibit Ras-dependent cell growth and re-
duce the phosphorylated MAPK levels, although their IC50 values
exceeded those of the SCH compounds (Peri et al., 2005; Palmioli
et al., 2009; Sacco et al., 2011).
Drawing from the work at Schering-Plough involving natural
products with hydroxy-containing moieties, Müller et al. (2009)
also developed new Ras inhibitors derived from levoglucosenone
(Fig. 11A), a nonplanar chiral bicyclic molecule derived from cellu-
lose. The α,β-unsaturated carbonyl group of levoglucosenone was
made to react with aromatic nitrones by 1,3-dipolar cycloaddition
and was further modified by fusion with an isoxazolidine ring to pro-
duce a tricyclic scaffold (Fig. 11A). The novel compounds based on
the tricyclic scaffold varied in terms of the substituents R1, R2, and
R3 . The two most potent compounds (compounds 1 and
4) (Fig. 11B) were tested in vitro at 50 μM for their inhibitory effect
on the GEF-mediated nucleotide exchange in human Ras protein,
p21hRas (Müller et al., 2009). When testing their inhibitory effects
on various cancer cells, compounds 6, 7, and 8 (Fig. 11B) showed
high potency in human ovarian (A2780) and breast adenocarcinoma
(MCF-7) cancer cell lines (range of IC50 values = 3.1–7.1 μM) (Müller
et al., 2009). The research team further tested compounds 1 and 4
against wild-type and mutant K-Ras-transformed NIH3T3 mouse fi-
broblasts. In contrast to compound 1, which had almost no effect
on both types of cells, compound 4 led to partial and complete
growth inhibition in normal and mutant K-Ras-transformed cells, re-
spectively. Thus, it was proven to be selective in its inhibitory effect
on the Ras-dependent growth of mutant cells.

Fig. 9. Chemical structure of SCH compounds and related D-arabinose-derived compounds. (A) Core scaffold of SCH compounds contain three main functional groups: the naphthyl group
(dashed box), the phenylhydroxylamine group (solid box), and the sulfonamide group (dash-dotted box). The naphthyl group of these compounds was inserted into a hydrophobic
binding pocket close to the switch II region. The phenylhydroxylamine group was located close to the bivalent Mg2+ ion and the β-phosphate group of the bound GDP in the binding site.
The interaction occurred via hydrogen bonds with Thr58. The sulfonamide group is significant as it forms hydrogen bond interactions with Gly60, which is important for the restructuring
of switch II and for stabilizing the formation of Ras-GTP. (B) On the basis of SCH compound structures, Peri et al. (2005, 2006) developed the first (compounds 1–4) and second (com-
pounds 5–10) generation of inhibitors containing the bicyclic scaffold (solid circle), by retaining the functional phenylhydroxylamine and benzyl moieties.
48 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
Glucose
N-acetyl
glucosamine
Compound 2 Compound 3
Fig. 10. Chemical structures of glycosylated Ras inhibitors developed by Sacco et al. (2011) by conjugating sugar moieties found most abundantly in plants and animals, which are glucose,
N-acetyl glucosamine, and lactose (solid boxes).
Lai et al. (2009) identified at least 11 hydrophobic phytoconstituents
including pheophorbide a (an epimer), pheophorbide a' (an epimer),
pyropheophorbide a (an epimer), methyl pyropheophorbide a,
hexadecanoic acid, oleic acid, linoleic acid, linolenic acid, campesterol,
stigmasterol, and β-sitosterol from a Malaysian medicinal plant
known as Typhonium flagelliforme (Keladi tikus). These compounds
did not exhibit a significant anticancer effect when administered indi-
vidually, but in combination they showed a synergistic antiproliferative
effect on NCI-H23 lung cancer cell line. The first four aforementioned
compounds are chlorophyll catabolites and were proven to be
photosensitive; that is, they were found to inhibit cell growth with
high potency upon light activation. With mutant K-Ras considered the
important prognostic biomarker for NSCLCs (Marks & Pao, 2009),
five of the potent compounds (pheophorbide a, pheophorbide a',
pyropheophorbide a, hexadecanoic acid, and 1-hexadecene; Fig. 12) re-
ported by Lai et al. (2009) were evaluated for their mechanism of action
using an in silico approach (Fatima & Yee, 2014). The researchers exam-
ined the SOS-dependent nucleotide-exchange inhibitory activity by
docking the compounds, either directly or allosterically, onto different
reported binding sites on mutant K-Ras proteins (K-RasG12D protein
– PDB ID: 4DST with co-crystallized ligand 4,6-dichloro-2-methyl-
3aminoethyl-indole (DCAI) bound to the SOS pocket (Maurer et al.,
2012); and K-RasG12C proteins – PDB ID: 4LUC with a sulfonamide as
a co-crystallized ligand and 4LYF with a vinyl sulfonamide as a co-
crystallized ligand (Ostrem et al., 2013) bound to allosteric sites).
Fatima and Yee demonstrated that pheophorbide epimers bound
strongly to two binding sites, one of which was present at the region
of the Ras–SOS interacting surface whereas the other was located
at the switch II region. By contrast, long-chained hydrocarbon
hexadecanoic acid inserted itself well into the binding site similar to
the co-crystallized ligand, by folding into a U-shape. This compound,
as well as the unsaturated hydrocarbon 1-hexadecene, can easily slide
into and bind with low affinity (range of −4.5 to −5.0 kcal/mol) a pock-
et that was previously reported by Ostrem et al. (2013). Consequently,
the binding disrupted the switch regions. Overall, these compounds
can block oncogenic Ras proteins by inhibiting SOS-initiated nucleotide
exchange and subsequent Ras–GTP formation. It should be noted, how-
ever, that such docking studies conducted without MD simulation and
bioassays are speculative (Chen, 2015). Moreover, Ras cannot be con-
cluded readily as the target of these compounds. This is because confir-
mations with bioassays are needed to show the inhibitory effect of these
compounds on the K-Ras pathway and to further determine their effect
on the GEF-mediated nucleotide exchange.
Rasfonin (Fig. 13) is an unsaturated δ-lactone (or α-pyrone) deriva-
tive, isolated from different parts of a few fungal species, including
Lactose
Compound 4
Talaromyces sp. 3656-A1 (Tomikawa et al., 2000), Trichurus terrophilus
(Fujimoto et al., 2000; Akiyama et al., 2003; Fujimoto et al., 2005), and
Doratomyces sp. (Xiao et al., 2014). Compared to its effect on a non-
RAS-dependent cell line, this fungal secondary metabolite with an IC50
value of 0.37 μM was shown to induce apoptosis in Ba/F3-V12 cells
(transfected murine cells (Ba/F3) expressing H-RasG12V) (Tomikawa
et al., 2000). The target selectivity of the compound was tested in vitro
against two different human pancreatic cancer cell lines, BxPC-3
(wild-type KRAS) and Panc-1 (mutated KRAS), and in vivo against
Panc-1 (Xiao et al., 2014). It showed a selective inhibitory effect on
Panc-1 compared with BxPC-3 with an IC50 value of 5.5 μM versus
10 μM, respectively. Apart from inhibiting cell proliferation, the com-
pound also suppressed contact-independent growth, cell migratory ac-
tivity, and the invasiveness of Panc-1 cells in a dose-dependent manner
(Xiao et al., 2014). In these experiments, salirasib was used as a positive
control as it has a well-defined mechanism of action. Like salirasib,
rasfonin reduced EGF-induced Ras activation in a dose-dependent man-
ner. In addition, it also inhibited kinases downstream of Ras, such as c-
Raf, MEK, and ERK without affecting the expression of these proteins,
unlike salirasib, which reduces the cellular levels of the Ras protein
(Xiao et al., 2014). Moreover, based on the GTP/GDP exchange assay,
the suppression of Ras activity was attributed to the reduction in ex-
pression of SOS1, a GEF. In the in vivo experiment, rasfonin showed
greater potency than salirasib in suppressing Panc-1 tumor growth.
6. Downregulation of Ras proteins
The expression of RAS oncogene is crucial during tumor initiation
and maintenance, which is regulated by several Ras–effector pathways
including MAPK/ERK and PI3K/AKT pathways (Lim & Counter, 2005).
Natural compounds in food such as polyphenolics can serve as potential
anticancer agents by modulating cancer cell proliferation and survival.
Quercetin, also known as 3,3′,4′,5,7-pentahydroxyflavone (Fig. 14), is
a naturally occurring polyphenolic flavonoid molecule normally found
as a plant pigment in red wine, fruits, vegetables, and grains. Previous
studies have shown that quercetin is a promising chemopreventive
agent in many types of cancer, such as colorectal (Ranelletti et al.,
2000; Psahoulia et al., 2007), prostate (Slusarz et al., 2010), and breast
(Deng et al., 2013) cancer. Particularly in the Ras signaling pathway,
treatment of primary colon tumors and human colon cancer cell lines
with quercetin led to the inhibition of p21-Ras gene expression and de-
crease in the p21–Ras protein levels (Ranelletti et al., 2000). Notably,
Ras expression was significantly inhibited in two colon cancer cell
lines (Colo-320HSR and Colo-205) after a 48-h treatment with querce-
tin (IC50 = 0.5 μM) (Ranelletti et al., 2000). The compound was shown
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57 49

