Small molecule SOS1 agonists modulate MAPK and PI3K signaling

Subscriber access provided by MIDWESTERN UNIVERSITY

Letter

Small molecule SOS1 agonists modulate MAPK and PI3K signaling via independent cellular responses Denis T. Akan, Jennifer E. Howes, Jiqing Sai, Allison L. Arnold, Yugandhar Beesetty, Jason Phan, Edward T Olejniczak, Alex G. Waterson, and Stephen W. Fesik ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.8b00869 • Publication Date (Web): 08 Feb 2019 Downloaded from http://pubs.acs.org on February 9, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Small molecule SOS1 agonists modulate MAPK and PI3K signaling via independent cellular responses Denis T. Akan◊, Jennifer E. Howes◊, Jiqing Sai◊, Allison L. Arnold◊, Yugandhar Beesetty◊, Jason Phan◊, Edward T. Olejniczak◊, Alex G. Waterson#,¥ and Stephen W. Fesik◊, #, ¥ ◊Department

of Biochemistry, Vanderbilt University School of Medicine, 2215 Garland Avenue, 607 Light Hall, Nashville, Tennessee 37232-0146, USA #Department

of Pharmacology, Vanderbilt University School of Medicine, 2215 Garland Avenue, 607 Light Hall, Nashville, Tennessee 37232-0146, USA ¥Department

of Chemistry, Vanderbilt University, 2215 Garland Avenue, 607 Light Hall, Nashville, Tennessee 37232-0146, USA Corresponding Author: Stephen W. Fesik Email: [email protected] Tel.: +1 (615) 322-6303 Fax: +1 (615) 875-3236 Key Words: SOS1, RAS, agonist, small molecule, cancer Conflict of Interest Statement: These SOS1 agonist compounds are the subject of a collaboration between Vanderbilt University and Boehringer Ingelheim.

ACS Paragon Plus Environment

1

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 18

Abstract Activating mutations in RAS lead to oncogenesis by enhancing downstream signaling, such as through the MAPK and PI3K pathways. Therefore, therapeutically targeting RAS may perturb multiple signaling pathways simultaneously. One method for modulating RAS signaling is to target the activity of the guanine nucleotide exchange factor SOS1. Our laboratory has discovered compounds that bind to SOS1 and activate RAS. Interestingly, these SOS1 agonist compounds elicit biphasic modulation of ERK phosphorylation and simultaneous inhibition of AKT phosphorylation levels. Here we utilized multiple chemically distinct compounds to elucidate whether these effects on MAPK and PI3K signaling by SOS1 agonists were mechanistically linked. In addition, we used CRISPR/Cas9 gene-editing to generate clonally-derived SOS1 knockout cells and identified a potent SOS1 agonist that rapidly elicited on-target molecular effects at substantially lower concentrations than those causing off-target effects. Our findings will allow us to further define the on-target utility of SOS1 agonists.


2

Page 3 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Main Manuscript The three RAS protein isoforms are part of a family of membrane anchored, small GTPase proteins that participate in the transduction of extracellular stimuli to intracellular signaling. RAS proteins cycle from a GDP-bound ‘inactive’ state to a GTP-bound ‘active’ state.(1) In their GTP-bound state, RAS proteins activate multiple effectors implicated in the development and progression of cancer. The activation of RAS regulates a number of cellular processes including growth, survival and proliferation.(1–3) Indeed, activating mutations in RAS are responsible for approximately a third of all tumors, and mutations in RAS can also generate resistance to available cancer therapeutics.(4,5) Thus, RAS is an important target for anti-cancer pharmacological intervention. Although the RAS proteins can intrinsically regulate nucleotide exchange, guanine nucleotide exchange factor proteins (GEFs) such as Son of Sevenless homolog 1 (SOS1) catalyze the exchange between GDP to GTP on RAS at a higher rate and under finer temporal and spatial control.(6) Our laboratory has discovered multiple chemical series of small molecules that bind to SOS1 and activate nucleotide exchange on RAS.(7–10) Treatment with compounds from all three series led to the elevation of RAS-GTP levels in cancer cells and the biphasic modulation of downstream MAPK pathway phosphorylation, which is characterized by an increase in phospho-ERK1/2T202/Y204 levels at lower concentrations of compound and a decrease below baseline at higher treatment concentrations.(7–11) Using a tool compound from the indole series of SOS1 agonists, we previously discovered that this decrease in ERK phosphorylation was achieved via induction of a negative feedback loop that overrides compound-mediated activation of RAS-GTP.(12) In addition to the effects on RAS and downstream ERK signaling, treatment with this indole compound also led to the modulation of AKT phosphorylation, a key downstream effector of PI3K signaling.(11) In contrast to the biphasic pattern of ERK phosphorylation, however, AKT phosphorylation was inhibited in a dose-dependent manner.(11,12) The therapeutic promise of a small molecule that could achieve dual modulation of two oncogenic signaling pathways led us to address whether these signaling events were mechanistically related and achieved through a biological mechanism dependent on SOS1.


