Direct Attack on RAS: Intramolecular Communication and Mutation-Specific Effects.

CCR FOCUS Direct Attack on RAS: Intramolecular Communication and Mutation-Specific Effects Kendra Marcus and Carla Mattos

Abstract The crystal structure of RAS was first solved 25 years ago. In spite of tremendous and sustained efforts, there are still no drugs in the clinic that directly target this major driver of human cancers. Recent success in the discovery of compounds that bind RAS and inhibit signaling has fueled renewed enthusiasm, and in-depth understanding of the structure and function of RAS has opened new avenues for direct targeting. To succeed, we must focus on the molecular details of the RAS structure and understand at a high-resolution level how the oncogenic mutants impair function. Structural networks of intramolecular communication between the RAS active site and membrane-

interacting regions on the G-domain are disrupted in oncogenic mutants. Although conserved across the isoforms, these networks are near hot spots of protein–ligand interactions with amino acid composition that varies among RAS proteins. These differences could have an effect on stabilization of conformational states of interest in attenuating signaling through RAS. The development of strategies to target these novel sites will add a fresh direction in the quest to conquer RAS-driven cancers.

Introduction

(GAP), such as p120GAP and neurofibromin (NF1), that increase the very slow RAS intrinsic GTP hydrolysis rate to turn the signal off (6). This cycle is represented in Fig. 1 in the article by Cox and colleagues (3). Although this canonical cycle is the regulatory holy grail for RAS and related GTPases, a high-resolution view reveals variations that embellish and complement this cycle based on detailed structural features. RAS is composed of the core Gdomain containing the catalytic machinery and a lipidated Cterminal hypervariable region with low homology between isoforms, which tethers the G-domain to the membrane and drives localization to specific microdomains (7). The G-domain of RAS is composed of a 6-stranded b-sheet flanked by 5 a-helices and 10 connecting loops with two switch regions, switch I and switch II, that have different conformations in the GDP (guanosine diphosphate) and GTP-bound states (Fig. 1 and Fig. 2; ref. 8). Residues in the P-loop interact closely with the nucleotide. Within the range of conformations accessed by the switch regions in RAS–GTP, switch I has two major states: One in which it is open, away from the nucleotide (state 1) and the other where it is closed over the nucleotide (state 2; Fig. 2A; ref. 9). Switch I must be in state 2 to interact with effector proteins (10). Switch II contains the catalytic residue Q61 and is normally highly disordered (11), resulting in a very slow hydrolysis rate in the absence of GAPs (12). We have shown an association between allosteric modulation and intrinsic hydrolysis in the presence of RAF that would be expected to take place independent of GAPs (13–16). In our current model, the GTP hydrolysis rate is expected to be increased in the RAS/RAF complex by binding of Ca2þ at a remote allosteric site, resulting in a conformational shift from a catalytically impaired T state (disordered active site) to the competent R state in which the catalytic residue Q61 is ordered in the active site (Fig. 2B). Note that accompanying this disorder to order transition is a shift in helix3/loop 7 toward helix 4. Given the canonical GTPase cycle, major points of interference expected to result in signal attenuation through RAS are the inhibition of exchange of GDP to GTP catalyzed by GEFs and the inhibition of interaction between

RAS GTPase is a master signaling protein at the hub of numerous signal transduction pathways, responding to upstream signals from a variety of sources and interacting with a wide range of effector proteins to regulate cell proliferation, survival, migration, and apoptosis (1). Signal transduction pathways involving RAS are illustrated in McCormick's overview of this CCR Focus series (2). The three major isoforms of RAS, KRAS, NRAS, and HRAS together are mutated in about 20% of human cancers, primarily in the active site at residues G12, G13, and Q61 near the g-phosphate of the guanosine triphosphate (GTP) substrate (Fig. 1). RAS GTPases are, therefore, among the most urgent targets for the treatment of cancer. Yet, to date, there are no drugs in the clinic that directly target RAS, and thus, these proteins have come to be thought of as undruggable. The current state of directly drugging the Gdomain of RAS is the subject of this review. Targeting posttranslational modifications at the C-terminus is discussed by Cox and colleagues (3) in this CCR Focus series, while alternative strategies focused on altered metabolic pathways driven by RAS mutants and synthetic lethality approaches are reviewed by Kimmelman (4) and Downward (5), respectively. RAS functions as an on/off switch regulated by guanine nucleotide exchange factors (GEF), such as SOS, to facilitate loading of GTP turning the signal on, and by GTPase-activating proteins

Department of Chemistry and Chemical Biology, Northeastern University, Boston, Massachusetts. Corresponding Author: Carla Mattos, Department of Chemistry and Chemical Biology, Northeastern University, 102 Hurtig Hall, 360 Huntington Avenue, Boston, MA 02115. Phone: 617-373-6166; Fax: 617-373-8795; E-mail: [email protected] doi: 10.1158/1078-0432.CCR-14-2148 2015 American Association for Cancer Research.

Clin Cancer Res; 21(8); 1810–8. 2015 AACR.

See all articles in this CCR Focus section, "Targeting RASDriven Cancers."

1810 Clin Cancer Res; 21(8) April 15, 2015

Downloaded from clincancerres.aacrjournals.org on May 7, 2015. © 2015 American Association for Cancer Research.

Direct Attack on RAS

Figure 1. Location and identity of oncogenic mutations at G12, G13, and Q61 on RAS– GTP. The RAS structure is shown on the left in ribbon diagram bound to the GTP analogue 50 -guanylyl imidodiphosphate (GppNHp). The focus is on the active site with the P-loop, switch I, and switch II labeled in Roman type. The 5 a-helices, a1-a5, are also labeled in Roman type. Residues G12 (cyan), G13 (magenta), and Q61 (light orange) are drawn explicitly, with residue labels in boldface type and with N, Ca, C ¼ O, and side chain groups shown explicitly. N atoms are in blue, O atoms are in red, and the P atoms of the nucleotide are in dark orange. The identities of the mutations found in G12, G13, and Q61 in KRAS, NRAS, and HRAS are listed on the right in the order corresponding to how often they occur in human cancers (modified from ref. 1 and crosschecked with ref. 28).

α2 α3 Switch II α4

α5

G12 P-loop

RAS–GTP and its effector proteins (Fig. 3). Stabilization of distinct conformational states opens new opportunities for targeting RAS directly. Switch I stabilized in state 1 would have greatly diminished interaction with effectors and switch II stabilized in the ordered R state would promote intrinsic hydrolysis (Fig. 3). This review first focuses on the accomplishments achieved so far in terms of small-molecule inhibitors that directly interact with RAS. We then turn to an analysis of the structural details associated with RAS–GTP conformational states and communication between remote areas of the G-domain, as well as the effect of oncogenic mutations. This structural understanding suggests unexplored strategies to move forward in new directions.

