Dissecting the pH Sensitivity of Kinesin-Driven Transport

Abstract

Kinesin-1 is a crucial motor protein that drives the microtubule-based movement of organelles, vital for cellular function and health. Mostly studied at pH 6.9, it moves at approximately 800 nm/s, covers about 1 μm before detaching, and hydrolyzes one ATP per 8 nm step. Given that cellular pH is dynamic and alterations in pH have significant implications for disease, understanding how kinesin-1 functions across different pH levels is crucial. To explore this, we executed single-molecule motility assays paired with precise optical trapping techniques over a pH range of 5.5–9.8. Our results show a consistent positive relationship between increasing pH and the enhanced detachment (off rate) and speed of kinesin-1. Measurements of the nucleotide-dependent off rate show that kinesin-1 exhibits the highest rate of ATPase activity at alkaline pH, while it demonstrates the optimal number of ATP turnover and cargo translocation efficiency at the acidic pH. Physiological pH of 6.9 optimally balances the biophysical activity of kinesin-1, potentially allowing it to function effectively across a range of pH levels. These insights emphasize the crucial role of pH homeostasis in cellular function, highlighting its importance for the precise regulation of motor proteins and efficient intracellular transport.

Introduction

Kinesin-1 is a crucial molecular motor protein responsible for the transport of organelles along microtubules (MT) within cells, which is vital to maintaining cellular health and function.¹⁻⁴ Comprehending the regulatory mechanisms behind the operation of the kinesin-1 motor is essential for understanding disease pathogenesis and designing synthetic nanomachines. The optimal activity of kinesin motor proteins is characterized by their efficiency in transporting cellular cargo along MTs. This efficiency can be quantitatively measured by biophysical parameters, including velocity, travel distance, force production, and rates of dissociation from MT.⁵⁻⁸ Factors such as temperature,^9,10 salt concentration,¹¹⁻¹³ and hydrogen ion concentration (pH)¹⁴ have been demonstrated to influence these biophysical parameters of kinesin-1. Despite extensive research on kinesin motors, the exact mechanisms to achieve optimal functionality in varied environments remain to be more clearly defined.

Growing evidence indicates that the cytosolic pH within cells is highly dynamic rather than stable.^15,16 The variation in pH is critical during the cell cycle¹⁷ and is instrumental in regulating cell-substrate adhesion remodeling.¹⁸ Cardiac muscle cells exhibit beat-to-beat cellular acidification.¹⁹ Abnormalities in the regulation of pH are characteristic of various diseases; for example, increased intracellular pH is detected in cancer cells,²⁰ while a reduced cytosolic pH is associated with neurodegenerative conditions such as amyotrophic lateral sclerosis (ALS).²¹ Understanding how motor proteins function in the face of such a pH variability is essential. Studies conducted in vitro have demonstrated that the optimal activity of macromolecules is closely related to their structural stability, which is governed by the surrounding pH.²² Fluctuations in pH can significantly affect critical interactions, including protein–protein²³ and protein–ligand²⁴ that further affect the functional activities of proteins. Therefore, pH-dependent changes are crucial considerations in understanding the molecular dynamics that underpin cellular function and health. In vitro motility experiments have shown that low pH can affect the mechanics and kinetics of myosin motors.²⁵ Currently, our understanding of how the pH affects the kinetics of MT-based kinesin motors is limited. To study the effects of pH variability on motor protein function, we used single-molecule bead assays and optical trapping techniques. These methods allowed us to measure the biophysical parameters and transition rates of individual kinesin motors with pH serving as a modulating factor. In this work, we present a comprehensive analysis of the kinesin-1 motor’s biophysical parameters across a pH range from 5.5 to 9.8.

Structurally, kinesin-1 motors feature two highly conserved globular motor domains at their core, with each domain harboring an MT-binding site and an adenosine triphosphate (ATP)-binding site that powers motor activity.¹ In addition, kinesin-1 is equipped with two identical light chains that facilitate cargo recognition and attachment.^1,26 In vitro experiments conducted at the pH of 6.9 demonstrate that, in the presence of high ATP levels, kinesin-1 motors move at an average speed of 800 nm/s and can travel approximately 1 μm along microtubules before detaching. The movement of kinesin-1 along microtubules is driven by a “hand-over-hand” mechanism that involves ATP (T) hydrolysis.²⁷Figure 1A depicts the stepwise progression of kinesin-1 as it moves along an MT.²⁸ The sequence initiates with the motor in a one-head-bound state poised to bind ATP (state-1 in Figure 1A), wherein the forward (MT-bound) head is nucleotide-free (Ø) and the rear (tethered) head carries an adenosine diphosphate (ADP) (D) molecule. Following ATP binding at a rate k_b to the forward head (state-2), the motor executes partial neck-linker docking and moves the rear-head forward. ATP hydrolysis prompts the tethered head to swing forward further (state-3). The motor can then either secure itself to the next MT binding site, completing the mechanical step while releasing ADP at rate k_D from the now-tethered head (state-4), or prematurely undergo an off-pathway dissociation by releasing an inorganic phosphate (Pi) from the head containing ADP-Pi (Off). Upon successful forward stepping (state-4), the cycle concludes with Pi release and detachment of the rear head, resetting the motor to the ATP-waiting starting configuration (state-1) and advancing kinesin-1 by an 8.0-nm step toward the MT plus-end. The velocity at which the motor protein operates is determined by Michaelis–Menten kinetics, which depend on the concentration of ATP and the corresponding rates of enzymatic turnover rates (k_c).²⁹ This turnover rate results from a combination of different individual rates: ATP hydrolysis, ADP release, and phosphate (Pi) release. Currently, our understanding of how these kinetic parameters change under varying pH conditions is still limited.

