Real-Time Observation of Enzymatic Polyhydroxyalkanoate Polymerization Using High-Speed Scanning Atomic Force Microscopy.

The initial stage of in vitro polyhydroxyalkanoate (PHA) polymerization by PHA synthase from Ralstonia eutropha (PhaCRe) on a mica substrate in water was observed using high-speed scanning atomic force microscopy (HS-AFM). Before PHA polymerization, the adsorption-desorption cycle of the PhaCRe molecule on mica was observed in real time. For approximately 30 s after the addition of the PHA monomer, no significant change was observed on the mica substrate, but PhaCRe could be transformed into an active enzyme in water upon contact with the monomer during this period. Subsequently, linearly elongating rod-shaped objects were observed on the mica substrate, plausibly as a result of the polymerization reaction. The height of these elongating objects was considerably larger than the expected height for a single PHA chain. This observation suggests that PHA chains generated during the reported experiments might form some kind of a semiregular structure.

Introduction

Polyhydroxyalkanoates (PHAs) are carbon- and energy-storage materials found in various microorganisms and can be synthesized from renewable resources such as sugars, plant oils, and carbon dioxide.[1,2] PHAs are expected to be used as sustainable plastics and as materials for medical applications because of their thermoplasticity, biodegradability, and biocompatibility.[3,4] PHA synthase (PhaC), a key enzyme in PHA biosynthesis, catalyzes the polymerization of the acyl moiety of (R)-3-hydroxyacyl-coenzyme A [(R)-3HA-CoA] into PHA in aqueous solution under ambient conditions.[5] The molecular weights of the biosynthesized PHAs range from tens of thousands to several millions (g·mol–1). Among the known PhaCs, PHA synthase from the bacterium R. eutropha is the most studied enzyme.

To better understand the polymerization promoted by PhaC, in vitro studies, rather than in vivo experiments, with purified PhaC and chemically synthesized [(R)-3HA-CoA] have been conducted for quantitative analyses. The expected in vitro formation of PHA was previously observed using atomic force microscopy (AFM) in air using dried samples.[6−9] These studies provided high-resolution images of PhaCs and the resultant polymers in the initial stage of polymerization; however, these were not real-time observations and the mechanism of polymer formation has not been elucidated. To fully understand the polymerization process, it is essential to characterize the reaction dynamics during PHA polymerization by observing the process in real time.

Recently, a new-generation AFM technique, referred to as high-speed scanning atomic force microscopy (HS-AFM), has been developed and applied to real-time analyses of the dynamics of biomolecules, such as the “walking” of myosin V and the degradation of cellulose by “jamming” cellulases.[10,11] HS-AFM observations provide new insights into reaction mechanisms because scanning is achievable at a high time resolution (20 frames per second at the maximum) even in aqueous solutions. A previous study demonstrated the feasibility of using HS-AFM to analyze the enzymatic syntheses of hyaluronic acid polymers.[12] This new observation technique is applicable to a wide range of enzymatic polymerization reactions. We regarded PHA polymerization as an interesting target for this observation because of the unique feature of the polymerization reaction, generating a water-insoluble polymer from water-soluble monomers, and because of the industrial usefulness of PHA as a plastic material.

In the present study, the dynamics of PHA polymerization is assessed using HS-AFM to provide a better understanding of PHA polymerization catalyzed by PhaCRe. The observation raises the possibility of secondary structure formation of PHA chain during polymerization, which may help us understand its polymerization mechanisms.

Results and Discussion

The adsorption of PhaCRe molecules on highly oriented pyrolytic graphite (HOPG) was preliminarily evaluated in water using HS-AFM. Very few PhaCRe molecules could be observed on HOPG in water (data not shown), suggesting that the interaction between PhaCRe and the HOPG substrate in water may be weak. The same experiment was repeated using a mica substrate; the acquired images are presented in Figure . The observed width and height of the particles on the mica substrate were 15.0 ± 4.9 and 2.9 ± 0.7 nm, respectively, without correction for the AFM tip radius. The width of these particles is similar to that of a single PhaCRe molecule, as observed in previous reports.[6,7] Notably, the majority of particles adsorbed on the mica remained on the substrate, whereas several particles seemed to be dissociated (Figure , the particle in the dotted circle) and some were newly adsorbed (particles in the dotted squares). This observation indicated that the PhaCRe molecules on mica were equilibrated between the adsorbed and desorbed states. To the best of our knowledge, this is the first direct observation of dissociation and adsorption of PhaCRe on a substrate surface.