The State of Small Molecules That Directly Bind RAS The problem of drugging RAS for treatment of human cancers is a major challenge for the academic and industry research communities, and the limited success attained so far has recently converged in a call for renewed efforts to conquer this seemingly insurmountable problem (17, 18). Numerous reviews on this topic have been published in the last 2 years (19–27). These articles outline the diverse strategies used to target RAS and related signaling pathways, including direct targeting. The picomolar affinity of RAS for GDP and GTP, coupled with sub-millimolar concentrations of these nucleotides in the cell, make the prospect of developing drugs that effectively compete with the nucleotide generally unrealistic. Furthermore, the lack of other well-defined active site pockets have left us with limited options for directly targeting RAS. The exchange of GDP for GTP, catalyzed by GEFs, turns on signaling through Ras, promoting interaction with effector proteins (1). Hydrolysis of GTP to GDP is the key reaction through which RAS signaling is turned off and is the center of

www.aacrjournals.org

Q61

Switch I

G13

α1 GppNHp

KRAS

NRAS

G12

G12

HRAS G12

G12D G12V G12C G12A G12S G12R

G12D G12S G12C G12V G12A G12R

G12V G12S G12D G12C G12R G12A

G13

G13

G13

G13D G13C G13S G13R G13A G13V

G13D G13R G13V G13C G13A G13S

G13R G13D G13V G13S G13C

Q61

Q61

Q61

Q61H Q61L Q61R Q61K Q61P Q61E

Q61R Q61K Q61L Q61H Q61P Q61E

Q61R Q61L Q61K Q61H Q61P Q61E

© 2015 American Association for Cancer Research

action for oncogenic mutations at G12, G13, and Q61 that impair both intrinsic and GAP-catalyzed hydrolysis (28). The challenge of directly targeting the RAS G-domain, thus, amounts to three approaches, all of which have been attempted with little or moderate success: increasing hydrolysis rates of mutant RAS (29, 30), inhibiting the exchange of GDP for GTP (31–39), or interfering with effector binding and activation (40–49). Recent efforts based on high-throughput screening of fragments to identify compounds that affect downstream signaling, accompanied by crystal structures that allowed structure-based optimization in some cases, have for the first time provided a high-resolution view of how small molecules are accommodated in binding pockets near the active site to interfere with RAS signaling (33–35, 39). Within this cohort of structures are covalent inhibitors of KRASG12C (33, 35), a mutant commonly found in lung adenocarcinoma (28), bringing to the forefront the prospect of selectively targeting this specific mutant with minimal effect on wildtype RAS. This kind of specificity will be difficult to achieve with other oncogenic mutants due to their less reactive side chains. One of the two covalent series of inhibitors consists of nucleotide analogues that bind in place of GDP (33), and the other series of inhibitors bind in a pocket between switch II and helix 3 (35). Other compounds for which there are crystal structures are not selective for particular mutants and bind at a site on the outer side of switch II (34, 39). All of these compounds with high-resolution crystal structures bind to the GDP-bound form of RAS on either side of switch II, inhibiting nucleotide exchange catalyzed by GEFs (Fig. 4A). Given that oncogenic RAS mutants retain RAS in its GTPbound form or promote rapid exchange in the absence of GEFs (50), it has been proposed that nucleotide exchange inhibitors may be best used in cancers bearing wild-type RAS (22). However, an increasing understanding of the role of the wild-type protein in promoting oncogenic RAS-driven cancers (51, 52) suggests that

Clin Cancer Res; 21(8) April 15, 2015


1811

CCR FOCUS

A

B

T state

L7

a2

L7

a2 Allosteric site

a3 a4

R state

a3 a4

Switch II

Switch II

a1 State 1

State 2 GppNHp

GppNHp

Switch I

Switch I

T state vs. R state

State 1 vs. state 2


GEF inhibitors may have a role to play in these types of cancers as well. Targeting RAS–GTP has produced a few low-affinity inhibitors of RAS–effector interactions accompanied by NMR information on the general binding area for the compounds near switch I (Fig. 4B; refs. 44–46). One of the compounds has a second binding site near loop 7 and the C-terminus, visible in a crystal structure of RAS (44). These compounds stabilize switch I in an open conformation that disfavors interaction with effector proteins. This mode of binding illustrates a strategy associated with stabilization of conformational states that takes advantage of controlling accessibility to functionally important conformers. In RAS–GTP, when switch I is open in state 1 Y32 is far from the active site where it obstructs the binding of RAF–RBD (RASbinding domain) and other effectors, while in the closed state 2 Y32 becomes part of the active site, allowing interaction with effector proteins (9, 10). A recent crystal structure of HRAS in the GTP-bound form in which switch I is in state 1 reveals a binding pocket between switch I and the nucleotide that may be used for structure-based ligand development (53). This is only one angle on the more general strategy of stabilizing conformational states that may offer windows of opportunity for direct targeting. RAS is full of this type of opportunity throughout its small but complex structure, where intramolecular communication networks between distant sites on the RAS G-domain have been discovered in the past few years (13, 54–56), but remain largely unexplored. To open new opportunities for drugging RAS, we must understand these structural features in depth, their relative importance in KRAS, NRAS, and HRAS, as well as the ways in which specific mutations at G12, G13, and Q61 affect them in terms of hydrolysis rates and communication with the membrane.


Figure 2. Conformational states associated with RAS–GTP. The RAS structures are shown in ribbon diagram with switch I and switch II labeled. The elements of secondary structure discussed in the text are also labeled. All structures are bound to the GTP analogue GppNHp, shown explicitly. A, switch I in state 1 is represented by the HRAS structure with PDB code 4EFL (pink), and state 2 is represented by the HRAS structure with PDB code 3K8Y (green). B, switch II in the T state is represented by the HRAS structure with PDB code 2RGE (blue), and the R state is represented by the HRAS structure with PDB code 3K8Y (green). Note that the model in the T state (blue) does not contain the N-terminal end of switch II (including Q61) due to disorder in that part of the 2þ structure. Also note the Ca and acetate bound in the allosteric site in the R state structure (green), with accompanying shift in helix 3/loop 7 (a3/L7) toward helix 4 (a4).

