Polymerization and stability of actin conjugated with polyethylene glycol

Abstract

This study aimed to elucidate the impact of polyethylene glycol (PEG) conjugation on protein-protein interactions by investigating the properties of PEG-conjugated actin (PEG-actin). Various PEG molecules were covalently bound to actin monomers by reacting maleimide groups with Cys374 on the actin. The apparent polymerization rate constant (k_on^app) and the critical concentration (Cc) were measured by fluorescence spectroscopy using pyrene-labeled actin, as a function of the portion of PEG-actin and the molecular mass of the conjugated PEG (750 to 10000 Da). The k_on^app gradually decreased as the percentage of PEG-actin increased. At 90% PEG-actin, the k_on^app decreased substantially as the PEG size increased, resulting from a modulation of C-terminus by the conjugated PEGs and their steric hindrance. The Cc was slightly increased by PEG conjugation in the content by up to 50%. Meanwhile, 90% PEG-actin exhibited a substantial increase in Cc. The Cc was almost linearly related to the gyration radius of PEG. These results suggest that the PEG conjugation to actin impedes the association of actin with the filament in a PEG size-dependent manner. Furthermore, the stability of PEG-actin against an extrinsic factor was assessed. PEG-actins >2000 Da were more susceptible to digestion than intact actin when the PEG-actin monomer was subjected to α-chymotrypsin. Thus, conjugation of PEG to Cys374 on actin did not protect actin monomers against their proteolysis by α-chymotrypsin.

Significance

Conjugating PEG to proteins can improve their properties such as stability; therefore, conjugation can be applied in drug delivery fields. Here, we showed that actin molecules conjugated with PEGs exhibit an alteration in the polymerization properties of actin with a dependence on the content and size of the conjugated PEGs. These findings provide insights into the selection of PEG size to regulate the binding affinity between proteins.

Introduction

Polyethylene glycol (PEG) is used for protein modification and condensation because it acts as an inert polymer or osmolyte to prevent the non-specific association of biological substances. The conjugation of PEG to therapeutic agents, termed PEGylation, can improve their solubility and reduce immunogenicity in vivo; therefore, PEGylation of proteins has been exploited in pharmaceutical fields, including drug delivery [1–3]. However, the impact of PEG conjugation on protein function is still not fully understood owing to a lack of biophysical characterization. To clarify the impact of PEG conjugation on protein-protein interactions, we examined the polymerization of PEG-conjugated actin, that is, the association of actin monomers with actin filaments, as a model because the polymerization kinetics and structure of actin have been well studied [4,5].

The properties of actin filaments and their surrounding conditions are important for myosin binding and force generation. Hypermobile water around actin filaments, which moves faster than bulk water, is thought to be involved in energy transduction from ATP to mechanical work [6]. Hydration also influences the binding between actin and myosin, potentially by modulating associated entropy changes [7]. PEGs in solution can induce depletion forces between proteins because of the excluded volume changes around them [8,9]. PEG conjugation to proteins may influence hydration shells. Therefore, it would be interesting to investigate whether the conjugation of actin with PEG has exceptional effects such as facilitating polymerization and filament stabilization.

Conjugation of PEGs to actin has been previously reported (Figure 1) [10]. Considering the microdelivery system described in a previous study, PEGs were treated as virtual materials to be carried onto actin filaments driven by myosin motors in an in vitro motility system. This previous study showed that conjugation with PEG up to a molecular mass of 2 kDa, whose gyration radius is slightly smaller than that of the actin monomer, does not affect the sliding velocity of actin filaments. Conjugation with higher-molecular-mass PEGs drastically decreases the motile fraction of actin filaments containing only 20% PEG-conjugated actin. These results indicate a limitation in the carrying capacity of actin filaments as delivery vehicles [10]. Furthermore, improving actin stability and preservation remains a challenge for microdevice applications [11]. Therefore, if PEG conjugation stabilizes actin, this method can be useful for artificial microdevices.

Figure 1.