Figure 1

HS-AFM observation of PhaCRe on mica in water without the addition of (R)-3HB-CoA. The incubation time is shown in each figure. The dotted circle and the dotted squares indicate dissociating and adsorbing PhaCRes, respectively.

A model of in vitro PHA polymerization on hydrophobic surfaces (HOPG) involving the dissociation of PhaCRe from the HOPG substrate to initiate PHA polymerization has been previously proposed.[7] Although the substrate was not the same in this study, this model was partially validated using HS-AFM observation.

PHA polymerization was also observed using HS-AFM, where mica was used as a substrate with the addition of (R)-3-hydroxybutyryl-coenzymeA [(R)-3HB-CoA] as a monomer for the synthesis of poly[(R)-3-hydroxybutyrate], P(3HB). In two of the observed cases (cases 1 and 2), relatively high spatial resolution was successfully achieved. These videos are available as Videos S1 (case 1) and S2 (case 2). Snapshots from these videos are presented in Figures and 3.

Figure 2

HS-AFM observation of P(3HB) polymerization by PhaCRe on mica in water with the addition of (R)-3HB-CoA at 0:34 in Video S1 (case 1). The elapsed time from (R)-3HB-CoA addition is shown in each figure on the second time scale. (A) Full images of observation every 10 s. (B) Enlarged and higher time-resolution images of the area in the solid square of (A). The arrows indicate the growth direction of the elongating object. The HS-AFM video of this figure is available as Video S1. The time of monomer addition (0:34) is defined as 0 s.

Figure 3

HS-AFM observation of P(3HB) polymerization by PhaCRe on mica in water with the addition of (R)-3HB-CoA at 0:30 in Video S2 (case 2). The elapsed time from the (R)-3HB-CoA addition is shown in each figure on the second time scale. (A) Full image of observation at 90 s after the addition of monomer. (B) Enlarged and higher time-resolution images of the area in the dashed square of (A). The arrow indicates the growth direction of the elongating object. The globular object indicated by a triangle in the frame at 112 s might be the occasionally visualized PhaCRe. The dashed area in the frame at 118 s indicates the elimination of the elongating object during the scan. The HS-AFM video of this figure is available as Video S2. The time of monomer addition (0:30) is defined as 0 s.

Before the addition of monomer, the number of particles per view, corresponding to PhaCRe adsorbed on mica, was ∼10 (Figure A). After the addition of monomer at 0:34 (Video S1), several dozen elongating objects were observed in the same area. The position where elongation started differed from that of PhaCRe before the addition of monomer. This observation strongly suggests that the adsorbed PhaCRe would be inactive for polymerization because it could not be transformed to active synthase on the mica substrate; instead the active synthase was formed in water.

Elongating objects could scarcely be detected on the mica substrate shortly after the addition of monomer; these species appeared only ∼30 s after the addition of monomer (Figure A). This may be due to the lag phase at the start of polymerization that has been observed during in vitro spectrophotometric activity assay.[5,13−15] The average width of the elongating objects was 10.5 ± 2.7 nm without correction for the AFM tip radius, whereas the height of the objects was 2.2 ± 0.5 nm. This size is smaller than that of the PhaCRe molecule (see above); thus, it is natural to assume that the elongating objects are P(3HB) chains. However, the height is not consistent with that expected for an extended single polymer chain (usually less than 1 nm).[12,16,17] Such oblong objects seem to have semiregular undulations, which might be attributed to the formation of a secondary structure. It has been proposed that the cellular P(3HB) oligomer forms an exolipophilic endopolarophilic helical structure as a component of an ion channel.[18,19] Therefore, on the mica surface, the P(3HB) polymer chains might form a semiregular helical structure.

In most cases, the objects elongated linearly and promptly, and the direction of elongation remained constant after the initiation of polymerization (Figure A and Video S1). The robust elongation might suggest the formation of a rigid structure like the helical structure. The maximum rate of elongation of the objects was calculated to be 4.3 ± 1.1 nm·s–1 from 10 different elongating spots. Assuming that the elongating object forms a helical structure with 14 monomers per turn and a helical rise of 0.45 nm (the expected diameter is 2.4 nm), as proposed by Reusch and Sadoff,[18,19] the elongation rate is comparable to the polymerization rate of ∼130 monomers·per second. This rate is slightly higher than the catalytic turnover of PhaCRe of 100 monomers per second, determined using a comprehensive kinetic analysis.[14] However, some active synthases would individually exhibit higher catalytic turnover than the average value of the ordinary synthases.