Intramolecular Communication in the Structure of Wild-type Isoforms of RAS Bound to GTP The Ras G-domain is subdivided in two lobes (57). The first (residues 1-86), identical in sequence between the Ras isoforms, contains the nucleotide-binding site, including the P-loop (residues 10-17), as well as the switch I (residues 30-40) and switch II (residues 60-76) regions that recognize effector and regulatory proteins, and thus is referred to as the effector lobe (58). The second contains several membrane-interacting regions that link elements of allosteric communication between the active site and the membrane (13, 55, 59), and thus is referred to as the allosteric lobe (58). This lobe is 90% identical between the isoforms, with most differences residing in sites of protein–membrane interactions, thus conferring specificity to the way in which the Gdomain of KRAS, NRAS, and HRAS interacts and communicates with the membrane (60). A challenging aspect of directly targeting RAS in its GTPbound form is the fact that switch I and switch II are disordered in solution (11, 58, 61), as is also the case in RAS–GDP (62), making it difficult for small molecules to bind. When bound to RAF–RBD, switch I is in state 2 and Y32 interacts with the g-phosphate of GTP via a bridging water molecule (Fig. 5). This brings Y32 into reach of a water-mediated H-bonding network that leads to N85 at the N-terminal end of helix 3 (56) and onto nucleotide-sensing residues R161 and R164 in helix 5 and D47 and E49 in the effector lobe loop L3 (55). This watermediated H-bonding network is highly conserved in HRAS– GTP, but not present in HRAS–GDP, with switch I providing a structural venue for communication between the g-phosphate of GTP and the membrane through the allosteric lobe (56).

Clinical Cancer Research



GEF inhibitors

Figure 3. The GTPase cycle, including the nuances of conformational states. RAS is shown in cyan, with switch I in light blue and switch II in yellow. The GEF SOS is in a dark brick color and GAP is in pink-gray. Black crosses indicate opportunities for blocking interactions that result in signal propagation through RAS, and the thick black arrow associated with stabilization of the R state indicates enhanced hydrolysis of GTP. GEF and covalent inhibitors target RAS–GDP, while RAF inhibitors, inhibitors that stabilize state 1 (state 1 inhibitors), or potential inhibitors that stabilize the R state to stimulate hydrolysis, target RAS–GTP. Note the distinction between inhibitors that prevent binding of RAF to RAS in state 2 and the state 1 inhibitors that stabilize a RAS conformation that intrinsically does not bind RAF.

Covalent inhibitors

RAS-GDP Hydrolysis stimulation, R state

Intrinsic hydrolysis

RAS/GAP

RAS/GEF

State 1 inhibitors RAS-GTP, state 2 RAS/RAF inhibitors

RAS/RAF

RAS-GTP, state 1

Cell proliferation, survival


The GTP-bound orientation of RAS with respect to the membrane is mediated through R128 and R135 in helix 4 in response to the nucleotide sensor residues (55, 59), and thus this intramolecular network that connects switch I to the allosteric lobe may have an impact on the orientation of the G-domain. Switch II also has been observed in distinct conformations associated with T and R states of allosteric modulation of the active site, given a state 2 conformation of switch I (16). Although normally disordered in crystals in the T state, switch II becomes highly ordered by binding of small molecules at a remote allosteric site found between helix 3, loop 7, and helix 4 (Fig. 2B; ref. 13). In crystals in which switch I has the conformation observed in the RAS/RAF–RBD complex (15), binding of calcium at the allosteric site results in a shift of helix 3 toward helix 4 and away from switch II, making room for the switch II helix to become ordered (13). This results in positioning of R68 at the center of an H-bonding network that facilitates placement of catalytic residue Q61 in the active site. Thus, in its effector-bound R state, switch II is connected to the membrane-interacting allosteric lobe through an H-bonding network to the allosteric site. Residues in the intramolecular communication networks from switch I and switch II in the effector lobe to the membrane interacting allosteric lobe described above are conserved in the three isoforms (Fig. 5), implicating the interlobal connectivity of the G-domain in the fundamental mechanisms through which RAS functions.


Binding Pockets on the Allosteric Lobe of RAS Differ between the Isoforms Screening of large compound libraries resulted in the identification of molecules that inhibit RAS signaling, as summarized above. In a very different approach, a smaller group of generic compounds or probe molecules have been used with both experimental and computational methods to identify the location of binding site hot spots on RAS–GTP in search of pockets that can be explored in drug discovery (58, 63). In addition to identifying known sites on the effector lobe of RAS– GTP, this work identified hot spots of protein–ligand interactions on the allosteric lobe that are not far from the elements involved in the interlobal communication networks discussed above. These hot spots are key targets of opportunity, as they are not identical between the isoforms (60) and to date most have been unexplored. The locations of identified hot spots on the RAS–GTP effector lobe include the binding pocket near switch I known as the binding site for RAF–RBD (site 1) and the binding site on the outer side of switch II (site 2), while the binding site between switch II and helix 3 is found in the interlobal region (site 3; Fig. 5A). Sites 2 and 3 are also found to bind small molecules in RAS– GDP (Fig. 4A). These three sites, as well as the site near loop 7 mentioned above (Fig. 4B), are identified as high consensus sites because they are found repeatedly in several studies (25). Site 3 between switch II and helix 3 is of note because, in spite of it being



1813

CCR FOCUS A

B

State 1 inhibitor (44)

a5 a5 L7

L7

a2

GEF inhibitors (34, 39)

a2

a4 a4 a3

Switch II Covalent inhibitor (35)

Switch II

a3

State 1 inhibitor

a1

Switch I Switch I

GppNHp

Raf inhibitors

GDP


a strong hot spot of protein–ligand interaction, every small molecule known to bind at this site in RAS–GTP is observed to stabilize an anticatalytic conformation associated with the T state of the allosteric switch (58, 64). Thus, targeting the site between switch II and helix 3 might inadvertently result in prolonged signaling through RAS–RAF in some situations. Because RAF does not bind switch II (15), a compound at this site would be unlikely to have a significant effect on RAF binding while abrogating hydrolysis of GTP, working against the goal of inhibiting RAS signaling. Although no natural binding partners have been identified to interact with RAS at site 3, importin-b occupies it in RAN GTPase to halt hydrolysis during transport of cargo into the nucleus (65). All of the sites identified on the allosteric lobe other than the already mentioned site near loop 7 (site 5) have been classified as low consensus sites because they have been found in a single study (58) and no compounds from large library screens have been found to bind at those sites. Interestingly, however, the amino acid composition of these sites varies among the isoforms (60) and they are found conspicuously close to the elements of the interlobal communication networks in RAS–GTP. Figure 5 shows the sites of interest in the allosteric lobe and their relationship to the H-bonding networks that connect the two lobes. One of the sites is between helices 3 and 4, adjacent to the allosteric site (site 6; Fig. 5A), and compounds that bind there interact with the side chain hydroxyl group of Y137 (58), whose main chain carbonyl group coordinates the Ca2þ in the R state of the allosteric switch (13). Two other sites are between helices 4 and 5 and contain helix 4 residues R128 (site 7) and R135 (site 8), respectively (Fig. 5B). Residues R128 and R135 help maintain an orientation of the RAS G-domain with respect to the membrane conducive to effector binding and were shown to abrogate signaling when mutated in cells (59). Another low consensus hot spot identified by the same study with an experimental approach (58) is at the interface between the effector and allosteric lobes and is known to bind


