Exploring the Nanotoxicology of MoS2: A Study on the Interaction of

Dec 13, 2017 - Computational Biological Center, IBM Thomas J. Watson Research ... *E-mail: [email protected]., *E-mail: [email protected]...
0 downloads 0 Views 2MB Size
Subscriber access provided by READING UNIV

Article

Exploring the Nanotoxicology of MoS2: A Study on the Interaction of MoS2 Nanoflake and K+ Channels Zonglin Gu, Leigh D. Plant, Xuanyu Meng, Jose Manuel Perez-Aguilar, Zegao Wang, Mingdong Dong, Diomedes E. Logothetis, and Ruhong Zhou ACS Nano, Just Accepted Manuscript • DOI: 10.1021/acsnano.7b07871 • Publication Date (Web): 13 Dec 2017 Downloaded from http://pubs.acs.org on December 13, 2017

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 free 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 accessible to all readers and 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.

ACS Nano 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 38 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 Nano

Exploring the Nanotoxicology of MoS2: A Study on the Interaction of MoS2 Nanoflake and K+ Channels

Zonglin Gu1§, Leigh D. Plant2§, Xuan-Yu Meng1,*, Jose Manuel Perez-Aguilar3, Zegao Wang5, Mingdong Dong5,6, Diomedes E. Logothetis2* and Ruhong Zhou1,3,4,*

1. Institute of Quantitative Biology and Medicine, SRMP and RAD-X, Collaborative Innovation Center of Radiation Medicine of Jiangsu Higher Education Institutions, Soochow University, Suzhou 215123, China 2. Department of Pharmaceutical Sciences in the School of Pharmacy, Northeastern University Bouvé College of Health Sciences, Boston, MA 02115 USA. 3. Computational Biological Center, IBM Thomas J. Watson Research Center, Yorktown Heights, NY 10598, USA 4. Department of Chemistry, Columbia University, New York, NY 10027, USA 5. Interdisciplinary Nanoscience Center, Aarhus University, DK-8000 Aarhus C, Denmark 6. Department of Chemistry, Stanford University, Stanford, California 94305, United States.

§

These authors contributed equally to this work *Corresponding author, E-mail: [email protected]; [email protected]

[email protected],

ABSTRACT: Molybdenum disulfide (MoS2) nanomaterial has recently found various applications in the biomedical field mainly due to its outstanding physicochemical properties. However, little is known about its interactions with biological systems at the atomic level, which intimately relates to the biocompatibility of the material. To provide insights into the effects of MoS2 in biological entities, we investigated the interactions of MoS2 with proteins from a functionally important membrane family, the ubiquitous potassium (K+) channels. Here, we study four representative K+ channels ˗ KcsA, Kir3.2, the Kv1.2 paddle chimera, and K2P2 ˗ to investigate their interactions with a triangular MoS2 nanoflake using Molecular Dynamics (MD) simulations combined with electrophysiology experiments. These particular K+ channels were selected based on the diversity in their structure, that is, although these

ACS Paragon Plus Environment

ACS Nano 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 38

K+ channels display similar structural motifs, they also contain significant differences related to their particular function. Our results indicate that the MoS2 nanoflake is able to stably bind to three out of the four channels, albeit through distinct binding modes. The binding mode between each channel and MoS2 underlies the specific deleterious influence on the channel’s basic physiological function: For KcsA, MoS2 binds on the extracellular loops, which indirectly destroys the delicate structure of the selectivity filter causing a strong leak of K+ ions. In the binding mode with Kir3.2, the MoS2 nanoflake completely covers the entrance to the channel pore affecting the normal ion conduction. For the Kv1.2 chimera, the MoS2 nanoflake prefers to bind into a crevice located at the extracellular side of the Voltage Sensor Domain (VSD). Interestingly, the crevice involves the N-terminal segment of S4, a crucial transmembrane helix which directly controls the gating process of the Kv1.2 chimera channel by electromechanical coupling the VSD to the transmembrane electric field. MoS2 contact with S4 from the Kv1.2 chimera, potentially influences the channel’s gating process from open to closed states. In all three systems, the van der Waals contribution to the total energy dominates the binding interactions; also, hydrophobic residues contribute the most contact points, which agrees with the strong hydrophobic character

of

the

MoS2

nanomaterial.

Electrophysiology

recordings

using

two-electrode voltage-clamp (TEVC) show that currents of Kir3.2 and Kv1.2 are both blocked by the MoS2 nanoflakes in a concentration-dependent way. While the background K+ channel, K2P2 (TREK-1), identified as a negative control, is not blocked by the MoS2 nanoflakes. The large and rigid extracellular domain of K2P2 appears to protect the channel from disturbance by the nanoflakes. MoS2’s intrinsic chemical properties, together with the channels’ specific features, such as the electrostatic character and complex surface architecture, determine the critical details of the binding events. These findings might shed light on the potential nanotoxicology of MoS2 nanomaterials and help to understand the underlying molecular mechanism.

KEYWORDS: MoS2 nanoflake, potassium channels, nanotoxicology, molecular dynamics simulation, electrophysiology. ACS Paragon Plus Environment

Page 3 of 38 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 Nano

Carbon-based nanomaterials (CBNs), including graphene,1 carbon nanotubes (CNTs),2 fullerenes3 and their respective derivatives, are probably the most popular nanostructures and nanomaterials due to their outstanding mechanical, optical, and electrical properties.4-6 These nanomaterials have raised considerable interest in the biomedical field,7-8 for their potential applications in drug delivery,9 optical imaging,10 nanotherapeutics,11-13 and biosensing.14 Interestingly, molybdenum disulfide (MoS2), a prototypical member of the transition metal dichalcogenides (TMD), has also attracted great attention very recently in numerous applications including transistors, sensors, optoelectronics, filtration devices, catalytic agents, and in hydrogen storage media.15-21 Moreover, the MoS2 nanomaterial is deemed to share similar physicochemical properties with various CBNs and thus can potentially replicate the CBNs’ success in the biomedical field. For instance, MoS2 nanosheets have already been proposed as drug delivery platforms through surface modification or coating (e.g., PEG, chitosan, proteins).22-23 Two-dimensional MoS2 sheets are also found to possess detrimental properties against the growth of different microorganism (i.e., bacteria and fungi), hence acting as potential effective antimicrobial and antifungal agents.24 Remarkably, because of its highly efficient near-infrared (NIR) adsorption capacity, MoS2 has also been used as a photothermally-triggered drug delivery platform in anti-cancer therapy.25-26 Furthermore, MoS2 has been proposed as a field-effect biosensor for the detection of DNAs and proteins27-28 due to its semiconducting electronic properties (a moderate direct band gap of 1.8 eV). MoS2 has also been utilized as a medical contrast agent in X-ray Computed Tomography imaging because of the adsorption properties of the Mo element at this wavelength range.26

In spite of its excellent properties and potentially widespread applications in various medical and biological fields, there is limited and inconsistent information regarding MoS2’s cytotoxicity.29-30 It is known that polyethylene glycol (PEG) functionalized MoS2 nanosheets are enriched in reticuloendothelial systems (RES) for one month ACS Paragon Plus Environment

ACS Nano 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

after intravenous injection.31 The accessibility of MoS2 nanoflakes to the cell surface in multicellular organisms needs further investigation. The biosafety of evaluations for MoS2 nanomaterials are still in progress.32

In this study, we focus on small sized (a few nanometers to tens nanometers) MoS2 nanosheets, which have comparable dimensions with important macro-biomolecules such as membrane proteins and study if a strong adverse binding may occur between the two nanoscopic entities. If it does, what are the structural and functional consequences (e.g., toxicity) of such interactions? And, what are the specific binding features and binding modes of these MoS2-protein interactions? To address these interesting questions, we used an equilateral triangle MoS2 “nanoflake” with 9 Mo atoms in one side (n=9; side length of ~28 Å) based on Besenbacher and coworker’s recent study.33 They synthesized a series of equilateral triangle MoS2 nanoclusters with n=4-12 Mo atoms on one side, and in this study we chose a middle size with n=9 as a representative.33 As for the macromolecular counterpart, we selected the members of the important membrane protein family of potassium (K+) channels. This family was chosen not only for their central function but also for the ubiquitous distribution in different living cells.34-37 Three representative K+ channels, with comparable contact areas with the MoS2 nanoflake, were considered: the bacterial KcsA channel,38-39 the inwardly rectifying K+ channel Kir3.2 (also called GIRK2),40 a chimeric voltage-dependent K+ channel (Kv1.2-Kv2.1 paddle-chimera channel, termed Kv1.2 chimera)41, and a K2P2 channel (TREK-1).42 The binding complex systems were investigated by all-atom molecular dynamics simulations and TEVC recording to study the effect of nanoflakes on channels expressed in Xenopus laevis oocytes. We found that the MoS2 nanoflake can interact directly with all three different K+ channels, albeit by distinct binding modes, which suggest different perturbing effects of the MoS2 nanoflake on each channel’s physiological function. In contrast, the TREK-1 channel is identified as a negative control that uses its large and rigid extracellular domain to protect the channel from disturbance of the nanoflakes. Our results provide mechanistic insights into the potential cytotoxic properties of this ACS Paragon Plus Environment

Page 4 of 38

Page 5 of 38 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 Nano

MoS2 nanomaterial.