3


Page 4 of 18

Figure 1. Inhibition of phospho-AKTS473 by 1 is independent of changes in RAS-GTP and phospho-ERK signaling. (a) Schematic of potential mechanisms of compound-induced MAPK and PI3K pathway modulation including (i) cross-talk from MAPK pathway activation causing inhibition of PI3K signaling and (ii) relief of negative feedback from PI3K pathway on MAPK signaling.(13) (b) HeLa cells were treated with 50 μM compound 1 over a timecourse of up to 15 minutes. (c) SK-MEL-28 cells were treated with 12.5 μM compound 1 over a timecourse of up to 180 minutes. HeLa RASGTP samples were included as a positive control for RAS-GTP levels. (d) HeLa cells were treated with 0.5 μM pictilisib over a timecourse of up to 30 minutes. VC = vehicle control.

To this end, we first assessed whether compound 1 directly inhibited any of the four catalytic subunits of Class I PI3K (Supplementary Table S1). At the concentrations tested, we did not observe any inhibition of p110α or γ function, and compared to the effects of a PI3K inhibitor PI-103, compound 1 also had very minimal inhibitory effects on p110β and δ subunit activity.


4

Page 5 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The MAPK and PI3K pathways can regulate each other via cross-talk and feedback mechanisms.(13) As compound 1 only minimally inhibited PI3K directly (Supplementary Table S1), we hypothesized that SOS1 agonists may elicit downstream cross-talk or feedback between the PI3K and MAPK pathways to induce the simultaneous biphasic modulation of phospho-ERK and dose-dependent inhibition of AKT phosphorylation (Fig. 1a). Using the RAS activator indole compound 1 previously reported in our mechanistic studies,(12) we initially sought to determine whether inhibition of phospho-AKTS473 was dependent on RAS-GTP levels and downstream phospho-ERK signaling events. To accomplish this, we assessed the effects of compound 1 on the levels of RASGTP, phospho-ERK1/2T202/Y204, and phospho-AKTS473 over a timecourse of 15 minutes, using an activating concentration of compound 1. Consistent with our previous reports,(12) increases were observed in the levels of both RAS-GTP and correspondingly in phospho-ERK after 4–5 minutes of compound treatment. However, rapid inhibition of phospho-AKTS473 was observed after only 1–2 minutes of treatment with 50 μM compound 1, preceding the modulation of both RAS-GTP and phospho-ERK, suggesting that inhibition of AKTS473 phosphorylation was independent of compound-mediated RAS and ERK activation (Fig. 1b). To test this further, we assessed the effects of compound 1 over time in BRAFV600E mutant SK-MEL-28 cells. Consistent with our previous findings in the context of cells harboring a BRAF V600E activating mutation,(11) compound-mediated modulation of phospho-ERK levels was not observed (Fig. 1c). Furthermore, as anticipated, compound treatment also failed to elicit a detectable increase in RASGTP levels, likely due to constitutive negative feedback from phospho-ERK.(14) However, compound 1-mediated inhibition of phospho-AKTS473 was still observed, independent of changes in phospho-ERK and RAS-GTP levels (Fig. 1c). This provided additional evidence that inhibition of AKT phosphorylation by compound 1 was not due to cross-talk or feedback from modulation of MAPK signaling. We next determined whether the modulation of RAS-GTP and ERK phosphorylation were caused by inhibition of AKTS473 phosphorylation. Using the PI3K inhibitor pictilisib (GDC-0941),(15) we observed rapid and robust inhibition of phospho-AKTS473 in HeLa cells (Fig. 1d). However, pictilisib treatment did not cause the biphasic modulation of phospho-ERK levels or rapid activation of RAS-GTP levels that were observed after compound 1 treatment (Fig. 1d). This suggested that the MAPK pathway modulation observed in response to SOS1 agonist compounds was unlikely due to feedback from the effect of phospho-AKTS473 inhibition. Taken together, these data suggest that the compound-mediated inhibition of phospho-AKTS473 and induction of RAS-GTP and


