The higher level of complexity of K-Ras4B activation at the membrane

Abstract

Is nucleotide exchange sufficient to activate K-Ras4B? To signal, oncogenic rat sarcoma (Ras) anchors in the membrane and recruits effectors by exposing its effector lobe. With the use of NMR and molecular dynamics (MD) simulations, we observed that in solution, farnesylated guanosine 5'-diphosphate (GDP)-bound K-Ras4B is predominantly autoinhibited by its hypervariable region (HVR), whereas the GTP-bound state favors an activated, HVR-released state. On the anionic membrane, the catalytic domain adopts multiple orientations, including parallel (∼180°) and perpendicular (∼90°) alignments of the allosteric helices, with respect to the membrane surface direction. In the autoinhibited state, the HVR is sandwiched between the effector lobe and the membrane; in the active state, with membrane-anchored farnesyl and unrestrained HVR, the catalytic domain fluctuates reinlessly, exposing its effector-binding site. Dimerization and clustering can reduce the fluctuations. This achieves preorganized, productive conformations. Notably, we also observe HVR-autoinhibited K-Ras4B-GTP states, with GDP-bound-like orientations of the helices. Thus, we propose that the GDP/GTP exchange may not be sufficient for activation; instead, our results suggest that the GDP/GTP exchange, HVR sequestration, farnesyl insertion, and orientation/localization of the catalytic domain at the membrane conjointly determine the active or inactive state of K-Ras4B. Importantly, K-Ras4B-GTP can exist in active and inactive states; on its own, GTP binding may not compel K-Ras4B activation.-Jang, H., Banerjee, A., Chavan, T. S, Lu, S., Zhang, J., Gaponenko, V., Nussinov, R. The higher level of complexity of K-Ras4B activation at the membrane.

Keywords: GDP/GTP exchange; KRAS; farnesyl insertion; phospholipids; signaling.

Figure 1.

NMR CSPs of residues in K-Ras4B-GDP by the farnesyl and their mapping onto a crystal structure. A) Superimpositions of ¹H-¹⁵N HSQC spectra of 0.28 mM K-Ras4B-GDP with the PTMs (red) and 0.28 mM K-Ras4B-GDP with the non-PTMs (blue). Examples of CSPs as a result of farnesyl and methyl groups are marked. B) Mapping of the perturbed residues on the structure of the GDP-bound K-Ras4B catalytic domain. In the structure, residues with high CSPs are marked. The CSP residues are grouped; region 1 (blue surface; I21, V29, I36, D38, S39, Y40, L52), region 2 (green surface; Q61, E63, Y64), and region 3 (pink surface; N26, F28, T148, R149, A155).

Figure 2.

Snapshots representing relaxed structures of K-Ras4B with the PTMs in the aqueous environment for the GDP-bound (A) and GTP-bound (B) states. Cartoons for the catalytic domains are shown in green and pink for the GDP- and GTP-bound states, respectively. The HVR in the tube representation is blue, and the farnesyl as a stick is yellow. In the catalytic domain, the red sticks and green spheres represent the nucleotide and Mg²⁺ ions, respectively. Config., configuration.

Figure 3.

3D DCCM of the helix-to-helix motions, including α3-α4, α3-α5, and α4-α5 helices in the allosteric lobe for the GDP-bound (A) and GTP-bound K-Ras4B (B) in the aqueous environment. Positively correlated residues with correlation coefficient C(i,j) = 0.6 (red); anticorrelated (negatively correlated) residues with C(i,j) = −0.6 (dark purple); and uncorrelated residues with C(i,j) = 0 (dark green) are shown.

Figure 4.

Snapshots representing relaxed structures of K-Ras4B with the PTMs on the anionic lipid bilayer composed of DOPC:DOPS (mole ratio 4:1) for the GDP-bound (A) and GTP-bound (B) states. Cartoons for the catalytic domains are shown in green and pink for the GDP- and GTP-bound states, respectively. The HVR in the tube representation is blue, and the farnesyl as a stick is yellow. In the catalytic domain, the red sticks and green spheres represent the nucleotide and Mg²⁺ ions, respectively. For the lipid bilayer, white surface denotes DOPC, and gray surface represents DOPS.

Figure 5.

3D DCCM of the helix-to-helix motions, including α3-α4, α3-α5, and α4-α5 helices in the allosteric lobe for the GDP-bound (A) and GTP-bound K-Ras4B (B) on the lipid bilayer. Positively correlated residues with correlation coefficient C(i,j) = 0.6 (red); anticorrelated (negatively correlated) residues with C(i,j) = −0.6 (dark purple); and uncorrelated residues with C(i,j) = 0 (dark green) are shown.

Figure 6.

The deviation (Δz) from the bilayer surface for the center of mass of selected domains of K-Ras4B. A) Time average of Δz from the bilayer surface for the center of mass of the catalytic domain, HVR, and farnesyl terminus for the GDP-bound state. B) Time series (t) of Δz for the farnesyl terminus for the GDP-bound state. C, D) The same time average (C) and time series (D) of Δz for the GTP-bound state.

Figure 7.

Interaction map representing the percentage of K-Ras4B HVR interactions with surrounding environments, including its own catalytic domain, lipid, and water. The interaction energies of each HVR with the catalytic domain, lipid, and water are calculated and then averaged over the time.

Figure 8.

Probability distribution functions for helix θ, with respect to the bilayer normal for the effector lobe helices—α1 and α2—and the allosteric lobe helices—α3, α4, and α5—for the GDP-bound (A) and GTP-bound K-Ras4B (B). The helix tilt was calculated for the angle between 2 vectors formed by the helix axis and normal axis of the bilayer.

Figure 9.

Comparison of effector-binding activity of GTP-bound wild-type, G12D, and G12V K-Ras4B_1–188 by use of Raf-1 RBD-binding assay in the absence (A) and presence (B) of membrane fractions prepared from NIH 3T3 fibroblasts.

Figure 10.

Diagrams representing the inactive (left) and active (right) states of K-Ras4B at the anionic membrane. The balloon represents the K-Ras4B catalytic domain, the blue thread tied to the balloon represents the HVR, and the yellow sawtooth represents the farnesyl. In the inactive state, the membrane-interacting HVR hauls the effector lobe to the membrane surface, burying the effector-binding site. In the active state, the catalytic domain liberates the HVR, exposing the effector-binding site and fluctuating reinlessly.