Figure 4. Binding sites on RAS for currently published inhibitors. RAS is shown in ribbon diagram with labeled secondary structural elements. A, binding sites for inhibitors that bind RAS–GDP with representative molecules on either side of switch II. Protein residues that interact with inhibitors between switch II and helix 3 (a3) are colored blue, and those that interact on the outer side of switch II and the b-sheet are colored orange. B, binding sites for inhibitors that bind to RAS–GTP. The one inhibitor for which there is a crystal structure is shown near loop 7 (L7) and the C-terminal end of helix 5 (a5) with associated residues in green. There are no crystal structures for molecules bound to switch I, but residues with which these inhibitors are known to interact are shown in red. Numbers in parentheses are references to articles in which the representative inhibitor molecules shown in this figure were published.

RAF–CRD (cystein-rich domain), the second RAS-binding domain found in RAF (site 4; Fig. 5B; ref. 66). Binding at this site can potentially have an effect on the switch I to helix 5 communication network associated with nucleotide sensing residues in helix 5 near the membrane or affect the orientation of the G-domain at the membrane surface.

Effects of Oncogenic Mutations on Intramolecular Communication Networks in RAS–GTP It is well known that oncogenic RAS mutants are insensitive to GAPs (67–70). This is a difficult problem to overcome, as the mutations themselves severely perturb the active site and impair the ability of GAPs to enhance GTP hydrolysis even upon binding RAS (50, 71). If it is indeed the case that the duration of the RAS– RAF interaction is regulated by intrinsic hydrolysis activated by membrane interaction and calcium binding at the allosteric site (13, 14, 16, 56), there an exciting window of opportunity exists for targeting signaling through RAS–RAF enhanced by some oncogenic mutants. This is of interest because the RAS–RAF–MEK–ERK pathway is a potent mitogenic pathway and an aggressive driver of several types of cancers (72). Intrinsic hydrolysis is dependent on an ordered active site with an intact communication network to the allosteric site that requires the R state positioning of helix 3/ loop 7 and correct placement of the R68 link to Q61 through water-mediated H-bonds (Fig. 5A). It has been shown that both G12V and Q61L can attain the R and T states of the allosteric switch under circumstances where switch I is in state 2 (14), as it is when bound to RAF, and it is likely that other oncogenic mutants can as well. In structures available for the G12 and Q61 mutants, those in which switch I is in state 2 and the allosteric switch is in the R state (14, 34, 35, 67, 73, 74), with the exception of G12D, the bridging water molecule seen in the wild-type protein is not present and Y32




A D107

B

L7 5

L7

α3

Allosteric site

α2

R97

R68

7

SW II

3

H94

α4

L3

6

R128

E49

6

Allosteric site

α5

R164

2 Y137

D107

5

α4

E47 β2

R161

R135 8

8

4

1 N86

Q61

SW I

Y32

N86 α1

1


Figure 5. Interlobal communication networks and ligand-binding hot spots on RAS in its GTP-bound form. RAS is shown in ribbon diagram (PDB code 3K8Y), with residues conserved in the KRAS, NRAS, and HRAS isoforms shown in gray, those appearing in two of the isoforms in yellow, and those that are distinct in each isoform in red. The GTP analogue GppNHp is shown in the active site near switch I (SWI) and switch II (SWII). Calcium and acetate are green and shown in the allosteric site. Residues involved in the communication networks are shown explicitly. Two orientations are shown in A and B, rotated approximately 180 degrees relative to each other. The communication network from SWII to the allosteric site and from SWI to the nucleotide sensor residues in helix 5/loop 3 (a5/L3) is shown in black dashed lines. Binding site hot spots are shown in spheres. Two sites in the effector lobe are shown in olive green: the RAF-RBD–binding pocket near SWI (site 1) and the site on the outer side of SWII (site 2). Two sites in the interlobal region are shown in cyan: the site between SWII and helix 3 (a3; site 3) and the RAF-CRD–binding site between helix 1 (a1), beta sheet 2 (b2), and helix 5 (a5; site 4). Four sites in the allosteric lobe are shown in violet: the site near loop 7 (L7) opposite the allosteric site (site 5), the site between helices a3 and a4 near Y137 and H94 (site 6), and two sites between helices a4 and a5 near R128 and R135, respectively (sites 7 and 8). Amino acid residues are in boldface type, and secondary structural elements of proteins are in Roman type.

makes a direct H-bond with the g-phosphate of GTP (Fig. 6). The arrangement in wild-type RAS is shown in Fig. 6, with the bridging water molecule within H-bonding distance of Y32, the g-phosphate, and of Q61. RASG12V and RASQ61L, superimposed on the wild-type, are representative examples of the direct H-bond between Y32 and the g-phosphate in most oncogenic mutants (Fig. 6A). This results in a severed connection between switch I and the helix 5 nucleotide sensor residues associated with membrane orientation (54). In the G12D mutant, for which there are two crystal structures, the bridging water molecule is either replaced with one of the side-chain oxygen atoms of the D12 residue or is present and D12 replaces Q61 in its interaction with the bridging water molecule (Fig. 6B). Contrary to what is observed in all other oncogenic mutants, the water-mediated network from the g-phosphate to the nucleotide sensing residues in helix 5 is intact in both structures of RASG12D, and this may contribute to the effectiveness and frequency of this mutant in human cancers (20, 28). The two structures differ in their effects on switch II. When D12 replaces the bridging water molecule, the side chain of Q61 is rotated toward the solvent. The network from the allosteric site to R68 is intact, but the connection to Q61 is impaired because of steric hindrance in the active site. When D12 takes the position of Q61, the N-terminus of switch II moves out of place, completely severing the connectivity to the allosteric site. All of the mutants in structures where helix 3/ loop 7 is in the R state affect switch II in one of these two ways. G12V and G12C affect the connectivity to the active site with respect to the Q61 rotamer, and could potentially benefit from stabilization of the R state by ligand binding at the allosteric site or possibly in one