5


Page 6 of 18

ERK modulation are not due to feedback or crosstalk between the two signaling pathways. Although compounds do not directly inhibit the activity of PI3K (Supplementary Table S1), due to the rapidity of compound-mediated phospho-AKTS473 inhibition we hypothesized that compound treatment led to reduced PI3K activity, possibly by causing an increase in the stoichiometry of the RAS::RAF association and therefore sequestering RAS away from PI3K. Furthermore, positive feedback on the allosteric site of SOS1(16,17) by compound-mediated increases in RAS-GTP levels may result in the sequestration of SOS and RAS specifically at the intracellular membranes, thus decreasing phospho-AKT levels that are generally only modulated by PIP3 at the plasma membrane. This hypothesis was supported by live confocal microscopy experiments showing that GFP-tagged SOS1WT rapidly clusters in response to compound 1 treatment, but not after DMSO treatment alone (Supplementary Movies S1 and S2). These SOS1 clusters appeared evenly throughout the cell with some exclusion from the nuclear region, and therefore suggest that SOS1 localization is not restricted to one specific membrane after compound treatment. While signaling via RAF-MEK-ERK is independent of the site of SOS-mediated RAS activation, AKT signaling is dependent on PIP3-mediated SOS localization to the plasma membrane, thus the compound-mediated effects on SOS1 localization that we observe may contribute to the AKT inhibition phenotype. The positive feedback-loop invoked by activated RAS signaling may further contribute to the even cellular distribution of SOS1 and the temporal maintenance of phospho-AKT inhibition that we observe after compound treatment. Furthermore, when a CAAX-tagged SOS1 construct was used, that localizes to both the endomembranes and the external plasma membrane,(12,18) we observed that upon compound treatment, phospho-AKT levels were inhibited (Supplementary Fig. S1). Taken together, these data imply that compound-mediated phospho-AKTS473 inhibition may be due to membranespecific changes in SOS1 localization and activation. Two activating hotspot mutations in the p110α catalytic subunit of PI3K, encoded by the PIK3CA gene, have been reported that have different dependencies on RAS-GTP. The p110αE545K mutant is dependent on RAS-GTP for gain-of-function; whereas, p110αH1047R is active independently of RAS-GTP binding.(19,20) Therefore, we hypothesized that cell lines harboring the p110αH1047R mutant may be insensitive to compound-induced phospho-AKTS473 inhibition. To test this, we treated a selection of cell lines expressing wild-type p110α, p110αE545K, or p110αH1047R with up to 50 μM of SOS1 agonist compound 1 for 30 minutes and assessed phospho-AKTS473 levels (Fig. 2b). Interestingly, we observed inhibition


6

Page 7 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of phospho-AKTS473 levels in response to compound 1 treatment across all of the cell lines tested that was independent of p110α mutation status. In addition, compound 1 treatment caused inhibition of phospho-AKTS473 in an isogenic NIH/3T3 cell line model transiently overexpressing either HA-tagged wild-type p110α, p110αE545K, or p110αH1047R, despite elevated basal levels of phosphoAKTS473 (Supplementary Fig. S2). Collectively, these data show that PI3K hyperactivation does not prevent the inhibitory effects of compound treatment on phospho-AKT473 signaling.

Figure 2. Compound 1 elicits inhibition of phospho-AKTS473 independent of p110α mutation status. Cell lines expressing either wild-type p110α, p110αE545K, or p110αH1047R were treated with up to 50 μM compound 1 for 30 minutes. VC = vehicle control.

In principle, if multiple different compound chemotypes elicit the same signaling phenotype, it is likely that these effects on signaling are caused through the same on-target mechanism of action.(21) Compounds from three chemically distinct series of SOS1 agonists discovered in our laboratory all bind to SOS1 in the same pocket, and all cause activation of SOS1-mediated nucleotide exchange of RAS in vitro and in cancer cells.(7–10) Furthermore, each of these series elicit a biphasic modulation of phospho-ERK1/2T202/Y204 levels in a concentration dependent manner. Based on these observations, we tested whether compounds from each series could inhibit AKT phosphorylation in a similar fashion to the indole compound 1.


7


Page 8 of 18

To assess the levels of phospho-AKTS473 and total AKT in both HeLa and NCI-H727 cells, we optimized an In Cell Western (ICW) experiment, and generated compound dose response curves from which to determine IC50 values for AKTS473 phosphorylation (Supplementary Fig. S3). Compounds tested from each chemical series caused inhibition of phospho-AKTS473 levels in both HeLa and NCI-H727 cells (Table 1). This suggested that the inhibition of phospho-AKTS473 was not chemotype or cell line specific. In addition, the potencies of each individual compound on the inhibition of AKT phosphorylation were generally similar between cell lines and chemical series. These common observations between series and signaling phenotypes added weight to the hypothesis that therapeutically activating RAS may have simultaneous modulatory effects on multiple signaling pathways. However, despite similar potencies on phospho-AKTS473 inhibition, the three chemical series exhibited substantially different binding affinities to SOS1 measured by a Fluorescence Polarization Assay (FPA; Table 1). In particular, benzimidazole compounds possess Kd values below 200 nM, whereas Kd values for compounds from the indole and quinazoline series were greater than 1 μM. We expected that if phospho-AKTS473 inhibition was due to on-target engagement of SOS1 by our compounds that this would be reflected in improved binding potency. However, no correlation between Kd and phospho-AKTS473 IC50 was observed (Pearson correlation R squared ICW vs. FPA HeLa = 0.022 and NCI-H727 = 0.055; Supplementary Fig. S4). Due to this discrepancy, it was important to confirm that dual modulation of both ERK and AKT signaling by the SOS1 agonists was elicited via an on-target biological mechanism dependent on SOS1.