RESULTS In order to examine the structural and functional effects of MoS2 nanoflakes on K+ ion channels, three distinct K+ channels, KcsA, Kir3.2, and Kv1.2 chimera, were selected and investigated by MD simulations. The structures of the three K+ channels are shown in Figure 1, where some of their common and differential structural elements are indicated. All three K+ channels share the conserved structure of the central Pore Domain, consisting of two transmembrane helices, TM1 and TM2 that allow selective access to K+ ions. However, there are also structural domains that are specific to the individual channels which are related to their particular function and regulation. For example, Kir3.2 contains a large cytoplasmic domain that accommodates the binding sites for the channel’s molecular regulators, i.e., sodium ions (Na+) and the phosphoinositide PIP2 (see central panels in Fig 1a,b)40; on the other hand, Kv1.2 chimera contains a voltage sensor domain (VSD), which is responsible for sensing changes in the membrane potential of cells and thereby regulates the gating process in the pore domain (see right panels in Fig 1a,b) .41

The extracellular domains of the three K+ channels also display structural differences (Figure 1). In the three K+ channels, the functionally important Extracellular Loop 1 (EL1), which connects the helices from the pore domain (the outer helix TM1 and the pore-forming inner helix TM2), presents different sequence length, residue sequence identity, and tertiary structure. Additionally, the extracellular side of the VSD domain of Kv1.2 chimera, contains two long loops, S1-S2 Loop that connects the transmembrane helices S1 and S2, and S3-S4 Loop that connects transmembrane helices S3 and S4 (see Figure S1). Thus, the extracellular domain of each channel shows a complex surface architecture. Lastly, the extracellular domains of the three K+ channels are further distinguished by their electrostatic potential surface (Figure 1c): for KcsA, the EL1 and the short EL2 (connecting the selectivity filter with TM2) are mainly neutral with some scattered positive charges; for Kir3.2, the extracellular ACS Paragon Plus Environment

ACS Nano 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

elements show predominantly negative charges; for Kv1.2 chimera, the VSD extracellular surface is mainly neutral while its pore domain is similar to its counterpart in the Kir3.2 channel.

Here we explored the molecular basis related to possible binding mechanisms between the diverse and complex extracellular domain of specific K+ channels and the nanoscopic MoS2 fragment. We aim to provide answers not only about the similarities and differences in the binding modes of the MoS2 nanoflake with these distinct K+ channels, but also about the functional implications of such interactions as well as the factors that regulate the respective interacting modes.

KcsA Channel: The MoS2 Nanoflake Disturbs the Structure of the Selectivity Filter In each run of the simulated MoS2/KcsA complex binding system, the MoS2 nanoflake was initially positioned 1.2 nm away from the extracellular side of the channel. Figure 2a shows the final conformations obtained from three independent trajectories. Consistently, the MoS2 nanoflake structure preferentially interacts with residues located at the subunit interface (see Fig. 2a), specifically, with three distinct KcsA regions: the EL1 loop, EL2 loop, as well as part of the Pore Helix (a short helix connecting the EL1 with the selectivity filter). The selectivity filter (SF) region that surrounds the central ion pathway, was not directly involved in the KcsA binding mode. However, by binding to a neighboring peripheral region, the MoS2 nanoflake significantly changes the spatial arrangement between adjacent subunits and further reshapes the structure along the ion pathway. To quantify the structural changes at the SF, we monitored the distance between the Oxygen atoms from diagonal subunits. As shown in the Figure 2b, the distances in all three runs increase by 3-8 Å compared to the distance values in the control run (where the MoS2 nanoflake was not included), indicating a dilated SF structure in the presence of MoS2. Moreover, we also extended one representative trajectory to double the simulation time (300 ns), as shown in ACS Paragon Plus Environment

Page 6 of 38

Page 7 of 38 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 Nano

Figure S5, and found that the MoS2 nanoflake caused further structural distortion of the KcsA SF.

The ion permeation is concomitantly influenced by the enlarged SF structure. During the MoS2/KcsA simulation trajectories, K+ ions were still observed to pass through the SF but in a chaotic way, that is, the ions were no longer kept in line with the SF (following the so-called knock-on mechanism)43 as it was observed in the control run. The permeation events are described in Figure 2c, where the number of passing K+ ions is counted along the simulation time. The distorted SF structures allowed more K+ ions to pass through the pore, that is, as a consequence of the MoS2/KcsA binding, the K+ permeation across the channel occurred at a higher rate. In the most severe situation (run 2), the MoS2 binding led to ~70 K+ ions to pass through the channel during the 150 ns of simulation time, almost 4 times higher than that observed in the control run; in the other two runs, the number of passing K+ ions reached 2-3 times the level of the control run.

To understand the dynamical binding process in detail, the trajectory from run 3 was used for illustration (Figure 3). After 7 ns, the MoS2 nanoflake has initiated the contacts with KcsA, particularly residues P55, Y82, T85, and L86 from chain A; Q58 and R64 from chain B; and D80, Q56, R64, L81, and Y82 from chain D. Notably, the contacting residues consist of five hydrophobic/aromatic amino acids which account for ~45% of the total contacting residues. Thus, from the “early stage” of the direct interaction, hydrophobic and van der Waals (vdW) interactions dominate the MoS2 binding to KcsA.

From the period between 7 ns and 36 ns, termed “transition stage”, MoS2 no longer interacted with residues from chain B but instead strengthened its interaction with residues from chain A. The interactions with chain D also decreased by reducing the number of contacting residues from 5 to 2 (the total number of contacting residues decrease from 11 in the “early stage” to 7). Note that all the 7 remaining contacting ACS Paragon Plus Environment

ACS Nano 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

residues are hydrophobic (P55, V84, L86, P63), aromatic (Y82) or basic (R64, R89), implying that in addition to the dominant hydrophobic and vdW interactions in the early stage, electrostatic interactions also come into play.

After the transition stage and until ~41 ns, the MoS2 nanoflake further penetrated between chain A and chain D. As a consequence of the deep penetration, the number of contacting residues increased, from 7 to 10, indicating a stronger interaction between the MoS2 nanoflake and the KcsA channel. After 41 ns, the penetration process seemed to be completed since the nanostructure conformation remained steady for the rest of the simulation, maintaining the number of residue contacts and the interaction energy. The final complex conformation involves the following binding residues: A54, P55, Y82, P83, V84, T85, L86, and R89 from chain A, and P63, R64, L66, W67, E71, D80, and L81 from chain D. Interestingly, residues D80 and L81 are located next to the highly conserved ion-selective motif 77-GYG-79, while residues P63, R64, L66, W67, and E71, are part of the Pore Helix. Both residue segments belong to very sensitive regions where mutations have been found to significantly alter the behavior of the K+ channel.38-39, 44 Also, R64 and R89 are found to form a ‘sandwiching’ interaction with an anionic lipid to maintain the integrity of the KcsA tetramer.45-46 By directly contacting these protein segments, and hence significantly perturbing the protein structure, the MoS2 nanoflake successfully disturbed the normal function of the K+ channel (increasing the K+ ion passage up to 4-fold), which might result in a reduced MoS2 biocompatibility (i.e., increased toxicity).

Table 1 summarizes the interaction energy between the KcsA channel and the MoS2 nanoflake along the simulations. The average total interaction energy is -65.47 kcal/mol, which can be divided into vdW (-52.55 kcal/mol) and Coulombic (-12.92 kcal/mol) contributions. The vdW energy dominates the MoS2/KcsA interactions, as mentioned above. In addition, we also analyzed the contacting atom pair number (a contacting atom pair between protein and MoS2 was counted if the distance between ACS Paragon Plus Environment

Page 8 of 38

Page 9 of 38 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 Nano

any protein heavy atom is within 6.0 Å of any MoS2 atom) and contact residue types (i.e., hydrophobic/aromatic, hydrophilic, and charged residues) based on the equilibrated trajectories and the final snapshots, respectively. We found that the contacting atom pair numbers are more sensitive and closely relate to energy changes, while contact residues are relative stable in the equilibrated trajectories. Similarly, a contact residue is considered if at least one residue heavy atom is within 6.0 Å of the MoS2. In this case, KcsA has an average contact atom pair number of ~546, and a total number of contact residues of ~15. From the average number of contact residues, we found that hydrophobic residues account for two-thirds of the total contribution, which is in agreement with the intrinsic strong hydrophobicity of the MoS2 nanomaterial (the monolayered MoS2 displays a large water contact angle, which is even larger than that of graphene47-48).

Kir3.2 channel: MoS2 Covers the Extracellular Opening of the Channel Representative structures (final conformations) from three independent trajectories for the simulations of the MoS2/Kir3.2 system are shown in Figure 4. Remarkably, as indicated in run 2, the MoS2 nanoflake can bind across the extracellular pore domain of the channel. By blocking the extracellular domain of the Kir3.2 channel, the MoS2 fragment (positioned right above the SF) works as a ‘lid’ that covers the Selectivity Filter. The ‘lid’ conformations of the MoS2 nanoflake are stabilized by specific interactions with residues located at the EL1 in three out of the four channel subunits, and interestingly, some of the main protein contact are with the distal face of the MoS2 plane (see central panel in Fig 4). This mode of interaction between the EL1 segments and the distal face of the planar MoS2 nanoflake resembles some open-closing “buckle” mechanical systems. When we extended the trajectory to 300 ns (Figure S6) only minor changes occurred compared to 150 ns, suggesting that the interactions we observe are stable. The particular interaction details of the MoS2/Kir3.2 binding mode are predicted to cause dramatic and deleterious changes in the physiological function of this type of K+ channel.

