Zwitterion Moieties in Polypeptides Synergistically Enhance the Release of Cellulose and Amorphous Polysaccharides from Plant Cell Walls

Abstract

This study demonstrated that covalently localized zwitterionic moieties in zwitterionic polypeptides (ZIPs) effectively disrupt hydrogen bonds in cellulosic substrates, including filter paper and plant cell wall materials, without significant cytotoxicity. ZIPs with varying densities of zwitterionic side chains were synthesized via the postpolymerization modification of histidine-containing oligopeptides. The newly developed ZIPs predominantly comprised repeating units with zwitterionically converted side chains. Such ZIPs can cleave multiple hydrogen bonds by anchoring the zwitterionic structure at specific sites, thereby partially dissociating the polysaccharide chains in the cell wall. They are especially effective in dissolving amorphous cellulose, even at low concentrations in aqueous solutions. Importantly, this effect was achieved with minimal cellular toxicity, harnessing the advantages of ionic liquid-like properties while mitigating their high-toxicity limitations. This biofriendly approach to cell wall denaturation highlights a novel method for controlling hydrogen bond networks in polysaccharides and cell walls. These findings indicate a new approach for reducing biomass recalcitrance and developing next-generation biobased materials and fuels derived from plant cell walls.

Introduction

The plant cell wall, the largest terrestrial carbon reservoir, has been exploited for millennia in applications such as wood, cotton, and paper pulp. Recently, it has attracted attention as a critical raw material for next-generation biomaterials, including cellulose nanofibers (CNFs) and bioethanol. However, its efficient utilization is hindered by recalcitrance arising from its complex chemical composition and structural organization. The cell wall consists primarily of fibrous cellulose embedded in an amorphous polysaccharide matrix. These components contain abundant hydroxyl groups that interact cooperatively with neighboring ones and are stabilized by electron delocalization, resulting in the formation of highly crystalline cellulose fibers and a supramolecularly oriented hydrogen bond network. − Additionally, water molecules, which constitute up to 65% of the cell wall, interact with these hydroxyl groups in amorphous polysaccharides, acting as plasticizers that impart unique viscoelasticity and enable the loosening, extension or softening of the cell wall during plant growth. , Thus, the formation of hydrogen bonds between polysaccharides (both polysaccharide–polysaccharide and polysaccharide–water molecule interactions) plays a crucial role in maintaining cell wall integrity and extensibility. However, these well-organized networks also confer resistance to mechanical, chemical, and enzymatic degradation, limiting the utilization of cell walls in biorefineries. Therefore, developing cell wall engineering methods to manipulate these hydrogen bonds could be a powerful solution to overcome biomass recalcitrance and obtain biofuels and bioproducts more efficiently.

Recent advances in understanding the molecular mechanisms of hydrogen bond formation among polysaccharides during cell wall synthesis have enabled genetic or biological engineering approaches for cell wall modification in living plants. − Strategies such as genetically regulating the amount of polysaccharides and hydrogen bonding through the modulation of endoglucanase activity create voids and nicks in the cell wall to address the recalcitrance. The expression of a wall-loosening protein, expansin, is also a promising method to loosen or weaken the network. − However, these approaches require fine-tuning the expression of genes of interest; otherwise, the modified biosynthesis system affects other downstream biosynthetic pathways and compromises other critical traits, such as polysaccharide molecular weight, plant growth, tissue integrity and biomass yield. , Furthermore, genetic modification of the cell wall is hindered by the physical barrier of the cell wall itself, limiting the ability to obtain desired cell wall-modified plants. , All these factors limit the applicability of this genetic engineering approach across a wide range of plant species.

Chemical approaches to cell wall modification offer promising alternatives for effective hydrogen bond cleavage in the cell wall. , Zwitterionic liquids (ZILs), which consist of an imidazolium cation covalently conjugated with a single carboxylate anion, have been shown to efficiently disrupt intermolecular hydrogen bonds in cellulose, partially dissociating the polysaccharide network of the cell wall in various microorganisms (e.g., bacteria, yeast, and molds) and even a terrestrial plant (Arabidopsis thaliana). − This ability originates from their high hydrogen bond acceptability, as indicated by the β value, a Kamlet–Taft parameter, exceeding 0.8. However, cell wall denaturation by ZILs is limited by their water sensitivity, which restricts ZIL application in plant systems. Consequently, high concentrations are required for ZILs to disrupt polysaccharide interactions sufficiently for effective cell wall denaturation. However, since ZILs are used as antibacterial agents, they are often assumed to exhibit bacteriostasis at high concentrations. These findings suggest that ZILs may interfere with plant growth by disrupting metabolic or proliferation pathways, such as those of microorganisms. In summary, high concentrations promote cell wall denaturation but increase cytotoxicity, whereas low concentrations reduce toxicity but limit cell wall denaturation. Therefore, chemical substances that effectively function as hydrogen bond cleaving agents in aqueous environments without causing cellular damage are needed.