8

Page 9 of 18

Table 1. In Cell Western derived phospho-AKTS473 IC50 in HeLa and NCI-H727 cells and FPA Kd values for compounds from the indole, quinazoline and benzimidazole SOS1 agonist series.

Quinazolines(8)

Indoles(7,12)

Compounda

Benzimidazoles(10)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


pAKTS473 IC50 (μM)b

FPA Kd (μM)c

HeLa

NCI-H727

15e

12.3 ± 4.8

> 25.0

13.27 ± 1.909

15c

4.65 ± 2.5

> 25.0

322.2 ± 82.16

15f

4.62 ± 2.2

23.0 ± 0.1

23.70 ± 4.441

18

3.69 ± 1.7

14.3 ± 5.3

4.327 ± 0.106

19

2.83 ± 1.0

14.3 ± 5.0

2.663 ± 0.028

17

2.64 ± 0.9

10.8 ± 4.4

6.901 ± 1.088

16

2.42 ± 0.7

10.2 ± 1.5

22.94 ± 6.772

1

1.68 ± 0.6

10.2 ± 4.2

21.90 ± 4.307

12h

1.43 ± 0.4

6.86 ± 1.9

6.886 ± 0.464

32

14.9 ± 7.0

10.3 ± 5.6

3.122 ± 0.723

30

13.5 ± 5.4

10.0 ± 5.0

11.77 ± 1.824

35

12.7 ± 3.8

20.0 ± 7.4

1.981 ± 0.446

34

12.1 ± 6.4

11.9 ± 6.3

1.502 ± 0.169

33

8.54 ± 4.5

9.62 ± 0.9

4.467 ± 0.292

22

6.09 ± 1.6

7.69 ± 2.0

61.71 ± 35.04

59

> 25.0

> 25.0

0.012 ± 0.002

38

> 25.0

> 25.0

0.124 ± 0.034

65

19.8 ± 12

> 25.0

0.009 ± 0.002

63

17.9 ± 8.6

> 25.0

0.036 ± 0.006

61

14.1 ± 7.8

> 25.0

0.175 ± 0.003

60

8.39 ± 4.7

> 25.0

0.016 ± 0.004

56

7.48 ± 3.9

> 25.0

0.102 ± 0.011

55

7.48 ± 3.6

22.5 ± 9.5

0.113 ± 0.018

62

5.78 ± 1.2

12.8 ± 0.3

0.012 ± 0.001

58

4.42 ± 2.2

19.7 ± 10

0.093 ± 0.036

47

3.34 ± 0.6

14.5 ± 4.9

0.143 ± 0.009

42

2.35 ± 1.1

7.34 ± 1.5

0.140 ± 0.013

64

1.81 ± 0.5

10.4 ± 3.6

0.044 ± 0.007

Compound numbers are taken from previous publications.(7,8,10,12) and structures are provided in Supplementary Table S2. b Each value represents the mean ± S.D. of three independent replicates. c Each value represents the mean ± S.D. of two independent replicates. a