of the nearby allosteric lobe sites to promote the shift in helix 3. Full wild-type rates will not be achieved because of the perturbed active site. However, it has been argued that a 30- to 50-fold enhancement of the hydrolysis rate may result in significant anticancer activity (75), and it is worth testing whether this is possible for the G12V and G12C mutants. The G12R and Q61H mutants displace residue 61 from the active site, abrogating connectivity to the allosteric lobe, and are less likely to respond to ligand binding aimed at promoting the R state. The G12D and Q61 mutants directly impair the chemistry of GTP hydrolysis by either placing a negative charge right over the nucleotide or removing an important catalytic residue. Furthermore, Q61L is also unique in that it has a strong tendency to stabilize the T state in the anticatalytic conformation when bound to RAF, abolishing even the slow hydrolysis rate observed in the uncomplexed protein (14, 15, 73). In the absence of binding partners, both switch I and switch II are disordered in solution (11, 61), and they become stabilized by interactions with other proteins (15, 30, 76, 77). In the crystal structures we have discussed, switch I is stabilized in state 2 by crystal contacts in the conformation found in the RAS/RAF–RBD complex (15) or in the RAS/NORE1A complex (78) and switch II is less constrained by crystal contacts. This may not always be the case. Therefore, in general one must be vigilant in taking into account the effects of crystal contacts. Remarkably for RAS, most of the switch I conformations stabilized by crystal contacts are also observed in a biologically relevant complex (64). Given a careful structural analysis, the kind of information discussed herein will be invaluable in obtaining insights leading to specific targeting of



1815

CCR FOCUS A

B

SW II

SW II Q61

Q61

Q61 Q61 L61

α3

Q61

Bridging water α3

G/V12

G12D

Bridging water

SW I

SW I Y32

N86 Y32

N85

N86

N85

K117

GppNHp

K117

GppNHp


Figure 6. The active site of wild-type and mutant RAS proteins with switch I in the effector-bound conformation (state 2). Wild-type HRAS and its associated water molecules in the network are in green (PDB code 3K8Y). The network from N85 to the g-phosphate of GTP is shown in black dashed lines. Note that this network also extends to Q61 in wild-type RAS. A, the oncogenic mutants other than G12D, represented by HRASG12V shown in pink (PDB code 3OIW) and HRASQ61L in yellow (PDB code 3OID). Note the direct H-bond between Y32 and the g-phosphate of the nucleotide shown in red dashed lines. The bridging water molecule is missing in the mutant structures. B, the G12D mutant in HRAS and KRAS. HRASG12D is in cyan (PDB code 1AGP). The D12 residue takes the place of the bridging water molecule in this structure. KRASG12D is in orange (PDB code 4DSN). The bridging water molecule is present in this structure, and the D12 side chain takes the place of the Q61 side chain in making an H-bond to the bridging water molecule, displacing the N-terminal end of SWII near the active site. Red dashed lines are shown for the active site H-bonding interactions in the two G12D structures. In both structures Y32 is in a position to connect to the water network to N85, which in turn continues on to the nucleotide sensor residues in helix 5, R161, and R164 (shown in Fig. 5B). Amino acid residues are in boldface type, and secondary structural elements of proteins are in Roman type.

RAS mutants. Although focusing on the allosteric site and nearby pockets on the allosteric lobe (sites 5 and 6) may target RAS in a mutant-specific fashion, targeting the sites near R128 and R135 between helices 4 and 5 (sites 7 and 8) is likely to work more generally by disrupting the functional orientation of the Gdomain relative to the membrane, which is specific for each of the isoforms (54, 55, 79, 80), possibly superseding the effects of particular mutations.

Conclusions Targeting RAS against cancers has been a challenge for the past 25 years. Yet recent progress in the discovery of small molecules that bind directly to the G-domain with inhibitory effects on its function is exciting and encouraging. For RAS–GTP, focusing on the RAS–effector interaction at the active site by stabilizing state 1 is particularly promising, as this is the most direct way to inhibit RAS signaling, although nonspecific with respect to particular isoforms. In-depth understanding of the structure and function of RAS and its interaction with the membrane opens new avenues for targeting isoforms or specific mutations that are in need of further development. Targeting the allosteric lobe will require assays that detect binding at those sites in high-throughput screens, and the hypotheses derived from structural analysis of specific mutants regarding the activation of the allosteric switch need to be confirmed. Another unexplored area of opportunity is


interaction between the C-terminal hypervariable region and switch II in K-RAS4B with calmodulin, which results in signal attenuation (81–84). Although promotion of protein–protein interaction is not trivial, it is not unprecedented (85) and should be explored in the future. Although treatment of RAS-driven cancers will continue to require a cocktail of inhibitors of multiple pathways (86), the rapidly increasing knowledge of RAS structure and function fuels the vision that among available inhibitors in the future will be a drug that directly binds to RAS.

Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed.

Authors' Contributions Conception and design: K. Marcus, C. Mattos Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K. Marcus, C. Mattos Writing, review, and/or revision of the manuscript: K. Marcus, C. Mattos Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): K. Marcus

Grant Support This work was supported by a grant from the National Science Foundation (MCB-1244203; to C. Mattos). Received December 12, 2014; revised February 17, 2015; accepted February 24, 2015; published online April 15, 2015.




