Page 32 - Read Online

Conroy et al. Cancer Drug Resist 2021;4:543-58 https://dx.doi.org/10.20517/cdr.2021.07 Page 9

dual inhibition of the PI3K and the ERK pathway would appear to be a reasonable therapeutic target. With
this in mind combination strategies have been explored in RAS mutant disease using both mTOR inhibitors
[64]
along with RAF/MEK inhibitors . Direct PI3K inhibitors have also been trialled with MEK inhibitors.
[65]
However, in clinical trials, this combination was both toxic and only minimally efficacious .

AKT, which plays a key role in the activation of mTOR following its interaction with one of the 3 main
[66]
PI3Ks, is amplified in the setting of RAS-mutant cancers . Unfortunately no AKT inhibitors are approved
for use in clinical practice. The evaluation of AKT inhibition with MEK inhibitors is under investigation,
but similar challenging side effect profiles have been seen to that of combined mTOR and MEK
inhibition .
[67]

Direct inhibitors of RAS
Attempts at direct RAS inhibition have been hampered by numerous challenges. Firstly, the activity of RAS
is tightly governed by GEFs and GAPs, which control the high-affinity, picomolar interactions between
GDP, GTP and RAS. Putative inhibitors struggle to overcome this affinity. Secondly, there are structural
challenges. An effective inhibitor would typically require a deep hydrophobic pocket for binding. These
pockets were long understood to be absent, although recent research demonstrates that they may arise
dynamically as RAS goes through the GDP/GTP cycle [68,69] . Finally, there are selectivity and toxicity
challenges. The “switch region”, which changes conformation upon GTP binding and recruits effector
proteins, is highly conserved across other G-proteins in the body. Therefore, any agents which target this
region in RAS, would be associated with risk of toxicity elsewhere.

A number of agents investigated were compounds that competed directly with GDP for the nucleotide
binding site of RAS and compounds that bind to RAS at the RAF binding site and inhibit RAS/RAF
[70]
complex formation . However, none were sufficiently potent to be considered for further investigation.
[71]
More recently, however, direct targeting of RAS has been achieved for the G12C K-RAS mutation. This has
been realised through the development of a small molecule inhibitor that binds covalently to the cysteine
residue that results from the specific G12C mutation and has been shown to inhibit oncogenic RAS [72,73] . The
inherently reactive nature of cysteine which is found at codon 12 of K-RAS(G12C) can be exploited for
covalent small molecule inhibitors and the idea of targeting cysteine is one that is commonly exploited in
[74]
drug discovery . Another important feature of targeting this cysteine is that wild-type K-RAS lacks the
cysteine in the active site unlike the mutant K-RAS(G12C).

[73]
Ostrem et al. initially identified this novel allosteric binding pocket behind switch II referred to as the
switch-II pocket, which led to the development of the first compounds to irreversibly target G12C. These
compounds bind to K-RAS(G12C) in the GDP-bound inactive state, blocking SOS-catalysed nucleotide
exchange and ultimately inhibiting K-RAS(G12C) association with RAF . This switch II pocket is present
[73]
in the GDP-bound inactive form of K-RAS only. Therefore, targeting of G12C needs to occur in the GDP-
bound state. K-RAS(G12C) in its steady state is in the active GTP-bound state, but the presence of a high
level of GTPase activity leaves it open to covalent attack . The identification of this pocket led to a search
[75]
for covalent inhibitors of K-RAS. The first was a molecule ARS-853 , and its development led to the proof
[76]
of concept that this specific isoform of K-RAS could be targeted using a covalent inhibitor.

A number of advances in the area ultimately led to the development of AMG510 (sotorasib) . This
[73]
particular G12C inhibitor has succeeded where others had not as its potency and selectivity was optimised
through an interaction with a previously unexploited groove His95 . Preclinical studies have demonstrated
[77]
responses and regression of K-RAS mutant tumours treated with AMG510. The results of the Phase I study