9


H N

N Cl

μM

28

RAS-GTP 28

RAS

49

NH

Cl

12h

HeLa kDa

SOS1sgB6

NH2

O N H

EGF VC 3.1 6.3 12.5 25 50 VC 3.1 6.3 12.5 25 50

kDa

H N

N Cl

Pull-down

1

HeLa

Pull-down

b

NH

Input RAF1-RBD

N H

Input RAF1-RBD

NH2

O

SOS1sgB6

EGF VC 3.1 6.3 12.5 25 50 VC 3.1 6.3 12.5 25 50

a

μM

28

RAS-GTP 28

RAS 49

pERK1/2T202/Y204

38

pERK1/2T202/Y204

38

49

49

Total ERK

38

Total ERK

38

62

62

pAKTS473 62

pAKTS473

62

Total AKT

Total AKT

SOS1

98

SOS1 98

c

d

Me N N

NH

N N

Cl

N H

F

μM

28

RAS-GTP 28

RAS 49 38

62

62

98

SOS1sgB6

μM

28

RAS-GTP 28

RAS

49

pERK1/2T202/Y204

49 38

Me

64

EGF VC 0.001 0.01 0.1 1 10 VC 0.001 0.01 0.1 1 10

HeLa kDa

SOS1sgB6

Pull-down

34

EGF VC 3.1 6.3 12.5 25 50 VC 3.1 6.3 12.5 25 50

kDa

Pull-down

HeLa

N

Me

Input RAF1-RBD

N H

NH

N

Cl

F

N

Me

Input RAF1-RBD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 18

38

pERK1/2T202/Y204

49

Total ERK

38

Total ERK

62

pAKTS473

pAKTS473 62

Total AKT SOS1

Total AKT

98

SOS1

Figure 3. Compound-mediated inhibition of phospho-AKTS473 is achieved independent of SOS1 protein expression. Representative best-in-class compounds from the indole, quinazoline and benzimidazole chemical series all elicit dose-dependent inhibition of AKTS473 phosphorylation independent of SOS1 protein expression levels. RAS-GTP, phospho-ERK1/2T202/Y204 and phosphoAKTS473 levels were assessed in parental HeLa and SOS1-knockout HeLa-SOS1sgB6 cell lines in response to 30 minute treatment with (a) indole compound 1 (b) indole 12h (c) quinazoline 34, and (d) 10 minute treatment with benzimidazole 64. VC = vehicle control.


10

Page 11 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


To accomplish this, we used a CRISPR/Cas9 gene editing approach. CRISPR/Cas9 has proven useful for target selectivity studies in cancer cells,(22,23) therefore, to determine if modulation of ERK and AKT signaling were achieved through an on-target SOS1-dependent mechanism we generated clonal SOS1-knockout HeLa cell lines in which to assess the compound-mediated effects on RAS-GTP, phospho-ERK and phospho-AKT levels. We chose indole 12h, quinazoline 34 and benzimidazole 64 as representative best-in-class compounds from each series as they bound to SOS1 with high affinity and elicited potent modulation of both phospho-ERK and phospho-AKT signaling (Table 1).(7,8,10) For comparison and consistency, we also included the indole compound 1, which binds with lower affinity to SOS1 (Table 1).(11,12) Importantly, where the anticipated activation of RAS-GTP and associated modulation of phospho-ERK1/2T202/Y204 signaling was observed in parental HeLa cells treated with each compound, in two distinct clonally-derived SOS1 knockout HeLa cells, activation of RAS-GTP levels and phospho-ERK1/2T202/Y204 modulation was not observed to the same degree (Fig. 3 and Supplementary Fig. S5). This suggested that compound-mediated modulation of both RAS-GTP and phospho-ERK levels was achieved through an on-target SOS1 dependent mechanism. The benzimidazole compound 64 exhibits an approximately 100-fold higher binding affinity for SOS1 than the compounds tested from the other two chemical series (Table 1). Consistent with this, benzimidazole compound 64 elicited rapid, on-target activation of RAS-GTP and phospho-ERK1/2 at sub-micromolar concentrations, and represents our most potent SOS1 agonist in cells discovered to date (Fig. 3d). A small induction of RAS-GTP was observed after compound 1 treatment in the SOS1 knockout cells across the same treatment concentrations as the parental cells. In contrast, induction of RAS-GTP in the SOS1 knockout cells was only observed with benzimidazole compound 64 at the highest concentrations tested. This concentration was above the relevant dose required to activate RAS-GTP and ERK phosphorylation in a SOS1-dependent and dose-dependent manner in parental HeLa cells. These data show a clear separation between the on- and offtarget effects of SOS1 agonist treatment with the more potent benzimidazole compound 64. In addition, we have confirmed that compounds from each series all modulate RAS and phospho-ERK signaling through the same molecular mechanism, mediated through enhanced phosphorylation of SOS1 at Ser-Pro-Pro residues by phosphoERK1/2T202/Y204 (Supplementary Fig. S6).(12) We hypothesize that a potential