Figure 1.

Impact of pH on kinesin-1 motor motility at 1 mM ATP concentration: (A) The figure displays the stepwise ATPase cycle of kinesin-1 as it moves along a microtubule using a hand-over-hand mechanism. The green and gray ovals represent alpha and beta tubulin of the microtubule, respectively. The semicircles indicate the head domains (orange-rear and white-front), with D (ADP), T (ATP), Pi (phosphate), Φ (empty state), k_b (ATP-binding rate), and k_D (rate of ADP release). (B) The “binding fraction”, proportion of moving beads relative to the total beads assessed, is represented as blue dots. The red curve is the fit (1 – exp(−x/b)) of the binding fraction data, where “b” denotes a fixed parameter and “x” represents the kinesin-to-bead ratio. Error bars indicate the standard error among multiple experimental replicates conducted on different days. The inset graphic depicts the experimental motility scheme with zero load. (C) Comparison of distribution of run lengths of single kinesin-1 motors at pH 5.5 (orange, n = 64), pH 6.9 (green, n = 84), and pH 9.8 (blue, n = 54). Single exponential fits (solid lines) deliver average run lengths. (D) Velocity distributions at pH 5.5 (orange, n = 64), 6.9 (green, n = 84), and 9.8 (blue, n = 54). The average velocities were obtained from the Gaussian fits (solid lines) of the distributions. The error bars on the column plots were calculated using √(p∗(1 – p)/n, where p is the normalized fraction for each column. (E) Average run lengths (black) and average velocities (red) at the corresponding pH values. (F) Calculated off rates (v/R) (±standard error) for a given pH. R-squared values in curve fittings were maintained ≥ 0.9. The number of motile tracks analyzed is “n”. < > denotes the average values.

This study examines how changes in pH affect the function of the kinesin-1 motor protein at the single-molecule level. Our research employs high-resolution bead motility assays and optical trapping methods to closely examine the effects of varying pH levels on the dynamic properties of the motor protein, including its travel speed, processivity, and force production. The investigation of these dynamic properties is based on the Michaelis–Menten kinetic model, which allows us to dissect the complex chemomechanical behavior of the kinesin motor protein. We postulate that pH changes serve as a regulatory mechanism for the motor protein’s performance. Our data provide evidence that kinesin-1 motor proteins achieve a balance in functionality at a physiological pH of 6.9, which allows them to adapt effectively to varying pH conditions in diverse cellular environments. This study delineates how pH levels affect kinesin-1’s motility and ATPase activity. These insights enhance our understanding of the molecular underpinnings of intracellular movement and underscore their significance for cellular function and health.

Materials and Methods

Single-molecule motility assays with optical trapping were performed using a standard protocol that involved the preparation of anti-His-coated beads for specific recruitment of kinesin motors, MT polymerization, and preparation of flow chambers.³⁰ Streptavidin-coated polystyrene beads of 514 nm diameter (Spherotech, SVP-05–10) were coated with biotinylated anti-His antibodies using the protocol as described in ref (31). Taxol-stabilized MTs were polymerized as in ref (32) using tubulin proteins (Cytoskeleton, T240). Potassium hydroxide (KOH)-cleaned poly-L-lysine-coated coverslips, microscope glass slides, and double sided sticky tape were used to prepare flow chambers with a 15–20 μL volume. Taxol-stabilized MTs were immobilized on coverslips, and the surface was blocked with 5 mg/mL casein. Kinesin-1 (K560) (a gift from Gross Lab at the University of California, Irvine) was purified from bacteria as described in ref (33). For the recruitment of motors on beads, kinesin-1 was incubated with anti-His-coated beads in 50 μL of incubation buffer (80 mM Pipes pH 6.9, 4 mM MgCl₂, 1 mM EGTA, 0.5 mg/mL BSA, 2 mM DTT, 0.01 mM MgATP, 10 μM Taxol, 1 mg/mL casein) for 15 min at room temperature. Motor-bound bead solution in motility buffer (80 mM Pipes pH 6.9, 4 mM MgCl₂, 1 mM EGTA, 4 mg/mL BSA, 2 mM DTT, 10 μM Taxol, 50 μg/mL glucose, 68 μg/mL catalase, 0.1 mg/mL glucose oxidase, and 0.5% β-mercaptoethanol) was flowed into flow cells, and all measurements were performed at 24 °C. The concentration of ATP was varied as required, and in control experiments, 1 mM ATP was used. To achieve an alkaline/acidic medium, 250 mM sodium hydroxide (NaOH)/40 mM concentration of hydrogen chloride (HCl) of various volumes was added to the incubated bead–kinesin-1 mixture before infusing into the flow chamber. For each volume, a micro pH-meter (Thermo Scientific Orion 8103BNUWP ROSS Ultra pH Electrode, Catalog No. 13-642-238 with Thermo Scientific Orion Star A111 Benchtop pH Meter, Catalog No. 13-645-501) along with pH-paper strips was used to determine the pH values.