ACS Paragon Plus Environment

ACS Nano 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

In the case of run 1 and run 3, the general MoS2/Kir3.2 binding mode details (i.e., the contacting domains) are very similar to the aforementioned run 2, however, the contact regions are slightly different. In run 1, MoS2 interacts with the EL1 segment from only one subunit, while in run 3, the interaction is stabilized with the EL1 segment from two different subunits. Also in run 3, at the length of our simulations (150 ns), the MoS2 nanoflake only partially covers the extracellular opening of the Kir3.2 channel (Figure 4).

To examine the determinants of the MoS2/Kir3.2 binding mode, we selected run 2 as a representative trajectory for illustration (Figure 5). At early stages of the simulation (t=4.6 ns), the MoS2 nanoflake already contacted the extracellular loops in chain A (colored red in Fig. 5) and in chain D (colored orange in Fig. 5), where the residues in the EL1 from chain A contacted the distal face of the planar MoS2 while residues in the EL1 from chain D contacted the proximal face of the plane. At this stage, the MoS2 planar fragment was tilted relative to the lipid membrane by ~45 degrees. From 4.6 ns to 26 ns, the MoS2 triangle constantly adjusted its orientation at the extracellular side of the channel trying to fit into the space between the EL1s from chain A and chain D. As seen in Figure 5b, at around 26 ns, the MoS2 fragment reached again an orientation parallel to the membrane and partially covered the pore of the Kir3.2 channel. Also at this stage, the EL1 segments from chain A and D “buckled” the MoS2 by interacting with the distal face of the MoS2 plane. After 26.0 ns, the MoS2 fragment slowly slid toward the center of the extracellular opening of the Kir3.2 channel and at 78.0 ns, the nanoflake was already located right above the channel, completely covering its pore. At this site, the pose of the MoS2 nanoflake is highly stable with three of the four EL1 segments (from chain B, C and D) “buckled up”, that is, where residues from these protein segments interact with the distal face of the MoS2 plane. Additionally, three EL2 segments also contributed to further stabilize this particular MoS2/Kir3.2 binding mode (Figure 5h).

Figure 5d illustrates the time evolution of the trajectory where the MoS2 moves ACS Paragon Plus Environment

Page 10 of 38

Page 11 of 38 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 Nano

toward the final binding site at the extracellular surface of Kir3.2. The MoS2 nanoflake is simplified by using its center of mass (COM) and the color code corresponds to the time along the simulation. From the positions of the COM we can see that, due to the irregular extracellular surface of the protein, a mechanism where the channel’s pore is directly blocked by the nanoflake is not observed, instead, the MoS2 fragment first interacts with the periphery of the protein then subsequently drifts into the final binding site position, i.e., on top of the channel opening. In the final binding position of the MoS2/Kir3.2 binding mode, the flexibility of the EL1 segments to favorably accommodate the MoS2 nanoflake, is essential. Using the Cα position of residue P129, located on the top of the EL1 segment, we monitored the distance variation between diagonal subunits (dchain AD and dchain BC), whose difference (∆d = d chain AD – d chain BC) finely correlates with the movement of the MoS2 fragment (Fig 5e-g). Here, the distance of the EL1 segments from chain A and D (d chain AD) are expanded by the MoS2 during the first 6 ns, and then, the dchainBC becomes larger after 25 ns, that is, when the MoS2 nanoflake slides into the site and turns to interact with EL1 segments from chain B and C.

An energy analysis (Table 1) indicates that the total average interaction energy between the Kir3.2 channel and the MoS2 nanoflake is -59.34 kcal/mol, with a vdW and Coulombic contributions of -49.82 kcal/mol and -9.52 kcal/mol, respectively. Compared to the MoS2/KcsA system, the interaction between MoS2 and Kir3.2 is slightly smaller with a ~5 kcal/mol difference with respect to the KcsA case, which is consistent with the values for their contacting atom pair numbers (546 vs 483, see also Table 1). Moreover, as in the case of MoS2/KcsA system, the analysis of the contact residue types for the MoS2/Kir3.2 binding mode showed that hydrophobic/aromatic residues are also the predominant ones. However, contrary to the MoS2/KcsA case, where the hydrophilic and charged residues only contribute to ~6% and 26% respectively, the MoS2-Kir3.2 binding complex shows that hydrophilic and charged residue each account for ~30% of total contributions. This significant contribution is mainly due to the large distribution of hydrophilic residues in the extracellular surface ACS Paragon Plus Environment

ACS Nano 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 12 of 38

of the Kir3.2 channel (Figure S2). Calculation of the electrostatic interactions using a continuum electrostatic model (solving the Poisson-Boltzmann equation with the Adaptive Poisson-Boltzmann Solver, APBS)49 also illustrates the presence of more charges/hydrophilicity residues on the extracellular surface of the Kir3.2 than in the KcsA channel (Figure 1c).

Kv1.2 Chimera Channel: MoS2 Binds at the Voltage Sensor Domain In the simulations of the MoS2/Kv1.2 chimera system, the MoS2 fragment mainly interacts with the extracellular side of the channel’s voltage sensor domain (VSD), which involves the S1-S2 loop and the N-terminal segment of the S4 helix (Figure 6). Additionally, the loop that connects helix S5 and the Pore Helix (S5-PH loop) is also involved in MoS2 binding. The S5-PH loop is equivalent to the EL1 in the Kir3.2 and KcsA channels (herein we use EL1 for convenience). In the MoS2/Kv1.2 chimera simulations, we did not observe an obvious deleterious effect upon the MoS2 binding (Figure S3), as it was the case for both the KcsA and the Kir3.2 channels, where MoS2 either disturbs the selectivity filter (KcsA) or blocks the ion pathway (Kir3.2). We also extended one representative trajectory to 300 ns and found no obvious change versus the conformation at 150 ns (Figure S7). However, we did notice that the N-terminal segment of helix S4 closely contacts the edge of the MoS2 nanoflake; the helix S4 residues involved in the interaction (define as residues within 6 Å to the MoS2) are L280, Q283, and R286 (Figure 7). Helix S4 in the VSD, has been suggested to be closely involved in the gating process of Kv channels

38-41

.

Specifically, seven positively charged residues (Arg or Lys) located in helix S4 of the Kv1.2 chimera channel, including the aforementioned MoS2-contacting R286, are responsible for sensing voltage changes between the inner and outer cellular membrane.

Furthermore,

helix

S4

has

been

directly

implicated

in

the

electromechanical coupling mechanism that gates the channel in response to changes in transmembrane voltage50-53. In the presence of MoS2, the normal movement of helix S4, driven by the change in the membrane voltage, may be hampered as its N-terminus becomes engaged in interactions with the MoS2 nanoflake. By monitoring ACS Paragon Plus Environment

Page 13 of 38 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 Nano

the minimal distances between residues in the S4 N-terminus and the MoS2 fragment, we concluded that the MoS2 binding pose is very stable for the length of our simulations (see Figure 7). The mechanism involved in the MoS2/Kv1.2 chimera binding mode is illustrated in Figure 7a. The MoS2 binding event occurs at early stages (~14 ns) during the trajectory and remains stable for the rest of the simulation time, as suggested by the parameters in Figure 7b.

We conducted the MD simulations based on the open channel structure, where the S4 helix stretches towards the extracellular side of the protein. It has been proposed that in the closed state of the Kv channels, the S4 helix would move downward towards the intracellular side of the membrane, hence burying its positively charge residues deep inside the membrane.41 We speculate that the binding of the MoS2 nanoflake to the extracellular side of the protein, mediated by interactions with residues located at the S4 N-terminal segment, might inhibit the downward motion of helix S4 that is required during the gating process of the channel from its open to closed state. Thus far, no structural information is available for the closed conformations of the Kv channels. How the MoS2 nanoflake would interact with the closed state of the Kv channels, where the N-terminal segment of S4 is tentatively less accessible, remains unknown until additional structural information becomes available.

The average total interaction energy between the Kv1.2 chimera channel and MoS2 is -82.78 kcal/mol, with a vdW and Coulombic contributions of -67.40 kcal/mol and -15.38 kcal/mol, respectively (Table 1). From the three transmembrane systems investigated here, the MoS2 displays the strongest interaction for the Kv1.2 chimera channel, which is consequence of the largest contacting atom pair number of this system,

~692.

From

the

contact

residues,

the

Kv1.2

chimera

has

10

hydrophobic/aromatic, 2 hydrophilic, and 6 charged residues, again indicating a predominant involvement of hydrophobic residues (similar ratio to the KcsA system). Lastly, our analysis of the electrostatic interactions corroborates the strong preference of MoS2 to interact with extracellular regions largely populated by non-charged ACS Paragon Plus Environment

ACS Nano 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 14 of 38

residues (see white surface areas in Figure 1c).

PMF Calculations The potential of mean force (PMF) calculations combined with umbrella sampling were conducted to measure the binding free energies of MoS2 nanoflake to the K channels. The results indicate that the MoS2 nanoflake binding free energies are -163.68 kJ/mol to the KcsA, -78.54 kJ/mol to the Kir3.2, -131.82 kJ/mol to the Kv1.2 chimera (Figure 8). Thus, binding to any of the K+ channel studied provide an energetically favorable state for the MoS2 nanoflakes.