(A) Structure of PEG-maleimide. (B) The band pattern of actin conjugated with different PEGs (2000, 5000, and 10000 Da in molecular mass) detected by SDS-PAGE. (C) Comparison of sizes of actin monomers (PDB: 1ATN) composed of 4 SD (1, 2, 3, and 4) and PEGs in each molecular mass. R_g denotes the gyration radius. Each PEG was covalently bound to Cys374 located in C-terminus. Reproduced with permission from [10], copyright 2022 by Wiley. PEG: polyethylene glycol; SDS-PAGE: sodium dodecyl-sulfate polyacrylamide gel electrophoresis; SD: subdomain.

Here, we aim to clarify the properties of PEG-conjugated actin from the following perspectives: (1) the impact on the polymerization of actin revealed by fluorescence spectroscopy with pyrene-labeled actin, (2) the influence on the stability against extrinsic factors by evaluating the α-chymotryptic digestion products of actin, and (3) thermal stability determined by circular dichroism (CD) spectroscopy.

Materials and methods

Preparation of PEG-conjugated actin

The Institutional Animal Care and Use Committee of the Yamagata University approved all the procedures and protocols used in the animal experiments (approval number R5056). Actin was purified from rabbit skeletal muscles. Methoxy-PEG maleimide reagents with average molecular masses of 750 Da (product number 712558), 2 kDa (product number 731765), 5 kDa (product number 63187), and 10 kDa (product number 712469) were purchased from Sigma-Aldrich (St. Louis, MA, USA) (Figure 1A). To covalently bind PEG to Cys374 on actin monomers, PEG-maleimides and actin monomers were mixed at a molar ratio of 40:1 in a modified G-buffer (5 mM HEPES (pH 7.8), 0.1 mM CaCl₂, 0.2 mM ATP, and 0.1 mM TCEP) and incubated at 4°C for 12 h. After terminating the reaction with 50 mM DTT, the actin monomers in the mixture were polymerized into the filaments by adding 1/9 the volume of F-solution (1 M KCl and 20 mM MgCl₂). The actin filaments, including PEG-conjugated actin, were ultracentrifuged at 150,000×g to separate the unbound PEG-maleimide (ultracentrifuge CS100EX; Hitachi Co., Ltd., Japan). The precipitate was dialyzed against G-buffer (2 mM Tris-HCl (pH 8.0), 0.1 mM CaCl₂, 0.2 mM ATP, 0.02% 2-mercaptoethanol). Finally, unexpected aggregates in the sample were removed by ultracentrifugation. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) revealed that none of the actin molecules bound to more than one PEG molecule per actin molecule for all tested PEG sizes. The yield of PEG-conjugated actin was approximately 93% as determined by SDS-PAGE (Figure 1B). The PEG-conjugated actin is referred to as xx-PEG-actin (xx is the molecular mass of PEG).

Quantification of polymerized actin

Polymerization of PEG-actin was monitored by fluorescence spectroscopy (F-2500; Hitachi Co., Ltd.) using the pyrene-labeled actin method [12]. Pyrene-labeled actin was prepared by reacting actin with N-(1-pyrenyl)iodoacetamide (Setareh Biotech, Tyler, TX, USA). The fluorescence intensity of pyrene bound to Cys374 on actin increases several times when pyrene-labeled actin was polymerized [13]. Therefore, the amount of polymerized actin was determined by measuring the fluorescence intensity. Samples containing intact, PEG-, and pyrene-labeled actin were prepared in G-buffer. The percentage of pyrene-labeled actin was set at 10% for all measurements and the percentage of PEG-actin was varied at 0%, 20%, 50%, and 90%. Polymerization was initiated by adding 1/9 volume of 10×P-solution (1 M KCl, 20 mM MgCl₂, and 0.1 M HEPES (pH 7.4)), followed by the measurement of fluorescence intensity for 30 min at 25°C. The final actin concentration and volume were set to 0.2 mg/mL and 0.5 mL, respectively. The experimental conditions of the spectrometer were 365-nm excitation, 407-nm emission, 2.5-nm slit, and 700-V PMT voltage. A 0.8 mL quartz cuvette (type 52; Starna Scientific Ltd., UK) was used. The effect of phalloidin on polymerization was also examined. Phalloidin (Wako Pure Chemical Corp., Japan) was added to the samples at a 4-time molar ratio to actin immediately before measurement.