to be equally effective in inhibiting the expression of all three Ras iso-
forms: H-Ras, K-Ras, and N-Ras (Ranelletti et al., 2000). Similar results
were obtained in other studies reporting the antiproliferative and anti-
tumor effects of quercetin on colorectal cancer cells (Psahoulia et al.,
2007; Xavier et al., 2009). The responses to quercetin treatment include
reduced production of Ras oncoproteins, induced autophagy and apo-
ptosis, and inhibition of cell viability. Psahoulia et al. (2007) also dem-
onstrated the effect of quercetin in accelerating the degradation of Ras
oncoproteins, mediated by proteasomal function, although the
mechanism was unclear. The chemopreventive effect of quercetin can
be attributed to its binding to either the Raf-1 or MEK protein in the
Ras–MAPK pathway, as shown by Lee et al. (2008) in the skin epidermal
cell line with inhibition of tumor promoter-induced neoplastic transfor-
mation. In another study, Song et al. (2013) showed that quercetin sup-
pressed cancer cell invasion and migration by binding to PI3K, another
downstream kinase protein directly activated by Ras-GTP, at the ATP
binding site (Walker et al., 2000). Indeed, quercetin was also shown
to inhibit the proliferation of colon cancer cells through its effect on

A.
4
5 3a 3

6 9 2
8a
8 1
7

Levoglucosenone Tricyclic scaffold

B.

1/2 3

4 5
Fig. 11. Chemical structures of levoglucosenone and its derivatives. (A) Chemical structures of levoglucosenone and the tricyclic scaffold derived from levoglucosenone by 1,3-dipolar
addition. (B) Chemical structures of tricyclic compounds 1–9. Compounds 1 and 2 differ in the stereo-configuration at carbon 3 (C3, solid arrow). The presence of ketone group (solid
box) in compounds 1 and 4 could result in decreased cell penetration and, thus, metabolic instability. Compounds 6 and 7 differ in the stereo-configuration at carbon 4 (C4, dotted
arrow). They show potency in inhibiting cell proliferation, probably due to the presence of the benzyl ether (dashed box) at the C4, which increases the lipophilicity that favors cell
penetration.
50 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

6/7 8

9
Fig. 11 (continued).

K-Ras by regulating both the Ras-MAPK and PI3K pathways (Xavier
et al., 2009). The latter pathway is critical for maintaining transformed
growth in RAS mutant cell lines both in vitro and in RAS-driven tumor
xenografts in mice (Lim & Counter, 2005). Quercetin displayed selectiv-
ity for cells harboring oncogenic forms of Ras, thus preferentially induc-
ing autophagy, cell cycle arrest, and degradation of mutant activated
Ras proteins over their inactive wild-type counterparts (Psahoulia
et al., 2007). Ranelletti et al. (2000) observed a similar phenomenon,
whereby a significant reduction of Ras protein levels was observed in
both colon cancer cell lines and primary colorectal tumors carrying
Ras oncogenes. Delgado et al. (2000) attributed this selectivity to the ar-
rest of oncogenic RAS-induced growth and an increase in the messenger
RNA (mRNA) and protein levels of p21WAF1, a cyclin-dependent kinase
(CDK) inhibitor regulated by the tumor suppressor protein p53. Accord-
ing to Delgado et al. (2000), the p21WAF1 gene promoter was signifi-
cantly activated in transfected NIH3T3 cells with mutant RAS genes in

Pheophorbide a Pheophorbide a Pyropheophorbide a

Hexadecanoic acid 1-Hexadecene

Fig. 12. Chemical structures of selected active compounds found in Typhonium flagelliforme (Keladi tikus).
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
Fig. 13. Chemical structure of rasfonin.
a p53-independent manner, with less potent effects on K562 cells carry-
ing wild-type RAS.
7. Discussion and conclusion
Natural products represent a rich source of lead compounds for drug
discovery. These natural compounds can either be used directly as ther-
apeutic agents or be chemically modified to enhance their biological
activity.
Ras mutations are linked to some of the most fatal cancers, such as
those of the pancreas, lungs, and colon. However, researchers have
failed to discover a clinically efficacious drug that directly targets the on-
cogenic Ras protein and its signaling despite several decades of research.
Thus, oncogenic Ras is considered “undruggable” because it lacks bind-
ing pockets for the interaction between Ras and biologically active small
molecules, thus making the inhibition of oncogenic Ras activity impos-
sible. Alternative approaches to blocking oncogenic Ras signaling have
been explored by targeting the regulatory proteins of Ras activation,
but research efforts have been unsuccessful to date. Blocking Ras activa-
tion with FTIs is an effective method of abrogating oncogenic Ras signal-
ing by targeting other key regulatory proteins in the pathway. Although
these small molecules initially showed potential as inhibitors of Ras
membrane association, a key step of Ras activation, clinical trials involv-
ing these compounds for anticancer treatment have been largely
unsuccessful.
Nevertheless, researchers continue to explore new strategies to target
oncogenic Ras and its signaling. As FTIs were found to be clinically inef-
fective largely due to an alternative mechanism of Ras lipidation, namely
geranylgeranylation, simultaneous inhibition of both farnesylation
and geranylgeranylation with a combinatory therapy of FTI and a
geranylgeranyltransferase inhibitor (GGTI) or a dual prenylation inhibi-
tor (DPI) was considered. Lobell et al. (2001) followed through on this
concept in preclinical models: mice and dogs. Although FTI–GGTI combi-
natory therapy and DPI were known to effectively cause deprenylation of
K-Ras in vivo, the effect was not selective to K-Ras in either normal or
tumor mouse tissue. They proved that GGTIs and DPIs caused systemic
toxicity at effective doses, mainly through an unknown mechanism in ad-
dition to myelosuppression, which was ultimately fatal. As a result, the
therapeutic potential of FTI–GGTI and DPIs was significantly limited by
the inherent toxicity, which might be a response to FTI–GGTI and DPI
Fig. 14. Chemical structure of quercetin. The chromone structure (dashed box) is critical
for its binding to the ATP-binding site of PI3K, which is a protein directly downstream of
Ras.
51
treatments that involved prenylated proteins other than K-Ras. Because
of the narrow therapeutic window, FTI-GGTI and DPIs may not be a
promising strategy for blocking oncogenic Ras signaling.
Recently, several novel natural compounds and their derivatives that
target oncogenic Ras and its signaling have emerged, showing antican-
cer activity in the preclinical stage, thus leading to further development
of drugs targeting RAS-driven cancers (Table 1). In this timely review,
we elaborated on natural compounds and their derivatives with an in-
hibitory effect on Ras signaling and anticancer activity at different stages
of drug discovery. We categorized these compounds based on the four
known strategies for blocking Ras signaling: inhibition of Ras–effector
interactions, interference of Ras membrane association, prevention of
Ras–GTP formation, and downregulation of Ras proteins. We also
discussed the anticancer potential of a novel prospective anti-Ras strat-
egy: enhancement of the intrinsic Ras GTPase activity.
Sulindac sulfide is a Ras binder with a dual inhibitory effect on Ras
signaling. The compound simultaneously impairs the nucleotide ex-
change of Ras and inhibits Ras–Raf-1 protein–protein interaction, thus
blocking Ras signaling. Due to its dual inhibitory effect on Ras-
mediated signaling, this compound is a pharmacologically attractive
compound for targeting RAS-driven cancers. Nevertheless, studies
have shown that the biological activity of sulindac sulfide is rather
weak and its anticancer activity has not been investigated in depth.
Therefore, a more potent compound must be derived from sulindac sul-
fide to strengthen the anti-Ras activity of its existing pharmacophore.
Another biologically active metabolite of sulindac, sulindac sulfone, is
a promising candidate that must be examined for its role in Ras signal-
ing. It has been studied in RAS-driven colorectal and mammary cancers;
however, none of these studies showed that the antitumor effect of
sulindac sulfone arose from the direct inhibition of Ras signaling. Al-
though a direct physical interaction between Ras and sulindac sulfone
has not been documented, the structural similarity of these two com-
pounds suggests the possibility of the direct binding of sulindac sulfone
to Ras. However, this hypothesis must be validated by in vitro binding
assays and crystallography studies.
Despite the failure of FTIs in clinical trials, blocking oncogenic Ras
signaling by targeting Ras membrane localization remains an effective
approach, because any key regulatory protein in the Ras membrane lo-
calization process serves as a potential target. Salirasib, a farnesyl mimic
that competes with farnesylated Ras for membrane association, can be
potentially developed into a drug targeting PDAC, 90% of which harbors
K-Ras mutations (Almoguera et al., 1988; Laghi et al., 2002). Salirasib
was found to dislodge active GTP-bound H-Ras and K-Ras-4B mutants
from the plasma membrane by competitively binding with the docking
protein of the oncoprotein, galectin (Fig. 4). Subsequently, the active cy-
toplasmic H-Ras and K-Ras-4B proteins were found to be degraded. The
earliest phase II clinical trial was conducted in patients with lung adeno-
carcinoma. Unfortunately, the trials were unsuccessful as salirasib was
found to be clinically ineffective. In a later study, the clinical efficacy of
salirasib was reevaluated in combination with gemcitabine in an animal
model (Laheru et al., 2012). The combination therapy exhibited good
anticancer activity against patient-derived PDAC xenografts and im-
proved the survival rate in nude mice, thus indicating the efficacy of a
combination therapy in targeting RAS-driven cancers (Haklai et al.,
2008; Laheru et al., 2012). This promising preclinical result highlights
the need to evaluate the clinical efficacy of this combination therapy
in patients with advanced PDAC. Combination therapy involving
gemcitabine and other cytotoxic drugs has been used in cancer treat-
ment previously. Erlotinib and gemcitabine was found to significantly
improve the survival rate of advanced pancreatic cancer (Moore et al.,
2007).
PFTS was found to be more potent than its parental compound,
salirasib, in vitro and in vivo (Mackenzie et al., 2013). Salirasib was
found to be partially metabolized into salirasib glucuronide, whereas
PFTS was completely metabolized into salirasib and salirasib glucuro-
nide in the blood plasma after oral administration. Compared with
 52
Table 1
Summary of natural compounds and their derivatives as anti-Ras agents.