Next, we asked if MoS2 nanoflakes would bind to all K+ channels. Identifying a negative control would confirm that the experiments described were not influenced by a systematic artifact. We studied the interaction between the MoS2 nanoflakes and the structurally distinct K+ channel, K2P2 using the same protocol. The 150 ns-long MD simulation shows that MoS2 interacts with the extracellular top of K2P2 without changing the global or local structure of the channel (Figure S8). The large and rigid extracellular domain of K2P2, called the cap-domain, protects the channel from disturbance by the nanoflakes. The cap-domain itself shows strong hydrophilicity and mainly negative charges distribute in its surface (Figure S9-10), implying a disadvantage of binding with MoS2 nanoflakes. Furthermore, compared to the other K+ channels studied, the PMF calculation indicates a significantly weaker binding affinity of -39.23 kJ/mol (Figure 8). Of note, the electrophysiology experiments further confirm that MoS2 nanomaterials does not influence the normal function of K2P2 channels (see below).

Electrophysiology results Next, we tested the effect of MoS2 nanoflakes on Kir3.2 and Kv1.2 channels expressed in Xenopus oocytes by TEVC.

Kir3.2 is an inwardly-rectifying K+

channel and thus passes limited outward currents at membrane potentials more positive than the equilibrium potential for K+ ions (EK). ACS Paragon Plus Environment

Here, we used

Page 15 of 38 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 Nano

Kir3.2-E152D (Kir3.2*), a variant that passes larger currents54.

Inward currents

mediated by Kir3.2* are assessed at -80 mV and become larger as the concentration of external K+ is raised and EK moves towards 0 mV (Fig. 9A-B).55

Consistent with

expectations from the MD simulations, MoS2 nanoflakes blocked Kir3.2* channels. The effect was concentration dependent, with an IC50 of 3.3 ± 1 µg/ml.

Greater

than >98% of Kir3.2* current was blocked when the concentration of MoS2 nanoflakes reached 1000 µg/ml (Fig. 9C). voltage-gated K+ channel, Kv1.2 (Fig. 9D-E).

MoS2 nanoflakes also blocked the Kv1.2 channels pass outward K+

currents that increase in magnitude with membrane depolarization.

As with Kir3.2*,

block of Kv1.2 was concentration-dependent, with an IC50 of 392 ± 7 µg/ml. Although the IC50 of block for Kv1.2 was ~ 10 time high than for Kir3.2*, the concentration-dependence was steeper, such that ~95% of the Kv1.2-mediated current was also ablated by 1000 µg/ml MoS2 nanoflakes (Fig. 9F).

In contrast to the effect on Kir3.2 and Kv1.2, MoS2 nanoflakes did not block K2P2, a background K+ channel that gates independently of the membrane potential and thus, passes currents across the physiological voltage-range.56

Outward K2P2 channel

currents were assessed at 40 mV and remained unaltered by application of MoS2 nanoflakes at up to 2000 µg/ml (Fig. 9G-H).

To confirm that the currents under

investigation were mediated by K2P2 channels, ~90% of the current was blocked channels by 3 mM BaCl2, in accord with work published by others.57

CONCLUSION Our hundreds of nanosecond-length MD simulations indicate that the MoS2 nanoflake is capable of binding to the three representative K+ channels, KcsA, Kir3.2, and Kv1.2 chimera, through different binding modes, which may lead to different effects in the channel’s biological functions.

ACS Paragon Plus Environment

ACS Nano 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

In the case of the KcsA channel, the MoS2 fragment interacts with the EL1 and EL2 segments, and with the Pore Helices from adjacent subunits, which indirectly disrupts the delicate SF structure that controls the selective permeation of K+ ions through the channel. The binding of the MoS2 nanoflake have deleterious effects in the normal ion selectivity function of the KcsA channel, dilating the SF structure and increasing the flux of K+ ions.

In the case of Kir3.2 channel, the MoS2 nanoflake prefers to interact with the extracellular EL1 loops from two or three different subunits. The direct consequence of the simultaneous interaction is that the MoS2 fragment entirely covers the channel pore, which probably blocks the K+ ion influx. We noticed that the identity and long extension of the EL1 loop in the Kir3.2 channel offers suitable flexibility to accommodate the MoS2 nanoflake in a “buckle up” manner.

For the Kv1.2 chimera channel, the MoS2 triangle binds onto an extracellular surface crevice constituted by the S1-S2 loop and the N-terminal segment of helix S4. We propose that MoS2 might delay or disturb the normal gating process of the channel from its open to closed states because the intimate contact with S4 would affect the mobility of this important helix, which has been related to the channel’s gating process through the S4-S5 linker. In accord with our simulations, electrophysiological studies showed that both Kir3.2* and Kv1.2 channels are blocked by the MoS2 nanoflakes in a concentration-dependent way with IC50 values of 3.3 ± 1 µg/ml and 392 ± 7 µg/ml, respectively.

Despite distinctive binding modes between the MoS2 nanoflake and the K+ channels, the underlying binding mechanism in the three systems is unified by inherent general properties: (i) MoS2 prefers to bind at regions with the minimum number of charged residues, particularly the negatively charged residues that are relatively abundant at the channel’s extracellular side. Such predominantly hydrophobic surface patches are usually not found near the central pore domain, thus for the KcsA and Kv1.2 chimera ACS Paragon Plus Environment

Page 16 of 38

Page 17 of 38 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 Nano

cases, the MoS2 fragment interacts with peripheral regions that surround the channel’s central pore. While for the Kir3.2 channel, where positively charged residues are more distributed on the extracellular surface, the closest contacting points are concentrated on the relatively charge-neutral EL1 loops. Nevertheless, the spatial compatibility between the MoS2 nanoflake and the Kir3.2 extracellular surface also play a significant role in their binding mode. (ii) MoS2 binds to the three channels with similar interaction energy magnitudes with a dominant contribution from the vdW energies (contributing ~80 to 84% to the total energy). Moreover, consistent with vdW predominance, hydrophobic residues account for most of the MoS2 contacting residues, especially in the case of the KcsA (~67%) and Kv1.2 chimera (~56%) systems. This trend of hydrophobic contribution is relatively lower in case of the Kir3.2 channel (~39%), since its extracellular distribution of non-hydrophobic residues is larger than the other two K+ channels. The general features identified in the binding modes of the MoS2 nanoflake/K+ channel systems are in good agreement with the intrinsic chemical properties of the MoS2 nanomaterial, which has been found to exhibit a strong hydrophobic character. 47-48, 58 In practice, MoS2 nanoflakes with other sizes (e.g. 4 to 12 Mo atoms on one side of the triangle) may all exist and they might also show some specific binding modes depending on the exact size; however, we believe the underlying driving forces and basic binding features derived from the current MoS2 with 9 Mo atoms on one side should still hold.

Despite the diversity and variability of the extracellular side of the K+ channels investigated here, we observed stable binding modes for the MoS2 fragment that causes significant and deleterious structural changes, and concomitantly, abnormal function of the K+ channels. We suggest that the MoS2 biocompatibility may be significantly compromised and thus, this investigation provides mechanistic insights into the potential toxic properties of the MoS2 nanomaterial via a family of important ion channels. Some other studies also found that the interactions of nanomaterials and channel proteins can cause interference with ion channel activities, such as interactions of C60 with K+ channels59-60 and gold nanoparticles with the hERG ACS Paragon Plus Environment

ACS Nano 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

channel.61 Moreover, a chitosan (CS) functionalized MoS2 nanostructure was found to serve as a antimicrobial agent for combating bacterial infections.62 Based on our data, we speculate that the interaction of MoS2 with microbial ion channels (e.g. KcsA in this study) might affect the natural intracellular ionic equilibrium and disrupt the physiology of the organism.

One the other hand, compared the KcsA, Kir3.2 and Kv channels, the TREK-1 channel is insusceptible to MoS2 nanoflakes due to its large and rigid extracellular domain (Figure S8), which extend our understanding on the interactions between membrane proteins and MoS2 nanoflakes. It is the MoS2’s intrinsic chemical properties, and the channels’ specific features, such as the electrostatic character and complex surface architecture, to determine the critical details of the binding events and specifically influence the function of the channels. .