Determination of apparent polymerization rate constant (k_on^app) and the critical concentration (Cc)

To evaluate the polymerization properties, the rate constant was determined using a simple model of the G-F transformation of actin, as shown in Equation (1): G- and F-actin denote the actin monomers and filaments, respectively.

$$\frac{d\left[A_F\right]}{dt}=k_{\mathrm{on}}\left[A_G\right]\left[A_{\mathrm{end}}\right]-k_{\mathrm{off}}\left[A_{\mathrm{end}}\right]$$ (1)

where [A_F] is the concentration of actin constituting F-actin, [A_G] is the concentration of G-actin, [A_end] is the concentration of the end of actin filament, and k_on and k_off are the polymerization and depolymerization rate constants, respectively. This model does not consider the distinction between the pointed and barbed ends of actin filaments. In addition, differences in the rate constant between ADP-actin and ATP-actin as well as between Mg-actin and Ca-actin were not considered.

At equilibrium, the left-hand side of Equation (1) becomes zero. Therefore, the concentration of G-actin, which represents the ratio of the k_off to the k_on, is constant, as shown in Equation (2):

$$Cc\equiv \left[A_G\right]=\frac{k_{\mathrm{off}}}{k_{\mathrm{on}}}$$ (2)

where Cc is the critical concentration of actin [14].

Equation (1) can be transformed into Equation (3), which represents the relationship between the concentration of actin constituting F-actin and elapsed time.

$$\left[A_F\right]=\left[A_0\right]-\frac{k_{\mathrm{off}}}{k_{\mathrm{on}}}+(\frac{k_{\mathrm{off}}}{k_{\mathrm{on}}}-\left[A_0\right])\exp \left(-k_{\mathrm{on}}\left[A_{\mathrm{end}}\right]t\right)$$ (3)

where [A₀] is the total concentration of actin and t is the elapsed time after the initiation of actin polymerization. Since [A_F] is proportional to the fluorescence intensity of pyrene, the time courses of the fluorescence intensity in the polymerization process were fitted into a modified Equation (3), i.e. intensity=α–β exp(–k_on [A_end](t–t_lag)), to determine the rate constant, where α and β are also fitting parameters, and t_lag is the lag time. However, in the cases of higher contents and larger sizes of PEG conjugates, the intensity did not saturate within 1800 s, and fitting was achieved. In this study, [A_end] was not controlled; therefore, we obtained only k_on[A_end], which is treated as the k_on^app. The k_on^app value for each PEG-actin concentration was normalized to that of the control (0% PEG-actin).

The Cc was determined by gradually diluting the F-actin solution from 0.125 to 0.05 mg/mL except for 90% of 10k-PEG-actin, which was tested in a concentration range of 0.3 to 0.1 mg/mL. Because a linear relationship between actin concentration and fluorescence intensity was observed, the Cc was obtained at the point where the fluorescence intensity was zero.

Resistance of actin monomers to α-chymotryptic digestion

To evaluate the stability of actin against an extrinsic factor, PEG-actin monomer was subjected to α-chymotrypsin (C-4129; Sigma-Aldrich). To initiate the α-chymotryptic digestion, actin and α-chymotrypsin were mixed in G-buffer at 25°C at 0.25 mg/mL and 0.01 mg/mL, respectively. Samples were collected at 1, 10, and 30 min and the reaction was terminated using an SDS-containing solution to denature the proteins. The amount of digested components was analyzed using SDS-PAGE. The gels were stained with Coomassie Brilliant Blue (CBB stain one; Nacalai Tesque, Japan).

Thermal stability of actin monomers

The CD spectra of the PEG-actin monomers were recorded using a CD spectrometer (J-820; Jasco Co., Japan). The solution conditions were 5 mM HEPES (pH 7.4), 0.1 mM CaCl₂, 0.1 mM DTT, 0.01 mM ATP, and 0.05 mg/mL actin. The thermal unfolding transition was monitored by measuring the ellipticity at 222 nm as a function of temperature in the range of 25–85°C (heating rate: 2°C/min) [15].