Compound Class of compound Origin Ras-targeting Specific Phases of drug Inhibitory activity Reference(s)
strategy mechanism development

Sulindac sulfide Phenylpropanoid Derived from safrole, a
compound found in sassafras
plants
Salirasib Phenolic acids Derived from salicylic acid, a
compound isolated from
willow tree, prokaryotic and
eukaryotic organisms
Inhibition of Ras–effector
interactions and inhibition of
Ras–GEF nucleotide exchange
Interference of Ras membrane
association
Binds Ras In vitro Focus formation – Barreiro & Lima, 1992;
H-rasG12V-transformed Herrmann et al., 1998;
REF (IC50 = 10–50 μM) Brunell et al., 2011
Binds Phase I clinical trials in MIA PaCa-2 (IC50 ≥ 80 μM); Marom et al., 1995;
galectin patients with advanced H-ras-transformed EJ cells Aharonson et al., 1998; Egozi
PDAC and refractory (IC50 = 5–15 μM); et al., 1999; Weisz et al.,
hematologic malignancies. Panc-1 1999; Gana-Weisz et al.,

S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57

Phase II clinical trial in (IC50 = 35 μM); 2002; Kloog & Cox, 2004;
patients with lung U87 Haklai et al., 2008; Chen
adenocarcinomas harboring (IC50 = 50 μM) et al., 2009; Goldberg et al.,
K-Ras mutations 2009; Riely et al., 2011;
Laheru et al., 2012; Baker &
Der, 2013; Mackenzie et al.,
2013; Badar et al., 2015
S-prenyl derivatives Phenolic acids Derived from salirasib Interference of Ras membrane Binds In vitro H-ras transformed EJ cells Aharonson et al., 1998
association galectin (IC50 = 10–N50 μM); Goldberg et al., 2009
Panc-1 (IC50 = 10–95 μM);
U87 (IC50 = 10–52 μM)
Carboxyl-modified Phenolic acids Derived from salirasib Interference of Ras membrane Binds In vivo — mice MIA PaCa-2 (IC50 = Mackenzie et al., 2013
derivatives, association galectin 44.1 ± 2.8 μM)
including PFTS
TLN-4601 Alkaloid Marine actinomycete Interference of Ras membrane Depletes Raf Phase I and II clinical trials MIA PaCa-2 (IC50 = 10 μM) Charan et al., 2004; Boufaied
bacterium, Micromonospora association in patients with GBM et al., 2010; Campbell et al.,
sp. 2010; Mason et al., 2010,
2012
Spermatinamine and Alkaloids Australian Verongida marine Interference of Ras membrane Bind ICMT In vitro • Spermatinamine – ICMT Buchanan et al., 2007, 2008;
aplysamine 6 sponge, Pseudoceratina sp. association (IC50 = 1.9 μM); K562 and Kottakota et al., 2012
HT29 (GI50 = 0.23 μM)
• Aplysamine 6 –ICMT
(IC50 = 14 μM)
Palmostatin B Lactone Derived from lipstatin, a Interference of Ras membrane Binds APT1 In vitro APT 1 (IC50 = 670 nM) Weibel et al., 1987; Dekker
compound isolated from the association et al., 2010; Cox, 2010;
sctinobacterium Dekker & Hedberg, 2011;
Streptomyces toxytricini Hernandez et al., 2013; Zhu
et al., 2014
SRJ09, SRJ10, and SRJ23 Terpenoids Derived from AGP, a Inhibition of Ras–GEF Bind Ras In vitro • SRJ09 – MCF-7 Jada et al., 2008; Hocker et al.,
compound isolated from an nucleotide exchange (GI50 = 4.2 μM); MDA-231 2013
Asian herb, Andrographis (GI50 = 5.0 μM); PC-3
paniculata (GI50 = 12.4 μM); HCT-116
(GI50 = 3.1 μM)
• SRJ10 – MCF-7
(GI50 = 5.3 μM); HCT-116
(GI50 = 4.7 μM)
• SRJ23 – MCF-7
(GI50 = 5.2 μM); MDA-231
(GI50 = 5.8 μM); PC-3