METHODS Molecular dynamics simulation. The MoS2 nanoflake was built as an equilateral triangle with side length of 2.86 nm and with force field parameters obtained from a refitted work58 based on the experimental monolayered water contact angle63 (force field parameters are listed in Table S1). The data obtained here validate the refitted parameters. The refit is also supported by our previous study64 that focused on the comparison of this refitted parameter and another force field (which had higher vdW contributions and was also recognized by other studies65-67). More importantly, the refitted force field provides a better consideration of water molecules in bio-simulations. Three K+ channels, KcsA, Kir3.2 and Kv1.2 chimera, were obtained from RCSB Protein Data Bank (PDB code: 3FB7,38 3SYA40 and 2R9R,41 respectively). Each channel was immersed in a palmitoyloleoylphosphatidylcholine (POPC) bilayer generated by the CHARMM-GUI v1.7 (http://www.charmm-gui.org). Three simulations for the MoS2/KcsA, MoS2/Kir3.2, and MoS2/Kv1.2 chimera systems were set up with box sizes of 9.98×9.98×9.94 nm3, 14.96×14.96×16.84

ACS Paragon Plus Environment

Page 18 of 38

Page 19 of 38 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 Nano

nm3 and 14.90×14.90×9.92 nm3, respectively. The MoS2 nanoflake was placed above each protein channels with a minimum distance of 1.2 nm as shown in Figure S4. Water molecules were then included in the systems and those located inside the lipid bilayer were removed. Finally a total 19205, 89639 and 41850 water molecules were included in the MoS2/KcsA, MoS2/Kir3.2, and MoS2/Kv1.2 chimera systems, respectively. K+ and Cl- ions were included so as to emulate physiological conditions (KCl solution 0.15 M). The solvated systems were then investigated by atomistic Molecular Dynamics (MD) simulations.68-73 In addition, control simulations, where the MoS2 nanoflake was not included in the system, were also set up and conducted following similar simulation conditions. A negative control system that MoS2 nanoflake had no functional effect on a K+ channel – the K2P2 channel (PDB code: 5VK542) – was also conducted, in which its box size was 11.08×11.08×13.57 nm3 including 38150 water molecules.

GROMACS software package (version 4.6.6) was used to perform MD simulations74 using the CHARMM36 force field75 and the TIP3P water model.76 The K+ ion used the standard parameters in CHARMM force field. Periodic boundary conditions (PBC) were applied in all directions. The systems’ temperature was kept at 300 K using a v-rescale thermostat77 and pressure was maintained at 1 atm by coupling the semi-isotropic (X+Y, Z) directions of the system using the Parrinello-Rahman’s algorithm.78 The van der Waals (vdW) interactions were computed with a cutoff distance of 1.2 nm while long-range electrostatic interactions were handled with the particle mesh Ewald (PME) method.79 Water molecules were constrained by the SETTLE algorithm80 and solute hydrogen bonds were constrained to their equilibrium values employing the LINCS algorithm.81 In production runs, a 2 ps time step was used and the coordinates were saved every 20 ps. The 0.1 V/nm electric field was applied along z coordinate (perpendicular to the membrane) in all simulations, which generated ~460 mV membrane potential (with a ~4.6 nm thickness of the POPC bilayer). Three independent (150 ns in length) simulations were carried out for each system. One representative trajectory of every system was extended to 300 ns to ACS Paragon Plus Environment

ACS Nano 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 20 of 38

verify the binding stability.

Potential of mean force (PMF). Umbrella sampling simulations were conducted to calculate the PMF values of MoS2 nanoflakes along the directions against the binding sites in the K+ channels.82-84 The distance (d) to its binding site was restrained at a reference distance (d0) with a harmonic force F = k × (d - d0) where k is the force constant (2000 kJ mol−1 nm−2). The spacing of the sampling windows was 0.05 nm. An equilibration of 2 ns followed by a 10-ns production run was conducted for each system at d0. The free energy profiles were generated using the Weighted Histogram Analysis Method.85-87

Electrophysiology Xenopus laevis oocyte expression.

Mouse Kir3.2-E152D, rat Kv1.2 and mouse

K2P2 were handled in pGEMHE, linearized using the Nhe 1 restriction enzyme, and in vitro transcribed using the mMessage mMachine® Kit (Ambion).

The

concentration of complementary RNA (cRNA) was quantified by optical density. Xenopus oocytes were surgically extracted, dissociated and defolliculated by collagenase treatment, and microinjected with 50 nl of a water solution containing the desired cRNAs.

To study Kir3.2* and Kv1.2, 2 ng of cDNA was injected per oocyte,

for K2P2, 5 ng was injected.

In all cases, oocytes were incubated for 2 to 4 days at

18°C.

Two-electrode voltage-clamp and data analysis.

Whole-oocyte currents were

measured by two-electrode voltage clamp (TEVC) with GeneClamp 500 (Molecular Devices), or TEC-03X (NPI) amplifiers.

Electrodes were pulled using a

Flaming-Brown micropipette puller (Sutter Instruments) and were filled with 3 M KCl in 1.5% (w/v) agarose to give resistances between 0.5 and 1.0 MΩs.

ACS Paragon Plus Environment

The

Page 21 of 38 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 Nano

oocytes were bathed in ND96 recording solution comprising, in mM: KCl 2, NaCl 96, MgCl2 1 and HEPES 5, buffered to pH 7.4 with KOH.

Where indicated, Kir3.2

channel currents were assessed in a high K+ recording solution comprising, in mM: KCl 96, NaCl 2, MgCl2 1 and HEPES 5, buffered to pH 7.4 with KOH.

Currents

were digitized using a USB interface (National Instruments) and recorded using WinWCP software (University of Strathclyde).

To study Kir3.2, oocytes were held

at 0 mV, and currents were assessed by 100 ms ramps from −80 to +80 mV that were repeated every second.

The effect of MoS2 nanoflakes was determined at −80 mV.

Cells expressing Kv1.2 or K2P2 were held at -80 mV, and the currents were evoked by 200- or 500-ms step depolarizations between -80 and 80 mV, as indicated.

Block

by MoS2 nanoflakes, or BaCl2 was studied by pulses to 40 mV repeated every 5 seconds.

Block was expressed as the percent-current block normalized to the

maximum current. Between 8-12 oocytes from different Xenopus frogs were studied per experiment.

Data were analyzed using Clampfit 10 (Molecular Devices) and

Origin (Microcal) software.

MoS2 nanoflakes.

The MoS2 flakes were prepared by lithium intercalation.

Typically, 1 g MoS2 (Sigma) is added into the flask which has 10 mL butyl-lithium (1.6 mol/L in hexane). The flask is always filled with argon gas. After immersing 2 days, the mixture was filtered and washed by hexane to remove excess residues. The intercalated powder was then exfoliated in water with 8.2 mg/mL. For electrophysiology studies, MoS2 nanoflakes were diluted in an aqueous buffer that contained the constituents of the recording solutions at concentrations such that the final ionic composition of the appropriate recording solution was not altered by the addition of the nanoflakes.

Each solution was sonicated for 30 minutes prior to use.

Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: The structural information of potassium channels, distances variation of Selectivity ACS Paragon Plus Environment

ACS Nano 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

Filter, simulation setups, extending simulations of three trajectories, final configurations of MoS2/TREK-1 simulations, electrostatic surface of TREK-1 and force field parameters of MoS2.

Acknowledgements The authors gratefully acknowledge the help from Weifeng Li and Zaixing Yang. We are also grateful to Heikki Vaananen for his technical support throughout this project. This work was partially supported by the National Natural Science Foundation of China under Grant Nos. 11374221, 11574224, and 21503140. RZ acknowledges the support from IBM Blue Gene Science Program (W125859, W1464125, W1464164). A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and Jiangsu Provincial Key Laboratory of Radiation Medicine and Protection. MD acknowledges the funding from Danish National Research Foundation and Aarhus Universitets Forskningsfond,

Author Contributions R.Z. and X.Y.M. conceived and designed the research. Z.G. and X.Y.M. carried out the molecular dynamics simulations. L.P. and D.E.L. designed the electrophysiology experiments and analyzed the data. Z.G., X.Y.M., and R.Z. analyzed the simulation data. Z.G., L.P., X.Y.M., J.M.P-A., D.E.L. and R.Z. co-wrote the manuscript. Z.W. and M.D. synthesized MoS2 nanoflakes. All authors discussed the results and commented on the manuscript.

Additional Information The authors declare no competing financial interests.

ACS Paragon Plus Environment

Page 22 of 38

Page 23 of 38 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 Nano

REFERENCES 1.

Novoselov, K. S.; Geim, A. K.; Morozov, S.; Jiang, D.; Zhang, Y.; Dubonos, S. a.; Grigorieva, I.;

Firsov, A., Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666-669. 2.

Iijima, S., Helical Microtubes of Graphitic Carbon. Nature 1991, 354, 56-58.

3.

Kroto, H. W.; Heath, J. R.; Obrien, S. C.; Curl, R. F.; Smalley, R. E., C60 : Buckminsterfullerene.

Nature 1985, 318, 162-163. 4.

Feng, L.; Liu, Z., Graphene in Biomedicine: Opportunities and Challenges. Nanomedicine 2011, 6,

317-324. 5.

Geim, A. K., Graphene: Status and Prospects. Science 2009, 324, 1530-1534.

6.

Sanchez, V. C.; Jachak, A.; Hurt, R. H.; Kane, A. B., Biological Interactions of Graphene-Family

Nanomaterials: An Interdisciplinary Review. Chem. Res. Toxicol. 2012, 25, 15-34. 7.

Cha, C.; Shin, S. R.; Annabi, N.; Dokmeci, M. R.; Khademhosseini, A., Carbon-Based

Nanomaterials: Multifunctional Materials for Biomedical Engineering. ACS Nano 2013, 7, 2891-2897. 8.

Lee, J. S.; Joung, H.-A.; Kim, M.-G.; Park, C. B., Graphene-Based Chemiluminescence

Resonance Energy Transfer for Homogeneous Immunoassay. ACS Nano 2012, 6, 2978-2983. 9.

