Nature - USA (2020-01-16)

424 | Nature | Vol 577 | 16 January 2020

Article

of KRAS(G12C) is sufficient for the divergent response. However,
drug-bound KRAS(G12C) cannot undergo nucleotide exchange to
the active state^3 –^5. We therefore considered how KRAS(G12C) might
be reactivated, when almost the entire initial pool is covalently bound
and inhibited by the drug.
The G12Ci treatment induced KRAS mRNA and KRAS protein expres-
sion (Fig. 4b, c, Extended Data Fig. 9a). This induction was heteroge-
neous across the population (Fig. 2a, b, Extended Data Fig. 6). It was
inversely proportional to KRAS–RAF–MEK–ERK signalling activity
(Extended Data Fig. 9a), and was most pronounced in clusters of quies-
cent cells with maximal inhibition of G12C output (Fig. 4d, e). Inhibiting
new KRAS synthesis with the transcription inhibitor actinomycin D
or KRAS-specific siRNAs prevented the KRAS–GTP rebound during
the G12Ci treatment (Fig. 4b, c, Extended Data Fig. 9b). Conversely,
doxycycline-induced KRAS(G12C) expression attenuated the effect
of drug treatment (Fig. 4f, Extended Data Fig. 9c).
To confirm that KRAS(G12C) is sufficient for the divergent response
to the G12Ci treatment, we used the KRASG12C siRNA to mimic the ini-
tial inhibitory phase, and doxycycline-induced expression of siRNA-
resistant KRASG12C (siRes-G12C) to mimic the adaptive phase triggered
by new KRAS(G12C). As evidenced in Fig. 4g and Extended Data Fig. 9d,
expression of siRes-G12C led to a bimodal distribution in cells trans-
fected with KRASG12C siRNA, with about 30% of cells escaping quiescence.
This phenocopies the effect of the G12Ci, even though no drug was
added in this experiment.
The data suggest that some newly synthesized KRAS(G12C) under-
goes nucleotide exchange to the active, drug-insensitive state before it
is bound and inhibited by the drug. Indeed, EGF stimulation attenuated
the inhibition of KRAS(G12C) when EGF was added before the G12Ci,
but not when it was added afterwards (Extended Data Fig. 9e). This sug-
gests that exposure to growth factors is the initial stimulus that affects
the inhibitory fate of cells with new KRAS(G12C). AURKA probably
operates later, to maintain active KRAS(G12C) and effector signalling.
Our study thus sheds light on why treatment with a KRAS(G12C)
inhibitor results predominantly in partial responses in patients with
lung cancer^30. We identify an adaptive fitness mechanism that allows
groups of cancer cells within a population to rapidly escape inhibition
(Extended Data Fig. 10, Supplementary Discussion). The synthesis of new
KRAS(G12C) and its distribution between the active or inactive states
modulates the divergent response. Drug-induced quiescent cells with-
out adequate expression of new KRAS(G12C) are eliminated from the

population by treatment. Cells with new KRAS(G12C)—which is rapidly

converted to the active, drug-insensitive state—escape inhibition and

resume proliferation. This occurs through signals that act upstream

of KRAS: receptor tyrosine kinases trigger nucleotide exchange, and

AURK signalling facilitates effector activation and cell-cycle progres-

sion. In cells in which these signals are not active (or in cells in which the

signals are pharmacologically suppressed), new KRAS(G12C) spends a

longer time in its inactive conformation, in which it can be bound and

inhibited by the drug. These cell-intrinsic events are sufficient for a rapid,

multifactorial and non-uniform adaptive process that limits the thera-

peutic potential of conformation-specific KRAS(G12C) inhibition. This

mechanism must be suppressed for complete and durable responses

in the clinic.

Online content

Any methods, additional references, Nature Research reporting sum-

maries, source data, extended data, supplementary information,

acknowledgements, peer review information; details of author con-

tributions and competing interests; and statements of data and code

availability are available at https://doi.org/10.1038/s41586-019-1884-x.

1. Pylayeva-Gupta, Y., Grabocka, E. & Bar-Sagi, D. RAS oncogenes: weaving a tumorigenic
   web. Nat. Rev. Cancer 11 , 761–774 (2011).


2. Li, S., Balmain, A. & Counter, C. M. A model for RAS mutation patterns in cancers: finding
   the sweet spot. Nat. Rev. Cancer 18 , 767–777 (2018).