Results

Apparent polymerization rate constant of PEG-actin

The amount of polymerized actin corresponds to the fluorescence intensity of pyrene; therefore, the time course of the fluorescence intensity was used to determine the k_on^app (Figure 2A and 2B). In this study, the k_on^app was evaluated to examine the effects of PEG conjugation. The k_on^app gradually decreased with an increase in the percentage of PEG-actin (Figure 2C). The dependence on the PEG size was obscured by up to 50%. At 90% PEG-actin, 5 k-PEG-actin considerably decreased k_on^app. A slight increase in the fluorescence intensity was observed for 90% 10k-PEG-actin (Figure 2B). The lag time needed for nucleation, that is, the formation of actin oligomers, increased slightly with increasing percentage of PEG-actin up to 50% (Figure 2D). Addition of 90% PEG-actin resulted in a longer lag time. However, the PEG size dependence was obscured. Actin trimers are likely to form via the lateral binding of incoming monomers to dimers [16]. Therefore, the sensitivity of the binding of incoming actins to the PEG size may differ between the nucleation and elongation of actin filaments.

Figure 2.

The polymerization of PEG-actin at 0.2 mg/mL. (A) An example of time courses of pyrene fluorescence intensity of actin solutions with a parameter of the percentage of 750-PEG-actin. Black, blue, purple, and brown lines indicate 0, 20, 50, and 90% 750-PEG-actin, respectively. (B) An example of time courses of pyrene fluorescence intensity of 90% PEG-actin solutions with a parameter of the molecular mass of conjugated PEG. The intensity of each PEG-actin was normalized to the maximum value of its respective control. Blue, purple, pink, and green lines indicate 750-, 2k-, 5k-, and 10k-PEG-actin, respectively. The k_on^app (C) and lag time (D) as a function of the percentage of PEG-actin and the molecular mass of conjugated PEG. Blue, purple, pink, and green circles indicate the cases of 750-, 2k-, 5k-, and 10k-PEG-actin, respectively. The k_on^app and lag time of 90% 10k-PEG-actin were not determined because this exhibited little intensity changes for 1800 s. Each symbol and bar indicates the average and standard error of 3 experiments, respectively. The k_on^app was normalized to the control value, which was determined in each experiment (see the data point at 0% PEG-actin). PEG: polyethylene glycol; k_on^app: apparent polymerization rate constant.

Critical concentration of PEG-actin

The Cc was determined from the x-intercept by linearly fitting the relationship between the fluorescence intensity and actin concentration (Figure 3A–D). Figure 3E shows that the Cc of 20% PEG-actin was similar to that of intact actin (0% PEG-actin). PEG-actin at 50% slightly increased the Cc; however, no significant difference was observed between intact actin and 750-, 2k-, or 10k-PEG-actin. In addition, the dependence on PEG size was obscure within this range. Meanwhile, 90% PEG-actin considerably increased the Cc, and the dependence on the PEG size was obvious. The Cc exhibited a linear relationship with the square root of the PEG mass, which also reflects the gyration radius (Figure 3F).

Figure 3.

The Cc of PEG-actin. Examples (one of 3 experiments) of pyrene fluorescence intensity related to actin concentration in the cases of 750- (A), 2k- (B), 5k- (C), and 10k- (D) PEG-actins. In the case of 90% 10k-PEG-actin, the range of actin concentrations was taken up to 0.3 mg/mL (right panel). Black, blue, purple, and brown symbols indicate the case of 0, 20, 50, and 90% of PEGs, respectively. The difference of ranges at the y-axis among panels is due to distinct data sets. (E) The critical concentration as a function of the percentage of PEG-actin and the molecular mass of 750- (blue), 2k- (purple), 5k- (pink), and 10k- (green) PEG. Each symbol and bar indicates the average and the standard error of 3 experiments, respectively. No significant difference (ns) between intact actin (control) and 750-, 2k-, or 10k-PEG-actin at 50% content: p>0.05, Welch’s t-test. (F) At 90% PEG-actin, a relationship between the critical concentration and the square root of the molecular mass of PEG. Cc: critical concentration; PEG: polyethylene glycol.