11


Page 12 of 18

mechanism for compound-mediated RAS-GTP activation is through allosteric stabilization of the RAS:SOS1 association, and we intend to assess this in the future. In contrast, compounds from all three chemical series consistently elicited inhibition of phospho-AKTS473 levels in the parental HeLa cell line and in both clonally-derived SOS1 knockout cell lines tested, indicating that inhibition of phospho-AKT signaling was achieved in a SOS1 independent manner (Fig. 3 and Supplementary Fig. S5). Indeed, cellular treatment with high micro-molar concentrations of compound could be responsible for the off-target inhibition of phospho-AKT levels through a pleiotropic off-target stress response,(21) however it was striking to us that compounds from three chemically distinct series could all coincidentally elicit the same off-target effects. Therefore, in addition to targeting SOS1, these compounds may modulate the activity of an additional unknown target that modulates PI3K pathway signaling to elicit the observed inhibition of phospho-AKTS473 levels. The alternative SOS isoform, SOS2, is highly homologous to SOS1 (69% similarity) and shares a degree of functional redundancy,(24) and was therefore an obvious candidate target protein. Therefore, we used NMR to test whether representative compounds from each chemical series also bound to SOS2. As anticipated, compounds from all series caused multiple shifts in the SOS1 NMR spectra, indicative of compound binding to SOS1 (Supplementary Fig. S7). However, no distinct chemical shift changes were observed in the SOS2 NMR spectra upon the addition of compounds. These data suggest that compounds from the three chemical series do not bind to SOS2, even at concentrations much higher than those used in cell-based studies. Therefore, SOS1 agonists are unlikely to cause SOS2 activation. Lack of detectable binding of compounds to SOS2 makes it unlikely that SOS2 is the target protein responsible for compoundmediated AKT signaling effects across each of the chemical series. We hypothesized that an alternative guanine nucleotide exchange factor may be responsible for the observed off-target effects. To test this, we conducted a small siRNA screen of 17 GEFs with homology to SOS1 and, after GEF knockdown in HeLa cells, we assessed the effects of compound on phospho-AKTS473 levels in response to compound 1 treatment at two different concentrations (Supplementary Fig. S8). We predicted that if an alternative GEF was responsible for modulation of phospho-AKT signaling, then we would observe a preservation of phospho-AKT levels after siRNA-mediated GEF knockdown. However, no significant reversal of the phospho-AKT inhibition phenotype was observed consistently with three or


12

Page 13 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


more of the four oligonucleotides targeting the expression of any GEF tested, and at different concentrations of compound treatment. Overall, these data imply that the inhibition of phospho-AKT is not due to the compound-mediated modulation of an alternative GEF with homology to SOS1. Therefore, compound-mediated inhibition of phospho-AKT is likely due to a more distinct protein or set of proteins, or even a result of pleiotropic off-target stress at high concentrations.(21) Interestingly, using lower concentrations (0.01–1 μM) of the benzimidazole 64, we observed an on-target induction of RAS-GTP and phospho-ERK signaling, independent of any inhibitory effect on phospho-AKTS473 levels (Fig. 3d and Supplementary Fig. S5d). However, minimal separation of the effects on phosphoERK and phospho-AKT signaling were observed with the indole or quinazoline compounds tested (Fig. 3a–c). These series-specific data are important as they highlight the necessity for high-affinity SOS1 ligands to elicit potent on-target cellular effects, and support the use of the SOS1 FPA as the primary assay used to drive our SAR campaign in the future.(10) We anticipate that the discovery of more potent benzimidazole compounds with enhanced binding affinity will allow us to fully parse the on- and off-target effects of SOS1 agonists in cancer cells. This is important because we believe that a validated SOS1 agonist that invokes phospho-ERK hyperactivation in the absence of off-target phospho-AKT inhibition may provide therapeutic potential if utilized to provoke synthetic lethal events in the presence of another activating oncogene such as mutant EGFR.(25) In conclusion, our laboratory has identified three independent series of SOS1 agonists that elevate RAS-GTP levels and modulate ERK phosphorylation in cancer cells by an on-target, SOS1-dependent mechanism. We discovered that compound-mediated inhibition of phospho-AKTS473 is independent of SOS1. Importantly, we show that with marked enhancements in binding affinity, a compound from the lead benzimidazole series potently modulates RAS-GTP and phospho-ERK signaling at concentrations substantially lower than those causing inhibition of AKT phosphorylation. These findings will allow us to progress in our drug-discovery efforts and to precisely address the on-target anti-tumoral effects of a potent SOS1 agonist. Materials and Methods See Supporting Information. Associated Content Supporting Information:


13


Page 14 of 18

The supporting information associated with this manuscript is available free of charge via the ACS Publications website. Files include: Materials and Methods; Supplementary Tables S1–S4 showing compound structures, sgRNA and PCR oligonucleotide sequences, and PI3K inhibition data; Supplementary Figures S1–S8 include overexpression of CAAX-tagged SOS1 and mutant PI3K isoforms, In Cell Western optimization and correlations, molecular signaling in SOS1sgE4 clonally-derived cells, pSPP SOS1 immunoprecipitation, SOS1 and SOS2 NMR spectra, and siRNA GEF screening data; Supplementary Movies S1–S2 showing live confocal microscopy of GFP-SOS1WT; and Supplementary References. Acknowledgements This work was supported by a Research Investigator Award from the Lustgarten Foundation awarded to S.W. Fesik in 2015, and a National Cancer Institute SPORE Grant in GI Cancer (5P50A095103-09) to R. J. Coffey. We thank our colleagues in the Fesik Laboratory (Vanderbilt University) for helpful discussions. We acknowledge J. Abbott, T. Hodges, A. Little, and P. Patel (Vanderbilt University, Fesik Lab) for synthesizing the SOS1 agonist compounds. We acknowledge The Vanderbilt University Biomolecular NMR facility for instrumentation. This facility receives support from an NIH SIG Grant (1S-10RR025677-01) and Vanderbilt University matching funds. Confocal imaging experiments and data analysis were performed in part through use of the Vanderbilt Cell Imaging Shared Resource (supported by NIH grants CA68485, DK20593, DK58404, DK59637, and EY08126.) The Zeiss LSM880 confocal microscope was obtained through an NIH S10 grant (1S-10OD021630-01). References [1] Pylayeva-Gupta, Y., Grabocka, E., and Bar-Sagi, D. (2011) RAS oncogenes: weaving a tumorigenic web, Nat. Rev. Cancer 11, 761–774. [2] Papke, B., and Der, C. J. (2017) Drugging RAS: Know the enemy, Science 355, 1158–1163. [3] Castellano, E., and Downward, J. (2011) RAS Interaction with PI3K: More Than Just Another Effector Pathway, Genes Cancer 2, 261–274. [4] Cox, A. D., Fesik, S. W., Kimmelman, A. C., Luo, J., and Der, C. J. (2014) Drugging the undruggable RAS: Mission possible?, Nat. Rev. Drug Discov. 13, 828–851. [5] Prior, I. A., Lewis, P. D., and Mattos, C. (2012) A comprehensive survey of Ras mutations in cancer, Cancer Res. 72, 2457–2467. [6] Boriack-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D., and Kuriyan, J. (1998) The structural basis of the activation of Ras by Sos, Nature 394, 337–343.


14

Page 15 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


[7] Abbott, J. R., Hodges, T. R., Daniels, R. N., Patel, P. A., Kennedy, J. P., Howes, J. E., Akan, D. T., Burns, M. C., Sai, J., Sobolik, T., Beesetty, Y., Lee, T., Rossanese, O. W., Phan, J., Waterson, A. G., and Fesik, S. W. (2018) Discovery of Aminopiperidine Indoles That Activate the Guanine Nucleotide Exchange Factor SOS1 and Modulate RAS Signaling, J. Med. Chem. 61, 6002–6017. [8] Abbott, J. R., Patel, P. A., Howes, J. E., Akan, D. T., Kennedy, J. P., Burns, M. C., Browning, C., Sun, Q., Rossanese, O. W., Phan, J., Waterson, A. G., and Fesik, S. W. (2018) Discovery of Quinazolines That Activate SOS1Mediated Nucleotide Exchange on RAS, ACS Med. Chem. Lett. 9, 941– 946. [9] Burns, M. C., Howes, J. E., Sun, Q., Little, A. J., Camper, D. V., Abbott, J. R., Phan, J., Lee, T., Waterson, A. G., Rossanese, O. W., and Fesik, S. W. (2018) High-throughput screening identifies small molecules that bind to the RAS:SOS:RAS complex and perturb RAS signaling, Anal. Biochem. 548, 44–52. [10] Hodges, T. R., Abbott, J. R., Little, A. J., Sarkar, D., Salovich, J., Howes, J. E., Akan, D. T., Sai, J., Arnold, A., Browning, C., Burns, M. C., Sobolik, T., Sun, Q., Beesetty, Y., Coker, J., Scharn, D., Stadtmueller, H., Rossanese, O. W., Phan, J., Waterson, A. G., McConnell, D., and Fesik, S. W. (2018) Discovery and Structure-Based Optimization of Benzimidazole-Derived Activators of SOS1-Mediated Nucleotide Exchange on RAS. J. Med. Chem. 61, 8875–8894. [11] Burns, M. C., Sun, Q., Daniels, R. N., Camper, D., Kennedy, J. P., Phan, J., Olejniczak, E. T., Lee, T., Waterson, A. G., Rossanese, O. W., and Fesik, S. W. (2014) Approach for targeting Ras with small molecules that activate SOS-mediated nucleotide exchange, Proc. Natl. Acad. Sci. U. S. A. 111, 3401–3406. [12] Howes, J. E., Akan, D. T., Burns, M. C., Rossanese, O. W., Waterson, A. G., and Fesik, S. W. (2018) Small Molecule-Mediated Activation of RAS Elicits Biphasic Modulation of Phospho-ERK Levels that Are Regulated through Negative Feedback on SOS1, Mol. Cancer Ther. 17, 1051–1060. [13] Mendoza, M. C., Er, E. E., and Blenis, J. (2011) The Ras-ERK and PI3KmTOR pathways: cross-talk and compensation, Trends Biochem. Sci. 36, 320–328. [14] Yao, Z., Torres, N. M., Tao, A., Gao, Y., Luo, L., Li, Q., de Stanchina, E., Abdel-Wahab, O., Solit, D. B., Poulikakos, P., and Rosen, N. (2015) BRAF Mutants Evade ERK-Dependent Feedback by Different Mechanisms that Determine Their Sensitivity to Pharmacologic Inhibition, Cancer Cell 14, 370–383. [15] Folkes, A. J., Ahmadi, K., Alderton, W. K., Alix, S., Baker, S. J., Box, G., Chuckowree, I. S., Clarke, P. A., Depledge, P., Eccles, S. A., Friedman, L. S., Hayes, A., Hancox, T. C., Kugendradas, A., Lensun, L., Moore, P., Olivero, A. G., Pang, J., Patel, S., Pergl-Wilson, G. H., Raynaud, F. I., Robson, A., Saghir, N., Salphati, L., Sohal, S., Ultsch, M. H., Valenti, M., Wallweber, H. J., Wan, N. C., Wiesmann, C., Workman, P., Zhyvoloup, A.,