SCH-53870, SCH-53239,
and SCH-54292
Compounds 1–10 derived
from D-arabinose
Glycosylated Compounds
2–4
Compounds 1–9 derived
from levoglucosenone
Pheophorbide a,
pheophorbide a’,
pyropheophorbide a,
hexadecanoic acid,
1-hexadecene
Rasfonin
Quercetin
Abbreviations: AGP, Andrographolide; APT1, acyl protein thioesterase 1; GBM, Glioblastoma multiforme; GEF, Guanine nucleotide exchange factor; ICMT, Isoprenylcysteine carboxymethyltransferase; K-Ras, Kirsten-Ras; MEK, Mitogen-activated
protein/extracellular signal-regulated kinase kinase; PDAC, Pancreatic ductal adenocarcinoma; PFTS, Phospho-farnesylthiosalicylic acid; PI3K, Phosphoinositide-3-kinase; Raf, Rapidly accelerated fibrosarcoma; Ras, Ras sarcoma; REF, Rat embryo
fibroblasts; SOS1, Son of Sevenless 1.
Cancer Cell Lines: A549, lung; A2780, ovary; BxPC-3, pancreas; Colo-205, colon; Colo-320HSR, colon; HCT-116, colon; HT29, colon; K562, bone marrow; MCF-7, breast; MDA-231, breast; MIA PaCa-2, pancreas; Panc-1, pancreas; PC-3, prostate; U87,
brain.
• SCH-53,239 – • SCH-53,239 – guanine nu-
nucleobase cleotide
• SCH-54,292 – sugar • SCH-54,292 – glucose
Sugars D-arabinose
Sugars • Compound 2 — glucose
• Compound 3 — N-acetyl
glucosamine
• Compound 4 — lactose
Carbohydrates Levoglucosenone (cellulose)
• Pheophorbide a, Typhonium flagelliforme
pheophorbide a’, (Keladi tikus)
pyropheophorbide
a–alkaloids
• Hexadecanoic
acid–carboxylic
acid
• Hexadecene–alkene
Lactone From fungi: Talaromyces
sp. 3656-A1, Trichurus
terrophilus, and Doratomyces
sp.
Flavonoid Red wine, many fruits,
vegetables, and grains
Inhibition of Ras–GEF
nucleotide exchange
Inhibition of Ras–GEF
nucleotide exchange
Inhibition of Ras–GEF
nucleotide exchange
Inhibition of Ras–GEF
nucleotide exchange
Inhibition of Ras–GEF
nucleotide exchange
Reduction of SOS1 (GEF)
protein expression
Downregulation of Ras
proteins
(GI50 = 6.6 μM); HCT-116
(GI50 = 3.8 μM)
Bind Ras In vitro • SCH-53,870 and SCH-53,239 – Taveras et al., 1997; Palmioli,
Nucleotide-exchange 2009
(IC50 = 0.5 μM)
• SCH-54,292 - Nucleotide
exchange (IC50 = 0.7 μM)
Bind Ras In vitro • Compound 1 – Nucleotide Peri et al., 2005, 2006
exchange (IC50 = 76.0 μM)
• Compound 2 – Nucleotide
exchange (IC50 = 57.3 μM)
• Compound 3 – Nucleotide
exchange (IC50 = 35.5 μM)
• Compound 4 – Nucleotide
exchange (IC50 = 62.3 μM)
• Compounds 5–10 have no
clearly defined IC50 values
Bind Ras In vitro • Compound 2 – Nucleotide Sacco et al., 2011
exchange (IC50 = 80 μM)
• Compound 3 – Nucleotide
S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
exchange (IC50 = 80 μM)
• Compound 4 – Nucleotide
exchange (IC50 = 100 μM)
Bind Ras In vitro A2780 (IC50 = 3.1 to N40 μM); Müller et al., 2009
MCF-7 (IC50 = 5.5 to N40 μM);
A549 (IC50 = 15.5 to N40 μM;
HCT-116 (IC50 = 17 to N40 μM )
Bind Ras In silico - Lai et al., 2009; Fatima &
Yee, 2014
Is not clearly In vivo — mice Ba/F3-V12 Tomikawa et al., 2000;
defined (IC50 = 0.37 μM); Panc-1 (IC50 Fujimoto et al., 2000;
= 5.5 μM); BxPC-3 (IC50 = 10 Akiyama et al., 2003;
μM) Fujimoto et al., 2005;
Xiao et al., 2014
Binds Raf or In vitro Ras expression – Ranelletti et al., 2000;
MEK and Colo-320HSR and Psahoulia et al., 2007; Xavier
PI3K proteins Colo-205 et al., 2009; Slusarz et al.,
(IC50 = 0.5 μM) 2010; Deng et al., 2013; Song
et al., 2013
53
54 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
salirasib metabolism, an equimolar concentration of PFTS led to higher
levels of salirasib and lower levels of salirasib glucuronide. Based on
these findings, salirasib can be considered the primary contributor to
the high potency of PFTS.
The farnesyl group of salirasib is the core pharmacophore responsi-
ble for the reported anti-Ras activity of this compound. The farnesyl
group acts by competing with GTP-bound H-Ras and K-Ras-4B proteins
for binding to the binding domain on galectin at the plasma membrane.
Notably, the length of the isoprenoid group on a compound determines
the inhibitory effect of this compound on Ras activity and Ras-mediated
cell proliferation and survival (Aharonson et al., 1998). A compound
with a 10-carbon geranyl group was found to lack the inhibitory effect
found in compounds with a 15-carbon farnesyl group and a 20-carbon
geranylgeranyl group (Aharonson et al., 1998). Although the mecha-
nism of Ras inhibition by salirasib derivatives was not elucidated previ-
ously, we propose that they may demonstrate the same anti-Ras
mechanism as salirasib as they share a common 15-carbon farnesyl
group. Structural modifications of salirasib, such as halogenation of the
benzene ring, esterification, or amidation of the carboxyl group of the
thiosalicylic acid, generate new compounds with varying anti-Ras activ-
ity in vitro and in vivo. Among the salirasib derivatives, amide deriva-
tives showed improved efficacy compared with their parental
compound, although these derivatives retained the inhibitory selectivi-
ty of salirasib toward the GTP-bound Ras protein. Therefore, amide de-
rivatives of salirasib may be used in place of salirasib as promising Ras
inhibitors for cancer therapy.
TLN-4601 also shares the same farnesyl group with salirasib. This
compound was reported to reduce the GTP-bound K-Ras levels in addi-
tion to depleting the total Raf-1 protein in pancreatic cancer cells and
transformed cells, both of which express oncogenic K-Ras (Campbell
et al., 2010). We postulate that TLN-4601 inhibits Ras activity by inter-
fering with its association to the plasma membrane. Similar to salirasib,
this compound may compete with GTP-bound H-Ras and K-Ras-4B pro-
teins for binding to galectin, thus increasing the unbound activated Ras
that can be readily degraded in the cytoplasm.
The posttranslational modification of Ras, so as to transport Ras intra-
cellularly to the plasma membrane and embed it to the inner leaflet of
the plasma membrane, is regulated by several enzymes. These enzymes
are alternative targets for blocking oncogenic Ras signaling via the pre-
vention of Ras maturation in a subsequent GEF-mediated activation at
the plasma membrane. In this review, we described several natural
compounds that target these enzymes. Among these compounds,
spermatinamine emerged as the leading compound with significant inhi-
bition on Ras activation. Although spermatinamine is structurally large, it
can serve as an effective inhibitor of Ras activation because of its relative-
ly strong inhibitory effect on the catalytic activity of ICMT and promising
sub-micromolar anticancer effect on leukemia and colon cancer cell lines.
Future studies that specifically focus on modifying the chemical structure
of spermatinamine are needed to improve its drug likeness. Neverthe-
less, changing the chemical structure of spermatinamine without affect-
ing its inherent anticancer activity is challenging.
Palmostatin B led to sub-micromolar inhibition of APT1 activity
in vitro, and its inhibitory effect on Ras signaling was proven in a cell
line that stably expressed oncogenic Ras (Dekker et al., 2010). The com-
pound was found to prevent palmitoylated Ras from relocalizing from
the endomembranes to the plasma membrane, thus inhibiting GEF-
dependent Ras activation. Thus, it can serve as an effective inhibitor of
Ras signaling. However, it is primarily limited by its poor substrate se-
lectivity. The anti-depalmitoylation activity of palmostatin B targets a
wide range of palmitoylated proteins in addition to Ras (Zeidman
et al., 2009). These palmitoylated proteins are distributed in various tis-
sues, which are crucial for maintaining the functioning of an organism
(Zeidman et al., 2009). The poor selectivity poses a limitation to tests
in animal models. To overcome this limitation, palmostatin B can be
modified to increase its selectivity toward oncogenic Ras, thus minimiz-
ing its systemic toxicity.
For wild-type Ras to be activated, GDP must be removed from the