Effect of phalloidin on the polymerization of PEG-actin

The addition of phalloidin, an actin filament stabilizer, increased the polymerization rate of 5k-PEG-actin to the same level as that of intact actin (Figure 4). The Cc of 5k-PEG- and 10k-PEG-actins also decreased with the addition of phalloidin (Supplementary Figure S1). Therefore, phalloidin stabilizes actin filaments by tethering the subunits, regardless of PEG conjugation.

Figure 4.

Effect of phalloidin on the polymerization of PEG-actin at 0.2 mg/ml. Time courses of pyrene fluorescence intensity after initiation of the polymerization. Gray, blue, black, and red lines indicate in the absence of phalloidin, intact actin, 90% 5k-PEG-actin, and in the presence of phalloidin, intact actin, and 90% 5k-PEG-actin, respectively. PEG: polyethylene glycol.

Resistance of PEG-actin to α-chymotryptic digestion

To determine whether PEG conjugation protects actin monomers from proteases, PEG-actin was subjected to α-chymotryptic digestion. Bands at 33 kDa and 35 kDa for digested actin appeared 10 min after the initiation of the reaction (Figure 5A). The amount of digested PEG-actin conjugated with 2–10 kDa PEGs was significantly higher than that of intact actin, 10 min after the initiation of digestion (Figure 5B). Meanwhile, no significant difference in the digestion amount was observed between intact and 750-PEG-actin. Therefore, the conjugation of PEGs to Cys374 does not protect actin from α-chymotrypsin digestion.

Figure 5.

Alpha-chymotryptic digestion of PEG-actin. (A) An example image of the band pattern of digested components of intact actin (left) and 2k-PEG-actin (right) on an SDS-PAGE gel. Reaction times are shown at the top of the gel. (B) The digestion ratio of intact actin (black), 750- (blue), 2k- (purple), 5k- (pink), and 10k- (green) PEG-actins over time, calculated from the ratio of the decrease in the amount of undigested actin in each reaction time. The curves indicate a hyperbolic fit. Each symbol and bar indicates the average and the standard error of 3 experiments. Star symbol indicates statistical significance between intact actin and 2k-, 5k-, or 10k-PEG-actins at 10 min: *p<0.05 (n=3), Welch’s t-test. There was no significant difference between intact and 750-PEG-actin. PEG: polyethylene glycol; SDS-PAGE: sodium dodecyl-sulfate polyacrylamide gel electrophoresis.

Thermal stability of PEG-actin

The thermal stability of PEG-actin was evaluated by measuring its CD spectrum. Figure 6 shows the ellipticity of the PEG-actin monomers during the unfolding process at elevated temperatures. In the process from native to unfolded actin, the ellipticity was decreased owing to the decrease in helix, and the transition temperature was estimated to be 61°C for intact actin monomers. 2k-PEG- and 5k-PEG-actin monomers exhibited transition temperatures of 63°C and 62°C, respectively. PEG conjugation slightly increased the thermal stability of the actin monomers.

Figure 6.

Temperature dependence of the ellipticity of intact (black), 2k-PEG- (purple), and 5k-PEG- (pink) actins, measured at 222 nm. Solid line indicates a fitting curve for the determination of the transition temperature (T_m) at the midpoint between the upper and the lower limits of observed ellipticity, as previously reported [15]. PEG: polyethylene glycol.

Discussion

Impact of PEG conjugation on the polymerization of actin

Dynamic processes such as the development and decay of actin filaments are complicated by various factors such as nucleation, severing, and dependence on cations and nucleotides [17]. Therefore, we considered the apparent rate constant instead of the exact rate constant for the polymerization of actin filaments. To clarify the properties of PEG-actin, the normalized rate constants were compared as a function of the PEG-actin content and the size of the conjugated PEG.