Bao, H.; Pan, Y.; Ping, Y.; Sahoo, N. G.; Wu, T.; Li, L.; Li, J.; Gan, L. H.,

Chitosan-Functionalized Graphene Oxide as a Nanocarrier for Drug and Gene Delivery. Small 2011, 7, 1569-1578. 10. Li, B.; Cheng, Y.; Liu, J.; Yi, C.; Brown, A. S.; Yuan, H.; Tuan, V.-D.; Fischer, M. C.; Warren, W. S., Direct Optical Imaging of Graphene in Vitro by Nonlinear Femtosecond Laser Spectral Reshaping. Nano Lett. 2012, 12, 5936-5940. 11. Li, M.; Yang, X.; Ren, J.; Qu, K.; Qu, X., Using Graphene Oxide High Near-Infrared Absorbance for Photothermal Treatment of Alzheimer's Disease. Adv. Mater. 2012, 24, 1722-1728. 12. Yang, K.; Wan, J.; Zhang, S.; Tian, B.; Zhang, Y.; Liu, Z., The Influence of Surface Chemistry and Size of Nanoscale Graphene Oxide on Photothermal Therapy of Cancer Using Ultra-Low Laser Power. Biomaterials 2012, 33, 2206-2214. 13. Yang, Z.; Kang, S.-g.; Zhou, R., Nanomedicine: De Novo Design of Nanodrugs. Nanoscale 2014, 6, 663-677. 14. Shao, Y.; Wang, J.; Wu, H.; Liu, J.; Aksay, I. A.; Lin, Y., Graphene Based Electrochemical Sensors and Biosensors: A Review. Electroanal. 2010, 22, 1027-1036. 15. Late, D. J.; Liu, B.; Matte, H. S. S. R.; Dravid, V. P.; Rao, C. N. R., Hysteresis in Single-Layer MoS2 Field Effect Transistors. ACS Nano 2012, 6, 5635-5641. 16. Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A., Single-Layer MoS2 Transistors. Nat. Nanotechnol. 2011, 6, 147-150. 17. Ataca, C.; Ciraci, S., Dissociation of H2O at the Vacancies of Single-Layer MoS2. Phys. Rev. B 2012, 85, 195410. 18. Jaramillo, T. F.; Jorgensen, K. P.; Bonde, J.; Nielsen, J. H.; Horch, S.; Chorkendorff, I., Identification of Active Edge Sites for Electrochemical H-2 Evolution from MoS2 Nanocatalysts. Science 2007, 317, 100-102. 19. Li, W.; Yang, Y.; Weber, J. K.; Zhang, G.; Zhou, R., Tunable, Strain-Controlled Nanoporous MoS2 Filter for Water Desalination. ACS Nano 2016, 10, 1829-1835. 20. Li, Y.; Wang, H.; Xie, L.; Liang, Y.; Hong, G.; Dai, H., MoS2 Nanoparticles Grown on Graphene:

ACS Paragon Plus Environment

ACS Nano 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 24 of 38

An Advanced Catalyst for the Hydrogen Evolution Reaction. J. Am. Chem. Soc. 2011, 133, 7296-7299. 21. Shanmugam, M.; Bansal, T.; Durcan, C. A.; Yu, B., Molybdenum Disulphide/Titanium Dioxide Nanocomposite-Poly 3-Hexylthiophene Bulk Heterojunction Solar Cell. Appl. Phys. Lett. 2012, 100, 153901. 22. Liu, T.; Shi, S.; Liang, C.; Shen, S.; Cheng, L.; Wang, C.; Song, X.; Goel, S.; Barnhart, T. E.; Cai, W., et al., Iron Oxide Decorated MoS2 Nanosheets with Double Pegylation for Chelator-Free Radio Labeling and Multimodal Imaging Guided Photothermal Therapy. ACS Nano 2015, 9, 950-960. 23. Liu, T.; Wang, C.; Cui, W.; Gong, H.; Liang, C.; Shi, X.; Li, Z.; Sun, B.; Liu, Z., Combined Photothermal and Photodynamic Therapy Delivered by Pegylated MoS2 Nanosheets. Nanoscale 2014, 6, 11219-11225. 24. Yang, X.; Li, J.; Liang, T.; Ma, C.; Zhang, Y.; Chen, H.; Hanagata, N.; Su, H.; Xu, M., Antibacterial Activity of Two-Dimensional MoS2 Sheets. Nanoscale 2014, 6, 10126-10133. 25. Wang, S.; Li, K.; Chen, Y.; Chen, H.; Ma, M.; Feng, J.; Zhao, Q.; Shi, J., Biocompatible Pegylated Mos2 Nanosheets: Controllable Bottom-up Synthesis and Highly Efficient Photothermal Regression of Tumor. Biomaterials 2015, 39, 206-217. 26. Yin, W.; Yan, L.; Yu, J.; Tian, G.; Zhou, L.; Zheng, X.; Zhang, X.; Yong, Y.; Li, J.; Gu, Z., et al., High-Throughput

Synthesis

of

Single-Layer

Mos2

Nanosheets

as

a

Near-Infrared

Photothermal-Triggered Drug Delivery for Effective Cancer Therapy. ACS Nano 2014, 8, 6922-6933. 27. Wang, L.; Wang, Y.; Wong, J. I.; Palacios, T.; Kong, J.; Yang, H. Y., Functionalized MoS2 Nanosheet-Based Field-Effect Biosensor for Label-Free Sensitive Detection of Cancer Marker Proteins in Solution. Small 2014, 10, 1101-1105. 28. Zhu, C.; Zeng, Z.; Li, H.; Li, F.; Fan, C.; Zhang, H., Single-Layer MoS2-Based Nanoprobes for Homogeneous Detection of Biomolecules. J. Am. Chem. Soc. 2013, 135, 5998-6001. 29. Teo, W. Z.; Chng, E. L. K.; Sofer, Z.; Pumera, M., Cytotoxicity of Exfoliated Transition‐Metal Dichalcogenides (MoS2, WS2, and WSe2) Is Lower Than That of Graphene and Its Analogues. Chem.-A Eur. J. 2014, 20, 9627-9632. 30. Chng, E. L. K.; Sofer, Z.; Pumera, M., MoS2 Exhibits Stronger Toxicity with Increased Exfoliation. Nanoscale 2014, 6, 14412-14418. 31. Hao, J.; Song, G.; Liu, T.; Yi, X.; Yang, K.; Cheng, L.; Liu, Z., In Vivo Long‐ Term Biodistribution, Excretion, and Toxicology of Pegylated Transition‐Metal Dichalcogenides MS2 (M= Mo, W, Ti) Nanosheets. Adv. Sci. 2017, 4, 1600160. 32. Chen, Y.; Tan, C.; Zhang, H.; Wang, L., Two-Dimensional Graphene Analogues for Biomedical Applications. Chem. Soc. Rev. 2015, 44, 2681-2701. 33. Lauritsen, J. V.; Kibsgaard, J.; Helveg, S.; Topsøe, H.; Clausen, B. S.; Lægsgaard, E.; Besenbacher, F., Size-Dependent Structure of MoS2 Nanocrystals. Nat. Nanotechnol. 2007, 2, 53-58. 34. van der Cruijsen, E. A. W.; Nand, D.; Weingarth, M.; Prokofyev, A.; Hornig, S.; Cukkemane, A. A.; Bonvin, A. M. J. J.; Becker, S.; Hulse, R. E.; Perozo, E., et al., Importance of Lipid-Pore Loop Interface for Potassium Channel Structure and Function. P. Natl. Acad. Sci. USA 2013, 110, 13008-13013. 35. Lange, A.; Giller, K.; Hornig, S.; Martin-Eauclaire, M. F.; Pongs, O.; Becker, S.; Baldus, M., Toxin-Induced Conformational Changes in a Potassium Channel Revealed by Solid-State Nmr. Nature 2006, 440, 959-962. 36. Ader, C.; Schneider, R.; Hornig, S.; Velisetty, P.; Wilson, E. M.; Lange, A.; Giller, K.; Ohmert, I.; Martin-Eauclaire, M.-F.; Trauner, D., et al., A Structural Link between Inactivation and Block of a K+