nucleotide-binding pocket by GEF, thus providing space for the intracel-
lularly abundant GTP to bind to Ras. Whether oncogenic Ras, mutations
of which cause so-called constitutive activation, would ever return to its
GDP-bound inactive state and require GEF for its reactivation remains to
be elucidated. In a recent study, contrary to popular opinion, oncogenic
Ras was not found to be constitutively active and independent of GEF-
mediated activation (Huang et al., 2014). Although Ras mutations inter-
fere with Ras–GAP interaction and impair the GTP hydrolysis of active
Ras, its intrinsic GTPase activity remains unaffected. Therefore, the
GTP hydrolysis of active oncogenic Ras can still occur in the absence of
GAP, but more slowly, leading to prolonged Ras activation. Once onco-
genic Ras returns to its GDP-bound inactive form due to its intrinsic
GTPase activity, it relies on GEF for its activation.
To this end, we were able to confirm the feasibility of targeting Ras–
GEF interaction and GTP loading of Ras as a strategy to block oncogenic
Ras signaling. As oncogenic Ras is activated in a GEF-dependent manner,
similar to wild-type Ras, interfering with Ras–GEF interaction and
inhibiting GTP loading of Ras remains an effective approach to target on-
cogenic Ras signaling. AGP and its benzylidene derivatives (SRJ09 and
SRJ23) were found to block the nucleotide exchange of wild-type and
oncogenic Ras by abrogating Ras–GEF interaction. Based on the SBDD
approach and ensemble docking, Hocker et al. (2013) reported Ras as
a potential target of AGP and its benzylidene derivatives. AGP is a natu-
ral compound with significant anticancer activity. The benzylidene de-
rivatives of AGP exhibit excellent anticancer activity against several
cancer cell lines. Along with in silico docking results, SRJ09 and SRJ23
were shown to inhibit the EGF-induced activation of endogenous
wild-type Ras, indicating that SRJ09 and SRJ23 inhibited GEF-mediated
Ras activation by directly binding to Ras. Prolonged treatment of cells
stably expressing exogenous oncogenic Ras with SRJ09 and SRJ23 re-
duced the levels of GTP-bound oncogenic Ras. This showed that onco-
genic Ras slowly returned to its inactive state due to its intrinsic
GTPase activity, and AGP derivatives prevented the reactivation of onco-
genic Ras by GEF. It is interesting to note that AGP derivatives selectively
showed low micromolar growth inhibitory activity against cancer cell
lines harboring Ras mutations, in spite of unbiased docking of these
compounds into wild-type and oncogenic Ras in silico. Altogether, AGP
and its derivatives may serve as effective anti-Ras drugs with minimal
adverse effects on the normal cells.
Recently, a promising natural compound, rasfonin, was found to tar-
get oncogenic Ras signaling by downregulating the protein expression
of GEF. Rasfonin was expected to equally inhibit both signaling path-
ways mediated by wild-type and oncogenic Ras, considering that
these pathways are activated in a GEF-dependent manner. Neverthe-
less, Xiao et al. (2014) reported that rasfonin was selective in inhibiting
the growth of a PDAC cell line harboring oncogenic K-Ras over a cell line
harboring wild-type K-Ras. This finding is indeed interesting, as is the
molecular mechanism proposed to explain the greater dependency of
RAS oncogene-driven cell proliferation on GEF. The inhibitory effect of
rasfonin on RAS oncogene-driven cell growth, migration, and invasion
was also shown to be confined to inhibiting the Ras–MAPK pathway.
Xiao et al. (2014) investigated the effect of rasfonin on several EGFR-
mediated pathways, among which the MAPK pathway showed signifi-
cant response to rasfonin treatment. These findings suggested that
rasfonin may serve as a potential inhibitor of Ras signaling with signifi-
cant target specificity, thus overcoming the limitation of poor selectivity
commonly encountered in drug discovery.
Quercetin is another promising natural candidate for the treatment
of RAS-driven cancers. Studies have shown that quercetin inhibits RAS
oncogene expression and promotes the degradation of Ras oncoprotein
in primary colon tumors and human colon cancer cell lines. The antican-
cer effect of quercetin on colorectal cancer is clear as Ras signaling is im-
plicated in cell proliferation and survival. Quercetin was also found to
inhibit cancer cell growth, survival, invasion, and migration via physical
interaction with the Ras effector, thus blocking oncogenic Ras signaling.
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
Ranelletti et al. (2000) and Psahoulia et al. (2007) showed that the se-
lectivity of quercetin in inhibiting the growth of cancer cells harboring
oncogenic Ras can serve as the basis for the future development of
anti-Ras drugs with reduced off-target effects.
Oncogenic Ras shows an impaired response to GAP-mediated GTP
hydrolysis, leading to prolonged activation of oncogenic Ras. Neverthe-
less, it is yet to be examined whether the GTPase activity of oncogenic
Ras can be enhanced using a small molecule to compensate for its insen-
sitivity to GAP-mediated inactivation. Here, we propose a strategy ap-
plying computer-assisted screening on natural compounds and their
derivatives that readily dock into the active site of GTP hydrolysis on
Ras, wherein one or more pharmacophores of the compounds are in-
volved in stabilizing the transition state of GTP hydrolysis. Alternatively,
discovering a compound that can nullify the effect of Ras mutations
and facilitate Ras–GAP interaction is also feasible for counteracting
prolonged Ras activation.
Finding an effective therapeutic agent for RAS-driven cancers is not
impossible altogether, despite the challenges in drugging Ras. With
the development of popular anticancer agents from natural compounds,
we believe that natural products represent a rich source of novel anti-
cancer agents, particularly agents that target oncogenic Ras and its sig-
naling. Decades of research efforts and failures can inform future
effective methods of drugging Ras as well as blocking its signaling. It is
all the more imperative to discover agents that target oncogenic Ras
and its signaling.
Conflict of interest
The authors declare that there are no conflicts of interest.
Acknowledgements
The authors would like to thank the Malaysian Ministry of Higher
Education (MOHE) for funding projects through the Research Universi-
ty Grant Scheme (RUGS, 04-02-12-2017RU) and Fundamental Research
Grant Scheme (FRGS, 04-02-13-1324FR).
References
Aharonson, Z., Gana-Weisz, M., Varsano, T., Haklai, R., Marciano, D., & Kloog, Y. (1998).
Stringent structural requirements for anti-Ras activity of S-prenyl analogues.
Biochim Biophys Acta 1406(1), 40–50.
Akiyama, K., Kawamoto, S., Fujimoto, H., & Ishibashi, M. (2003). Absolute stereochemistry
of TT-1 (rasfonin), an α-pyrone-containing natural product from a fungus, Trichurus
terropilus. Tetrahedron Lett 44, 8427–8431.
Almoguera, C., Shibata, D., Forrester, K., Martin, J., Arnheim, N., & Perucho, M. (1988).
Most human carcinoma of the exocrine pancreas contain mutant c-K-Ras genes.
Cell 53, 549–554.
Badar, T., Cortes, J. E., Ravandi, F., O'Brien, S., Verstovsek, S., Garcia-Manero, G., et al.
(2015). Phase I study of S-trans,trans-farnesylthiosalicylic acid (salirasib), a novel
oral RAS inhibitor in patients with refractory hematologic malignancies. Clin
Lymphoma Myeloma Leuk 15(7), 433–438.
Baines, A. T., Xu, D., & Der, C. J. (2011). Inhibition of Ras for cancer treatment: the search
continues. Future Med Chem 3, 1787–1808.
Baker, N. M., & Der, C. J. (2013). Cancer: Drug for an ‘undruggable’ protein. Nature
497(7451), 577–578.
Barreiro, E. J., & Lima, M. E. (1992). The synthesis and anti-inflammatory properties of a
new sulindac analogue synthesized from natural safrole. J Pharm Sci 81(12),
