Probing ion binding in the selectivity filter of the KcsA potassium channel

Subscriber access provided by The University of Melbourne Libraries

Article

Probing ion binding in the selectivity filter of the KcsA potassium channel Cedric Eichmann, Lukas Frey, Innokentiy Maslennikov, and Roland Riek J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 11 Apr 2019 Downloaded from http://pubs.acs.org on April 11, 2019

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 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 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.

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 19 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

Journal of the American Chemical Society

Probing ion binding in the selectivity filter of the KcsA potassium channel Cédric Eichmann1, Lukas Frey1, Innokentiy Maslennikov2, Roland Riek1* 1Laboratory 2School

of Physical Chemistry, ETH Zurich, 8093 Zurich (Switzerland)

of Pharmacy, Chapman University, Orange, CA 92866 (USA)

*Correspondence should be addressed to R. R. ([email protected])

Abstract In potassium (K+) channels, permeation, selectivity, and gating at the selectivity filter are all governed by the thermodynamics and kinetics of the ion-protein interactions. Specific contacts between the carbonyl groups from the Thr-Val-Gly-Tyr-Gly signature filter sequence and the permeant ions generate four equidistant K+ binding sites, thereby defining the high ion selectivity and controlling the transport rate of K+ channels. Here, we used 15N-labeled ammonium (15NH4+) as a proxy for K+ to study ion interaction with the selectivity filter of the prototypical full-length K+ channel KcsA by solution state NMR spectroscopy in order to obtain detailed insights into the physicochemical basis of K+ gating. We found that in the closed inactive state of KcsA (at pH 7) four K+ binding sites are occupied over a wide range of

15NH + 4

concentrations, while in

intermediate closed-open conformations (at pH ~6) the number and occupancy of K+ binding sites are reduced to two. However, in presence of the scorpion toxin agitoxin II a total loss of 15NH + 4

binding is observed.

15NH + 4

titration studies allowed us to determine the dissociation

constants of the four binding sites with values around 10 mM in the closed state of KcsA. Moreover, kinetic NMR experiments measured in the steady state equilibrium detected an offand on-rate for 15NH4+ of ca. 102 s-1 and 103 s-1 between KcsA-bound 15NH4+ and the bulk. These findings reveal both the thermodynamics and kinetics of the ion binding sites and thus contribute to our understanding of the action of K+ channels.

1 Environment ACS Paragon Plus

Journal of the American Chemical Society 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 19

Introduction Potassium channels (K+) are essential for the passive transport of K+ across cell membranes, thereby regulating cellular ion concentrations and membrane potentials. Crystal structures 1,2 have revealed a highly conserved narrow selectivity filter composed of four successive layers of main chain carbonyl oxygen atoms from the Thr-Val-Gly-Tyr-Gly signature amino acid sequence, which forms the basis of the exquisite K+ selection against other ions, the high conduction rate in the open state and the lack of conduction in the closed state of the channel. The K+-coordinating carbonyl oxygens of the pentapeptide segment point into the four-fold symmetric pore to act as surrogate waters for dehydrated K+ ions when passing through the filter. This unique spatial geometry generates four equidistant K+ binding sites within the selectivity filter Crystal structures of the prokaryotic K+ channel KcsA

1,6

1,3–5

(Fig. 1a).

indicate a superposition of two states

with two occupied K+ binding sites alternated by water molecules that weaken the Coulomb repulsion between the cations 4,6. However, geometrical considerations allow a closer grouping of the K+ ions within the filter

6,7.

Indeed, re-refined KcsA crystal structures demonstrate the

possibility of direct cationic contacts, which enable much higher K+ transfer rates than in waterseparated patterns 8. In line with these findings, molecular dynamics simulations of K+ conduction through various K+ channels showed that three ions are bound in the selectivity filter at K+ concentrations of 10-400 mM 8. In the present study we used

15N-labeled

ammonium

(15NH4+) as a K+-mimic to characterize both thermodynamically and kinetically the ion binding sites within the selectivity filter of the K+ channel KcsA from Streptomyces lividans in its closed (at pH 7) as well as in partially open/inactivated states (between pH 7 and 6) by two-dimensional (2D) solution state NMR spectroscopy.

Results Identification of the NMR signals of 15N-labeled ammonium bound in the selectivity filter of KcsA At pH 4 KcsA is readily permeable to K+ (ionic radius 1.33 Å) as well as its analogues Rb+ (1.48 Å), NH4+ (1.44 Å), and Tl+ (1.47 Å), while the channel is closed at pH 7 9. We selected 15NH4+ to investigate directly ion binding to full-length KcsA in detail by solution state NMR spectroscopy because it is a spin ½ probe with a high gyromagnetic ratio yielding high sensitivity. 15NH4+ has 2 Environment ACS Paragon Plus

Page 3 of 19 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


previously been used to characterize K+ binding sites of DNA quadruplexes large enzymes

13

10–12

and 40 kDa

by solution state NMR. Tetrameric wild-type KcsA was produced and

solubilized in n-dodecylphosphocholine (DPC) detergent micelles at pH 7 as described by Baker et al. 14. Due to fast exchange of the ammonium protons with H2O 11,13 at pH 7, the signal of free bulk 15NH4+ is absent in a 15NH4+-selective 2D [15N,1H]-HMQC NMR spectrum (Fig. 1b, dashed black line). The presence of five