ACS Paragon Plus Environment

Page 25 of 38 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 Nano

Channel. Nat. Struct. Mol. Biol. 2008, 15, 605-612. 37. Schneider, R.; Ader, C.; Lange, A.; Giller, K.; Hornig, S.; Pongs, O.; Becker, S.; Baldus, M., Solid-State Nmr Spectroscopy Applied to a Chimeric Potassium Channel in Lipid Bilayers. J. Am. Chem. Soc. 2008, 130, 7427-7435. 38. Cuello, L. G.; Jogini, V.; Cortes, D. M.; Pan, A. C.; Gagnon, D. G.; Dalmas, O.; Cordero-Morales, J. F.; Chakrapani, S.; Roux, B.; Perozo, E., Structural Basis for the Coupling between Activation and Inactivation Gates in K+ Channels. Nature 2010, 466, 272-275. 39. Cuello, L. G.; Jogini, V.; Cortes, D. M.; Perozo, E., Structural Mechanism of C-Type Inactivation in K+ Channels. Nature 2010, 466, 203-208. 40. Whorton, M. R.; MacKinnon, R., Crystal Structure of the Mammalian Girk2 K+ Channel and Gating Regulation by G Proteins, Pip2, and Sodium. Cell 2011, 147, 199-208. 41. Long, S. B.; Tao, X.; Campbell, E. B.; MacKinnon, R., Atomic Structure of a Voltage-Dependent K+ Channel in a Lipid Membrane-Like Environment. Nature 2007, 450, 376-383. 42. Lolicato, M.; Arrigoni, C.; Mori, T.; Sekioka, Y.; Bryant, C.; Clark, K. A.; Minor, D. L., K2P2. 1 (Trek-1)–Activator Complexes Reveal a Cryptic Selectivity Filter Binding Site. Nature 2017, 547, 364-368. 43. Kopfer, D. A.; Song, C.; Gruene, T.; Sheldrick, G. M.; Zachariae, U.; de Groot, B. L., Ion Permeation in K+ Channels Occurs by Direct Coulomb Knock-On. Science 2014, 346, 352-355. 44. Wylie, B. J.; Bhate, M. P.; McDermott, A. E., Transmembrane Allosteric Coupling of the Gates in a Potassium Channel. P. Natl. Acad. Sci. USA 2014, 111, 185-190. 45. Valiyaveetil, F. I.; Zhou, Y.; MacKinnon, R., Lipids in the Structure, Folding, and Function of the Kcsa K+ Channel. Biochemistry 2002, 41, 10771-10777. 46. Weingarth, M.; Prokofyev, A.; van der Cruijsen, E. A.; Nand, D.; Bonvin, A. M.; Pongs, O.; Baldus, M., Structural Determinants of Specific Lipid Binding to Potassium Channels. J. Am. Chem. Soc. 2013, 135, 3983-3988. 47. Rafiee, J.; Mi, X.; Gullapalli, H.; Thomas, A. V.; Yavari, F.; Shi, Y.; Ajayan, P. M.; Koratkar, N. A., Wetting Transparency of Graphene. Nat. Mater. 2012, 11, 217-222. 48. Gaur, A. P. S.; Sahoo, S.; Ahmadi, M.; Dash, S. P.; Guinel, M. J. F.; Katiyar, R. S., Surface Energy Engineering for Tunable Wettability through Controlled Synthesis of MoS2. Nano Lett. 2014, 14, 4314-4321. 49. Dolinsky, T. J.; Nielsen, J. E.; McCammon, J. A.; Baker, N. A., Pdb2pqr: An Automated Pipeline for the Setup of Poisson-Boltzmann Electrostatics Calculations. Nucleic Acids Res. 2004, 32, 665-667. 50. Long, S. B.; Campbell, E. B.; MacKinnon, R., Crystal Structure of a Mammalian Voltage-Dependent Shaker Family K+ Channel. Science 2005, 309, 897-903. 51. Long, S. B.; Campbell, E. B.; MacKinnon, R., Voltage Sensor of Kv1.2: Structural Basis of Electromechanical Coupling. Science 2005, 309, 903-908. 52. Aggarwal, S. K.; MacKinnon, R., Contribution of the S4 Segment to Gating Charge in the Shaker K+ Channel. Neuron 1996, 16, 1169-1177. 53. Larsson, H. P.; Baker, O. S.; Dhillon, D. S.; Isacoff, E. Y., Transmembrane Movement of the Shaker K+ Channel S4. Neuron 1996, 16, 387-397. 54. Yi, B. A.; Lin, Y. F.; Jan, Y. N.; Jan, L. Y., Yeast Screen for Constitutively Active Mutant G Protein-Activated Potassium Channels. Neuron 2001, 29, 657-67. 55. Logothetis, D. E.; Mahajan, R.; Adney, S. K.; Ha, J.; Kawano, T.; Meng, X. Y.; Cui, M., Unifying Mechanism of Controlling Kir3 Channel Activity by G Proteins and Phosphoinositides. Int. Rev.

ACS Paragon Plus Environment

ACS Nano 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

Neurobiol. 2015, 123, 1-26. 56. Honore, E., The Neuronal Background K2P Channels: Focus on Trek1. Nat. Rev. Neurosci. 2007, 8, 251-61. 57. Ma, X. Y.; Yu, J. M.; Zhang, S. Z.; Liu, X. Y.; Wu, B. H.; Wei, X. L.; Yan, J. Q.; Sun, H. L.; Yan, H. T.; Zheng, J. Q., External Ba2+ Block of the Two-Pore Domain Potassium Channel Trek-1 Defines Conformational Transition in Its Selectivity Filter. J. Biol. Chem. 2011, 286, 39813-22. 58. Luan, B.; Zhou, R., Wettability and Friction of Water on a MoS2 Nanosheet. Appl. Phys. Lett. 2016, 108, 131601. 59. Kraszewski, S.; Tarek, M.; Treptow, W.; Ramseyer, C., Affinity of C60 Neat Fullerenes with Membrane Proteins: A Computational Study on Potassium Channels. ACS Nano 2010, 4, 4158-4164. 60. Calvaresi, M.; Furini, S.; Domene, C.; Bottoni, A.; Zerbetto, F., Blocking the Passage: C-60 Geometrically Clogs K+ Channels. ACS Nano 2015, 9, 4827-4834. 61. Leifert, A.; Pan, Y.; Kinkeldey, A.; Schiefer, F.; Setzler, J.; Scheel, O.; Lichtenbeld, H.; Schmid, G.; Wenzel, W.; Jahnen-Dechent, W., et al., Differential Herg Ion Channel Activity of Ultrasmall Gold Nanoparticles. P. Natl. Acad. Sci. USA 2013, 110, 8004-8009. 62. Zhang, W.; Shi, S.; Wang, Y.; Yu, S.; Zhu, W.; Zhang, X.; Zhang, D.; Yang, B.; Wang, X.; Wang, J., Versatile Molybdenum Disulfide Based Antibacterial Composites for in Vitro Enhanced Sterilization and in Vivo Focal Infection Therapy. Nanoscale 2016, 8, 11642-11648. 63. Gaur, A. P. S.; Sahoo, S.; Ahmadi, M.; Dash, S. P.; Guinel, M. J. F.; Katiyar, R. S., Surface Energy Engineering for Tunable Wettability through Controlled Synthesis of MoS2. Nano Lett. 2014, 14, 4314-4321. 64. Gu, Z.; De Luna, P.; Yang, Z.; Zhou, R., Structural Influence of Proteins Upon Adsorption to MoS2 Nanomaterials: Comparison of MoS2 Force Field Parameters. Phys. Chem. Chem. Phys. 2017, 19, 3039-3045. 65. Kou, J.; Yao, J.; Wu, L.; Zhou, X.; Lu, H.; Wu, F.; Fan, J., Nanoporous Two-Dimensional MoS2 Membranes for Fast Saline Solution Purification. Phys. Chem. Chem. Phys. 2016, 18, 22210-22216. 66. Heiranian, M.; Farimani, A. B.; Aluru, N. R., Water Desalination with a Single-Layer MoS2 Nanopore. Nat. Commun. 2015, 6, 8616. 67. Farimani, A. B.; Min, K.; Aluru, N. R., DNA Base Detection Using a Single-Layer MoS2. ACS Nano 2014, 8, 7914-7922. 68. Chong, Y.; Ge, C.; Yang, Z.; Garate, J. A.; Gu, Z.; Weber, J. K.; Liu, J.; Zhou, R., Reduced Cytotoxicity of Graphene Nanosheets Mediated by Blood-Protein Coating. ACS Nano 2015, 9, 5713-5724. 69. El-Sayed, R.; Bhattacharya, K.; Gu, Z.; Yang, Z.; Weber, J. K.; Li, H.; Leifer, K.; Zhao, Y.; Toprak, M. S.; Zhou, R., Single-Walled Carbon Nanotubes Inhibit the Cytochrome P450 Enzyme, CYP3A4. Sci. Rep. 2016, 6, 21316. 70. Gu, Z.; Yang, Z.; Chong, Y.; Ge, C.; Weber, J. K.; Bell, D. R.; Zhou, R., Surface Curvature Relation to Protein Adsorption for Carbon-Based Nanomaterials. Sci. Rep. 2015, 5, 10886. 71. Gu, Z.; Zhang, Y.; Luan, B.; Zhou, R., DNA Translocation through Single-Layer Boron Nitride Nanopores. Soft Matter 2016, 12, 817-823. 72. Gu, Z. L.; Yang, Z. X.; Wang, L. L.; Zhou, H.; Jimenez-Cruz, C. A.; Zhou, R. H., The Role of Basic Residues in the Adsorption of Blood Proteins onto the Graphene Surface. Sci. Rep. 2015, 5, 11. 73. Yue, H.; Wei, W.; Gu, Z.; Ni, D.; Luo, N.; Yang, Z.; Zhao, L.; Garate, J. A.; Zhou, R.; Su, Z., et al., Exploration of Graphene Oxide as an Intelligent Platform for Cancer Vaccines. Nanoscale 2015, 7,

ACS Paragon Plus Environment

Page 26 of 38

Page 27 of 38 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 Nano

19949-19957. 74. Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E., Gromacs 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4, 435-447. 75. Klauda, J. B.; Venable, R. M.; Freites, J. A.; O'Connor, J. W.; Tobias, D. J.; Mondragon-Ramirez, C.; Vorobyov, I.; MacKerell, A. D., Jr.; Pastor, R. W., Update of the CHARMM All-Atom Additive Force Field for Lipids: Validation on Six Lipid Types. J. Phys. Chem. B 2010, 114, 7830-7843. 76. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L., Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926-935. 77. Bussi, G.; Donadio, D.; Parrinello, M., Canonical Sampling through Velocity Rescaling. J. Chem. Phys. 2007, 126, 014101. 78. Parrinello, M.; Rahman,

