Lipid Profiles of RAS Nanoclusters Regulate RAS Function

Abstract

The lipid-anchored RAS (Rat sarcoma) small GTPases (guanosine triphosphate hydrolases) are highly prevalent in human cancer. Traditional strategies of targeting the enzymatic activities of RAS have been shown to be difficult. Alternatively, RAS function and pathology are mostly restricted to nanoclusters on the plasma membrane (PM). Lipids are important structural components of these signaling platforms on the PM. However, how RAS nanoclusters selectively enrich distinct lipids in the PM, how different lipids contribute to RAS signaling and oncogenesis and whether the selective lipid sorting of RAS nanoclusters can be targeted have not been well-understood. Latest advances in quantitative super-resolution imaging and molecular dynamic simulations have allowed detailed characterization RAS/lipid interactions. In this review, we discuss the latest findings on the select lipid composition (with headgroup and acyl chain specificities) within RAS nanoclusters, the specific mechanisms for the select lipid sorting of RAS nanoclusters on the PM and how perturbing lipid compositions within RAS nanoclusters impacts RAS function and pathology. We also describe different strategies of manipulating lipid composition within RAS nanoclusters on the PM.

Keywords: RAS; lipid acyl chain; nanoclusters; phosphatidylserine; phospholipids; plasma membrane.

Figure 1

Specific lipids within RAS nanoclusters directly participate in effector recruitment. RAS nanoclusters possess distinct lipid profiles. The specific lipids enriched within these nanoclusters not only contribute to the structural integrity of the nanoclusters, but also directly participate in effector recruitment. Most RAS effectors contain specific lipid-binding domains. For example, RAF, a major KRAS effector, possesses separate PS- and PA-binding domains. Efficient recruitment of RAF to the PM, a key step in the activation of RAF, requires synergistic binding to both the GTP-bound active KRAS and specific lipids including PS and PA.

Figure 2

RAS proteins form spatially distinct nanoclusters in isoform- and guanine nucleotide-specific manners. HRAS, KRAS and NRAS each form non-overlapping nanoclusters on the PM. For each isoform, the GDP-bound inactive and the GTP-bound active forms also form separate nanoclusters. The nanoclusters of the inactive GDP-HRAS contain PIP2 and cholesterol, while the active GTP-bound HRAS selectively associates with PIP3. The active and inactive KRAS contain similar lipid contents, enriching PS and PA. The nanoclusters of the inactive KRAS contain additional PIP2. The lipid environments of the inactive and active NRAS are less clear, except that cholesterol is more enriched in the nanoclusters of the active GTP-bound NRAS.

Figure 3

KRAS polybasic domain contributes to the selective sorting of lipid headgroups. The prenylation (15-carbon farnesyl or 20-carbon geranylgeranyl chain) and the polybasic domain of KRAS combine to contribute to the selectivity of lipid headgroups within KRAS nanoclusters. The constitutively active oncogenic mutant KRASG12V with different combinations of the prenyl anchors and the residues of polybasic domains selectively sort distinct lipid types with different headgroups.

Figure 4

The prenylated polybasic domain of KRASG12V possesses capability to selectively sort PS acyl chains. Despite highly charged, the prenylated polybasic domain of KRASG12V distinguishes the acyl chain structures of PS lipids. In particular, four KRASG12V constructs with equivalently charges but distinct prenyl anchors and/or basic residues within the polybasic domain sort different PS species with distinct acyl chains. This selective sorting of lipid acyl chains contributes to the distinct abilities of KRASG12V polybasic domain mutants to respond to different membrane properties, including transmembrane voltages, membrane curvature, cholesterol depletion, etc.

Figure 5

KRAS polybasic domain mutants with equivalent charges possess distinct preferences for membrane curvature. KRASG12V with the original farnesylated hexa-lysine or the geranylgeranylated hexa-arginine prefers to form nanoclusters on flatter membranes with low curvature. On the other hand, KRASG12V with the farnesylated hexa-arginine or the geranylgeranylated hexa-lysine polybasic domain favors to interact with more curved membranes. The ability of these KRAS polybasic domain mutants to selectively sort distinct PS species with different packing geometries contributes to the distinct membrane curvature sensing capabilities of KRAS.

Figure 6

The select lipid sorting of KRAS can be targeted to manipulate KRAS function and pathology. The nanoclusters of the constitutively active oncogenic mutant KRAS enrich PS lipids. Two different strategies can be used to compromise KRAS nanoclustering, thus effector recruitment and signal transduction: (1) perturbation of the PS content in the plasma membrane; (2) alteration of KRAS polybasic domain to change the lipid preferences of KRAS.