15NH + 4

signals at a

sample containing both 0.25 mM KcsA and 75 mM

15N

chemical shift around 20 ppm of a

15NH + 4

are indicative of

15NH + 4

species

bound to KcsA (Fig. 1b-e). This observation is strengthened by further addition of 75 mM KCl to a KcsA sample treated with 75 mM

15NH +, 4

which leads to a complete disappearance of the

15NH + 4

signals (Fig. 1b, dashed gray line) pointing towards successful competition of K+ with the

15NH + 4

occupied binding sites. Moreover, it suggests that both K+ and

15NH + 4

target the same

binding sites in KcsA with a higher affinity for K+ ions as reported 9. Next, we prepared a sample comprising 0.25 mM KcsA and 0.4 mM agitoxin II. Agitoxin II is a scorpion toxin that binds in the micromolar range to KcsA

15

and blocks the channel entrance.

Solid state NMR studies on a modified KcsA channel (KcsA-Kv1.3) indicate that the scorpion toxin kaliotoxin interferes with the activity of KcsA-Kv1.3 by plugging the channel pore with the positively charged side chain of lysine Lys27

16–18,

which is highly conserved among scorpion

toxins. Indeed, addition of 75 mM 15NH4+ to the KcsA-agitoxin II sample did not give rise to any detectable cross peak signals at 20 ppm (Fig. 1f and compare with Fig. 1b). Thus, these experiments show that the KcsA-DPC complex closely mirrors a functional membrane channel 14 and the five cross peaks observed at pH 7 with

15N

chemical shifts of ~20 ppm in the 2D

[15N,1H]-HMQC spectrum result from 15NH4+ ions bound in the selectivity filter (Fig. 1b). In order to investigate an assignment of the KcsA-bound 15NH4+ signals and correlate them with the five binding sites shown in the X-ray crystal structure of KcsA 3, it is rationalized that due to the exchange of bound 15NH4+ with free 15NH4+, the latter having an anticipated chemical shift of the water frequency due to its fast exchange with H2O, the chemical shifts of the observed 15NH4+ resonances are influenced by the chemical shift of H2O. Following this line of reasoning, higher solvent exposure of an ion-binding site gives rise to faster exchange with free 15NH4+ resulting in a KcsA-bound

15NH + 4

resonance closer to the water frequency. Thus, the NMR resonances are

assigned to S0-S4 from 6.152-6.640 ppm as indicated in Fig. 1b. This tentative assignment of the

15NH + 4

cross peaks to their corresponding binding sites in the

selectivity filter was further validated by recording a 2D [15N,1H]-HMQC spectrum of a KcsA sample treated with 75 mM

15NH + 4

in presence of 160 M paramagnetic gadodiamide (Gd3+)



Page 4 of 19

(Fig. 1g). The water soluble Gd3+ induces relaxation yielding distance-dependent signal loss of KcsA-bound

15NH + 4

in the 2D [15N,1H]-HMQC spectrum. Comparison of the 1D cross sections

of the 2D [15N,1H]-HMQC spectra in absence and presence of Gd3+ shows that the signals S0, S1, and S2 (6.152, 6.578, and 6.595 ppm) were attenuated most significantly by the paramagnetic Gd3+ due to enhanced relaxation indicating that these are the binding sites in closest proximity to the solvent (Fig. 1a-e,g). In attempt to further improve our assignment of the KcsA-bound

15NH + 4

signals we collected

paramagnetic relaxation enhancement (PRE) data of the MTSL (S-(1-oxyl-2,2,5,5-tetramethyl2,5-dihydro-1H-pyrrol-3-yl)methyl methanesulfonothioate)-labeled KcsA(Q58C) variant treated with 75 mM 15NH4+. Visualization of a cysteine residue at position 58 in the crystal structure of the closed inactive state of KcsA revealed distances of 12.2-22 Å between the Cys58 S atom and K+ ions at the binding sites S0-S4 (Fig. S1a). Given that magnetic dipolar interactions between a nucleus and the unpaired electron from the paramagnetic spin label result in line broadening of the NMR signal, the MTSL probe induces distance-dependent cross peak attenuation 19 of KcsAbound

15NH +. 4

Comparison of the

15NH + 4

signal heights in the 2D [15N,1H]-HMQC spectra

between unlabeled and paramagnetic MTSL-labeled KcsA(Q58C) revealed only cross peak broadening beyond detection for all KcsA-bound 15NH4+ resonances (Fig. S1b). To minimize the paramagnetic effect of the nitroxide spin label we added ascorbic acid yielding an NMR spectrum of KcsA-bound

15NH + 4

with partially decreased cross peaks intensities compared to unlabeled

KcsA(Q58C). Unfortunately, we did not succeed in a more detailed assignment of the KcsAbound

15NH + 4

signals because the 2D [15N,1H]-HMQC NMR spectrum of the ascorbic acid-

reduced MTSL-labeled KcsA(Q58C) sample only shows uniform attenuation of the resonances (Fig. S1b) attributed to the close distance of the spin label

19.