Our optical tweezer system is built on a Nikon TE300 microscope capable of collecting differential interference contrast (DIC) images as described in ref (34). For single-molecule bead motility assays, a single bead, coated with kinesin-1 motors, was trapped in the flow chamber using a 976 nm trapping laser (BL976-PAG900, Thorlabs) at a reduced optical trap stiffness (0.015 pN/nm). The bead was brought to the MT for motor interaction. A binding event was defined when kinesin-1 on the trapped bead showed movement along the MT. Diluted kinesin-1 was bound to beads to produce binding fractions of <40% to obtain a single molecule regime.⁶ For each experimental condition, at least 35 beads were tested for motility experiments.

A “run length” is defined as the distance traveled by a kinesin-coated bead along a microtubule (MT) from the point of initial binding to the point of detachment. All run length and velocity measurements were conducted under zero load by manually blocking the optical trapping beam as soon as the bead exhibited movement along the microtubule. Using the manual approach, the presence of optical trapping force during the initial movement of motors within the optical trap may introduce backload and cause early detachment, resulting in shorter runs. Runs affected by the laser force were eliminated using a standard deviation approach (Figure S2), as described below.

The movement of kinesin-1-coated beads was meticulously documented using differential interference contrast (DIC) microscopy, with an IR-2000 DAGE-MTI camera operating at 40 frames per second and a resolution of 23.6 nm/pixel. The position coordinates of the kinesin-1-coated beads were obtained with subnanometer precision by tracking their movements using the centroid position method with custom-made LabVIEW code.³⁵ The resolution of the position tracking was determined to be ∼17.0 nm (Figure S1). Through coordinate transformation, the x- and y-position coordinates were aligned to be parallel and orthogonal, respectively, to the long axis of the microtubule (Figure S2A). The standard deviation (SD) of the y-coordinate signal was calculated using a 10-point window (spanning 0.25 s). Within the optical trap, the SD of the y-signal was 5.8 ± 1.7 nm, which is significantly smaller than the average SD when the bead was outside the trap (Figure S2). For enhanced accuracy, tracks with an SD of less than 10 nm (dashed line on Figure S2B) were discarded from the analysis. This step effectively removed very short runs that detached prematurely due to the backload of the laser, ensuring the reliability of the results (Figure S2A). The velocity of the track was subsequently determined by fitting a line to the transformed x-coordinate data. Thereafter, the average values of run lengths and velocities were determined by fitting single exponential and Gaussian functions to their respective distributions.

For force generation and dwell-time measurements, position detection of a trapped object was accomplished through back-focal-plane detection of light using a position-sensitive diode (PSD).³⁶ The PSD signal in voltage was converted to the position signal in nanometers by comparing the PSD signal to the video-recorded bead position in nanometers. The stiffness of the optical trap (the conversion factor from position to force) was determined using the passive power-spectrum method.³⁷ For all force measurements, the optical trap stiffness was kept at 0.063 pN/nm.

The addition of NaOH to a motility buffer not only changes the pH but also alters the buffer’s ionic strength. This alteration can significantly impact the mechanical activities of kinesin-1 motor proteins.³⁸ Consequently, we conducted control experiments with different ionic strengths using potassium (K) acetate in a motility buffer to isolate and understand these effects. The ionic strength in a buffer was calculated using formula , where C_j and Z_j represent the concentration and charge of ion “j”, respectively. For example, for a concentration of 250 mM sodium ion from NaOH, the changed ionic strength is = 125 mM. In comparison, employing 125 mM potassium acetate yields a similar change in ionic strength, calculated as

Consequently, both methods result in an equivalent increase in the ionic strength within the buffer. To assess the influence of varying ionic strengths on kinesin-1 motility, we conducted control experiments at a 60% binding fraction using potassium (K) acetate to adjust the buffer’s ionic strength. We selected a 60% binding fraction because, at very high ionic strengths like 250 mM K acetate, single-molecule experiments were challenging due to very short run lengths. Detailed results of these control experiments are provided in the Supporting Information.

Results

Elevated pH Triggers Increased Speed and Detachment of Kinesin-1 Motors

To accurately track the effects of the surrounding pH on the mechanical actions of a kinesin-1 motor, we employed single-molecule bead motility assays (Figure 1B). Motors were specifically recruited onto streptavidin-coated polystyrene beads via biotinylated anti-His antibodies. All measurements were performed at a binding fraction smaller than 40% to ensure a single-motor regime (Figure 1B). The motility measurements were carried out at pH levels ranging from 5.5 to 9.8 with saturated ATP of 1 mM. At pH of 6.9, the kinesin-1 motor displayed an exponentially distributed run length with a mean value R_6.9 = 0.90 ± 0.03 μm (Figure 1C) and Gaussian velocity distribution with an average velocity v_6.9 = 0.78 ± 0.02 μm/s (Figure 1D). The established norms for the velocity and run length of a single kinesin-1 at this pH level correlate significantly with our results.³⁹ Interestingly, the average run length was longest at a low pH of 5.5, decreasing at higher pH values (Figure 1C). On the other hand, an increase in pH from 5.5 to 9.8 was correlated with an increase in average velocity from v_5.5 = 0.27 ± 0.02 μm/s to v_9.8 = 1.3 ± 0.04 μm/s (Figure 1D,E). The shapes of the probability distributions of velocity and run length remained unchanged. However, the standard deviation of the velocity distributions showed an increment with a corresponding increase in the pH level. This displayed the heterogeneity of the movement of kinesin, a finding that aligns with previous detections.¹⁴ The reduction in average run lengths with increasing pH deviated from the earlier reports conducted using fluorescence assay.¹⁴ The detachment rate of the motors from the MT, (referred to as the “off rate” k_off) determined by k_off = v/R, exhibited an increase as the pH level was raised (Figure 1F). Here, v and R, respectively, denote the average speed and average run lengths of the kinesin-1 molecule. The analysis of the two-sample t-test has revealed significant deviations in run length, velocity, and off rate at higher and lower pH levels relative to the neutral pH of 6.9, with a highly significant probability (p < 0.0001). Experimentally measured average run lengths and velocities of single kinesin-1 at different pH levels and ATP concentrations are listed in Table 1.