To address these challenges, we propose a novel approach using zwitterionic polypeptides (ZIPs) to cleave hydrogen bonds effectively at low concentrations under aqueous conditions. ZIPs are designed to contain zwitterionic-modified repeating units with similar structures to those of cellulose-solubilizing ZILs. ZIPs possess multiple repeating structures that act as hydrogen cleavage sites in a single molecule, whereas ZILs have only a single site. Localizing hydrogen bond cleavage sites in ZIPs using covalent bonds is expected to enable the cooperative and extensive cleavage hydrogen bonds within polysaccharides with synergistic effects. This property is expected to allow ZIPs to denature amorphous cellulose and polysaccharides in cell walls at low concentrations in an aqueous solution while maintaining low cytotoxicity. To explore the importance of the spatial localization of the zwitterionic structure along a polymer backbone, we investigated the relationship between the local zwitterionic concentration within the polymer chain and the cleavage of hydrogen bonds in the cell wall. Here, we synthesized three ZIP variants with varying densities of zwitterionic-modified side chains within single polymer chains to serve as an architectural controls (Figure ). The denaturing ability of these ZIPs was quantitatively assessed by measuring changes in the mechanical and morphological properties of ZIP-treated cell walls. This study demonstrates the potential of ZIPs as substances that effectively and selectively cleave hydrogen bonds to manipulate plant cell wall structures in aqueous environments while addressing the trade-off between hydrogen bond cleavability and toxicity.

1.

ZIPs with different zwitterionic-modified side chain densities in single polypeptide chains used in this study. The respective chemical structures are described in the box. Zwitterionic conversion ratios were determined by ¹H NMR and are expressed as the mean ± standard deviation (n = 3).

Experimental Section

Materials

Deuterated trifluoroacetic acid (TFA-d) was purchased from Sigma–Aldrich (cat#152005, St. Louis, MO). Calcofluor white (CelloFluor, cat#17353) was purchased from Polysciences, Inc. (Warrington, PA). Trypan blue solution (0.4%) was purchased from Thermo Fisher Scientific, Inc. (cat#15250061, Waltham, MA). Ethyl 4-bromobutyrate (cat#259–44562), N,N-dimethylformamide (DMF, cat#045–02911), diethyl ether (cat#055–01155), 1 N aqueous sodium hydroxide (cat#192–02175), iodoethane (cat#055–03953), N-methyl-2-pyrrolidone (NMP, cat#131–15111), deuterated dimethyl sulfoxide containing 0.05 vol % tetramethylsilane solution (DMSO-d ₆, cat#046–34291), 2-[4-(2-hydroxyethyl)­piperazin-1-yl]­ethanesulfonic acid (HEPES, cat#095462), and sodium dodecyl sulfate (cat#196–08675) were all purchased from FUJIFILM Wako Pure Chemical Co. (Osaka, Japan). Three polypeptide precursors, poly­[GlyHis­(n-Bu)­Gly], poly­[GlyHis­(BuCO₂Et)­Gly], and poly­[His­(BuCO₂Et)], were synthesized according to a previously described method. , All chemical reagents were purchased from commercial suppliers and used without further purification unless otherwise detailed.

Analysis

Fluorescence Staining of Released Cellulose from Filter Paper after ZIP Treatment

Filter paper (Whatman no. 3, cat#WHA1003185, GE HealthCare Technologies Inc., USA) was cut into 10 mm × 2 mm strips (4.3 mg each) with a sharp razor blade (cat#4–1644–11, AS One Co., Japan). Two milligrams of each ZIP was dissolved in 1 mL of 20 mM HEPES (pH 7.0) buffer. Five filter paper strips and 400 μL of each ZIP solution were added to a 1.5 mL microcentrifuge tube (cat#131–815C, Watson Co., Ltd., Japan) and treated for 4 h at 350 rpm on a thermomixer shaking block (cat#2230000113, Eppendorf Co., Ltd., Germany) at 28 °C. After incubation, the 1.5 mL microcentrifuge tubes were centrifuged at 6,000 rpm for 1 min with a D1008 compact centrifuge (DLAB Scientific Co. Ltd., China), and the supernatant was removed using a pipet. The collected supernatant (400 μL) from the filter paper treated with 2 mg/mL ZIPs in 20 mM HEPES buffer (pH 7.0) was stained with 4 μL of 3.5 mg/mL calcofluor white for 15 min. Fifty microliter (μL) aliquots from each treatment were placed in separate wells of a flat-bottomed black 384-well plate (cat#781900, Greiner Bio-One Co., Austria) and imaged using an Olympus MVX10 stereomicroscope (Center Valley, PA). Excitation was set at 405 nm, and the emission range was 410–500 nm (Figure a). The fluorescence signal from each well in the flat-bottomed black plate was also measured to quantify the amount of cellulose released as fluorescence intensity using an Infinite M1000 PRO plate reader (Tecan Group Ltd., Switzerland). Excitation was set at 405 nm, and the emission wavelength was set at 420 nm, with excitation and emission slits of 5 nm. The fluorescence signal was recorded as relative fluorescence units using iControl software (Tecan Group Ltd., Switzerland) (Figure b). Data points read in the respective solutions (n = 4 per treatment) are shown in the respective box and whisker plots. Statistical significance was determined by one-way ANOVA with a post hoc Tukey test (*P < 0.05, **P < 0.01, ***P < 0.001).

2.

Release of cellulose fragments from filter paper increases with increasing zwitterionic-modified side chain density of ZIP. (a) Fluorescence image of a black-bottom plate filled with the respective ZIP solution of 20 mM HEPES buffer (NegCntrl) in each column. Cellulosic fragments released from filter paper under ZIP solutions were stained with calcofluor white for observation. White scale bar: 2 mm. (b) Quantification of the amount of cellulosic fragments released in the respective aqueous ZIP solutions (2 mg/mL). Data points read in the respective solutions (n = 4 per treatment) are shown in the respective box and whisker plots. Statistical significance was determined by one-way ANOVA with a post hoc Tukey test (*P < 0.05, **P < 0.01, ***P < 0.001).