15


Page 16 of 18

Zvelebil, M. J., and Shuttleworth, S. J. (2008) The identification of 2-(1Hindazol-4-yl)-6-(4-methanesulfonyl-piperazin-1-ylmethyl)-4-morpholin-4-yl-t hieno[3,2-d]pyrimidine (GDC-0941) as a potent, selective, orally bioavailable inhibitor of class I PI3 kinase for the treatment of cancer, J. Med. Chem. 51, 5522–5532. [16] Margarit, S.M., Sondermann, H., Hall, B.E., Nagar, B., Hoelz, A., Pirruccello, M., Bar-Sagi, D., and Kuriyan, J., (2003) Structural evidence for feedback activation by Ras.GTP of the Ras-specific nucleotide exchange factor SOS, Cell 112, 685–695. [17] Jeng, H.H., Taylor, L., and Bar-Sagi, D., (2012) Sos-mediated cross-activation of wild-type Ras by oncogenic Ras is essential for tumorigenesis. Nature Communications 3: 1168, DOI:10.1038/ncomms2173. [18] Choy, E., Chiu, V. K., Silletti, J., Feoktistov, M., Morimoto, T., Michaelson D., Ivanov, I. E., and Philips, M.R. (1999) Endomembrane trafficking of ras: the CAAX motif targets proteins to the ER and Golgi, Cell 98, 69–80. [19] Zhao, L., and Vogt, P. K. (2008) Helical domain and kinase domain mutations in p110alpha of phosphatidylinositol 3-kinase induce gain of function by different mechanisms, Proc. Natl. Acad. Sci. U. S. A. 105, 2652–2657. [20] Zhao, L., and Vogt, P. K. (2010) Hot-spot mutations in p110alpha of phosphatidylinositol 3-kinase (PI3K): differential interactions with the regulatory subunit p85 and with RAS, Cell Cycle 9, 596–600. [21] Blagg, J., and Workman, P. (2017) Choose and Use Your Chemical Probe Wisely to Explore Cancer Biology, Cancer Cell 32, 9–25. [22] Neggers, J. E., Vercruysse, T., Jacquemyn, M., Vanstreels, E., Baloglu, E., Shacham, S., Crochiere, M., Landesman, Y., and Daelemans, D. (2015) Identifying drug-target selectivity of small-molecule CRM1/XPO1 inhibitors by CRISPR/Cas9 genome editing, Chem. Biol. 22, 107–116. [23] Lin, A., Giuliano, C. J., Sayles, N. M., and Sheltzer, J. M. (2017) CRISPR/Cas9 mutagenesis invalidates a putative cancer dependency targeted in on-going clinical trials, eLife 6, DOI: 10.7554/eLife.24179. [24] Rojas, J. M., Oliva, J. L., and Santos, E. (2011) Mammalian son of sevenless Guanine nucleotide exchange factors: old concepts and new perspectives, Genes Cancer 2, 298–305. [25] Unni, A. M., Lockwood, W. W., Zejnullahu, K., Lee-Lin, S. Q., and Varmus, H. (2015) Evidence that synthetic lethality underlies the mutual exclusivity of oncogenic KRAS and EGFR mutations in lung adenocarcinoma, eLife 4, DOI: 10.7554/eLife.06907.


16

Page 17 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


For Table of Contents Only


17


Graphic for Table of Content 79x39mm (300 x 300 DPI)


Page 18 of 18

Small molecule SOS1 agonists modulate MAPK and PI3K signaling

Recommend Documents