Table 1. Kinesin Motility Parameters at Various pH and ATP Levels^a.

pH (Log)	ATP (μM)	run length (μm)	velocity (μm s–1)	off rate (s–1)
5.5	1000	1.25 ± 0.04	0.27 ± 0.02	0.22 ± 0.04
6.2	25	0.88 ± 0.05	0.27 ± 0.01	0.31 ± 0.02
	100	0.93 ± 0.06	0.42 ± 0.01	0.45 ± 0.03
	250	0.99 ± 0.05	0.52 ± 0.01	0.53 ± 0.02
	500	0.97 ± 0.06	0.53 ± 0.01	0.55 ± 0.03
	1000	1.04 ± 0.06	0.52 ± 0.04	0.50 ± 0.05
6.9	18	0.66 ± 0.08	0.18 ± 0.01	0.26 ± 0.03
	40	0.75 ± 0.05	0.35 ± 0.02	0.47 ± 0.03
	200	0.83 ± 0.04	0.54 ± 0.03	0.64 ± 0.05
	1000	0.90 ± 0.03	0.78 ± 0.02	0.87 ± 0.06
8.1	1000	0.71 ± 0.04	1.1 ± 0.03	1.55 ± 0.10
9.8	12	0.43 ± 0.02	0.087 ± 0.003	0.20 ± 0.01
	40	0.42 ± 0.03	0.25 ± 0.01	0.59 ± 0.03
	100	0.44 ± 0.03	0.44 ± 0.01	0.99 ± 0.06
	200	0.45 ± 0.04	0.72 ± 0.08	1.62 ± 0.22
	1000	0.49 ± 0.04	1.30 ± 0.04	2.65 ± 0.23

^aExperimentally measured average run lengths and velocities of a single kinesin-1, along with the calculated off rate.

Accelerated Rate of ATP Turnover Increases Kinesin-1 Velocity at Elevated pH

To evaluate the effects of altered pH on the kinetic parameters associated with the chemomechanical cycle of kinesin-1 motors, we examined the unloaded velocity (v) of a single kinesin-1 molecule at different ATP concentrations ([T]) at low (pH 6.2), neutral (pH 6.9), and high (pH 9.8) pH levels.

Under all pH conditions, motor average velocity (v) increased with ATP concentration ([T]) and displayed saturation behavior consistent with Michaelis–Menten kinetics,⁴¹ as described by the following equation (Figure 2A):

1

In eq 1, the ATP-turnover rate, the ATP-binding rate, and the step size of 8 nm of kinesin-1 are, respectively, denoted as k_c, k_b, and d. The ATP turnover reflects the combined events of ATP hydrolysis, inorganic phosphate (Pi) release, and ADP (D) release. The calculated values for the maximum speed of the motor enzyme and the Michaelis–Menten constant (M-M constant) are, respectively, obtained by the equations v_max = (8.0 nm)*k_c and . Under neutral pH conditions, fitting the velocity plot against ATP concentration provides estimates of the Michaelis–Menten parameters (k_c,6.9 = 95.0 ± 6.5 s^–1, k_b,6.9 = 1.7 ± 0.37 μM^–1s^–1, v_max,6.9 = 0.76 ± 0.05 μm/s, and K_M,6.9 = 55.2 ± 12.5 μM) (Figure 2A–C). The parameters obtained at pH 6.9 are consistent with the published values.⁴⁰

Figure 2.

Increased ATP-turnover rates at elevated pH conditions: (A) Depicts the average speed of kinesin-1 motors (±standard error of the mean, SE) at varied ATP concentrations. Data points are color-coded to represent measurements under different pH conditions: pH 6.2 (orange), pH 6.9 (green), and pH 9.8 (blue). Each data set, comprising a minimum of n = 35 tracks, was subjected to nonlinear regression analysis using the Michaelis–Menten kinetics model (eq 1). The quality of the fits was confirmed by R² values of greater than 0.9. (B) Portrays the increase in the ATP-turnover rate (k_c) and the decrease in the ATP-binding rate (k_b) in response to pH changes. (C) Presents the Michaelis–Menten constant (M-M constant, K_M) and the maximum velocity (v_max) obtained from the model. (D) Displays the decreasing catalytic efficiency (v_max/K_M) of kinesin-1 with increasing pH. (E) The graph presents the exponential average run length of kinesin-1 (±standard error) across a spectrum of ATP concentrations at pH 6.2 (orange), 6.9 (green), and 9.8 (blue). (F) Displays the calculated off rate of kinesin-1 motors (±standard error) under various pH conditions: pH 6.2 (orange), pH 6.9 (green), and pH 9.8 (blue), with each data set modeled using nonlinear regression based on the off-rate equation (eq 2). (G) The graph displays the fitting parameters rate-limiting rate k₀ and rate of ATP binding k_T, obtained from the off-rate analysis. (H) Represents kinesin-1’s waiting time to complete a step calculated from t₀ =1/k₀. The error bars for panels (B), (C), (D), (G), and (H) were calculated based on the fits of models to the corresponding graphs. < > denotes the average values.