Quantification of Cellulose Release from the Cell Wall by Aqueous ZIP Solutions

The abaxial side of the fifth scale epidermis in onion was peeled from white onion bulbs (Allium cepa, cv. Cometa) purchased from a local grocery store (State College, PA). Afterward, the cell wall peel obtained from the abaxial side was washed with ddH₂O (5 mL) three times to remove residual debris. Next, the peel was a 10 mm × 10 mm strip with a sharp razor blade (cat#4–1644–11, AS ONE Co., Osaka, Japan). The obtained onion epidermal cell wall strips were incubated in 20 mM HEPES buffer with/without the respective ZIPs (400 μL, 2.0 mg/mL) in each well of a 24-well plate (Greiner Bio-One Co., Kremsmünster, Austria) for 4 h on a rotary shaker (Wave-SI, TAITEC Co., Saitama, Japan) at 20 rpm and 25 °C. Next, the cell wall-treated solution (2 μL) was mounted on a glass slide (cat#1–9646–12, AS One Co., Osaka, Japan) and dried at 35 °C for 24 h. Afterward, the residual cellulose on the glass slide was imaged by a Dimension Icon AFM (Bruker, Santa Barbara, CA) with a Scanasyst in PeakForce Tapping mode and with quantitative nanomechanical property mapping (QNM) using a calibrated ScanAsyst-AIR AFM tip (radius; 5 nm, spring constant; 0.4 N/m, resonant frequency; 70 kHz, Bruker, Santa Barbara, CA) under ambient atmosphere at room temperature (Figure a to e).

3.

AFM topographic images of cellulose released from plant cell walls under different ZIP aqueous solutions. Images of cell walls treated with (a) 2 mg/mL P1, (b) 2 mg/mL P2, (c) 2 mg/mL P3 and (d) 20 mM HEPES buffer solutions. (e) P3 solution without cell wall infiltration was also collected as a negative control for this experiment. (f) Quantification of the amount of cellulose released in cell wall-treated ZIP aqueous solutions (2 mg/mL) or 20 mM HEPES buffer (NegCntrl). Box and whisker plots represent the fluorescence readings in the respective solutions (n = 6 technical replicates). One-way ANOVA with post hoc Tukey’s test was performed to determine the statistical significance among the respective solutions (****P < 0.0001).

To quantify the amount of cellulose released, the cell wall-treated ZIP solution (400 μL) was stained with calcofluor white solution (4 μL, 3.5 mg/mL) for 15 min, after which the cellulose release was quantified as the fluorescence intensity. Next, the stained cellulosic solution (50 μL) was transferred to each well of a flat-bottomed black 384-well plate (Greiner Bio-One Co., Kremsmünster, Austria). Next, the fluorescence signals from each well in the flat-bottomed black plate were quantified using an infinite M1000 PRO plate reader (Tecan Group Ltd., Männedorf, Switzerland). The excitation and emission wavelengths were 405 and 420 nm, respectively, and the excitation and emission slits were 5 nm. The fluorescence signal was recorded in relative fluorescence units using SparkControl software (Tecan Group Ltd., Männedorf, Switzerland) (Figure f).

Small-Angle X-ray Scattering (SAXS) Measurements

The cell wall was prepared from the abaxial side of the fifth-scale epidermis peel in a white onion bulb (Allium cepa, cv. Cometa) purchased from a local grocery store (Kyoto, Japan). Afterward, the obtained cell wall peel was washed with ddH₂O (5 mL) three times to remove residual debris. Next, the peel was cut into 10 mm × 3 mm pieces (the lengthwise orientation matched the vertical axis of the onion scale) and successfully washed under 1% SDS in 20 mM HEPES buffer (pH 7.0) for 30 min on a rotary shaker (Wave-SI, TAITEC Co., Saitama, Japan) at 25 rpm and 25 °C. Next, the peel was rinsed with excess ddH₂O, followed by immersion in the respective 20 mM HEPES buffer with/without the respective ZIPs (400 μL, 2.0 mg/mL) for 16 h on a rotary shaker (Wave-SI, TAITEC Co., Saitama, Japan) at 25 rpm and 25 °C. Afterward, the peel was washed with ddH₂O (1 mL) 6 times and dried under 0.1 MPa vacuum for 12 h to remove the excess water in the peel to increase the scattering intensity derived from the embedded cellulose microfibrils. After drying, the peels were adhered onto washers (M6, Utsunomiya Rashi Co. Ltd., Tokyo, Japan) by taping both ends and then mounted into the X-ray scattering chamber with the longitudinal cell direction in the vertical direction. Synchrotron SAXS measurements of ZIP-treated onion peels were performed on the BL05XU beamline (SPring-8, Harima, Japan) using an X-ray energy of 12.4 keV for 1 s of irradiation. SAXS data were collected using a Pilatus 2 M detector 0.87 m away from the sample (Figure b to e). The 2D SAXS profiles were obtained after the background pattern of the air was subtracted. The obtained 2D SAXS images were converted to 1D profiles by azimuthally integrating the images ± 20° from the horizontal direction using Fit2D (Figure ).

4.

SAXS measurements of the epidermal cell walls of ZIP-treated onion. (a) Schematic illustration of the instrumental setup. SAXS 2D images of (b) 20 mM HEPES buffer (NegCntrl), (c) 2 mg/mL P1, (d) 2 mg/mL P2, and (e) 2 mg/mL P3 solution-treated cell walls. The embedded image shows the cell wall. The x- and y-axes indicate the orientation of the cell wall sample placed on the stage. (f) Scattering vector profiles of the ZIP-treated cell wall obtained by azimuthally integrating 2D images ± 20° from the horizontal direction. (g) Illustration of the effect of ZIPs on the orientation of cellulose microfibrils aligned in the longitudinal direction of the cell.