References 1. Cox AD, Der CJ. Ras history: the saga continues. Small GTPases 2010;1: 2–27. 2. McCormick F. KRAS as a therapeutic target. Clin Cancer Res 2015;21: 1797–801. 3. Cox AD, Der CJ, Philips MR. Targeting RAS membrane association: back to the future for anti-RAS drug discovery? Clin Cancer Res 2015;21:1819–27. 4. Kimmelman AC. Metabolic dependencies in RAS-driven cancers. Clin Cancer Res 2015;21:1828–34. 5. Downward J. RAS synthetic lethal screens revisited: still seeking the elusive prize? Clin Cancer Res 2015;21:1802–9. 6. Bos JL, Rehmann H, Wittinghofer A. GEFs and GAPs: critical elements in the control of small G proteins. Cell 2007;129:865–77. 7. Henis YI, Hancock JF, Prior IA. Ras acylation, compartmentalization, and signaling nanoclusters. Mol Membr Biol 2009;26:80–92. 8. Milburn MV, Tong L, deVos AM, Brunger A, Yamaizumi Z, Nishimura S, et al. Molecular switch for signal transduction: structural differences between active and inactive forms of protooncogenic ras proteins. Science 1990;247:939–45. 9. Geyer M, Schweins T, Herrmann C, Prisner T, Wittinghofer A, Kalbitzer HR. Conformational transitions in p21ras and in its complexes with the effector protein Raf-RBD and the GTPase activating protein GAP. Biochemistry 1996;35:10308–20. 10. Spoerner M, Hozsa C, Poetzl JA, Reiss K, Ganser P, Geyer M, et al. Conformational states of human rat sarcoma (Ras) protein complexed with its natural ligand GTP and their role for effector interaction and GTP hydrolysis. J Biol Chem 2010;285:39768–78. 11. Ito Y, Yamasaki K, Iwahara J, Terada T, Kamiya A, Shirouzu M, et al. Regional polysterism in the GTP-bound form of the human c-Ha-Ras protein. Biochemistry 1997;36:9109–19. 12. John J, Frech M, Wittinghofer A. Biochemical properties of Ha-ras encoded p21 mutants and mechanism of the autophosphorylation reaction. J Biol Chem 1988;263:11792–9. 13. Buhrman G, Holzapfel G, Fetics S, Mattos C. Allosteric modulation of Ras positions Q61 for a direct role in catalysis. Proc Natl Acad Sci U S A 2010;107:4931–6. 14. Buhrman G, Kumar VS, Cirit M, Haugh JM, Mattos C. Allosteric modulation of Ras-GTP is linked to signal transduction through RAF kinase. J Biol Chem 2011;286:3323–31. 15. Fetics S, Guterres H, Kearney BM, Buhrman G, Ma B, Nussinov R, et al. Allosteric effects of the oncogenic RasQ61L mutant on Raf-RBD. Structure 2015;23:1–12. 16. Johnson CW, Mattos C. The allosteric switch and conformational states in Ras GTPase affected by small molecules. Enzymes 2013;33:41–67. 17. Stephen AG, Esposito D, Bagni RK, McCormick F. Dragging ras back in the ring. Cancer Cell 2014;25:272–81. 18. Thompson H. US National Cancer Institute's new Ras project targets an old foe. Nat Med 2013;19:949–50. 19. Chandrashekar R, Adams PD. Prospective development of small molecule targets to oncogenic Ras proteins. Open J Biophys 2013;3:207–11. 20. Cox AD, Fesik SW, Kimmelman AC, Luo J, Der CJ. Drugging the undruggable RAS: mission possible? Nat Rev Drug Discov 2014;13: 828–51. 21. Ebi H, Faber AC, Engelman JA, Yano S. Not just gRASping at flaws: finding vulnerabilities to develop novel therapies for treating KRAS mutant cancers. Cancer Sci 2014;105:499–505. 22. Mattingly RR. Activated ras as a therapeutic target: constraints on directly targeting ras isoforms and wild-type versus mutated proteins. ISRN Oncol 2013;2013:536529. 23. Spiegel J, Cromm PM, Zimmermann G, Grossmann TN, Waldmann H. Small-molecule modulation of Ras signaling. Nat Chem Biol 2014;10: 613–22. 24. Vivanco I. Targeting molecular addictions in cancer. Br J Cancer 2014;111: 2033–8. 25. Wang W, Fang G, Rudolph J. Ras inhibition via direct Ras binding—is there a path forward? Bioorg Med Chem Lett 2012;22:5766–76. 26. Wang Y, Kaiser CE, Frett B, Li HY. Targeting mutant KRAS for anticancer therapeutics: a review of novel small molecule modulators. J Med Chem 2013;56:5219–30. 27. Downward J. RAS's cloak of invincibility slips at last? Cancer Cell 2014; 25:5–6.