PEG conjugation to Cys374 on actin inherently induced low polymerization ability, as 90% PEG-actin considerably decreased k_on^app and increased the Cc. The flexible C-terminus of actin, where PEG is bound, is located near the D-loop of the following actin constituent along the actin protofilament, and this plays an important role in the stability and integrity of the filament structure [18]. It was predicted that Phe375 at the C-terminus would move toward Val47 in the D-loop during polymerization [19]. Therefore, the conjugation of PEG to Cys374 may modulate the flexibility of the C-terminus and affect the interaction of Phe375 with the D-loop, resulting in impaired polymerization. In addition, the finding that larger PEGs induced lower k_on^app and higher Cc suggests that steric hindrance from PEGs on filament ends hampered the binding of incoming actin monomers.

Nevertheless, a half-and-half mixture of intact and PEG-actin exhibited a Cc at the same level as that of intact actin alone, and the Cc was independent of PEG size. Likewise, the k_on^app value of this mixture was also independent of the PEG size, although the k_on^app gradually decreased as the percentage of PEG-actin increased. The independence of PEG size cannot be simply explained by steric hindrance from the PEGs to the access of incoming actin to the C-terminus.

The gyration radius of PEG can be calculated from its relationship with the square root of the molecular mass according to polymer theory [20,21]. The gyration radii of the 750-, 2k-, 5k-, and 10k-PEGs were approximately 1.0, 1.8, 3.0, and 4.5 nm, respectively (Figure 1C). The distance between the neighboring actin constituents in the long helix of the filament is approximately 6 nm. At concentrations of up to 50% PEG-actin, conjugated PEGs can have a distance of 12 nm between alternating actin components (assuming a homogeneous assembly of intact and PEG-actin). This distance allowed 750–10k of PEGs to be arranged on the filaments (Figure 7A left). In this situation, steric hindrance from the conjugated PEGs may have little effect on polymerization because of the flexibility of the PEGs and adequate adaptation to the configuration of the filaments. In contrast, 90% PEG-actin may produce multiple steric interferences between neighboring PEGs in the longitudinal and lateral directions along the protofilaments (Figure 7A right). Because of the decreased space available for flexible PEGs, the binding site was masked by conjugated PEGs. Thus, the PEG size-dependent suppression of polymerization became evident.

Figure 7.

(A) Schematic diagram of part of the actin filament (PDB: 6BNO) copolymerized with 5k-PEG-actin and intact actin. An actin monomer is drawn based on an atomic model (PDB: 1J6Z). The PEG of 5 kDa has the gyration radius of 3 nm, and the distance between neighboring actin components in long-helix of the filament is approximately 6 nm. (B) Apparent external work versus gyration radius of PEG. PEG: polyethylene glycol.

Meanwhile, the decrease in k_on^app in the mixture may result from the necessity of alternate binding between intact and PEG-actin because consecutive binding between PEG-actins was unfavorable, as demonstrated by 90% PEG-actin.

When the F-actin solution was rapidly diluted below the Cc, the number of actin constituents in the filaments decreased due to predominant depolymerization. The time courses of the decreasing pyrene intensity were fitted to an exponential curve to determine the k_off^app (Supplementary Figure S2). The k_off^app increased approximately 2-fold as the percentage of PEG-actin increased. No significant differences were observed between the 750-PEG-actin and 10k-PEG-actin. Therefore, the increase in Cc may be due to a decrease in the polymerization rate rather than an increase in the depolymerization rate. However, because we treated the apparent rate constants without determining the concentration of the filament end, sophisticated methods are required to obtain accurate rate constants and precise interpretations, such as direct observation of single actin filaments [22].

The Cc corresponds to the dissociation constant or reciprocal of the equilibrium constant, which associates with the free energy changes. Here, we compare the equilibrium constants between intact actin and PEG-actins and consider a work (△W) that PEG does to resist the binding of actins as follows.

$$\mathrm{\Delta }W=k_{B}T{\log }_e\left(\frac{K_0}{K_{PEG}}\right)$$ (4)

where K₀, K_PEG, k_B, and T are the equilibrium constants for intact actin and PEG-actin, the Boltzmann constant, and absolute temperature, respectively. The relationship between the calculated value of the work and the PEG size is shown in Figure 7B, based on the data in Figure 3F. The work was 8×10^–21 J for 10k-PEG-actin (R_g 4.5 nm). Assuming that △W=force×3-nm overlap between the PEGs, the force is approximately 3 pN. This estimated value is in the same range as the force measured directly (1 pN) for actin polymerization [23].