Nanoindentation Measurements by Atomic Force Microscopy (AFM)

The abaxial side of the fifth scale epidermis in onion was peeled from white onion bulbs (Allium cepa, cv. Cometa) purchased from a local grocery store (State College, PA). Afterward, the obtained cell wall peel was washed with ddH₂O (5 mL) three times to remove residual debris. Next, the peels were cut into 10 mm × 10 mm strips with a sharp razor blade (cat#4–1644–11; AS ONE Co., Osaka, Japan) and were subsequently washed with 20 mM HEPES buffer (pH 7.0) supplemented with 2 w/v% sodium dodecyl sulfate (SDS; 1 mL) on a rotary shaker (Wave-SI; TAITEC Co., Saitama, Japan) at 20 rpm for 15 min under ambient air conditions (25 °C) to remove the debris on the cell wall. Afterward, excess ddH₂O was applied to rinse the cell wall peel with SDS solution. All the cell wall strips were incubated in 20 mM HEPES buffer (200 μL, pH 7.0) for 1 h before the nanoindentation measurements. The force curve of the onion epidermal cell wall surface was acquired by AFM as described and measured with a Dimension Icon AFM in ScanAsyst and PeakForce QNM mode in fluid (Bruker, CA). We used a calibrated ScanAsyst Fluid+ cantilever (radius; 5 nm; resonant frequency; 150 kHz; Bruker, Santa Barbara, CA). The spring constant ranged from 0.7–1.1 N/m in air at 25 °C, which was estimated in thermal tuning mode. Next, the strips were mounted on microscope glass slides (cat#1–9646–12, AS One Co., Osaka, Japan) by fixing the cuticle side facing down (inner side of the epidermis) with a nail polish (Top&base coat N, Shiseido Co., Ltd., Tokyo, Japan) layer around the perimeter. The target force was set at 8 nN for the nanoindentation measurements with a ramp size of 800 nm, and the ramp speed was 3 μm/s. After 1 h of treatment with 20 mM HEPES buffer ± ZIP solution at a concentration of 2 mg/mL (200 μL, pH 7.0), at least 200 different locations from 20 cells of each cell wall strip were indented by the AFM tip at a velocity of 1 μm/s to a force of 8 nN. Samples from at least 3 biologically independent peels were indented after treatment. To compare the effects of ZIP treatment on the cell wall nanoindentation modulus, the retraction force–distance curves were fitted by the Sneddon model to calculate the cell wall nanoindentation modulus using Nanoscope analysis software. The Poisson ratio of the samples was set to 0.3. Because of the large variation among onions, we combined all the data obtained from each individual peel treated with the same solution and conducted one-way ANOVA with a post hoc Tukey test. Statistical significance was determined using GraphPad Prism 9.4.1 (GraphPad Software, Inc., San Diego, CA) (Figure ).

5.

Effects of ZIPs on the indentation properties of the cell wall surface as determined by AFM. (a) Retraction curves obtained in 2 mg/mL ZIP aqueous solution- or 20 mM HEPES buffer (NegCntrl)-treated cell walls. (b) Modulus calculated by the Sneddon model using the retraction curves obtained from different cells and different biological replicates (255 ≤ n ≤ 510). Statistical significance was determined by one-way ANOVA with a post hoc Tukey test (*P < 0.05, **P < 0.01, ****P < 0.0001).

Tensile Strain/Stress Assay

The abaxial side of the fifth scale epidermis in onion was peeled from a white onion bulb (Allium cepa, cv. Cometa) purchased from a local grocery store to prepare cell wall strips (State College, PA). Afterward, the obtained cell wall peel was washed with ddH₂O (5 mL) three times to remove residual debris. Next, the peel was cut into 10 mm × 3 mm strips with a sharp razor blade (cat#4–1644–11; AS ONE Co. Osaka, Japan). Onion epidermal strips (10 mm × 3 mm, lengthwise orientation matching the vertical axis of the onion scale) were treated with 20 mM HEPES buffer ± ZIP solution at a concentration of 2 mg/mL (400 μL, pH 7.0) for 16 h on a rotary shaker (Wave-SI, TAITEC Co. Saitama, Japan) at 20 rpm and 25 °C. The wall strip was clamped in a custom-made extensometer with 5 mm between the clamps and extended at 3 mm/min until a force of 90 g (0.88 N) was reached. Nine to 16 cell wall strips were measured for each ZIP treatment. All strain–stress curves from the ZIP-treated cell wall strips shown in Figure were generated from cell wall strips prepared from the same onion (Figure ).

6.

Merged graphs of engineering stress–strain curves obtained from ZIP-treated cell walls and 20 mM HEPES buffer-treated cell walls as a negative control (NegCntrl). Stress–strain curves of (a) 2 mg/mL P1 (b) 2 mg/mL P2 and (c) 2 mg/mL P3 solution-treated cell walls with those of the buffer-treated cell wall (9 ≤ n ≤ 16). The enlarged inlet shows the elastic region (λ = 0–0.1) of each graph.