28. Prior IA, Lewis PD, Mattos C. A comprehensive survey of Ras mutations in cancer. Cancer Res 2012;72:2457–67. 29. Ahmadian MR, Stege P, Scheffzek K, Wittinghofer A. Confirmation of the arginine-finger hypothesis for the GAP-stimulated GTP-hydrolysis reaction of Ras. Nat Struct Biol 1997;4:686–9. 30. Scheffzek K, Ahmadian MR, Kabsch W, Wiesmuller L, Lautwein A, Schmitz F, et al. The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science 1997;277:333–8. 31. Colombo S, Peri F, Tisi R, Nicotra F, Martegani E. Design and characterization of a new class of inhibitors of ras activation. Ann N Y Acad Sci 2004;1030:52–61. 32. Hocker HJ, Cho KJ, Chen CY, Rambahal N, Sagineedu SR, Shaari K, et al. Andrographolide derivatives inhibit guanine nucleotide exchange and abrogate oncogenic Ras function. Proc Natl Acad Sci U S A 2013;110: 10201–6. 33. Lim SM, Westover KD, Ficarro SB, Harrison RA, Choi HG, Pacold ME, et al. Therapeutic targeting of oncogenic K-Ras by a covalent catalytic site inhibitor. Angew Chem Int Ed Engl 2014;53:199–204. 34. Maurer T, Garrenton LS, Oh A, Pitts K, Anderson DJ, Skelton NJ, et al. Smallmolecule ligands bind to a distinct pocket in Ras and inhibit SOS-mediated nucleotide exchange activity. Proc Natl Acad Sci U S A 2012;109:5299–304. 35. Ostrem JM, Peters U, Sos ML, Wells JA, Shokat KM. K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions. Nature 2013; 503:548–51. 36. Patgiri A, Yadav KK, Arora PS, Bar-Sagi D. An orthosteric inhibitor of the Ras-Sos interaction. Nat Chem Biol 2011;7:585–7. 37. Sacco E, Abraham SJ, Palmioli A, Damore G, Bargna A, Mazzoleni E, et al. Binding properties and biological characterization of new sugar-derived Ras ligands. Med Chem Commun 2011;2:396–401. 38. Schopel M, Jockers KF, Duppe PM, Autzen J, Potheraveedu VN, Ince S, et al. Bisphenol A binds to Ras proteins and competes with guanine nucleotide exchange: implications for GTPase-selective antagonists. J Med Chem 2013;56:9664–72. 39. Sun Q, Burke JP, Phan J, Burns MC, Olejniczak ET, Waterson AG, et al. Discovery of small molecules that bind to K-Ras and inhibit Sos-mediated activation. Angew Chem Int Ed Engl 2012;51:6140–3. 40. Barnard D, Sun H, Baker L, Marshall MS. In vitro inhibition of Ras–Raf association by short peptides. Biochem Biophys Res Commun 1998;247: 176–80. 41. Clark GJ, Drugan JK, Terrell RS, Bradham C, Der CJ, Bell RM, et al. Peptides containing a consensus Ras binding sequence from Raf-1 and theGTPase activating protein NF1 inhibit Ras function. Proc Natl Acad Sci U S A 1996;93:1577–81. 42. Karaguni IM, Glusenkamp KH, Langerak A, Geisen C, Ullrich V, Winde G, et al. New indene-derivatives with anti-proliferative properties. Bioorg Med Chem Lett 2002;12:709–13. 43. Kato-Stankiewicz J, Hakimi I, Zhi G, Zhang J, Serebriiskii I, Guo L, et al. Inhibitors of Ras/Raf-1 interaction identified by two-hybrid screening revert Ras-dependent transformation phenotypes in human cancer cells. Proc Natl Acad Sci U S A 2002;99:14398–403. 44. Rosnizeck IC, Graf T, Spoerner M, Trankle J, Filchtinski D, Herrmann C, et al. Stabilizing a weak binding state for effectors in the human ras protein by cyclen complexes. Angew Chem Int Ed Engl 2010;49:3830–3. 45. Shima F, Yoshikawa Y, Matsumoto S, Kataoka T. Discovery of smallmolecule Ras inhibitors that display antitumor activity by interfering with Ras. GTP-effector interaction. Enzymes 2013;34:1–23. 46. Shima F, Yoshikawa Y, Ye M, Araki M, Matsumoto S, Liao J, et al. In silico discovery of small-molecule Ras inhibitors that display antitumor activity by blocking the Ras-effector interaction. Proc Natl Acad Sci U S A 2013;110:8182–7. 47. Upadhyaya P, Qian Z, Habir NA, Pei D. Direct ras inhibitors identified from a structurally rigidified bicyclic peptide library. Tetrahedron 2014;70: 7714–20. 48. Waldmann H, Karaguni IM, Carpintero M, Gourzoulidou E, Herrmann C, Brockmann C, et al. Sulindac-derived Ras pathway inhibitors target the RasRaf interaction and downstream effectors in the Ras pathway. Angew Chem Int Ed Engl 2004;43:454–8. 49. Wu X, Upadhyaya P, Villalona-Calero MA, Briesewitz R, Pei D. Inhibition of Ras-effector interaction by cyclic peptides. Med Chem Commun 2013; 4:378–82.



1817

CCR FOCUS 50. Smith MJ, Neel BG, Ikura M. NMR-based functional profiling of RASopathies and oncogenic RAS mutations. Proc Natl Acad Sci U S A 2013;110:4574–9. 51. Hayes TK, Der CJ. Mutant and wild-type Ras: co-conspirators in cancer. Cancer Discov 2013;3:24–6. 52. Young A, Lou D, McCormick F. Oncogenic and wild-type Ras play divergent roles in the regulation of mitogen-activated protein kinase signaling. Cancer Discov 2013;3:112–23. 53. Muraoka S, Shima F, Araki M, Inoue T, Yoshimoto A, Ijiri Y, et al. Crystal structures of the state 1 conformations of the GTP-bound H-Ras protein and its oncogenic G12V and Q61L mutants. FEBS Lett 2012;586:1715–8. 54. Abankwa D, Gorfe AA, Inder K, Hancock JF. Ras membrane orientation and nanodomain localization generate isoform diversity. Proc Natl Acad Sci U S A 2010;107:1130–5. 55. Abankwa D, Hanzal-Bayer M, Ariotti N, Plowman SJ, Gorfe AA, Parton RG, et al. A novel switch region regulates H-ras membrane orientation and signal output. EMBO J 2008;27:727–35. 56. Kearney BM, Johnson CW, Roberts DM, Swartz P, Mattos C. DRoP: a water analysis program identifies Ras-GTP-specific pathway of communication between membrane-interacting regions and the active site. J Mol Biol 2014;426:611–29. 57. Gorfe AA, Grant BJ, McCammon JA. Mapping the nucleotide and isoformdependent structural and dynamical features of Ras proteins. Structure 2008;16:885–96. 58. Buhrman G, O'Connor C, Zerbe B, Kearney BM, Napoleon R, Kovrigina EA, et al. Analysis of binding site hot spots on the surface of Ras GTPase. J Mol Biol 2011;413:773–89. 59. Gorfe AA, Hanzal-Bayer M, Abankwa D, Hancock JF, McCammon JA. Structure and dynamics of the full-length lipid-modified H-Ras protein in a 1,2-dimyristoylglycero-3-phosphocholine bilayer. J Med Chem 2007;50:674–84. 60. Parker JA, Mattos C. The Ras-membrane interface: isoform-specific differences in the catalytic domain. Mol Cancer Res 2015 Jan 7. [Epub ahead of print]. 61. O'Connor C, Kovrigin EL. Global conformational dynamics in ras. Biochemistry 2008;47:10244–6. 62. Kraulis PJ, Domaille PJ, Campbell-Burk SL, Van Aken T, Laue ED. Solution structure and dynamics of ras p21.GDP determined by heteronuclear three- and four-dimensional NMR spectroscopy. Biochemistry 1994;33:3515–31. 63. Grant BJ, Lukman S, Hocker HJ, Sayyah J, Brown JH, McCammon JA, et al. Novel allosteric sites on Ras for lead generation. PLoS ONE 2011; 6:e25711. 64. Holzapfel G, Buhrman G, Mattos C. Shift in the equilibrium between on and off states of the allosteric switch in Ras-GppNHp affected by small molecules and bulk solvent composition. Biochemistry 2012; 51:6114–26. 65. Vetter IR, Arndt A, Kutay U, Gorlich D, Wittinghofer A. Structural view of the Ran-Importin beta interaction at 2.3 A resolution. Cell 1999;97: 635–46. 66. Thapar R, Williams JG, Campbell SL. NMR characterization of full-length farnesylated and non-farnesylated H-Ras and its implications for Raf activation. J Mol Biol 2004;343:1391–408. 67. Franken SM, Scheidig AJ, Krengel U, Rensland H, Lautwein A, Geyer M, et al. Three-dimensional structures and properties of a transforming and a nontransforming glycine-12 mutant of p21H-ras. Biochemistry 1993;32: 8411–20.