1219–1222.
Bernards, A., & Settleman, J. (2004). GAP control: Regulating the regulators of small
GTPases. Trends Cell Biol 14, 377–385.
Berndt, N., Hamilton, A. D., & Sebti, S. M. (2011). Targeting protein prenylation for cancer
therapy. Nat Rev Cancer 11, 775–791.
Blum, R., Jacob-Hirsch, J., Rechavi, G., & Kloog, Y. (2006). Suppression of survivin expres-
sion in glioblastoma cells by the Ras inhibitor farnesylthiosalicylic acid promotes
caspase-dependent apoptosis. Mol Cancer Ther 5(9), 2337–2347.
Boriack-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D., & Kuriyan, J. (1998). The structural basis
of the activation of Ras by SOS. Nature 394(6691), 337–343.
Bos, J. L., Rehmann, H., & Wittinghofer, A. (2007). GEFs and GAPs: Critical elements in the
control of small G Proteins. Cell 129(5), 865–877.
Boufaied, N., Wioland, M. A., Falardeau, P., & Gourdeau, H. (2010). TLN-4601, a novel an-
ticancer agent, inhibits Ras signalling post Ras prenylation and before MEK activation.
Anticancer Drugs 21(5), 543–552.
55
Brenke, R., Kozakov, D., Chuang, G. Y., Beglov, D., Hall, D., Landon, M. R., et al. (2009).
Fragment-based identification of druggable 'hot spots' of proteins using Fourier do-
main correlation techniques. Bioinformatics 25(5), 621–627.
Brunell, D., Sagher, D., Kesaraju, S., Brot, N., & Weissbach, H. (2011). Studies on the metab-
olism and biological activity of the epimers of sulindac. Drug Metab Dispos 39(6),
1014–1021.
Buchanan, M. S., Carroll, A. R., Fechner, G. A., Boyle, A., Simpson, M. M., & Addepalli, R.
(2008). Aplysamine 6, an alkaloidal inhibitor of isoprenylcysteine carboxyl methyl-
transferase from the sponge Pseudoceratina sp. J Nat Prod 71(6), 1066–1067.
Buchanan, M. S., Carroll, A. R., Fechner, G. A., Boyle, A., Simpson, M. M., Addepalli, R., et al.
(2007). Spermatinamine, the first natural product inhibitor of isoprenylcysteine
carboxylmethyltransferase, a new cancer target. Bioorg Med Chem Lett 17(24),
6860–6863.
Campbell, P. M., Boufaied, N., Fiordalisi, J. J., Cox, A. D., Falardeau, P., Der, C. J., et al. (2010).
TLN-4601 suppresses growth and induces apoptosis of pancreatic carcinoma cells
through inhibition of Ras–ERK MAPK signaling. J Mol Signal 5, 18.
Chandra, A., Grecco, H. E., Pisupati, V., Perera, D., Cassidy, L., Skoulidis, F., et al. (2012). The
GDI-like solubilizing factor PDEδ sustains the spatial organization and signaling of Ras
family proteins. Nat Cell Biol 14, 148–158.
Chandrashekar, R., & Adams, P. D. (2013). Prospective development of small molecule tar-
gets to oncogenic Ras proteins. J Biophys 3, 207–211.
Charan, R. D., Schlingmann, G., Janso, J., Bernan, V., Feng, X., & Carter, G. T. (2004).
Diazepinomicin, a new antimicrobial alkaloid from a marine Micromonospora sp.
J Nat Prod 67(8), 1431–1433.
Chen, Y. (2015). Beware of docking. Trends Pharmacol Sci 36(2), 78–95.
Chen, L., Morrow, J. K., Tran, H. T., Phatak, S. S., Du-Cuny, L., & Zhang, S. (2012). From lap-
top to benchtop to bedside: Structure-based drug design on protein targets. Curr
Pharm Des 18(9), 1217–1239.
Chen, Z., Zheng, Z., Huang, J., Lai, Z., & Fan, B. (2009). Biosynthesis of salicylic acid in plants.
Plant Signal Behav 4(6), 493–496.
Cherfils, J., & Zeghouf, M. (2013). Regulation of small GTPases by GEFs, GAPs, and GDIs.
Physiol Rev 93, 269–309.
Cox, A. D. (2010). Protein localization: Can too much lipid glue stop Ras? Nat Chem Biol
6(7), 483–485.
Cox, A. D., Der, C. J., & Philips, M. R. (2015). Targeting RAS Membrane Association: Back to
the future for anti-RAS drug discovery? Clin Cancer Res 21(8), 1819–1827.
Cox, A. D., Fesik, S. W., Kimmelman, A. C., Luo, J., & Der, C. J. (2014). Drugging the
undruggable RAS: Mission possible? Nat Rev Drug Discov 13, 828–851.
de Castro Carpeño, J., & Belda-Iniesta, C. (2013). KRAS mutant NSCLC, a new opportunity for
the synthetic lethality therapeutic approach. Transl Lung Cancer Res 2(2), 142–151.
de la Cruz, F. F., Gapp, B. V., & Nijman, S. M. B. (2015). Synthetic lethal vulnerabilities of
cancer. Annu Rev Pharmacol Toxicol 55, 513–531.
Dekker, F. J., & Hedberg, C. (2011). Small molecule inhibition of protein depalmitoylation
as a new approach towards downregulation of oncogenic Ras signaling. Bioorg Med
Chem 19(4), 1376–1380.
Dekker, F. J., Rocks, O., Vartak, N., Menninger, S., Hedberg, C., Balamurugan, R., et al.
(2010). Small-molecule inhibition of APT1 affects Ras localization and signaling. Nat
Chem Biol 6, 449–456.
Delgado, M. D., Vaqué, J. P., Arozarena, I., Ló-Ilasaca, M. A., Martínez, C., Crespo, P., et al.
(2000). H-, K- and N-Ras inhibit myeloid leukemia cell proliferation by a p21WAF1-
dependent mechanism. Oncogene 19, 783–790.
Deng, X., Song, H., Zhou, Y., Yuan, G., & Zheng, F. (2013). Effects of quercetin on the proliferation of
breast cancer cells and expression of surviving in vitro. Exp Ther Med 6(5), 1155–1158.
Egozi, Y., Weisz, B., Gana-Weisz, M., Ben-Baruch, G., & Kloog, Y. (1999). Growth inhibition
of Ras-dependent tumors in nude mice by a potent ras-dislodging antagonist. Int J
Cancer 80(6), 911–918.
Elad-Sfadia, G., Haklai, R., Balan, K., & Kloog, Y. (2004). Galectin-3 augments K-Ras activa-
tion and triggers a Ras signal that attenuates ERK but not phosphoinositide 3-kinase
activity. J Biol Chem 279(33), 34922–34930.
Elad-Sfadia, G., Haklai, R., Ballan, E., Gabius, H. J., & Kloog, Y. (2002). Galectin-1 augments
Ras activation and diverts Ras signals of Raf-1 at the expense of phosphoinositide 3-
kinase. J Biol Chem 277(40), 37169–37175.
Fatima, A., & Yee, H. F. (2014). In silico screening of mutated K-Ras inhibitors from
Malaysian Typhonium flagelliforme for non-small cell lung cancer. Adv Bioinformatics
2014. http://dx.doi.org/10.1155/2014/431696.
Forbes, S. A., Bindal, N., Bamford, S., Cole, C., Kok, C. Y., Beare, D., et al. (2011). COSMIC:
Mining complete cancer genomes in the catalogue of somatic mutations in cancer.
Nucleic Acids Res 39, D945–D950.
Fujimoto, H., Okamoto, Y., Sone, E., Maeda, S., Akiyama, K., & Ishibashi, M. (2005). Eleven
new 2-pyrones from a fungi imperfecti, Trichurus terrophilus, found in a screening
study guided by immunomodulatory activity. Chem Pharm Bull 53(8), 923–929.
Fujimoto, H., Sone, E., Okuyama, E., & Ishibashi, M. (2000). Annual meeting of the pharma-
ceutical society of Japan. Abstr Pap 2, 68.
Gana-Weisz, M., Halaschek-Wiener, J., Jansen, B., Elad, G., Haklai, R., & Kloog, Y. (2002).
The Ras inhibitor S-trans, trans-farnesylthiosalicylic acid chemosensitizes human
tumor cells without causing resistance. Clin Cancer Res 8(2), 555–565.
Gideon, P., John, J., Frech, M., Lautwein, A., Clark, R., Scheffler, J. E., & Wittinghofer, A.
(1992). Mutational and kinetic analyses of the GTPase-activating protein (GAP)-
p21 interaction: The C-terminal domain of GAP is not sufficient for full activity. Mol
Cell Biol 12(5), 2050–2056.
Goldberg, L., & Kloog, Y. (2006). A ras inhibitor tilts the balance between Rac and Rho and
blocks phosphatidylinositol 3-kinase-dependent glioblastoma cell migration. Cancer
Res 66(24), 11709–11717.
Goldberg, L., Haklai, R., Bauer, V., Heiss, A., & Kloog, Y. (2009). New derivatives of
farnesylthiosalicylic acid (salirasib) for cancer treatment: Farnesylthiosalicylamide
inhibits tumor growth in nude mice models. J Med Chem 52, 197–205.
56 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
Grant, B. J., Lukman, S., Hocker, H. J., Sayyah, J., Brown, J. H., McCammon, J. A., et al. (2011).
Novel allosteric sites on ras for lead generation. PLoS One 6(10), e25711.
Greer, J., Erickson, J. W., Baldwin, J. J., & Varney, M. D. (1994). Application of the three-
dimensional structures of protein target molecules in structure-based drug design.
J Med Chem 37(8), 1035–1054.
Guha, A., Feldkamp, M. M., Lau, N., Boss, G., & Pawson, A. (1997). Proliferation of human