Trypan Blue Staining of ZIP-Treated Cell Walls

To measure the cytotoxicity of ZIP, the adaxial epidermis from the fifth scale of the onion (Allium cepa, cv. Cometa) was peeled from onion bulbs and gently washed with ddH₂O to remove debris. Next, the peels were sliced into square strips (10 mm × 10 mm) with a razor blade (cat#4–1644–11; AS ONE Co., Osaka, Japan) and mounted on a glass slide (cat#1–9646–12; AS One Co., Osaka, Japan). Notably, the hydrophobic side (the outer side of the adaxial epidermis) was fixed face down on the glass slide to avoid curling. The strip of onion epidermis (10 mm × 10 mm) fixed on the glass slide was treated with 20 mM HEPES buffer with/without the respective ZIPs (25 μL, 2.0 mg/mL, pH 7.0) for 1 h. Next, the strip was washed with excess ddH₂O and stained with 0.4 w/v% trypan blue solution (25 μL) for 5 min and washed with excess ddH₂O to remove the excess stain. The morphologies of the cellular boundaries and nuclei in the stained onion epidermal cells were observed under an inverted microscope at 20× magnification (Olympus IX70; Olympus Inc., Tokyo, Japan) (Figure a). Cytotoxicity was quantitatively evaluated by the following procedure. First, the percentage of dead cells was calculated by manually counting stained (light or dark blue) and unstained cells in each low-magnification (4×) image obtained with an inverted microscope (Olympus IX70; Olympus Inc., Tokyo, Japan). Each image (4× magnification) contained approximately 520 cells, with fewer than 20 stained as dead in all onion peels. Next, single-cell viability was calculated by subtracting the percentage of dead cells from 100. To assess biological variation in ZIP cytotoxicity, peels from four different onions were used, and single-cell viability was calculated for each peel. Statistical comparisons of cell viability among onion peels treated with 20 mM HEPES buffer with/without the respective ZIPs (25 μL, 2.0 mg/mL, pH 7.0) were performed using Dunnett’s multiple comparison test (****P < 0.0001) in GraphPad Prism 9.4.1 (GraphPad Software Inc., San Diego, CA) (Figure b).

7.

Cytotoxicity of 2 mg/mL aqueous ZIP solutions to onion epidermal cells. HEPES buffer (20 mM) was used as a negative control (NegCntrl), while a 1 N NaOH served as a positive control (PostCntrl). a) Onion epidermal cells treated with the indicated solutions and trypan blue. White scale bars: 100 μm. b) The average cell viability of ZIPs among different biological replicates (n = 4 biological replicates). Statistical significance was set at ****P < 0.0001 on the basis of Dunnett’s multiple comparison test.

Results and Discussion

Synthesis of ZIPs with Different Side Chain Densities

To regulate the local concentration of zwitterionic moieties, we aimed to vary the zwitterionic-modified side chain density in the histidine (His)-containing oligopeptide. Previously, zwitterionic polypeptide P1 was synthesized from a periodically His-containing oligopeptide, followed by a postpolymerization modification process, as shown in Scheme a. The conversion ratio in this process was reported to be 55% in a previous study. By resynthesizing P1 three times and performing ¹H NMR, the average conversion ratio to zwitterionic-modified P1 was determined to be 48± 6% (by comparing the integral values between the methylene protons of butyl and butylate groups on the imidazole moiety, designated peaks g and k in Figure S1), which is approximately the same value previously reported (Figure S1). Owing to the sterically disfavored N-alkylation process, introducing bulky ethyl-4-bromobutyrate into the imidazole ring of His did not occur in all the repeating units of P1. Thus, P1 possessed a relatively low ionic concentration.

1. Synthetic Pathway of ZIPs Starting from His-Containing Polypeptides .

^a Scheme of chemical synthesis for (a) P1, (b) P2, and (c) P3.

To prepare ZIPs with higher local ionic concentrations, increasing the conversion ratio or regulating the proximity of the zwitterionic moieties was attempted. First, highly converted zwitterionic polypeptide P2, with almost all repeating units converted to zwitterionic-modified imidazole, was synthesized from tripeptide GlyHis­(BuCO₂Et)­Gly-based monomers (Scheme b). The N-alkylation step of the imidazole ring was conducted using iodoethane as a less sterically hindered electrophile in NMP to increase the conversion ratio. Afterward, the hydrolysis step of the ester group on the imidazole ring in a basic solution was performed for zwitterionic modification. The zwitterionic modification step made the resulting oligopeptide soluble in water, resulting in the formation of a transparent solution. This suggests that hydrophilic groups were successfully introduced into the polypeptides. A comparison of the integral values between the methylene protons of the ethyl and butylate groups on the imidazole moiety, designed peaks i and j in Figure S2, revealed that the conversion ratio was 92 ± 5%, suggesting the synthesis of tripeptide-derived ZIP with a higher conversion ratio (Figure S2).

Next, for the synthesis of P3, which has the highest local ionic concentration, the synthetic pathway shown in Scheme c was conducted. The zwitterionic conversion synthetic pathway was modified, and the proximity of the zwitterionic moieties was also modified by regulating the number of glycine (Gly) residues between the histidine moieties (P3 lacks one Gly in its respective repeating unit compared to P1 or P2). Similar to P2, the conversion ratio of the dipeptide His­(BuCO₂Et)­Gly-based monomer to zwitterionic-modified P3 was determined to be 91 ± 7% from a comparison of the integral values of peaks g and i in the ¹H NMR spectrum (Figure S3). The MD simulation predicted the secondary structure of ZIP to be random-coil structure in aqueous solution (Figure S4). Overall, three polypeptides with different local densities of zwitterionic units were successfully synthesized from His-containing monomers. These three polypeptides were subsequently characterized to elucidate the correlation between the local concentration and cell wall denaturation.

Ability of ZIPs to Weaken Filter Paper

Since cellulose is the major component of the cell wall, filter paper (a network of pure cellulose) is typically used as a cellulosic substrate to test the effects of molecules that interact with cell wall polysaccharides. , Here, as in previous studies, we used filter paper as a cellulosic substrate to observe the effect of the respective ZIPs on cellulose macro/micro fibrils. After the filter paper strips were incubated and sonicated in aqueous solutions of different ZIPs at low concentrations (2 mg/mL) for 1 min, they partially disintegrated into cellulose fibril fragments, whereas the filter paper in 20 mM HEPES remained square (Figure S5a to Figure S5b). This suggests that the synthesized ZIPs effectively cleaved intermolecular hydrogen bonds formed between cellulose fibers.