15NH + 4

However, the data

further adds evidence that the detected 15NH4+ signals originate from bound ions in the filter. Titration of 15NH4+ reveals the thermodynamics of the individual 15NH4+ binding sites in the selectivity filter of KcsA in its closed state In order to determine the dissociation constant Kd of the individual binding sites within the selectivity filter of KcsA in its closed state at pH 7, we monitored the titration of 0.25 mM KcsA with 5-75 mM

15NH + 4

at 36 °C by recording

15NH +-selective 4

2D [15N,1H]-HMQC NMR

experiments. The 1D cross sections of the 2D [15N,1H]-HMQC spectra (Fig. 1b-d) comprise four relatively strong 15NH4+ signals assigned to the four K+ binding sites S1-S4 of the selectivity filter and a very weak cross peak attributed to the binding site S0 (Fig. 1b,e; see above). Qualitative 4 Environment ACS Paragon Plus

Page 5 of 19 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


inspection of the 15NH4+ concentration-dependent signal increase indicates very similar Kd values for S1-S4 in the millimolar range (Fig. S2). An overall quantitative fit using the Hill equation for n-binding sites allowed the determination of an average dissociation constant Kd,av ~ 11 mM with a Hill coefficient nav close to 1 as expected (Fig. 2a and Table S1). The individual analysis of the cross peak buildups S1-S4 with Kd ~ 8-12 mM and n~1 are distributed about the average (Table S1). However, the experimental data come along with quite large error bars due to the limited signal to noise and the small chemical shift dispersion of the cross peaks. Thus, the data are not investigated further in more detail. Occupancy of the individual binding sites in the selectivity filter of KcsA in its closed state Crystal structures of KcsA at pH ~7 and high concentrations of K+ indicate that on average between 2 and 2.5 conducting ions reside in the selectivity filter at four sites at once with an occupancy at each position of about 50%

3,4.

To obtain insights into how many binding sites are

occupied in solution, i.e. in absence of crystal constraints, the volume of the integrated

15NH + 4

cross peaks of a 0.25 mM KcsA sample at pH 6.5 and 36 °C in presence of 50 mM 15NH4+ was compared with cross peak volumes of 15N-1H moieties that show similar line shapes as the 15NH4+ signals derived from the same 2D [15N,1H]-HMQC experiment but with the 15N radio-frequency set to the amide region of the NMR spectrum. Taking into account that (i) KcsA is a tetramer, (ii) the steady state magnetization of the 15NH4+ ions originates from four protons, (iii) the different 1J HN

scalar couplings (i.e. 92 and 110 Hz for

15NH +, 4

15N-1HN

and Trp indole moieties, 73.5 Hz for

respectively), and (iv) the slow exchange rate constants of the 15NH4+ moieties in the 102-

103 s-1 range (Fig. 2b and c, see below), the occupancy of the selectivity filter was calculated to be 4.1. However, due to the complex dependency of the cross peak volume on relaxation, scalar couplings and exchange this number must be regarded rather as an estimate (possibly in the range of 3-4 per filter). Kinetics of the binding sites in the selectivity filter of KcsA in its closed state To assess average exchange rate constants for

15NH + 4

ion movement between KcsA-bound and

bulk 15NH4+, we performed water/NH4+-defocusing 2D [15N,1H]-HMQC experiments with water flip-back implementation. Due to the rapid exchange of free

15NH + 4

with H2O, the signal

originating either from exchangeable 15NH4+ between free and KcsA-bound species (Fig. S3a) or from bound

15NH + 4

(Fig. S3b) is suppressed in the corresponding pulse sequence. The signal

decay and increase of bound

15NH + 4

is then monitored at various exchange times. Using a 0.25



Page 6 of 19

mM KcsA sample in presence of 50 mM 15NH4+ at pH 6.5 and 36 °C with exchange time delays of 1.1-11 ms (Fig. 2b) and 1.4-12 ms (Fig. 2c) revealed that the kinetics between bound and unbound 15NH4+ is characterized by average off/on-exchange rates of koff,av ~ 91 s-1 (leading to a tav,1/2 ~ 8 ms) and kon,av ~ 852 s-1, respectively (Table S1). Please note, due to limitations of the setup of the experiment for the determination of the on-rate a dead time of 1.4 ms is present in the pulse sequence (Fig. S3b). Thus, kon,av ~ 852 s-1 needs to be considered as a lower boundary rather than an exact value. Occupancy, thermodynamics, and kinetics of the individual binding sites of the selectivity filter in the partially open-inactivated state of KcsA KcsA is closed at pH 7, starts opening by lowering the pH value and is entirely open at pH 4, which is followed by inactivation of the channel 20. To explore the number of occupied binding sites, the thermodynamics, and kinetics of KcsA in its open-inactivated state we performed pH titration studies on 75 mM measuring

15NH +-selective 4

15NH + 4

treated KcsA (0.25 mM) from pH 7 to pH 4 at 36 °C by

2D [15N,1H]-HMQC spectra. Interestingly, reducing the pH value

resulted in an overall decrease of the cross peak intensities of KcsA-bound

15NH + 4

(Fig. 3a,b)

suggesting a pH-dependent decrease of the Kd value. At pH 6 we detected the occupation of only two binding sites within the filter. No cross peak signals from KcsA-bound 15NH4+ were present at pH 5.8 and lower (data not shown) indicating that for each binding site the Kd lowered by at least a factor of 15 based on signal-to-noise measurements of the KcsA sample at pH 7 (please note, the intrinsic exchange of 15NH4+ with H2O decreases at lower pH). The relative population of the binding site S1-S4 was then obtained from the normalized total volumes of the integrated 15NH + 4

cross peaks as a function of the pH value. The selectivity filter positions S1-S4 are fully

occupied at pH 7, while linear fitting of the data predict two bound 15NH4+ ions at pH 6.4 (Fig. 3c) reminiscent of the occupied K+ binding sites described in the collapsed

3,4

and C-type

inactivated 5 KcsA crystal structures with one K+ ion residing at two sites of the filter (Fig. 3d and Fig. 4).