Although 90% of the 5k-PEG-actin had a lower polymerization ability, phalloidin facilitated polymerization and stabilized the PEG-actin filaments. This result indicates that the intrinsic polymerization ability of 5k-PEG-actin is retained even with the restricted movement of Phe375 and the steric hindrance imposed by the 6-nm size of the 5k-PEG moiety. Because of the effect of phalloidin, tetramethylrhodamine-phalloidin-labeled actin filaments conjugated with 750–10k PEGs were clearly observed under a fluorescence microscope as shown in our previous work, which also demonstrated that intact actin and PEG-actin were homogeneously copolymerized into filaments [10]. Stabilized PEG-actin filaments showed low bundling ability even in the presence of higher concentrations of MgCl₂ (Supplementary Figure S3). This suppression of the side-to-side association of filaments may be useful for controlling the interactions between filaments, as demonstrated in a previous study on semiflexible polymer networks [24].

Alpha-chymotryptic digestion and thermal stability of PEG-actin

PEGylation of therapeutic agents is widely used in pharmacological applications because PEG conjugation can improve the stability and duration of drug degradation in vivo [2]. If actin PEGylation improves the preservation and stabilization of its steric structure, it may be valuable for microdevices driven by actin filaments and myosin motors [11,25]. In this study, the α-chymotryptic digestion assay showed that actin digestion was promoted by conjugation with 2–10 kDa PEGs. Because we used PEG-maleimide, the PEGs covalently bound to Cys374 in subdomain 1 of the actin monomer [26]. Alpha-chymotrypsin can cleave multiple actin sites, including subdomain 2 [27]. Although the larger PEGs protected subdomain 1, the other domains were easily targeted by α-chymotrypsin because of the lack of PEG protection.

In CD spectroscopy, the transition temperature between folding and unfolding slightly increased from 61 to 63°C by PEGylation of actin. Therefore, PEGylation is expected to improve the thermal stability of actin.

Conclusion

PEG conjugation of actin decreased the k_on^app and increased the Cc, indicating a shift in equilibrium from filamentous to monomeric forms. The k_on^app decreased almost linearly as the percentage of PEG-actin increased. At 90% PEG-actin, a larger PEG induced a lower k_on^app and a higher Cc. Modulating the C-terminus of actin by PEG conjugation may weaken the filament stability owing to a lack of connectivity between the C-terminus and the D-loop. In addition, a high content of conjugated PEGs is likely to hamper the binding of incoming actin to the filament ends, owing to steric hindrance. Meanwhile, the mixture of intact actin and PEG-actin at 50% retained a control level at the Cc, independent of the PEG size. This suggests an allowance for the configuration of flexible PEG moieties along the actin filaments. Despite the high content and large size of PEG-actin, phalloidin stabilized the PEG-actin filaments, suggesting that its ability to tether the subunits together overcame the weak associations between PEG-actin molecules due to steric hindrance. Alpha-chymotryptic digestion of actin monomers containing D-loop cleavage was enhanced by the conjugation of larger PEGs to Cys374. Therefore, the PEG conjugation did not improve the preservation of actin against α-chymotryptic digestion. Overall, the tested PEG conjugation suppressed the association of actin with filaments in a manner dependent on PEG content and size, with no exceptional effects. These findings provide insights into the control of protein associations.

Conflict of interest

The authors declare no conflict of interest.

Author contributions

Masaya Sagara: Conceptualization, Methodology, Data curation, Investigation, Visualization, Writing, review, and editing. Kuniyuki Hatori: Conceptualization, Methodology, Data curation, Investigation, Formal analysis, funding acquisition, Supervision, Visualization, Writing the original draft, writing the review, and editing.

Data availability

The evidence data generated and/or analyzed during the current study are available from the corresponding author on reasonable request.