The cellulose microfibril network woven on the surface of the filter paper was imaged via scanning electron microscopy (SEM) after 4 h of 20 mM buffer or ZIP treatment (Figure S6a). After the filter paper was incubated with aqueous ZIP solutions, the pores between the fibrils gradually diminished, and film-like features formed on the surface (Figure S6b–d). The surface of the filter paper also became smoother, resembling that treated with (1-butyl-3-methylimidazolium chloride; [BMIM]­Cl), because of the high solubility of amorphous cellulose in the cellulose microfibrils. Short fibrils that were about to dissociate from the large bundles were observed in P2 and P3 solution-treated filter paper but not in P1 solution-treated filter paper (Figure S6c, d). This suggests that the dissociation of cellulose bundles in filter paper was especially promoted by treatment with highly side chain-modified zwitterionic polypeptides. FT-IR analysis of the surface of the ZIP-treated filter paper suggest that treatment with highly side-chain-modified ZIP effectively dissociates amorphous cellulose in filter paper via the cleavage of hydrogen bonds. Following this treatment, a decrease was observed in the peak areas at 2,900 cm^–1 (corresponding to C–H stretching in amorphous cellulose) and 3,400 cm^–1 (corresponding to the O–H stretching vibrations involved in hydrogen bonding), as shown in Figure S7. This can be attributed to the dissolution or disruption of amorphous cellulose, which inhibits the formation of hydrogen bonding networks. Conversely, filter paper treated with P1 showed an increase in the peak area around 3,400 cm^–1, likely due to the P1 molecules remaining attached to the filter paper (Figure S7).

To confirm the defibrillation of the cellulose bundle on the surface of the filter paper, the ZIP solution remaining after application to the filter paper was examined. Cellulosic fragments released from the filter paper surface were uniformly dispersed in the ZIP solutions, whereas only a small amount of cellulosic debris floated in 20 mM HEPES buffer (Figure a). To quantify the amount of cellulose in the ZIP aqueous solutions, we stained the cellulose in the respective solutions with calcofluor white and measured the fluorescence intensity (Figure b). Surprisingly, the fluorescence intensity of solutions of the highly side chain-modified peptides P2 and P3 was almost 6 times greater than that of the 20 mM HEPES solution and 2.5-fold higher than the fluorescence intensity of the P1 solution. This suggests that increasing the local ionic concentration (increasing the zwitterionic-modified side chain density) in the ZIP is essential for promoting the release of cellulosic fragments from filter paper.

The morphology of the cellulose released by P2 and P3 differed gradually over time, according to the confocal laser scanning microscopy (CLSM) observations. Indeed, the incubation of filter paper in P2 or P3 solution for 4 h led to the generation of similar images of cellulose macrofibrils (Figure S8a, b). After 24 h of incubation of filter paper with P2 solution, however, cellulose defibrillation increased slightly, resulting in the formation of fibrils approximately 50 μm in size (Figure S8c). In contrast, a longer incubation time effectively loosened the fibril structure in the P3 solution, and short fibrils (<20 μm) or particle-like cellulose fragments were observed (Figure S8d). Owing to the resolution limitation of CLSM, the shortened fibrils (<5 μm) had a particle-like blurred appearance (Figure S8e).

Dynamic light scattering (DLS) measurements were conducted to quantify the size distribution of the cellulose fibers in the P2 or P3 solution after 24 h of treatment. Without ZIP treatment, the cellulosic fragments were large with a broad size distribution (hydrodynamic diameter; 821 ± 56 nm; Figure S9a). When filter paper was treated with P2 or P3, the particle size distributions were measured as 266 ± 27 nm or 300 ± 10 nm, respectively, quantitatively suggesting that smaller cellulosic particles formed after the ZIP treatments (Figure S9b, c). A comparison of the size distributions obtained from DLS measurements of P2 and P3 solutions in 20 mM HEPES buffer to those obtained from filter paper-treated ZIP solutions revealed interesting results. In P2 aqueous solution, the particle size distribution obtained before and after the incubation of filter paper changed from 411 ± 19 nm to 266 ± 27 nm (Figure S10a to Figure S9b). However, particles smaller than 50 nm were not detected after processing the filter paper. This observation is consistent with the CLSM observations, indicating that the cellulose fibrils are not defibrillated to a size as small as a few nanometers in P2 solution. On the other hand, in the P3 aqueous solution, a wide size distribution was observed both before and after processing the filter paper, with size distributions of 379 ± 7 nm and 300 ± 10 nm, respectively (Figure S10b and Figure S9c). In contrast to the P2 solution, particles smaller than 50 nm were detected in the P3 aqueous solution after filter paper treatment. On the basis of the CLSM images, it can be inferred that more finely defibrillated cellulose is present in the solution.

Given these results, at a low concentration (2 mg/mL), aqueous solutions of ZIP, particularly those with highly modified side chains, exhibited strong hydrogen bond cleavage ability in aqueous environments, resulting in increased leaching of cellulose. Furthermore, the cellulose leached from the highly side chain-modified ZIP treatment was shorter in size, suggesting that these solutions can more effectively defibrillate and disperse cellulose.