maglinant astrocytomas is dependent on Ras activation. Oncogene 15(23),
2755–2765.
Guo, W., Wu, S., Liu, J., & Fang, B. (2008). Identification of a small molecule with synthetic
lethality for K-RAS and protein kinase C iota. Cancer Res 68, 7403–7408.
Haklai, R., Elad-Sfadia, G., Egozi, Y., & Kloog, Y. (2008). Orally administered FTS (salirasib)
inhibits human pancreatic tumor growth in nude mice. Cancer Chemother Pharmacol
61(1), 89–96.
Hanson, J. R. (2003). The classes of natural products and their isolation. In E. W. Abel (Ed.),
Natural products: The secondary metabolites (pp. 1–34). Cambridge, United Kingdom:
Royal Society of Chemistry Publishing.
Hernandez, J. L., Majmudar, J. D., & Martin, B. R. (2013). Profiling and inhibiting reversible
palmitoylation. Curr Opin Chem Biol 17(1), 20–26.
Herrmann, C. (2003). Ras-effector interactions: After one decade. Curr Opin Struct Biol 13,
122–129.
Herrmann, C., Block, C., Geisen, C., Haas, K., Weber, C., & Winde, G. (1998). Sulindac sulfide
inhibits Ras signaling. Oncogene 17, 1769–1776.
Hocker, H. J., Cho, K., Chen, C. K., Rambahal, N., Sagineedu, S. R., Shaari, K., et al. (2013).
Andrographolide derivatives inhibit guanine nucleotide exchange and abrogate on-
cogenic Ras function. Proc Natl Acad Sci 110, 10201–10206.
Huang, H., Daniluk, J., Liu, Y., Chu, J., Li, Z., Ji, B., & Logsdon, C. D. (2014). Oncogenic K-Ras
requires activation for enhanced activity. Oncogene 33, 532–535.
Jada, S. R., Matthews, C., Saad, M. S., Hamzah, A. S., Lajis, N. H., Stevens, M. F. G., et al. (2008).
Benzylidene derivatives of andrographolide inhibit growth of breast and colon cancer
cells in vitro by inducing G1 arrest and apoptosis. Br J Pharmacol 155, 641–654.
John, J., Rensland, H., Schlichting, I., Vetter, I., Borasio, G. D., Goody, R. S., et al. (1993). Ki-
netic and structural analysis of the Mg(2+)-binding site of the guanine nucleotide-
binding protein p21H-ras. J Biol Chem 268, 923–929.
Kapoor, A., & Travesset, A. (2015). Differential dynamics of RAS isoforms in GDP- and
GTP-bound states. Proteins 83(6), 1091–1106.
Kloog, Y., & Cox, A. D. (2004). Prenyl-binding domains: potential targets for Ras inhibitors
and anti-cancer drugs. Semin Cancer Biol 14(4), 253–261.
Konstantinopoulos, P. A., Karamouzis, M. V., & Papavassiliou, A. G. (2007). Post-
translational modifications and regulation of the RAS superfamily of GTPases as anti-
cancer targets. Nat Rev Drug Discov 6, 541–555.
Kottakota, S. K., Benton, M., Evangelopoulos, D., Guzman, J. D., Bhakta, S., & McHugh, T. D.
(2012). Versatile routes to marine sponge metabolites through benzylidene
rhodanines. Org Lett 14(24), 6310–6313.
Kozakov, D., Hall, D. R., Jehle, S., Luo, L., Ochiana, S. O., Jone, E. V., et al. (2015). Ligand de-
construction: Why some fragment binding positions are conserved and others are
not. Proc Natl Acad Sci 112(28), E2585–E2594.
Laghi, L., Orbetegli, O., Bianchi, P., Zerbi, A., Di Carlo, V., Boland, C. R., et al. (2002). Com-
mon occurrence of multiple K-RAS mutations in pancreatic cancers with associated
precursor lesions and in biliary cancers. Oncogene 21(27), 4301–4306.
Laheru, D., Shah, P., Rajeshkumar, N. V., McAllister, F., Taylor, G., & Goldsweig, H. (2012).
Integrated preclinical and clinical development of S-trans, trans-farnesylthiosalicylic
acid (FTS, salirasib) in pancreatic cancer. Invest New Drugs 30, 2391–2399.
Lai, C. S., Mas, R. H., Nair, N. K., Manson, S. M., & Navaratnam, V. (2009). Chemical constit-
uents and in vitro anticancer activity of Typhonium flagelliforme (Araceae).
J Ethnopharmacol 127(2), 486–494.
Ledford, H. (2015). Cancer: The Ras renaissance. Nature 520(7547), 278–280.
Lee, K. W., Kang, N. J., Heo, Y., Rogozin, E. A., Angelo, P., Hwang, M. K., et al. (2008). Raf and
MEK Protein kinases are direct molecular targets for the chemopreventive effect of
quercetin, a major flavonol in red wine. Cancer Res 68(3), 946–955.
Lim, K., & Counter, C. M. (2005). Reduction in the requirement of oncogenic Ras signaling
to activation of PI3K/AKT pathway during tumor maintenance. Cancer Cell 8,
381–392.
Lobell, R. B., Omer, C. A., Abrams, M. T., Bhimnathwala, H. G., Brucker, M. J., Buser, C. A.,
et al. (2001). Evaluation of farnesyl:protein transferase and geranylgeranyl:protein
transferase inhibitor combinations in preclinical models. Cancer Res 61, 8758–8768.
Lodish, H., Berk, A., Zipursky, S. L., Matsudaira, P., Baltimore, D., & Darnell, J. (2000).
Section 20.4, receptor tyrosine kinases and Ras. molecular cell biology (4th ed.). New
York: W. H. Freeman.
Mackenzie, G. G., Bartels, L. E., Xie, G., Papayannis, I., Alston, N., Alston, N., et al. (2013). A
novel Ras inhibitor (MDC-1016) reduces human pancreatic tumor growth in mice.
Neoplasia 15, 1184–1195.
Marom, M., Haklai, R., Ben-Baruch, G., Marciano, D., Egozi, Y., & Kloog, Y. (1995). Selective
inhibition of Ras-dependent cell growth by farnesylthiosalisylic acid. J Biol Chem
270(38), 22263–22270.
Mason, W. P., Belanger, K., Nicholas, G., Vallières, I., Mathieu, D., Desjardins, A., et al.
(2010). A phase II trial of TLN-4601 in patients with glioblastoma multiforme
(GBM) at first progression. J Clin Oncol 28, 2094.
Mason, W. P., Belanger, K., Nicholas, G., Vallières, I., Mathieu, D., & Kavan, P. (2012). A
phase II study of the Ras-MAPK signaling pathway inhibitor TLN-4601 in patients
with glioblastoma at first progression. J Neurooncol 107(2), 343–349.
Maurer, T., Garrenton, L. S., Oh, A., Pitts, K., Anderson, D. J., Skelton, N. J., et al. (2012).
Small-molecule ligands bind to a distinct pocket in RAS and inhibit SOS-mediated nu-
cleotide exchange activity. Proc Natl Acad Sci 109, 5299–5304.
Meder, D., & Simons, K. (2005). Ras on the roundabout. Science 307(5716), 1731–1733.
Moore, M. J., Goldstein, D., Hamm, J., Figer, A., Hecht, J. R., Gallinger, S., et al. (2007). Erlo-
tinib plus gemcitabine compared with gemcitabine alone in patients with advanced
pancreatic cancer: a phase III trial of the National Cancer Institute of Canada Clinical
Trials Group. J Clin Oncol 25(15), 1960–1966.
Müller, C., Frau, M. A. G., Ballinari, D., Colombo, S., Bitto, A., Martegani, E., et al. (2009). De-
sign, synthesis, and biological evaluation of levoglucosenone-derived Ras activation
inhibitors. ChemMedChem 4, 524–528.
Newman, D. J., & Cragg, G. M. (2012). Natural products as sources of new drugs over the
30 years from 1981 to 2010. J Nat Prod 75(3), 311–335.
Noonan, T., Brown, N., Dudycz, L., & Wright, G. (1991). Interaction of GTP derivatives with
cellular and oncogenic ras-p21 proteins. J Med Chem 34(4), 1302–1307.
Ostrem, J. M., Peters, U., Sos, M. L., Wells, J. A., & Shokat, K. M. (2013). K-Ras(G12C) inhibitors
allosterically control GTP affinity and effector interactions. Nature 503, 548–551.
Palmioli, A. (2009) Synthesis and biological characterization of new pharmacologically active
molecules derived from natural compounds (published doctoral thesis). University of
Milano-Bicocca, Milano, Italy.
Palmioli, A., Sacco, E., Abraham, S., Thomas, C. J., Di Domizio, A., De Gioia, L., et al. (2009).
First experimental identification of Ras-inhibitor binding interface using a water-
soluble Ras ligand. Bioorg Med Chem Lett 19, 4217–4222.
Peri, F., Airoldi, C., Colombo, S., Mari, S., Jiménez-Barbero, J., Martegani, E., et al. (2006).
Sugar-derived Ras inhibitors: Group epitope mapping by NMR spectroscopy and bio-
logical evaluation. Eur J Org Chem, 3707–3720.
Peri, F., Airoldi, C., Colombo, S., Martegani, E., van Neuren, S., Stein, M., et al. (2005). De-
sign, synthesis and biological evaluation of sugar-derived Ras inhibitors.
Chembiochem 6, 1839–1848.
Plowman, S. J., & Hancock, J. F. (2005). Ras signaling from plasma membrane and
endomembrane microdomains. Biochim Biophys Acta 1746, 274–283.
Prime, version 3.4, Schrödinger, LLC, New York, NY, 2013.