An increase in pH resulted in an enhancement in the ATP-turnover rate k_c and reduction in the ATP-binding rate k_b (k_c,9.8 = 182.7 ± 24.5 s^–1 and k_b,9.8 = 0.85 ± 0.06 μM^–1 s^–1) (Figure 2B). This led to a 2-fold increase in maximum velocity (v_max,9.8 = 1.50 ± 0.20 μm/s) and approximately 4-fold enhanced value of the M-M constant (K_M,9.8 = 214 ± 33 μM) than the conditions at neutral pH. Alternatively, acidic conditions resulted in a reduction in ATP-turnover rate k_c and enhancement in ATP-binding rate k_b (k_c,6.2 = 68.9 ± 1.8 s^–1 and k_b,6.2 = 2.6 ± 0.36 μM^–1s^–1). This caused a lower maximum velocity (v_max,6.2 = 0.55 ± 0.01 μm/s) and a 2-fold reduced value of the M-M constant (K_M,6.2 = 26.3 ± 3.7 μM) than the conditions at the neutral pH. The observations suggest that pH changes significantly influence various biochemical transitions during the motor’s mechanical cycle. As the environment changes from acidic to basic, there is a significant decrease in the ATP-binding rate (k_b), while the ATP-turnover rate (k_c) shows an increase (Figure 2B). This is in line with the finding that the speed increases with higher ATP levels (Figure 2A). Greater M-M constant values suggest that the motor’s affinity for ATP decreases as the pH rises (Figure 2C). The catalytic efficiency of the motor, calculated as the ratio of v_max/K_M, demonstrates that the motor becomes less efficient as the pH changes from acidic to basic (Figure 2D).

To further assess the detailed effects of pH, we analyzed the distance traveled at various ATP concentrations under different pH conditions. Previous studies have indicated that at neutral pH, the average run length is independent of ATP concentration.^41,42 Consistent with these findings, our empirical investigations across a range of ATP concentrations showed that the average run length of kinesin-1 motors remained unaffected, irrespective of whether the pH conditions were low, neutral, or high (Figure 2E). Hence, using average run length R = d*A, where step size d = 8 nm and A is the average number of catalytic cycles preceding detachment, we determined at pH 6.9, A = 100.0 ± 11. When the pH was raised to 9.8, the number of catalytic cycles, A, was roughly halved to 56 ± 3, while under acidic conditions, A increased to 120 ± 7. Consequently, we propose a model where the kinesin-1 off-rate k_off = v/R is modulated by ATP concentration in a multipart mechanistic framework that relies on the pH level.

To model the ATP dependence of the off rate, we first ascertained the duration t_on, which is the time the kinesin-1 motor remains bound to the microtubule during its processive movement. Within the chemomechanical cycle, the attachment time t_on depends on the motor’s waiting time at various stages to complete a full cycle (see Figure 1A). We therefore express t_on = t_T + t₀, where t_T is the waiting time for ATP binding (state-1 in Figure 1A) and t₀ represents the waiting times for ATP hydrolysis (state-2), ADP release (state-3), and phosphate (Pi) release (state-4) (Figure 1A). Neglecting the faster processes, t₀ reflects the waiting time for the rate-limiting step in the cycle. Given the ATP binding rate k_T and the rate for subsequent rate-limiting step k₀ = 1/t₀, the duration of attachment before falling off from microtubule can be formulated as t_on = A*(1/([T] * k_T)) + (1/k₀). The off rate of kinesin-1 was thus calculated as follows (eq 2)

2

The number of catalytic cycles (“A”) was deduced from run length analyses. By fitting the experimental off rate (k_off = v/R) as a function of ATP concentration ([T]) using eq 2 at various pH levels, we were able to estimate the constants for ATP binding (k_T) and the rate-limiting step (k₀). As illustrated in Figure 2G, the highest ATP-binding rate was observed at low pH (3.3 ± 0.7 μM^–1s^–1), and this rate decreased under both neutral (2.2 ± 0.4 μM^–1s^–1) and basic conditions (0.97 ± 0.1 μM^–1s^–1). On the contrary, the rate of the rate-limiting step (k₀) increased as pH moved from acidic to basic levels (Figure 2G). Consequently, the time required to complete a chemomechanical cycle after ATP binding was calculated to be () 15.8 ± 0.5 ms at pH 6.2, 11.0 ± 0.4 ms at pH 6.9, and significantly shorter at 5.8 ± 0.4 ms at pH 9.8 (Figure 2H). The analysis further provides the motor’s maximal off rate (k₀/A) as 0.5 ± 0.04 s^–1 at pH 6.2, 0.9 ± 0.1 s^–1 at pH 6.9, and 3.1 ± 0.3 s^–1 at pH 9.8, suggesting increased affinity for microtubule at lower pH values. The fitting parameters from Michaelis–Menten kinetics are listed in Table 2.

Table 2. Rate Constants from Michaelis–Menten Kinetics^a.

pH (Log)	kb(μM–1s–1)	kc(s–1)	KM(μM)	vmax(μm s–1)	k0(s–1)	kT(μM–1s–1)
6.2	2.62 ± 0.23	68.9 ± 1.8	26.3 ± 3.6	0.55 ± 0.02	63.2 ± 2.1	3.33 ± 0.50
6.9	1.65 ± 0.13	105.9 ± 2.5	64.0 ± 9.4	0.85 ± 0.02	91.3 ± 3.8	2.18 ± 0.36
9.8	0.85 ± 0.06	183 ± 15	214 ± 33	1.46 ± 0.19	171 ± 12	0.97 ± 0.05

^aATP-binding rate (k_b) and ATP-turnover rate (k_c) were determined by fitting eq 1 to the plot of average speed versus ATP concentration depicted (Figure 2A). Michaelis–Menten constant (K_M) and maximum velocity (v_max) were subsequently calculated using the values of k_b and k_c. Rate-limiting rate (k₀) and rate of ATP binding (k_T) were obtained through off-rate analysis (Figure 2F) using eq 2.