Discussion K+ ions appear to be present in a single file in the closed state of KcsA


Page 7 of 19 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 closed state at pH ~7 crystallographic data of the KcsA selectivity filter have shown two distinct conformations associated with low and high concentrations of K+

3,4.

In presence of K+

concentrations below 20 mM the filter is collapsed in the middle with up to a single K+ ion distributed at either of the two sites near the end of the filter

3,4,21.

As the K+ concentrations are

increased above 20 mM a second ion is present in the filter accompanied with a conformational change of the selectivity filter towards a form with two ions and intervening water molecules distributed over four sites at once, each with approximately half occupancy 4,6. The presented NMR studies indicate that in the closed conformation of KcsA at pH 7 all four ion binding sites in the selectivity filter can be occupied with

15NH + 4

ions over a

15NH + 4

concentration range of 5-75 mM (Fig. 1). At the lower concentration range, each of the ion binding site is occupied only partly. At high concentration, the NMR intensity measurements indicate the presence of a high

15NH + 4

occupancy per ion binding site. In first sight, our

experimental results conflict with crystal structures of KcsA. However, our findings are in line with the early suggestion that K+ ions are constrained in a single file with several K+ present in the closed state of the channel at any moment 22 and with recent molecular dynamics simulations by Köpfer et al.

8

that revealed three ions in the selectivity filter at a given time and at KCl

concentrations of 10-400 mM. Moreover, these molecular dynamics simulations have shown that re-refined crystal structures are consistent with single-file chains of cations in a direct contact 8. The thermodynamics and kinetics of the ions in the selectivity filter of KcsA in the closed state A comprehensive description of the K+ affinity of the selectivity filter is central to understand how K+ channels simultaneously achieve both ion selectivity and fast conduction. At pH 7 the 15NH + 4

titration NMR experiments on the KcsA-DPC complex in the closed state show Kd values

in the range of 10 mM for the binding sites S1-S4 with Hill coefficients n close to 1 implying no cooperative 15NH4+ binding and similar energetics for each binding site. Comparable K+ binding affinities of 3 and 6.1 mM were reported in solution NMR studies of KcsA solubilized in sodium dodecyl sulfate

23

(SDS) and dodecyl maltoside

24

(DDM) at pH 6 and 6.7, respectively, while

solid state NMR experiments of KcsA in a lipid bilayer system

25

reported Kd values for K+

binding in the micromolar regime. This divergence of the dissociation constants may be attributable to differences in the membrane mimetics used (i.e. lipids versus detergents) since key mechanistic details in membrane proteins show changes in detergent compared to lipid milieus



26,27

Page 8 of 19

including long-lasting C-type inactivation (at the millisecond to second timescale) and fast

flicker transitions (microsecond range) of KcsA 28–30. The weak

15NH + 4

binding affinity in the closed state is accompanied by slow off- and on-

exchange rates for 15NH4+ between the selectivity filter and the bulk with average koff,av and kon,av values of ca. 90 and 850 s-1 (Fig. 2b and c, Table S1), respectively. This results roughly in a Kd of 2.5 mM (Kd = [KcsA] kon,av / koff,av), which is close to the 11 mM determined from

15NH + 4

titration experiments, in particular when considering that the kon,av rate must be regarded as an estimate. Taking into account that similar low Kd values are observed for each of the four ion binding sites in the selectivity filter (Fig. S2), both the thermodynamics and kinetics of the binding sites are thus guaranteeing the lack of ion conductance through the membrane in the closed state of the channel. The thermodynamics and kinetics of the ions in the selectivity filter of KcsA from the closed towards the open state Thompson et al. used mutation studies to propose a molecular mechanism of pH sensing in KcsA K+ channels 31, which is based on a proton-dependent disruption of the salt bridge and hydrogenbonding network formed by the two glutamates Glu118 and Glu120 interacting with arginine Arg122. In addition, histidine His25 Glu118

31.

32

makes a bundle-crossing inter-subunit interaction with

At neutral pH these interactions connect the transmembrane helices TM1 and TM2

thereby stabilizing the closed conformation, while at acidic pH (at a pH of ~4), the glutamate and histidine residues protonate, disrupt these connections, and allow the channel to open

31.