Cellulose Release from Plant Cell Walls

To evaluate whether the ZIPs maintain their interaction with cellulose in plant cell wallsa more complex system comprising various amorphous polysaccharides and water moleculeswe performed a fluorescence assay similar to that used for filter paper. Cell wall strips were prepared from onion (Allium cepa) epidermis as previously reported. , The treatment with highly side-chain-modified ZIP yielded a decrease in the ratio of amorphous cellulose in cell wall surface, as evidenced by the reduced peak area at 2,900 cm^–1 (corresponding to C–H stretching in amorphous regions). Similarly, this trend was observed in the cell wall samples. These results are likely attributable to the dissolution or disruption of amorphous cellulose, which inhibits the formation of the hydrogen-bonding network (Figure S11). To verify whether the disruption of hydrogen bonds led to the actual release of cellulosic fragments into the solution, atomic force microscopy (AFM) observation was performed. This revealed that each cell wall-treated ZIP aqueous solution contained rod-like crystals (Figures a, b, c), which cannot be observed in cell wall-treated buffer (Figure d) or the polypeptide solution itself (Figure e). Considering that the crystalline components in the cell wall are mainly cellulose, we attributed these nanocrystals to the ability of the ZIPs to successfully leach cellulose nanocrystals from the cell wall at low concentrations in aqueous solution (2 mg/mL). These nanocrystals were approximately 3–10 nm in height, which corresponds to the previously reported height of cellulose nanocrystals. Wide-angle X-ray scattering (WAXS) measurements of pure cellulose nanocrystals after treatment with aqueous ZIP solutions (2 mg/mL) revealed that the crystal structure of cellulose was unaffected by ZIPs and that the cellulose Iβ structure remained unaffected (Figure S12). This suggests that aqueous solutions of ZIP (2 mg/mL) were able to release cellulose nanocrystals from cell walls while maintaining their cellulose Iβ-like crystalline structure. In the cell wall-treated P3 solution, some structures were observed where the crystalline structure had partially collapsed and the rod-like structure was not maintained (Figure c). This suggests that an increase in the local concentration of ionic structures led to a partial collapse of the cellulose microcrystals and other periodic features, smoothing the surface roughness, as previously reported. , Furthermore, to quantify the amount of cellulose leached from the onion epidermal cell wall in the respective ZIP solutions, the fluorescence intensities of calcofluor white-stained cellulose were measured. As shown in the boxplot, the measured fluorescence intensity increased as the local ionic concentration in the ZIPs increased (Figure f). Notably, the fluorescence intensity of the cell wall-treated P3 solution was 2 times greater than that of the 20 mM HEPES buffer or P1 solution. Although the increase in fluorescence intensities was smaller than that in the filter paper assay, as with the filter paper, higher fluorescence intensities were measured for highly side chain-modified ZIPs (P2 and P3 solutions). These findings suggest that highly side chain-modified ZIPs are also effective for hydrogen bond cleavage and enhancing the release of cellulose from plant cell walls (primary cell walls of onion epidermal cells) (Figure f).

ZIPs Distort the Alignment of the Cellulose Microfibrils

Small-angle X-ray scattering (SAXS) measurements were performed to elucidate the effect of ZIPs on the alignment of the cellulose microfibrils after cellulosic release. First, the samples were positioned so that the long axis of the cells in each cell wall section aligned with the vertical axis of the sample stage. Next, X-rays were irradiated perpendicular to the edge of the sample for edge-view measurements (Figure a). As a result, anisotropic scattering images, characterized by high scattering in the direction perpendicular to the cell’s long axis (equatorial direction), were observed from the onion epidermal cell walls treated with ZIP or 20 mM HEPES buffer (Figures b, c, d, e). Considering that electron densities tend to be higher in crystalline polymers than in amorphous polymers, these images are derived primarily from the scattering patterns of embedded crystalline cellulose microfibrils aligned along the longitudinal direction of the cells rather than from amorphous polysaccharides, as previously reported. SAXS equatorial scattering profiles that plot the normalized scattering intensity versus the scattering vector q were analyzed to quantitatively evaluate the effect of the respective ZIPs on the periodic structure of the cellulose microfibrils embedded in the cell wall (Figure f). Scattering profiles of 20 mM HEPES-treated cell walls showed a peak at q = 0.85 nm^–1, which corresponds to the center-to-center distance (7.4 nm wide spacing) of the cellulose microfibrils. This value was consistent with that reported in a previous study, indicating that HEPES salt did not affect the orientation of cellulose microfibrils or embedded cellulose microfibrils. In contrast, for the peak at approximately q = 0.85 nm^–1, the scattering intensity gradually diminished with increasing zwitterionic-modified side chain density of the ZIPs. In particular, when the cell wall was treated with P3, the peak completely disappeared, indicating that the orientation and alignment of the cellulose microfibrils along the longitudinal direction of the cells were distorted after P3 treatment. This may be due to the effective release of cellulose from the cell wall.