68. Gibbs JB, Schaber MD, Allard WJ, Sigal IS, Scolnick EM. Purification of ras GTPase activating protein from bovine brain. Proc Natl Acad Sci U S A 1988;85:5026–30. 69. Gremer L, Gilsbach B, Ahmadian MR, Wittinghofer A. Fluoride complexes of oncogenic Ras mutants to study the Ras-RasGap interaction. Biol Chem 2008;389:1163–71. 70. Vogel US, Dixon RA, Schaber MD, Diehl RE, Marshall MS, Scolnick EM, et al. Cloning of bovine GAP and its interaction with oncogenic ras p21. Nature 1988;335:90–3. 71. Smith MJ, Ikura M. Integrated RAS signaling defined by parallel NMR detection of effectors and regulators. Nat Chem Biol 2014;10:223–30. 72. McCubrey JA, Steelman LS, Chappell WH, Abrams SL, Wong EW, Chang F, et al. Roles of the Raf/MEK/ERK pathway in cell growth, malignant transformation, and drug resistance. Biochim Biophys Acta 2007;1773: 1263–84. 73. Buhrman G, Wink G, Mattos C. Transformation efficiency of RasQ61 mutants linked to structural features of the switch regions in the presence of Raf. Structure 2007;15:1618–29. 74. Krengel U, Schlichting I, Scherer A, Schumann R, Frech M, John J, et al. Three-dimensional structures of H-ras p21 mutants: molecular basis for their inability to function as signal switch molecules. Cell 1990;62: 539–48. 75. Ahmadian MR, Zor T, Vogt D, Kabsch W, Selinger Z, Wittinghofer A, et al. Guanosine triphosphatase stimulation of oncogenic Ras mutants. Proc Natl Acad Sci U S A 1999;96:7065–70. 76. Huang L, Hofer F, Martin GS, Kim SH. Structural basis for the interaction of Ras with RalGDS. Nat Struct Biol 1998;5:422–6. 77. Pacold ME, Suire S, Perisic O, Lara-Gonzalez S, Davis CT, Walker EH, et al. Crystal structure and functional analysis of Ras binding to its effector phosphoinositide 3-kinase gamma. Cell 2000;103:931–43. 78. Stieglitz B, Bee C, Schwarz D, Yildiz O, Moshnikova A, Khokhlatchev A, et al. Novel type of Ras effector interaction established between tumour suppressor NORE1A and Ras switch II. EMBO J 2008;27:1995–2005. 79. Abankwa D, Gorfe AA, Hancock JF. Mechanisms of Ras membrane organization and signalling: Ras on a rocker. Cell Cycle 2008;7:2667–73. 80. Kapoor S, Weise K, Erlkamp M, Triola G, Waldmann H, Winter R. The role of G-domain orientation and nucleotide state on the Ras isoform-specific membrane interaction. Eur Biophys J 2012;41:801–13. 81. Abraham SJ, Nolet RP, Calvert RJ, Anderson LM, Gaponenko V. The hypervariable region of K-Ras4B is responsible for its specific interactions with calmodulin. Biochemistry 2009;48:7575–83. 82. Lopez-Alcala C, Alvarez-Moya B, Villalonga P, Calvo M, Bachs O, Agell N. Identification of essential interacting elements in K-Ras/calmodulin binding and its role in K-Ras localization. J Biol Chem 2008;283:10621–31. 83. Villalonga P, Lopez-Alcala C, Bosch M, Chiloeches A, Rocamora N, Gil J, et al. Calmodulin binds to K-Ras, but not to H- or N-Ras, and modulates its downstream signaling. Mol Cell Biol 2001;21:7345–54. 84. Wu LJ, Xu LR, Liao JM, Chen J, Liang Y. Both the C-terminal polylysine region and the farnesylation of K-RasB are important for its specific interaction with calmodulin. PLoS One 2011;6:e21929. 85. Peyroche A, Antonny B, Robineau S, Acker J, Cherfils J, Jackson CL. Brefeldin A acts to stabilize an abortive ARF-GDP-Sec7 domain protein complex: involvement of specific residues of the Sec7 domain. Mol Cell 1999;3:275–85. 86. Nussinov R, Tsai CJ, Mattos C. "Pathway drug cocktail": targeting Ras signaling based on structural pathways. Trends Mol Med 2013;19: 695–704.



Direct Attack on RAS: Intramolecular Communication and Mutation-Specific Effects Kendra Marcus and Carla Mattos Clin Cancer Res 2015;21:1810-1818.

Updated version

Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/21/8/1810

Cited Articles

This article cites by 85 articles, 26 of which you can access for free at: http://clincancerres.aacrjournals.org/content/21/8/1810.full.html#ref-list-1

Citing articles

This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/21/8/1810.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].


Oncogenes: direct hit on mutant RAS.

The loss of NHS Direct is an attack on nursing.

Effects of RAS inhibitors on diabetic retinopathy.

Ras: structural details to guide direct targeting.

Topological computation based on direct magnetic logic communication.

Intramolecular Communication in Anionic Oxidized Phosphanes through a Chelated Proton.

Ras p21: effects and regulation.

Influencing feelings of cancer risk: direct and moderator effects of affectively laden phrases in risk communication.

Physician communication coaching effects on patient experience.

Hydrolysis of 3-chloro-3-cephems. Intramolecular nucleophilic attack in cefaclor.

Controlled bidirectional quantum secure direct communication.

Direct communication of results to patients.

Effects of RAS inhibitors on diabetic retinopathy - Authors' reply.

The idea of innateness: effects on language and communication research.

The idea of innateness: effects on language and communication research.

Direct communication between a coronary artery and the right ventricle.

An attack on tracheostomy.

How Intrinsic Molecular Dynamics Control Intramolecular Communication in Signal Transducers and Activators of Transcription Factor STAT5.

Changes in visual responsiveness following intercollicular lesions and their effects on avoidance and attack.

Direct electrochemistry and intramolecular electron transfer of ascorbate oxidase confined on L-cysteine self-assembled gold electrode.

Effects of lindane on membrane fluidity: intramolecular excimerization of a pyrene derivative and polarization of diphenylhexatriene.

Studies on the intramolecular and intermolecular kinetic isotope effects in pyruvate carboxylase catalysis.

A Nuclear Attack on Thrombosis and Inflammation.

Effects of steroidal hormones on sexual, attack, and search behavior in the isolated male chick.