Influence of pH Variations on the Force Generation of Kinesin-1 Motor Proteins

The process of quantifying the force production of the kinesin-1 molecules was performed using a single-beam optical tweezer, as shown in Figure 3A. In this setup, a bead was captured and propelled along an MT by a kinesin-1 motor under the counteracting force of the optical trap. The forward movement of the bead persists until the motor detaches from the MT. This disengagement is indicated by the emergence of a stable force signal or “stall force” (F_s), which is maintained over a discernible “stalling period”. For kinesin-1 motors, at neutral pH, a stall force condition is confirmed when the force plateau exhibits fluctuations within ± 30 nm/s, and the stall duration is at minimum 100 ms. Detachment of the motors can also occur before the stall regime is reached (Figure 3B). We accessed the force at detachment in both situations with and without reaching the stall regime. Under acidic conditions (pH 6.2) and basic conditions (pH 9.8), 98% of detachment events occurred at forces below 5.85 pN, while at neutral pH, only 80% of events required less than 5.85 pN to detach (as shown in Figure 3C). Under acidic and neutral pH conditions, 70% of the force events resulted in a plateau at an average stall force of 4.9 ± 0.1 and 5.6 ± 0.1 pN, respectively (depicted in Figure 3D). On the contrary, for the basic medium at pH 9.8, plateau events were rarely observed (as indicated in Figure 3B). When considering a 50 ms pause before detachment as the stall duration in a basic medium with high pH, the average force at stall was calculated to be 4.3 ± 0.1 pN. This represents a 23% decrease in the average stall force compared to that under pH 6.9.

Figure 3.

Optimal force generation by kinesin-1 occurs at physiological pH: (A) Illustration of the single-beam optical trap used to measure force generation by a single kinesin-1 motor. Trap calibration was performed utilizing passive power spectrum methods, with an established trap stiffness of 0.063 pN/nm. (B) Representative force event traces generated by single kinesin-1 motors at pH 6.9 (orange), pH 6.9 (green), and pH 9.8 (blue). The gray signals at the zero level indicate the bead’s fluctuations normal to the microtubule. (C) Presents the distribution of forces at which kinesin-1 motors detach from MTs within the optical trap. Data indicate that motor detachment probability escalates with increasing force; the detachment probabilities of a single kinesin-1 at 5.8 pN force for pH 6.2 (orange, n = 205), pH 6.9 (green, n = 424), and pH 9.8 (blue, n = 210) are 0.99, 0.80, and 0.98 respectively. (D) Comparison of stall force distributions for kinesin-1 motors at pH conditions 6.2 (orange, n = 150), 6.9 (green, n = 350), and pH 9.8 (blue, n = 180). Gaussian fits to these distributions yield average stall force values, illustrating a decrease in average stall force as the pH is altered from 6.9. Motor stall is defined by the condition where the motor velocity diminishes to a level within ± 30 nm/s, and the motor remains attached to the MT for a minimum of 75 ms (stall duration). The number of force events analyzed is represented by “n” from 10 independent beads. The stall forces at different pH are statistically significant as verified using the student’s two sample t-test with probability p < 0.0001.

Increasing pH Negatively Affects the Interaction Time and Dwell Time of Kinesin-1 Motors

To further investigate the details of the kinetics of kinesin-1 under load at varying pH, we analyzed the binding time (T_B), stall duration (T_s), and dwell time (T_dwell), the pause duration between steps obtained in the optical trapping experiment. Because stalling events are rare at high pH, we opted to analyze the total time the motor is bound to the microtubule (T_B), in addition to the length of time it remains stalled (T_s), (Figure 4A). We found that both the binding time and duration of the stall decreased as the pH increased (Figure 4B,C).

Figure 4.

Study of stall duration and dwell times of kinesin-1 under varied pH: (A) A typical stalling event, illustrating the displacement of a kinesin-1 motor within an optical trap; arrows indicate the binding time (T_B) and stall duration (T_S) for such stalling events. (B) Compares T_B of kinesin-1 motor proteins at pH 6.2 (orange), 6.9 (green), and pH 9.8 (blue). Single exponential cumulative distribution fitting yields average binding times T_B,6.2 = 1089 ± 11 ms (pH 6.2, orange, n = 205), T_B,6.9 = 750 ± 10 ms (pH 6.9, green, n = 424), and T_B,9.8 = 110 ± 10 ms (pH 9.8, blue, n = 210). (C) Comparison of T_S of kinesin-1 motors. The exponential curve fittings yield average stall duration T_S,6.2 = 1077 ± 21 ms (pH 6.2, orange, n = 150), T_s,6.9 = 420 ± 9 ms (pH 6.9, green, n = 350), and T_s,9.8 = 90 ± 2.8 ms (pH 9.8, blue, n = 180). (D) Step detection algorithm was used to trace bead movement in the laser trap before reaching the stall (force < 4.2 pN). The algorithm reveals kinesin-1 taking ∼8.0 nm steps across the pH levels. (E) Distribution of dwell times for pH 6.9 at 1 mM ATP (green dots) and 0.25 mM ATP (gray dots). The distribution of dwell times at 1 mM ATP is modeled by the convolution of two exponential decays (eq 3) (green line), suggesting the presence of two waiting times t₁ and t₂, while at low ATP, the distribution fits to a single exponential decay (gray line). The fittings provide t_1,6.9 = 19.5 ± 1.4 ms, t_2,6.9 = 3.3 ± 0.5 ms for pH 1 mM ATP (green, n = 3503), and t_6.9 = 28.4 ± 3.0 ms at 0.25 mM ATP. (F) Distribution of dwell times for different pH on a logarithm scale. The fittings yield t_1,6.2 = 24.2 ± 2.0 ms, t_2,6.2 = 4.2 ± 0.8 ms for pH 6.2 (orange, n = 2200), and t_1,9.8 = 9.1 ± 0.8 ms, and t_2,9.8 = 4.5 ± 0.6 ms for pH 9.8 (blue, n = 2500). Here, n represents the number of dwells analyzed. The statistical significance of binding times, stall duration, and dwell times at different pH values was analyzed using the student’s two-sample t-test, which showed probability p < 0.0001.