In the

open state the selectivity filter flickers thereby in the millisecond time range between ioncoordinating configurations of the closed state and a non-coordinating state 14. The presented spectroscopic pH studies provide evidence that already a small pH change from 7 to 6, for which the channel is open only to a few percent 14, is sufficient to decrease the affinity of 15NH + 4

to the selectivity filter by about 15-fold, thereby reducing the number of occupied K+

binding sites from five to two along with chemical shift changes (Fig. 3 a,b). This apparent inconsistency between loss of binding affinity and channel opening (Fig. 4) can be explained by a kinetic argument. Upon reduction of the pH from 7 to 6 the KcsA channel starts to fluctuate between an open non ion-coordinating and a closed ion-coordinating state (Fig. 3d). This fluctuation may be a result of a partial protonation of Glu118, Glu120, and His25 that is sufficient enough to disrupt slightly the hydrogen-bonding network that stabilizes the closed state of the channel yielding an open state with an average population of only a few percent. The presence of 8 Environment ACS Paragon Plus

Page 9 of 19 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


a fast dynamic exchange between an ion-coordinating closed and open state, which resembles the configuration of the crystal structures at low K+ concentrations

3

(Fig. 3d) and open-inactivated

KcsA 5 (Fig. 4) with ion occupancies only at positions S1 and S4, has been observed by solution state NMR

14

studying protein dynamics. It is interesting to note that the observed pH 7 to 6-

induced chemical shift changes of the

15NH + 4

signals may be attributable to a conformational

change from the closed towards the open state similar to the reported shifts of the 15N-1H moieties of the selectivity filter 14. Below pH 6, no 15NH4+-bound state can be detected anymore by solution state NMR (Fig. 4) due to fast kinetics between

15NH + 4

and the KcsA channel, thereby highlighting the presence of a

selectivity filter with conductance capacity towards the near diffusion limit. Based on detailed structural work using EPR

33,

fluorescence

34,

and NMR spectroscopy

14,

ion

conductance in KcsA is governed by a large conformational change at the cytoplasmic helix bundle crossing yielding fast fluctuation between ion-coordinating and non-coordinating states in the filter, which guarantees in addition to rapid ion flux also selectivity. These findings are corroborated here by thermodynamic and kinetic data acquisition directly on the cation ligand.

Methods Expression and purification of wild-type full-length KcsA Wild-type full-length 2H (80%), 15N-labeled KcsA from Streptomyces lividans was expressed and purified as described previously 14. Briefly, fractions containing the protein were pooled together and the buffer was exchanged to 20 mM Bis-Tris-HCl pH 7 and 3 mM DPC in presence of 15NH

4Cl

using a PD-10 (GE Healthcare) desalting column at 4 °C. For NMR measurements

KcsA was concentrated with a 50 kDa molecular weight cut-off Centricon (Amicon) concentrator to a final tetrameric sample concentration of ~0.25 mM. 3% of D2O was added for NMR measurements. NMR spectroscopy 2D [15N,1H]-HMQC experiments

35

with selective excitation burp pulses and the water/NH4+-

defocusing versions thereof with exchange time delays of 1.1-11 ms (Fig. S3a) and 1.4-12 ms (Fig. S3b), respectively, were all collected at 36 °C on a Bruker 700 MHz Avance III spectrometer equipped with a triple-resonance cryoprobe. Binding of 9 Environment ACS Paragon Plus

15NH + 4

to 0.25 mM 2H


Page 10 of 19

(80%), 15N-labeled KcsA was studied by performing series of titrations with 5-75 mM 15NH4Cl at pH 7. Agitoxin II (Alomone Labs) driven inhibition of 15NH4+ binding to KcsA was investigated by addition of 75 mM

15NH + 4

to a sample comprising 0.25 mM tetrameric KcsA and 0.4 mM

agitoxin II at pH 7. Solvent paramagnetic perturbation of the binding sites within the selectivity filter of KcsA at pH 7 was probed in presence of 160 M gadodiamide (Omniscan) in the NMR buffer. For MTSL (Toronto Research Chemicals) labeling 19,36, we introduced a cysteine mutation of wild-type KcsA at Gln58 using QuickChange mutagenesis (Stratagene). KcsA(Q58C) was incubated for 2 hours at room temperature with a three-fold excess of paramagnetic MTSL compared to the monomeric protein concentration. After spin label incorporation, free MTSL was removed with a PD-10 column followed by additional buffer exchange with a 50 kDa molecular weight cut-off concentrator. PRE of the 15NH4+ binding sites in MTSL-labeled KcsA(Q58C) was investigated at pH 7 in presence of 75 mM 15NH4+. The paramagnetic spin label was reduced with ten-fold excess of ascorbic acid. Average exchange rate constants between KcsA-bound and bulk 15NH + 4

were determined from a 0.25 mM 2H (80%),

presence of 50 mM

15NH +. 4

The pH-influence on

15N-labeled

15NH + 4

KcsA sample at pH 6.5 in

binding to KcsA was investigated by

changing the pH value from 7 to 6 in 0.25 unit steps using 0.25 mM 2H (80%), 15N-labeled KcsA in presence of 75 mM 15NH4+. Signal losses due to addition of salts (15NH4Cl or pH change) were quantified and corrected by recording 2D [15N,1H]-TROSY

37

spectra of the 2H (80%),

15N-

labeled KcsA samples prior to each 15NH4+-selective 2D [15N,1H]-HMQC NMR experiment. The spectra were processed with the program PROSA 38 and analyzed with the program XEASY 39.