Prior, I. A., & Hancock, J. F. (2012). Ras trafficking, localization and compartmentalized sig-
nalling. Semin Cell Dev Biol 23(2), 145–153.
Prior, I. A., Lewis, P. D., & Mattos, C. (2012). A comprehensive survey of Ras mutations in
cancer. Cancer Res 72(10), 2457–2467.
Psahoulia, F. H., Moumtzi, S., Roberts, M. L., Sasazuki, T., Shirasawa, S., & Pintzas, A. (2007).
Quercetin mediates preferential degradation of oncogenic Ras and causes autophagy
in H-RAS-transformed human colon cells. Carcinogenesis 28(5), 1021–1031.
Ranelletti, F. O., Maggiano, N., Serra, F. G., Ricci, R., Larocca, L. M., Lanza, P., et al. (2000).
Quercetin inhibits p21-RAS expression in human colon cancer cell lines and in prima-
ry colorectal tumors. Int J Cancer 85(3), 438–445.
Rees, D. C., Congreve, M., Murray, C. W., & Carr, R. (2004). Fragment based lead discovery.
Nat Rev Drug Discov 3, 660–672.
Riely, G. J., Johnson, M. L., Medina, C., Rizvi, N. A., Miller, V. A., & Kris, M. G. (2011). A phase
II trial of salirasib in patients with lung adenocarcinomas with KRAS mutations.
J Thorac Oncol 6(8), 1435–1437.
Riely, G. J., Marks, J., & Pao, W. (2009). KRAS mutations in non-small cell lung cancer. Proc
Am Thorac Soc 6, 201–205.
Rocks, O., Peyker, A., Kahms, M., Verveer, P. J., Koerner, C., & Lumbierres, M. (2005). An ac-
ylation cycle regulates localization and activity of palmitoylated ras isoforms. Science
307(5716), 1746–1752.
Rodriguez-Viciana, P., Warne, P. H., Dhand, R., Vanhaesebroeck, B., Gout, I., Fry, M. J., et al.
(1994). Phosphatidylinositol-3-OH kinase as a direct target of Ras. Nature 370(6490),
527–532.
Sacco, E., Abraham, S. J., Palmioli, A., Damore, G., Bragna, A., Mazzoleni, E., et al. (2011).
Binding properties and biological characterization of new sugar-derived ligands.
Med Chem Commun. http://dx.doi.org/10.1039/c0md00264j.
Scheffzek, K., Ahmadian, M. R., Kabsch, W., Wiesmuller, L., Lautwen, A., Schmitz, F., et al.
(1997). The Ras-RasGAP complex: structural basis for GTPase activation and its loss
in oncogenic Ras mutants. Science 277(5324), 333–338.
Scherer, A., John, J., Linke, R., Goody, R. S., Wittinghofer, A., Pao, E. F., et al. (1989). Crystal-
lization and preliminary X-ray analysis of the human c-H-Ras-oncogene product p21
complexed with GTP analogues. J Mol Biol 206, 257–259.
Schmidt, A., & Hall, A. (2002). Guanine nucleotide exchange factors for Rho GTPases: turn-
ing on the switch. Genes Dev 16, 1587–1609.
Sever, R., & Brugge, J. S. (2015). Signal transduction in cancer. Cold Spring Harb Perspect
Med 5(4). http://dx.doi.org/10.1101/cshperspect.a006098.
Slusarz, A., Shenouda, N. S., Sakla, M. S., Drenkhahn, S. K., Narula, A. S., MacDonald, R. S.,
et al. (2010). Common botanical compounds inhibit the hedgehog signalling pathway
in prostate cancer. Cancer Res 70(8), 3382–3390.
Song, N. R., Chung, M., Kang, N. J., Seo, S. G., Jang, T. S., Lee, H. J., et al. (2013). Quercetin
suppresses invasion and migration of H-Ras-transformed MCF10A human epithelial
cells by inhibiting phosphatidylinositol 3-kinase. Food Chem 142, 66–71.
Spiegel, J., Cromm, P. M., Zimmermann, G., Grossmann, T. N., & Waldmann, H. (2014). Direct
targeting of rab-GTPase-effector interactions. Angew Chem Int Ed Engl 53, 2498–2503.
Spoerner, M., Graf, T., Konig, B., & Kalbitzer, H. R. (2005). A novel mechanism for the mod-
ulation of the ras-effector interaction by small molecules. Biochem Biophys Res
Commun 334, 709–713.
Sun, Q., Burke, J. P., Phan, J., Burns, M. C., Olejniczak, E. T., Waterson, A. G., et al. (2012).
Discovery of small molecules that bind to K-Ras and inhibit SOS-mediated activation.
Angew Chem Int Ed Engl 51, 6140–6143.
Taveras, A. G., Remiszewski, S. W., Doll, R. J., Cesarz, D., Huang, E. C., Kirschmeier, P., et al.
(1997). Ras oncoprotein inhibitors: the discovery of potent, ras nucleotide exchange
inhibitors and the structural determination of a drug-protein complex. Bioorg Med
Chem 5(1), 125–133.
Taylor, M. T., Lawson, K. R., Ignatenko, N. A., Marek, S. E., Stringer, D. E., Skovan, B. A., &
Gerner, E. W. (2000). Sulindac sulfone inhibits K-ras-dependent cyclooxygenase-2
expression in human colon cancer cells. Cancer Res 60(23), 6607–6610.
Thompson, H. J., Jiang, C., Lu, J., Mehta, R. G., Piazza, G. A., Paranka, N. S., et al. (1997).
Sulfone metabolite of sulindac inhibits mammary carcinogenesis. Cancer Res 57(2),
267–271.
 S.Y. Quah et al. / Pharmacology & Therapeutics 162 (2016) 35–57
Tomikawa, T., Shin-Ya, K., Furihata, K., Kinoshita, T., Miyajima, A., Seto, H., et al. (2000).
Rasfonin, a new apoptosis inducer in ras-dependent cells from Talaromyces sp.
J Antibiot 53, 848–850.
Valkov, E., Sharpe, T., Marsh, M., Greive, S., & Hyvönen, M. (2012). Targeting protein–
protein interactions and fragment-based drug discovery. Top Curr Chem 317,
145–179.
Vetter, I. R., & Wittinghofer, A. (2001). The guanine nucleotide-binding switch in three di-
mensions. Science 294, 1299–1304.
Vivanco, I., & Sawyers, C. L. (2002). The phosphatidylinositol 3-kinase AKT pathway in
human cancer. Nat Rev Cancer 2(7), 489–501.
Walker, E. H., Pacold, M. E., Perisic, O., Stephens, L., Hawkins, P. T., Wymann, M. P., et al.
(2000). Structural determinants of phosphoinositide 3-kinase inhibition by
wortmannin, LY294002, quercetin, myricetin, and staurosporine. Mol Cell 6, 909–919.
Wang, Y., Kaiser, C. E., Frett, B., & Li, H. Y. (2013). Targeting mutant KRAS for anticancer
therapeutics: A review of novel small molecule modulators. J Med Chem 56(13),
5219–5230.
Weibel, E. K., Hadvary, P., Hochuli, E., Kupfer, E., & Lengsfeld, H. (1987). Lipstatin, an inhib-
itor pancreatic lipase, produced by Streptomyces toxytricini. I. Producing organism,
fermentation, isolation and biological activity. J Antibiot 40(8), 1081–1085.
Weisz, B., Giehl, K., Gana-Weisz, M., Egozi, Y., Ben-Baruch, G., & Marciano, D. (1999). A
new functional Ras antagonist inhibits human pancreatic tumour growth in nude
mice. Oncogene 18(16), 2579–2588.
Wennerberg, K., Rossman, K. L., & Der, C. J. (2005). The Ras superfamily at a glance. J Cell
Sci 118(Pt 5), 843–846.
Whyte, D. B., Kirschmeier, P., Hockenberry, T. N., Nunez-Oliva, I., James, L., Catino, J. J., et al.
(1997). K-Ras and N-Ras are geranylgeranylated in cells treated with farnesyl protein
transferase inhibitors. J Biol Chem 272, 14459–14464.
57
Williams, C. S., Goldman, A. P., Sheng, H., Morrow, J. D., & DuBois, R. N. (1999). Sulindac
sulfide, but not sulindac sulfone, inhibits colorectal cancer growth. Neoplasia 1(2),
170–176.
Wright, L. P., Court, H., Mor, A., Ahearn, I. M., Casey, P. J., & Philips, M. R. (2009). Topology
of mammalian isoprenylcysteine carboxyl methyltransferase determined in live cells
with a fluorescent probe. Mol Cell Biol 29(7), 1826–1833.
Xavier, C. P. R., Lima, C. F., Preto, A., Seruca, R., Fernandes-Ferreira, M., & Pereira-Wilson, C.
(2009). Luteolin, quercetin and ursolic acid are potent inhibitors of proliferation and
inducers of apoptosis in both KRAS and BRAF mutated human colorectal cancer cells.
Cancer Lett 281, 162–170.
Xiao, Z., Li, L., Li, Y., Zhou, W., Cheng, J., Liu, F., et al. (2014). Rasfonin, a novel 2-pyrone de-
rivative, induces RAS-mutated Panc-1 pancreatic tumor cell death in nude mice. Cell
Death Dis 5. http://dx.doi.org/10.1038/cddis.2014.213.
Zartler, E. R., & Mo, H. (2007). Practical aspects of NMR-based fragment discovery. Curr
Top Med Chem 7, 1592–1599.
Zeidman, R., Jackson, C. S., & Magee, Al (2009). Protein acyl thiotransferases (review). Mol
Membr Biol 26(1), 32–41.
Zhang, F. L., & Casey, P. J. (1996). Protein prenylation: Molecular mechanisms and func-
tional consequences. Annu Rev Biochem 65, 241–269.
Zhu, T., Wang, L., Wang, W., Hu, Z., Yu, M., & Wang, K. (2014). Enhanced production of
lipstatin from Streptomyces toxytricini by optimizing fermentation conditions and
medium. J Gen Appl Microbiol 60(3), 106–111.