3. Ostrem, J. M., Peters, U., Sos, M. L., Wells, J. A. & Shokat, K. M. K-Ras(G12C) inhibitors
   allosterically control GTP affinity and effector interactions. Nature 503 , 548–551 (2013).


4. Patricelli, M. P. et al. Selective inhibition of oncogenic KRAS output with small molecules
   targeting the inactive state. Cancer Discov. 6 , 316–329 (2016).


5. Lito, P., Solomon, M., Li, L. S., Hansen, R. & Rosen, N. Allele-specific inhibitors inactivate
   mutant KRAS G12C by a trapping mechanism. Science 351 , 604–608 (2016).


6. Simanshu, D. K., Nissley, D. V. & McCormick, F. RAS proteins and their regulators in human
   disease. Cell 170 , 17–33 (2017).


7. Hunter, J. C. et al. Biochemical and structural analysis of common cancer-associated
   KRAS mutations. Mol. Cancer Res. 13 , 1325–1335 (2015).


8. Zeng, M. et al. Potent and selective covalent quinazoline inhibitors of KRAS G12C. Cell
   Chem. Biol. 24 , 1005–1016 (2017).


9. Janes, M. R. et al. Targeting KRAS mutant cancers with a covalent G12C-specific inhibitor.
   Cell 172 , 578–589 (2018).


10. Corcoran, R. B. et al. EGFR-mediated re-activation of MAPK signaling contributes to
    insensitivity of BRAF mutant colorectal cancers to RAF inhibition with vemurafenib.
    Cancer Discov. 2 , 227–235 (2012).


11. Prahallad, A. et al. Unresponsiveness of colon cancer to BRAF(V600E) inhibition through
    feedback activation of EGFR. Nature 483 , 100–103 (2012).


12. Lito, P., Rosen, N. & Solit, D. B. Tumor adaptation and resistance to RAF inhibitors.
    Nat. Med. 19 , 1401–1409 (2013).

a b c

f

e

g

Control G12Ci, 2 h

Dox–KRASG12C

KRAS

ERK

pERK

Dox, 24 h

KRAS–GTP

d

76829103541

KRAS

mRNA,

z

G12C score

4

2

0

–2

0.20.40.60.8

All cells

Time (h) — Trend + 95% CI

0

1.0

0.5

8

Fold change ofKRAS

mRNA

0244872

G12Ci

G12Ci + AcD

AcD

4

Cells

2345

0

200

400

600

p27K–, log(AU)

• Dox+ Dox (siRes-G12C)

siG12C:

424 4872

KRAS β-Actin

h

AcD

G12Ci

G12Ci + AcD

KRAS–GTP

0 0424 4872 0424 4872

2345

p27K–, log(AU)

Cells^250

0

500

siNT

siG12C

G12Ci

siKRAS 100

50

0

NT

siG12CsiKRAS

KRAS

pERK/ERK

Intensity

(%)

G12C score

Cells from 24 h and 72 h

Mode

KRAS mRNA

Quiescent

0.6 0.2

Prolif.

–2 03

Fig. 4 | Newly synthesized KR AS(G12C) escapes trapping by the drug. a, Cells
expressing the quiescence biosensor (H358 p27K−, KRASG1 2 C+/−) were transfected
with KRAS-specific siRNAs targeting both wild-type and G12C alleles (siKR AS),
or only the G12C allele (siG12C), for 72 h and analysed by FACS. The effect of
a 72-h G12Ci treatment is shown. Inset, cell extracts were immunoblotted
and quantified to determine the intensity of KR AS expression and ERK
phosphorylation. b, Effect of the indicated treatments on KRAS mRNA.
AcD, actinomycin D. c, Inhibitor-treated cell extracts were analysed by
immunoblotting. d, e, Normalized KR AS expression across single cells as a

function of KR AS(G12C) output score (d) or in quiescent versus proliferating

(prolif.) cells (e). CI, confidence interval. f, H358 cells engineered to express

haemagglutinin (HA)-tagged KR AS(G12C) under a doxycycline (dox) -inducible

promoter were treated with the G12Ci in the presence of doxycycline

(0–2 μg ml−1). g, Cells expressing the quiescence biosensor, engineered to

stably express doxycycline-inducible siRes-G12C, were transfected with

KRASG12C siRNA, followed by doxycycline treatment (100 ng ml−1).

A representative of three independent experiments is shown in a–c, f, g.