PhaCRe molecules are thought to be present on the growing edge of the objects; they are difficult to observe because active PhaCRe would be suspended in water instead of being fixed on the mica substrate. Nonetheless, in a local site (Video S2), PhaCRe might occasionally be observable on the growing edge of the object (Figure B). This globular object, indicated by a triangle (frame at 112 s), is thought to be PhaCRe. In the last image in Figure B, plausibly due to physical obstruction of PhaCRe, the entire elongating object was removed from the mica surface by the AFM tip (indicated by a dashed ellipse in the frame at 118 s).

Several unexpected behaviors were found on careful observation of the AFM videos. The boxed areas in Figure B seemed to elongate in two directions from the same site. There might be two different synthases that elongate P(3HB) from very close starting points. As the reaction proceeded, strong adsorption of the PhaCRe–P(3HB) complex on the AFM tip would take place, thereby reducing the image resolution remarkably after 64 s in case 1 (Figure A) by increasing the tip radius as the elongation proceeded.

Notably, in this study, elongating objects were observed only in the tip scanning area (Figure ). This localized event might be attributed to the pinning effect of the P(3HB) chain synthesized by PhaCRe in water, which acts on the mica substrate because of the tapping force of the AFM tip.

Figure 4

HS-AFM wide view of scanning area during P(3HB) polymerization (case 2). The HS-AFM video of this figure is available as Video S2.

On the basis of the observations, the process of P(3HB) polymerization on the mica surface can be explained as follows: Initially, PhaCRe is adsorbed on the mica surface in an inactive form; however, before polymerization, PhaCRe undergoes cyclic desorption and adsorption on the mica substrate. Some dissociated PhaCRes species were activated for P(3HB) polymerization in water by the addition of (R)-3HB-CoA. After the lag phase at the start of polymerization, the P(3HB) chains were polymerized by drifting active synthases. The growing P(3HB) chain was fixed to the mica surface by the pinning effect of the AFM tip, given that the elongating objects were observed only in the scanned area. The elongating objects might form a semihelical structure,[18,19] a hypothesis corroborated by the dimensions (observed diameter is 2.2 ± 0.5 nm, whereas theoretical diameter is 2.4 nm) and the elongating behavior of the object. As polymerization proceeded, strong adsorption of the PhaCRe–P(3HB) complex on the AFM tip reduced the image resolution.

In conclusion, real-time observation of the polymerization by PhaCRe in water via HS-AFM was successfully executed. Following are the noteworthy findings: (i) PhaC undergoes cyclic adsorption and desorption on mica, and the desorption process is necessary for the P(3HB) polymerization; (ii) the P(3HB) chain might assume a helical structure during polymerization by PhaCRe, based on the observed height and elongating behavior; and (iii) P(3HB) polymerization drastically increases the adsorption of the PhaCRe–P(3HB) complex on mica and the AFM cantilever tip. The observations in this study are highly relevant for understanding the process of PHA polymerization.

Methods

For the preparation of PHA synthase, His-tagged PhaCRe was produced in Escherichia coli BL21(DE3) strain (Novagen, Madison, WI) harboring pET-15b::phaCRe.[13] The protein was also purified and desalted as previously reported.[14,15] The enzyme purity was confirmed using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE), according to the standard procedure. Protein concentrations were determined using a Quant-iT protein assay kit (Invitrogen, Carlsbad, CA). Aliquots of all proteins were subsequently frozen in liquid nitrogen and stored at −80 °C before other intended uses.

(R)-3HB-CoA was chemically synthesized and purified as a substrate for PhaCRe, according to previously reported techniques.[14]

The HS-AFM (NanoExplorer, RIBM, Tsukuba, Japan) imaging experiment was conducted using a silicon nitride cantilever (BL-AC10EGS or BL-AC10DS, Olympus, Tokyo, Japan, tip radius <15 nm). For the observation of nonpolymerizing PhaCRe, 2 μL of the PhaCRe solution (1 nM PhaCRe in ultrapure water) was placed on freshly cleaved mica (1.5 mm in diameter) and was incubated in a moist chamber at room temperature for 5 min. The mica was washed twice with 10 μL of ultrapure water and scanned in 110 μL of ultrapure water at room temperature. To observe polymer elongation, R-3HB-CoA (100 μM in final concn) was added to the scanning cell during the HS-AFM observation. The scale of the images was 500 × 500 nm2, and the scan rate of imaging was 0.5 frames per second.