A., Polymorphic Transitions in Single-Crystals - a New

Molecular-Dynamics Method. J. Appl. Phys. 1981, 52, 7182-7190. 79. Darden, T.; York, D.; Pedersen, L., Particle Mesh Ewald - an N.Log(N) Method for Ewald Sums in Large Systems. J. Chem. Phys. 1993, 98, 10089-10092. 80. Miyamoto, S.; Kollman, P. A., Settle - an Analytical Version of the Shake and Rattle Algorithm for Rigid Water Models. J. Comput. Chem. 1992, 13, 952-962. 81. Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J., Lincs: A Linear Constraint Solver for Molecular Simulations. J. Comput. Chem. 1997, 18, 1463-1472. 82. Kumar, S.; Rosenberg, J. M.; Bouzida, D.; Swendsen, R. H.; Kollman, P. A., Multidimensional Free-Energy Calculations Using the Weighted Histogram Analysis Method. J. Comput. Chem. 1995, 16, 1339-1350. 83. Torrie, G. M.; Valleau, J. P., Non-Physical Sampling Distributions in Monte-Carlo Free-Energy Estimation - Umbrella Sampling. J. Comput. Phys. 1977, 23, 187-199. 84. Roux, B., The Calculation of the Potential of Mean Force Using Computer-Simulations. Comput. Phys. Commun. 1995, 91, 275-282. 85. Hub, J. S.; de Groot, B. L.; van der Spoel, D., G_Wham-a Free Weighted Histogram Analysis Implementation Including Robust Error and Autocorrelation Estimates. J. Chem. Theory Comput. 2010, 6, 3713-3720. 86. Kirkwood, J. G., Statistical Mechanics of Fluid Mixtures. J. Chem. Phys. 1935, 3, 300-313. 87. Efron, B., 1977 Rietz Lecture - Bootstrap Methods - Another Look at the Jackknife. Ann. Stat. 1979, 7, 1-26.

ACS Paragon Plus Environment

ACS Nano 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

Figure 1. (a-b) Side and top views of the three K+ channels. K+ channels are tetramers with each subunit containing two (KcsA and Kir3.2) or six (Kv1.2 chimera) transmembrane helices (TM). The different subunits are colored as: chain A in red, chain B in purple, chain C in green, and chain D in orange. The transmembrane segment is also indicated. The structure of the KcsA channel (left panels) is the simplest one and is composed by two transmembrane helices (TM1 and TM2); Kir3.2 (central panels) contains a large cytoplasmic domain in addition to the two transmembrane helices domain (TM1 and TM2); the Kv1.2 chimera (right panels) has an extra voltage sensor domain (VSD) that consists of four transmembrane helices (S1-S4) as well as a linker helix (S4-S5 linker) connecting the VSD and the central Pore Domain (S5 and S6 in Kv1.2 chimera are equivalent to TM1 and TM2 in KcsA and Kir3.2, respectively). The Extracellular Loop 1 (EL1) connects TM1 (S5 in Kv channels) and the Pore Helix (PH). In the VSD of the Kv channels, the S1-S2 loop links the TM helices S1 and S2. (see also Figure S1) (c) The electrostatic surface of the three channels, calculated by the Adaptive Poisson-Boltzmann Solver (APBS) method, is displayed.

ACS Paragon Plus Environment

Page 28 of 38

Page 29 of 38 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 Nano

Figure 2. (a) Final configurations from the three independent trajectories (run 1-3) of the MoS2/KcsA simulation system. The MoS2 nanoflake is shown as spheres (Mo and S atoms colored orange and yellow, respectively), KcsA channel in ribbon representation, and the K+ ions are shown as blue spheres. (b) Distance variation between the backbone Oxygen atoms of residue V76 from diagonal subunits, which are part of the SF of the KcsA channel; (c) Cumulative permeation events in the three trajectories.

ACS Paragon Plus Environment

ACS Nano 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

Figure 3. Sequential mechanistic processes involved in the MoS2 nanoflake binding onto the KcsA channel. The MoS2 triangle is shown in VDW sphere following the same color code as in figure 2; The KcsA structure is shown in ribbon with chain A in red, chain B in purple, chain C in green, and chain D in orange. Contact residues at each time are labeled. The residue in the selectivity filter from each of the four subunits, are also depicted as sticks.

ACS Paragon Plus Environment

Page 30 of 38

Page 31 of 38 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 Nano

Figure 4. Final configurations from three trajectories (run 1-3) for the MoS2/Kir3.2 system.

ACS Paragon Plus Environment

ACS Nano 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

Figure 5. (a-c) Sequential mechanistic processes involved in the MoS2 nanoflake binding onto the Kir3.2 channel. (d) Moving trajectory for the Center of Mass (COM) of the MoS2 nanoflake along the 150 ns simulation time using the color-code from blue (initial) to red (final). The Kir3.2 channel is rendered as a mesh surface. (e) Difference in the distance between the Ca atoms from residue P129 for the case of diagonal subunits: ∆d = dchainAD - dchainBC. The variation of ∆d between 0-10 ns (red line) and 21-31 ns (blue line) are zoomed in the (f) and (g), respectively. (h) Residues contacting the MoS2 nanoflake, are shown. Here, residues Y159, D164, and K165 are from the EL2 segment while the rest of the labelled residues are from the EL1 segment.

ACS Paragon Plus Environment

Page 32 of 38

Page 33 of 38 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 Nano

Figure 6. Final configurations from the three trajectories (run 1-3) of the MoS2/Kv1.2 chimera simulation systems.

Figure 7. (a) Sequential mechanistic processes involved in the MoS2 nanoflake binding onto the Kv1.2 chimera channel. (b) Minimal distances variation between the heavy atoms of residues L280/Q283/R286 and the MoS2 nanoflake throughout the simulation time.

ACS Paragon Plus Environment

ACS Nano 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

Figure 8. Binding free energies of MoS2 on the KcsA, Kir3.2, Kv1.2 chimera and TREK-1 channels based on the PMF calculations.

ACS Paragon Plus Environment

Page 34 of 38

Page 35 of 38 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 Nano

Figure 9. MoS2 nanoflakes block Kir3.2 and Kv1.2 but not K2P2 potassium channels. Kir3.2, Kv1.2 or K2P2 channels were expressed in Xenopus oocytes and studied by TEVC, as described in the Methods. For a-c, the E152D variant of Kir3.2 (Kir3.2*) was studied because it passes higher current than wild type Kir3.2 channels. Data are means ± SEM for 8-12 cells per study. a. An example time-course showing block of Kir3.2* currents by MoS2 nanoflakes. Inward currents were assessed by repeated depolarizations to -80 mV, at the time-point indicated by the arrow shown in (b). b. Representative traces show that inward Kir3.2* current, evoked in 2 mM K+ (black), is increased in 96 mM external K+ (gray) and blocked by application of 10 µg/ml (green) and then 100 µg/ml (blue) MoS2 nanoflakes. The arrow indicates where currents were assessed for the time-course studies in (a). c. The concentration-response curve of Kir3.2* channels blocked by MoS2 nanoflakes. The data

ACS Paragon Plus Environment

ACS Nano 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

d. e.

f. g. h.

i.

are normalized and fit with the Hill equation; the mean IC50 is 3.3 ± 1 µg/ml. An example time-course showing block of Kv1.2 channels by MoS2 nanoflakes. Data are obtained from repeated step depolarizations to 40 mV. Representative current-voltage families for Kv1.2 channel evoked in response to step depolarizations between -80 and 80 mV, before (black) and after (red) application of 1000 µg/ml MoS2 nanoflakes. For the time-course studies in (d), data were collected at the time point indicated by the arrow. The concentration-response curve for Kv1.2 channel block by MoS2 nanoflakes. The data are normalized and fit with the Hill equation; the mean IC50 is 392 ± 7 µg/ml. An example time-course showing that K2P2 channel currents, evoked by repeated steps to 40 mV, are not altered by treatment with 1000 µg/ml MoS2 nanoflakes. Representative current traces for K2P2 channels, shown before (black) and after (red) block by 1000 µg/ml MoS2 nanoflakes. Currents were subsequently blocked by 3 mM BaCl2 (pink). For the time-course studies in (g), data were collected at the time point indicated by the arrow. A bar chart of the mean percent-block of K2P2 channels shows that currents are unaltered by MoS2 nanoflakes (gray). Treatment with 3 mM BaCl2 (pink), blocks 89 ± 4 % of the current.

ACS Paragon Plus Environment

Page 36 of 38

Page 37 of 38 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

ACS Nano

Table 1. Statistic on Interaction energy, binding free energy, contact pair number and contact residue type in the three simulation systems

vdW

Coulombic

Total

Contact pair number

KcsA

-52.55±3.54

-12.92±3.66

-65.47±5.63

546±42

Kir3.2

-49.82±7.54

-9.52±3.48

-59.34±9.70

483±75

11

9

8

Kv1.2

-67.40±4.07

-15.38±4.48

-82.78±6.26

692±55

10

2

6

Channel

Interaction energy (kcal/mol)

Contact residues Hydrophobic (including aromatic)

Hydrophilic

Charged

10

1

4

37

ACS Paragon Plus Environment

ACS Nano 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

TOC Figure

38

ACS Paragon Plus Environment

Page 38 of 38