Specifically, in an acidic solution, the binding time was 45% longer, and the stall duration was 157% longer, compared to those at pH 6.9. In contrast, in a basic solution, the binding time was reduced by 85% and the stall duration by 79%, compared to pH 6.9. At acidic pH, the motors stalled at lower forces and exhibited prolonged stalling periods, whereas at basic pH, the motors detached prematurely and rarely stalled. In comparison, at pH of 6.9, the kinesin-1 motors spent about 50% of their time reaching a plateau phase before stalling, suggesting that the motors’ velocity in response to an opposing load varies with the pH.

The T_dwell was obtained by fitting a step detection algorithm⁴³ to the bead position up to a force ∼4 pN (<plateau region) (Figure 4D). At all pH conditions, the probability density functions (Figure 4E,F) for dwell times are described by two sequential exponential decay processes with different time constants, t₁ and t₂.^44,45

3

Here, the constant parameter “b” represents the bin size of the distribution. The presence of two exponential decay processes suggests that the observed process is the result of two sequential rate-determining steps with different kinetics. At first, the transition process is more influenced by the faster component time t₂. As time increases, the decay component with time t₁ becomes significant. At a neutral pH with a saturated ATP concentration of 1 mM, we determined the time constants to be t₁= 19.5 ± 1.4 ms for the slower component and t₂= 3.3 ± 0.5 ms for the faster component. However, at a reduced ATP concentration of 0.25 mM, the faster component was no longer detectable, resulting in a singular exponential decay with a time constant of 28.4 ± 3.0 ms (Figure 4E). We interpret t₂ as the duration required for ATP binding, while t₁ represents the ATP-turnover time. The discrepancy between the measured (t₁) value of 19.5 ms and the ATP-turnover time of 11 ms, obtained from off-rate analysis, is likely due to the backward loading effect caused by the optical trap force. With increasing pH, the ATP-turnover time t₁ declined to t₁= 9.1 ± 0.8 ms at pH 9.8, while the ATP-binding time t₂ increased to t₂= 4.5 ± 0.6 ms. On the contrary, in an acidic environment of pH 6.2, ATP-binding and turnover times were prolonged, ATP-turnover time t₁= 24.2 ± 2.0 ms, and ATP-binding time t₂ = 4.2 ± 0.8 ms (Figure 4F). This comprehensive analysis has indicated that by increasing the pH level from 6.2 to 9.8, there is a discernible reduction in the total net dwell time (t₁ + t₂). Table 3 displays the parameters from the force measurements and dwell time analysis.

Table 3. pH-Dependent Properties of Kinesin in Optical Trap^a.

pH (Log)	<FS> (pN)	TS(s)	TB(s)	t1(ms)	t2(ms)
6.2	4.9 ± 0.1	1.08 ± 0.02	1.09 ± 0.01	24.2 ± 2.0	4.2 ± 0.8
6.9	5.6 ± 0.1	0.42 ± 0.01	0.75 ± 0.01	19.5 ± 1.4	3.3 ± 0.5
9.8	4.3 ± 0.1	0.090 ± 0.003	0.11 ± 0.01	9.1 ± 0.8	4.5 ± 0.6

^aAverage stalling force (<F_S>), average stall duration (T_S), and average binding time (T_B) were determined from force measurements. Fitting parameters t₁ and t₂ were determined from dwell-time analysis using eq 3 (Figure 4F).

Discussion

This study analyzes the effects of pH on the function and kinetics of kinesin-1 motor proteins. Kinesin-1 shows a balanced performance at a physiological pH of 6.9 in comparison to the acidic and alkaline medium. At higher pH levels, an increase in the velocity is accompanied by higher detachment rates, resulting in shorter run lengths. Conversely, at lower pH levels, a slower detachment rate extends the run length, although the speed is reduced compared to that at neutral pH.