Acknowledgments C.E. is supported by a SNSF Advanced Postdoc Mobility fellowship (P300PA_160979). Note The authors declare no competing financial interest. Supporting information Fig. S1. Paramagnetic MTSL-labeled KcsA(Q58C) produces line-broadening of bound 15NH4+ in the NMR spectrum. Fig. S2. 15NH4+ dissociation constants of the individual binding sites S1-S4 in the closed state of the KcsA selectivity filter at pH 7. 10 Environment ACS Paragon Plus

Page 11 of 19 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


Fig. S3. Water/NH4+-defocusing 2D [15N,1H]-HMQC pulse sequences for the determination of average (a) koff and (b) kon exchange-rates of 15NH4+ to KcsA. References (1) (2)

(3) (4)

(5)

(6) (7) (8) (9) (10) (11) (12) (13)

(14)

Doyle, D. A.; Morais Cabral, J.; Pfuetzner, R. A.; Kuo, A.; Gulbis, J. M.; Cohen, S. L.; Chait, B. T.; MacKinnon, R. The Structure of the Potassium Channel: Molecular Basis of K+ Conduction and Selectivity. Science 1998, 280 (5360), 69–77. Uysal, S.; Vásquez, V.; Tereshko, V.; Esaki, K.; Fellouse, F. A.; Sidhu, S. S.; Koide, S.; Perozo, E.; Kossiakoff, A. Crystal Structure of Full-Length KcsA in Its Closed Conformation. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (16), 6644–6649. https://doi.org/10.1073/pnas.0810663106. Zhou, Y.; Morais-Cabral, J. H.; Kaufman, A.; MacKinnon, R. Chemistry of Ion Coordination and Hydration Revealed by a K+ channel–Fab Complex at 2.0 |[angst]| Resolution. Nature 2001, 414 (6859), 43–48. https://doi.org/10.1038/35102009. Zhou, Y.; MacKinnon, R. The Occupancy of Ions in the K+ Selectivity Filter: Charge Balance and Coupling of Ion Binding to a Protein Conformational Change Underlie High Conduction Rates. J. Mol. Biol. 2003, 333 (5), 965–975. https://doi.org/10.1016/j.jmb.2003.09.022. Cuello, L. G.; Jogini, V.; Cortes, D. M.; Pan, A. C.; Gagnon, D. G.; Dalmas, O.; CorderoMorales, J. F.; Chakrapani, S.; Roux, B.; Perozo, E. Structural Basis for the Coupling between Activation and Inactivation Gates in K+ Channels. Nature 2010, 466 (7303), 272– 275. https://doi.org/10.1038/nature09136. Morais-Cabral, J. H.; Zhou, Y.; MacKinnon, R. Energetic Optimization of Ion Conduction Rate by the K+ Selectivity Filter. Nature 2001, 414 (6859), 37–42. https://doi.org/10.1038/35102000. Aqvist, J.; Luzhkov, V. Ion Permeation Mechanism of the Potassium Channel. Nature 2000, 404 (6780), 881–884. https://doi.org/10.1038/35009114. Köpfer, 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 (6207), 352–355. https://doi.org/10.1126/science.1254840. LeMasurier, M.; Heginbotham, L.; Miller, C. Kcsa. J. Gen. Physiol. 2001, 118 (3), 303– 314. Hud, N. V.; Schultze, P.; Feigon, J. Ammonium Ion as an NMR Probe for Monovalent Cation Coordination Sites of DNA Quadruplexes. J. Am. Chem. Soc. 1998, 120 (25), 6403–6404. https://doi.org/10.1021/ja9811039. Hud, N. V.; Schultze, P.; Sklenář, V.; Feigon, J. Binding Sites and Dynamics of Ammonium Ions in a Telomere Repeat DNA quadruplex1. J. Mol. Biol. 1999, 285 (1), 233–243. https://doi.org/10.1006/jmbi.1998.2327. Šket, P.; Plavec, J. Tetramolecular DNA Quadruplexes in Solution: Insights into Structural Diversity and Cation Movement. J. Am. Chem. Soc. 2010, 132 (36), 12724–12732. https://doi.org/10.1021/ja104889t. Werbeck, N. D.; Kirkpatrick, J.; Reinstein, J.; Hansen, D. F. Using 15N-Ammonium to Characterise and Map Potassium Binding Sites in Proteins by NMR Spectroscopy. Chembiochem Eur. J. Chem. Biol. 2014, 15 (4), 543–548. https://doi.org/10.1002/cbic.201300700. Baker, K. A.; Tzitzilonis, C.; Kwiatkowski, W.; Choe, S.; Riek, R. Conformational Dynamics of the KcsA Potassium Channel Governs Gating Properties. Nat. Struct. Mol. Biol. 2007, 14 (11), 1089–1095. https://doi.org/10.1038/nsmb1311. 11 Environment ACS Paragon Plus


