Coarse-Grained Simulations Suggest the Epsin N-Terminal Homology Domain Can Sense Membrane Curvature without Its Terminal Amphipathic Helix

Nanoscale membrane curvature is a common feature in cell biology required for functions such as endocytosis, exocytosis and cell migration. These processes require the cytoskeleton to exert forces on the membrane to deform it. Cytosolic proteins contain specific motifs which bind to the membrane, connecting it to the internal cytoskeletal machinery. These motifs often bind charged phosphatidylinositol phosphate lipids present in the cell membrane which play significant roles in signaling. These lipids are important for membrane deforming processes, such as endocytosis, but much remains unknown about their role in the sensing of membrane nanocurvature by protein domains. Using coarse-grained molecular dynamics simulations, we investigated the interaction of a model curvature active protein domain, the epsin N-terminal homology domain (ENTH), with curved lipid membranes. The combination of anionic lipids (phosphatidylinositol 4,5-bisphosphate and phosphatidylserine) within the membrane, protein backbone flexibility, and structural changes within the domain were found to affect the domain’s ability to sense, bind, and localize with nanoscale precision at curved membrane regions. The findings suggest that the ENTH domain can sense membrane curvature without the presence of its terminal amphipathic α helix via another structural region we have denoted as H3, re-emphasizing the critical relationship between nanoscale membrane curvature and protein function.

suggesting that H0 adds a significant amount of stability to the domains ability to bind neutral membranes.
We also observe that flexible ENTH domains seeded onto PIP2-containing membranes show an enhancement of localization between 0 and 5 nm. As well as an increase in fold change, this is also consistent with the observed concentration of PIP2 to concave membrane curvature in the interior of the pore (as shown in Figure 2b ii and in Figures S3e and S3g). This suggests that protein binding to this region can be attributed to local PIP2 enrichment. Note that the scalloped features in Supplementary Figure S3 at 16 nm and 20 nm are artefacts due to the rectangular simulation cell.
Interestingly, when the domains are modelled with the rigid elastic network, while localization to convex membrane curvature is still observed, this behavior is more dependent upon membrane lipid density (as indicated by the lower peak fold changes at around 12 nm in Figures 3b, 3d, 3e and 3h). In the elastic network ENTH/PC/PIP2 case, we observe almost total dependence on PIP2 (indicated by a fold change of ~1 across the entire radius, as shown in Figure 3f). This suggests that restricting the flexibility of the domain increases the strength with which the domain can bind to PIP2, while hampering its ability to sense curvature.
We also find that the ENTH (w/o H0) domain consistently shows enhanced localization to all of the membranes studied here (PC-, PC/PIP2-and PC/PS-containing) irrespective of whether the domain is modelled as flexible or with an elastic network, see Supplementary Figure S4 for PC/PS data. The flexible ENTH (w/o H0) domain is so strongly localized to curvature that the domain remained bound to the curved region during our simulations. In order to confirm the consistency of this behavior, multiple simulations were made from two starting positions, one starting above the central pore and the other above curvature, with the combined results shown in Figures 3g and 3h (an illustration of this process is shown in Figure S5). We also replicated this profile by performing a "mutated" simulation case whereby a snapshot from the flexible ENTH/PC/PIP2 case was taken as a starting position and truncated, to remove H0, before running simulations, also shown in Figure S5. This is also observed for elastic network ENTH domains interacting with PC-only membranes, where the presence or absence of H0 also makes little difference to the fold change profile of the proteins (Figure 3b and 3d). When PIP2 is added, however, the localization of elastic network domains with H0 is mainly driven by PIP2 localization (again, as indicated by the broadly constant fold change of ~1 in Figure 3f). For the same system but without H0, a minor curvature sensing enhancement is observed, reinforcing the notion that H0 plays a relatively minor role. When interacting with PS-containing membranes, we observed similar behavior, flexible ENTH and ENTH (w/o H0) domains strongly sense curvature while the elastic network ENTH domain is more dependent on PIP2. In addition, on PS-membranes, ENTH (w/o H0) domains sense curvature very similarly both when flexible and with a rigid elastic network (Supplementary Figure S4).

Lipid Binding
We analyzed how the residues of each domain bind to PC-, PC/PIP2-and PC/PS-containing membranes. Figure 4 shows these contacts for the first 100 residues of the domains on PC and  (Figure 4a i), the flexible ENTH domain shows hydrophobic interactions at both H0 and H3 with the overall contacts at H3 being elevated. However, the activity of H3 is completely removed for ENTH domains with an elastic network. When H0 is removed (Figure 4a ii), we conserve the hydrophobic action of H3 in flexible domains. When an elastic network is added, we still observe hydrophobic contacts around the same region except that it is localized approximately to the edge of H3. On PC/PIP2 membranes, the activity of H3 is greatly reduced for flexible ENTH domains and is again not present with an elastic network (Figure 4a iii). For the ENTH (w/o H0) domains however, the hydrophobic activity of H3 is conserved both when flexible and with a rigid elastic network (Figure 4a iv). The consistent hydrophobic activity of H3 suggests innate membrane curvature sensitivity of ENTH enabling it to be guided to regions of membrane curvature even before the formation of H0. Combined with the localization behavior in Figure 3, these results suggest that H0 alone is insufficient to sustain membrane curvature sensitivity in the ENTH domain.
On PC/PS-containing membranes, we see similar binding behavior, with H3 being        Crystal structure of the ENTH domain with hydrophobic residues of helices H0 and H3 highlighted in yellow and hydrophobic surfaces in transparent yellow along with helical wheels for helices H3 and H0 showing hydrophobic moment of each helix produced by HELIQUEST. 1 Residues 48, 50 and 51 are marked on the wheel and structure as these residues are hydrophobically active in the elastic network ENTH (w/o H0) case (see Figure 4a ii and iv).