At the pH of 6.9, the rates of ATP turnover (k_c) and ATP binding (k_b) align with the observed maximum velocity and run length of the kinesin-1 motor proteins.⁴⁶ Importantly, the ATP-turnover rate (k_c) is found to double when the pH is increased to 9.8, suggesting a concomitant increase in the maximum velocity (Figure 3B,C). This heightened rate of ATP turnover implies an increase in the rate of subtransition events within the kinesin’s chemomechanical cycle, including ATP hydrolysis, ADP release, and phosphate ion (Pi) release (Figure 1A). The analysis of the off rate versus ATP and the dwell time displays roughly 2-fold faster completion of the rate-limiting steps at a high pH of 9.8. The pathway governing Pi release from the bound head could precipitate premature motor detachment, while the release of ADP from the leading tethered head causes faster completion of the cycle (state-3, Figure 1A).²⁸

ATP-independent average run lengths at varying pH levels suggest that the pH of the medium can regulate the travel distances. Furthermore, at low ATP concentrations, kinesin-1 motors remain attached to the microtubule while waiting for ATP binding. This observation suggests that the ATP waiting state (state-1, Figure 1A) exhibits a stronger affinity for the microtubule compared to other states in the kinesin-1 cycle. Our results support that state-1 involves a two-head bound configuration, as proposed in ref (47), which may decrease the likelihood of detachment compared to a one-head bound state, as depicted in Figure 1A. This interpretation is inconsistent with recent findings published in ref (48). In addition, with a lower detachment rate for the ATP waiting state-1, the accelerated kinetics observed at high pH levels result in an increased probability of phosphate ion release, thereby heightening the likelihood of kinesin-1 detachment in its sensitive state-3 (Figure 1A). Under acidic conditions, a slower ATP turnover rate (k_c) suggests that the cycle involving kinesin-1 motors takes longer to complete. This is further evidenced by 45% and 24% increases in waiting times after ATP bonds with the motor, as indicated by the off-rate fitting parameter (t₀) and the dwell-time analysis parameter (t₁), respectively. This slowdown aligns with the decreased speed of kinesin-1 motors at low pH. A decrease of about 50% in the Michaelis–Menten constant (K_M) implies that kinesin-1 has a greater tendency to hold onto ATP in an acidic environment, reflecting a stronger affinity. With lower probability of detachment in the ATP waiting state, the motor travels increased travel distances under acidic conditions.

Our findings propose a model for the multimodal regulation of kinesin states by pH. However, some of these observations could be attributed to modifications in the tubulin structure induced by pH.⁴⁹ The C-terminal region of tubulin mediates the transition between the strong and weak binding states of kinesin-1 to the MT, which in turn affects kinesin-1 processivity.^13,50 Furthermore, pH levels influence the post-translational modifications of tubulin,^51,52 which in turn regulates the processivity of kinesin motors. Elevating pH by adding NaOH not only induces changes in protein conformation but also increases buffer ionic strength.³⁸ Adjusting ionic strength by regulating salt concentration affects kinesin motility; varying Pipes concentration increases average velocity and reduces the average run length,⁵³ while increasing potassium chloride (KCl) reduces both average velocity and average run length under zero load.³⁸ However, KCl does not affect run length with assisting load.²⁸ In our control motility assays, performed at a 60% binding fraction with increased ionic strength from K-acetate, we observed a reduction in the average run length of kinesin-1, although its average velocity remained unchanged (Figure S3). A comparable ionic strength did not precipitate an equivalent reduction in average run length as observed at high pH (Figure S3). This suggests that the influence of pH on the run length can be attributed to the synergistic effects of both the ionic strength and the structural destabilization caused by elevated pH. The binding energy between the motor domain of KIF5/Eg-5 and MT exhibits a minimal value around pH 5 and increases by approximately 4 kcal/mol when the pH shifts to 10.⁵⁴ Considering that there is a comparable pattern for the binding energy between kinesin-1 and MT, under acidic and alkaline conditions, the kinesin-MT interaction transitions from stronger to weaker. Alteration in force sensitivity as well as structural transition displacements during the ATPase cycle can regulate detachment kinetics of kinesin-1 under load.⁵⁵ Hence, the experimentally observed variations in binding time, stall duration, detachment force, and stall force across different pH conditions suggest significant structural changes and shifts in interaction energy, profoundly impacting kinesin motility.

Conclusions

Our study illustrates that the kinesin-1 motor regulates its travel distances by adjusting velocity and off rates in response to changes in pH levels. Any changes in an ambient pH result in a compromise between these two aspects, thereby influencing the overall performance of the kinesin-1 motor. This balance between the speed of kinesin-1 and its rate of release underscores a sophisticated adaptation mechanism, which is particularly important in physiological contexts where the pH can vary. Additionally, our findings could pave the way for further investigation into the regulatory processes that this motor may employ or require to adjust its functionality when faced with changes in intracellular pH. Organelles, such as endosomes, lysosomes, mitochondria, and the Golgi apparatus, maintain different pH levels, which requires a hydrogen ion leakage pathway that could, in turn, affect the surrounding cytoplasmic pH.⁵⁶ Such shifts in local pH levels could also influence motor-driven transport activities. Future research should consider whether this pH-sensitive behavior is also present in vivo, which would significantly enhance our understanding of intracellular transport mechanisms in a variety of cellular states and pathologies associated with pH dysregulation.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jpcb.4c03850.

• Additional methods and results for the motility assays to assess the effect of ionic strength on the biophysics of kinesin-1 proteins (PDF)

Author Contributions

S.K.T conceptualized and designed the project; S.K.T, F.B, M.B, and H.M.B performed the motility and optical trapping (OT) experiments and analyzed the data. S.K.T. wrote the paper. All authors approved the final version of the manuscript.

This work is supported by funds provided to S.K.T. by the Department of Natural Sciences and Office of Research and Sponsored Programs (ORSP) and the UM-Dearborn-UM-Flint Collaborative Research Grants, University of Michigan-Dearborn.

The authors declare no competing financial interest.