(15) Takeuchi, K.; Yokogawa, M.; Matsuda, T.; Sugai, M.; Kawano, S.; Kohno, T.; Nakamura, H.; Takahashi, H.; Shimada, I. Structural Basis of the KcsA K(+) Channel and agitoxin2 Pore-Blocking Toxin Interaction by Using the Transferred Cross-Saturation Method. Struct. Lond. Engl. 1993 2003, 11 (11), 1381–1392. (16) 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 SolidState NMR. Nature 2006, 440 (7086), 959–962. https://doi.org/10.1038/nature04649. (17) Miller, C. The Charybdotoxin Family of K+ Channel-Blocking Peptides. Neuron 1995, 15 (1), 5–10. (18) Garcia, M. L.; Gao, Y.-D.; McManus, O. B.; Kaczorowski, G. J. Potassium Channels: From Scorpion Venoms to High-Resolution Structure. Toxicon 2001, 39 (6), 739–748. https://doi.org/10.1016/S0041-0101(00)00214-2. (19) Battiste, J. L.; Wagner, G. Utilization of Site-Directed Spin Labeling and High-Resolution Heteronuclear Nuclear Magnetic Resonance for Global Fold Determination of Large Proteins with Limited Nuclear Overhauser Effect Data. Biochemistry (Mosc.) 2000, 39 (18), 5355–5365. (20) Cordero-Morales, J. F.; Cuello, L. G.; Zhao, Y.; Jogini, V.; Cortes, D. M.; Roux, B.; Perozo, E. Molecular Determinants of Gating at the Potassium-Channel Selectivity Filter. Nat. Struct. Mol. Biol. 2006, 13 (4), 311–318. https://doi.org/10.1038/nsmb1069. (21) Lockless, S. W.; Zhou, M.; MacKinnon, R. Structural and Thermodynamic Properties of Selective Ion Binding in a K+ Channel. PLoS Biol. 2007, 5 (5). https://doi.org/10.1371/journal.pbio.0050121. (22) Hodgkin, A. L.; Keynes, R. D. The Potassium Permeability of a Giant Nerve Fibre. J. Physiol. 1955, 128 (1), 61–88. (23) Chill, J. H.; Louis, J. M.; Miller, C.; Bax, A. NMR Study of the Tetrameric KcsA Potassium Channel in Detergent Micelles. Protein Sci. Publ. Protein Soc. 2006, 15 (4), 684–698. https://doi.org/10.1110/ps.051954706. (24) Imai, S.; Osawa, M.; Takeuchi, K.; Shimada, I. Structural Basis Underlying the Dual Gate Properties of KcsA. Proc. Natl. Acad. Sci. U. S. A. 2010, 107 (14), 6216–6221. https://doi.org/10.1073/pnas.0911270107. (25) Bhate, M. P.; Wylie, B. J.; Tian, L.; McDermott, A. E. Conformational Dynamics in the Selectivity Filter of KcsA in Response to Potassium Ion Concentration. J. Mol. Biol. 2010, 401 (2), 155–166. https://doi.org/10.1016/j.jmb.2010.06.031. (26) Cady, S. D.; Schmidt-Rohr, K.; Wang, J.; Soto, C. S.; Degrado, W. F.; Hong, M. Structure of the Amantadine Binding Site of Influenza M2 Proton Channels in Lipid Bilayers. Nature 2010, 463 (7281), 689–692. https://doi.org/10.1038/nature08722. (27) Lim, H.-H.; Miller, C. It Takes Two to Transport, or Is It One? Nat. Struct. Mol. Biol. 2012, 19 (2), 129–130. https://doi.org/10.1038/nsmb.2239. (28) Ader, C.; Schneider, R.; Hornig, S.; Velisetty, P.; Vardanyan, V.; Giller, K.; Ohmert, I.; Becker, S.; Pongs, O.; Baldus, M. Coupling of Activation and Inactivation Gate in a K+Channel: Potassium and Ligand Sensitivity. EMBO J. 2009, 28 (18), 2825–2834. https://doi.org/10.1038/emboj.2009.218. (29) Piasta, K. N.; Theobald, D. L.; Miller, C. Potassium-Selective Block of Barium Permeation through Single KcsA Channels. J. Gen. Physiol. 2011, 138 (4), 421–436. https://doi.org/10.1085/jgp.201110684. (30) Cuello, L. G.; Jogini, V.; Cortes, D. M.; Perozo, E. Structural Mechanism of C-Type Inactivation in K(+) Channels. Nature 2010, 466 (7303), 203–208. https://doi.org/10.1038/nature09153.


Page 12 of 19

Page 13 of 19 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


(31) Thompson, A. N.; Posson, D. J.; Parsa, P. V.; Nimigean, C. M. Molecular Mechanism of pH Sensing in KcsA Potassium Channels. Proc. Natl. Acad. Sci. U. S. A. 2008, 105 (19), 6900–6905. https://doi.org/10.1073/pnas.0800873105. (32) Takeuchi, K.; Takahashi, H.; Kawano, S.; Shimada, I. Identification and Characterization of the Slowly Exchanging pH-Dependent Conformational Rearrangement in KcsA. J. Biol. Chem. 2007, 282 (20), 15179–15186. https://doi.org/10.1074/jbc.M608264200. (33) Perozo, E.; Marien, D.; Cortes; Cuello, L. G. Structural Rearrangements Underlying K+Channel Activation Gating. Science 1999, 285 (5424), 73–78. https://doi.org/10.1126/science.285.5424.73. (34) Blunck, R.; Cordero-Morales, J. F.; Cuello, L. G.; Perozo, E.; Bezanilla, F. Detection of the Opening of the Bundle Crossing in KcsA with Fluorescence Lifetime Spectroscopy Reveals the Existence of Two Gates for Ion Conduction. J. Gen. Physiol. 2006, 128 (5), 569–581. https://doi.org/10.1085/jgp.200609638. (35) Riek, R.; Wider, G.; Pervushin, K.; Wüthrich, K. Polarization Transfer by Cross-Correlated Relaxation in Solution NMR with Very Large Molecules. Proc. Natl. Acad. Sci. U. S. A. 1999, 96 (9), 4918–4923. (36) Bibow, S.; Polyhach, Y.; Eichmann, C.; Chi, C. N.; Kowal, J.; Albiez, S.; McLeod, R. A.; Stahlberg, H.; Jeschke, G.; Güntert, P.; Riek R. Solution Structure of Discoidal HighDensity Lipoprotein Particles with a Shortened Apolipoprotein A-I. Nat. Struct. Mol. Biol. 2017, 24 (2), 187–193. https://doi.org/10.1038/nsmb.3345. (37) Pervushin, K.; Riek, R.; Wider, G.; Wüthrich, K. Attenuated T2 Relaxation by Mutual Cancellation of Dipole-Dipole Coupling and Chemical Shift Anisotropy Indicates an Avenue to NMR Structures of Very Large Biological Macromolecules in Solution. Proc. Natl. Acad. Sci. U. S. A. 1997, 94 (23), 12366–12371. (38) Güntert, P.; Dötsch, V.; Wider, G.; Wüthrich, K. Processing of Multi-Dimensional NMR Data with the New Software PROSA. J. Biomol. NMR 1992, 2 (6), 619–629. https://doi.org/10.1007/BF02192850. (39) Bartels, C.; Xia, T. H.; Billeter, M.; Güntert, P.; Wüthrich, K. The Program XEASY for Computer-Supported NMR Spectral Analysis of Biological Macromolecules. J. Biomol. NMR 1995, 6 (1), 1–10. https://doi.org/10.1007/BF00417486.



Figures Fig. 1


Page 14 of 19

Page 15 of 19 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


Fig. 2



Fig. 3


Page 16 of 19

Page 17 of 19 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


Fig. 4

Table of Contents Graphic



Page 18 of 19

Figure captions Fig. 1. K+ binding in the closed state of KcsA. Crystal structure of the (a) conductive state (Protein Data Bank (PDB) 1K4C) of KcsA. (b) 2D [15N,1H]-HMQC spectrum of 75 mM 15NH4+ in presence of 0.25 mM KcsA at pH 7, 36 °C, indicates five (S0-S4) occupied binding sites in the selectivity filter. Shown above, 1D cross sections of 2D [15N,1H]-HMQC spectra of 75 mM 15NH

4Cl

(dashed black line) at 36 °C dissolved in NMR buffer. At 75 mM 15NH4+ and 0.25 mM

KcsA addition of 75 mM KCl leads to a complete disappearance of the

15NH + 4

signals (dashed

gray line). (c-e) 1D cross sections of the 2D NMR spectra upon addition of 5 and 75 mM 15NH4+ to 0.25 mM KcsA. The experimental data were simultaneously deconvoluted by Gaussian functions (green lines). (f) Addition of 75 mM 0.4 mM agitoxin II inhibits binding of

15NH +. 4

15NH + 4

to a 0.25 mM KcsA sample treated with

(g) 75 mM

15NH +-treated 4

KcsA (0.25 mM) in

presence of 160 M Gd3+ indicates that S0 and S1 (6.152 and 6.578 ppm) are in closest proximity to the solvent. Fig. 2. 15NH4+ binding kinetics of the KcsA selectivity filter. (a) The normalized volume of the integrated 15NH4+ cross peaks from the 2D [15N,1H]-HMQC NMR spectra of KcsA-bound 15NH4+ at pH 7 and 36 °C were fitted by a hyperbolic binding function r = n.[15NH4+]n/(Kd+[15NH4+]n). Determination of average (b) off- (koff,av) and (c) on-exchange rates (kon,av) between KcsA-bound and bulk 15NH4+ at pH 6.5 and 36 °C obtained from water/NH4+-defocusing 2D [15N,1H]-HMQC experiments. Using a 0.25 mM KcsA sample in presence of 50 mM 15NH4+ resulted in 91 s-1 and 852 s-1 for koff,av and kon,av, respectively. Fig. 3. Reduced number of occupied binding sites in the selectivity filter of KcsA. (a), (b) 1D cross sections of 2D [15N,1H]-HMQC spectra of 0.25 mM KcsA in presence of 75 mM

15NH + 4

between pH 7 and 6 at 36 °C. Green lines represent the simultaneously deconvoluted experimental data by Gaussian functions. (c) Relative population of the binding site S1-S4 obtained from the normalized total volume of the integrated 15NH4+ cross peaks as a function of pH value. Linear data fitting shows that the selectivity filter positions S1-S4 are half occupied at pH 6.44. (d) The NMR data may indicate that the selectivity filter starts to fluctuate between a closed (four occupied binding sites) and an open state (two occupied sites) reminiscent of the conductive (PDB 1K4C) and collapsed (PDB 1K4D) KcsA crystal structures.


Page 19 of 19 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


Fig. 4. Correlation between selectivity filter ion occupancy and partially open KcsA crystal structures. Comparison of the individual 1D cross sections of the 2D [15N,1H]-HMQC spectra from KcsA-bound 15NH4+ between pH 7 and 4 with open KcsA crystal structures 30 of 14, 23, and 32 Å (PDB 3FB5, 14 Å; 3F7V, 23 Å; 3F5W, 32 Å).


Probing ion binding in the selectivity filter of the KcsA potassium channel

Recommend Documents