ZIPs Affect the Nanomechanical Properties of the Plant Cell Wall

To further quantify the cell wall denaturing abilities of the respective ZIPs, the modulus of the ZIP-treated cell wall surface was used as an indicator. Nanoindentation measurements combined with AFM techniques are typically used to quantify the mechanical properties of cell wall surfaces. Thus, we conducted nanoindentation measurements on ZIP-treated cell walls by pressing the tip 150–200 nm into the respective wall surfaces before the target force of 8 nN was reached, and we obtained retraction curves (Figure a). The indentation depth became shallower on the basis of the obtained retraction curves from the ZIP-treated cell wall. Nanoindentation measurements were performed using various onion epidermal cell walls, and the Sneddon model was applied to the obtained retraction curves to calculate the surface modulus of the respective walls. The elastic modulus of the 20 mM HEPES-treated cell wall was determined to be 3.2 ± 1.8 MPa. This value is relatively consistent with the indentation modulus observed after water incubation, as previously reported, suggesting that the addition of HEPES salt had little effect on the cell wall surface. In contrast, the calculated modulus of the ZIP-treated cell walls was higher than that of the 20 mM HEPES-treated cell wall, with the modulus increasing as the side chain density of the ZIPs increased. The indentation modulus of the P1-treated cell wall was 3.8 ± 2.0 MPa, which was slightly greater than that of the 20 mM HEPES buffer-treated cell wall. The cell wall moduli of P2- and P3-treated cell walls were calculated to be 4.0 ± 1.8 MPa and 4.9 ± 3.2 MPa, respectively, both of which exceeded that of the 20 mM HEPES buffer-treated cell wall (Figure b). A previous report indicated that the indentation modulus generally reflects the compliance of an amorphous polysaccharide (such as pectin) on the cell wall surface. In addition, ZIP itself is considered to reduce the indentation modulus when merely absorbed on the cellulose surface^.28 The observation that the indentation modulus increased with increasing side chain density suggests that the high hydrogen bond cleavability of highly side chain-modified ZIPs promoted interactions, not only with cellulose microfibrils embedded within the cell wall but also with the amorphous polysaccharides on the cell wall surface.

ZIP Treatment Does Not Weaken the Initial Cell Wall Elastic Properties or Loosening Properties

The mechanical properties of the denatured cell walls treated with ZIPs were measured to reveal the effect of hydrogen bond cleavage by ZIPs. To obtain stress–strain curves, cell wall strips treated with ZIPs or 20 mM HEPES buffer were clamped onto a handmade extensometer and stretched until breakage or until the elongation limit of the machine was reached. The stress was divided by the cross-sectional area (3 mm × 7 μm) of the initial wall strips as previously reported. , The strain was calculated by dividing the elongation length by the initial length of the cell wall strip (5 mm). No clear differences were observed between the stress–strain curves obtained from ZIP-treated or buffer-treated cell walls (Figures a, b, c). In the initial linear region of the strain stress curves showing elastic deformation, as shown in the merged graph in the inset in Figure , there was no significant difference between ZIP-treated cell walls and 20 mM HEPES buffer-treated cell walls. This suggests that the initial rigidity of the cell wall itself was retained, even though the hydrogen bonding was cleaved by ZIPs. Despite hydrogen bond cleavage, the minimal effect of ZIPs on the initial modulus of the cell wall is consistent the results of molecular dynamics (MD) simulations from a previous study, which indicated that hydrogen bonds make a relatively minor contribution of only 11% to the strength of cellulose. However, in the plastic deformation region (λ = 0.3–0.4), the P2- or P3-treated cell wall required more stress, suggesting that it became stiffer than the 20 mM HEPES buffer-treated cell wall. This may be due to the spatial distance effect of the cellulose microfibrils induced by ZIPs.

Furthermore, cell wall “loosening”, a sliding of cellulose microfibrils induced by a few nonenzymatic wall proteins such as expansin, is well-known as a characteristic behavior of cell walls that facilitates cell expansion during plant growth. Only a few proteins are reported to have a loosening effect, which can be interpreted as the ability to increase the creep activity of the wall, and are considered valuable materials for reducing cell wall recalcitrance in the biorefinery field. Cleavage of the inter/intracellular hydrogen bonds formed in cell wall polysaccharides is considered an essential step in the induction of wall creep. Since ZIPs, especially P3, contain abundant hydrogen bond-cleaving moieties, we hypothesized that P3 would be a “synthetic loosening protein”. After onion epidermal cell wall strips were incubated with P3 solution for 40 min, a slight decrease in the extension rate from 0.30 μm/min to 0.16 μm/min was observed (Figure S13). Both the wall incubated with 20 mM HEPES buffer and P3 showed a monotonic increase in the creep length (indicated in the graph as creep in position). This suggests that the highly side chain-modified ZIP P3 had no effect for loosening. Overall, ZIPs cleaved hydrogen bonds in the cell wall without causing loosening or weakening of the initial elastic properties derived from crystalline cellulose. This suggests that ZIPs enable the manipulation of hydrogen bonds in the cell wall while preserving its mechanical stiffnessa challenge for genetic modification. This characteristic is crucial for the development of high-stiffness biomaterials.

Effects of ZIPs on Living Plant Cells

The cytotoxicity of ZIPs to live plant cells was evaluated using onion epidermal cells to assess their potential for effective cell wall denaturation in a plant system. After the plant cell strips obtained from the adaxial side of the onion epidermis, which maintained intact living cells, were incubated with aqueous ZIP solutions (2 mg/mL), 20 mM HEPES, and 1 N NaOH for 1 h, the strips were immersed in trypan blue staining solution for 5 min to stain the dead cells blue (Figure a). To confirm that the dead cells could be stained with trypan blue, the plant cells were treated with 1 N NaOH solution, which is known for its high cytotoxicity due to interactions with cell membranes and cell walls, and then stained. Most cells were stained blue, confirming that trypan blue effectively stained the dead cells. In contrast, transparent cells were observed among onion epidermal cells treated with ZIP or 20 mM HEPES, with few dead cells. To quantify the cytotoxicity of ZIPs in different onion samples, similar treatments were performed on epidermal cell sections from each onion, and the cell viability was calculated by dividing the number of live cells by the total number of cells (approximately 500 cells). The cell viability was greater than 90% in sections treated with ZIP and 20 mM HEPES, indicating that 2 mg/mL aqueous ZIP solution, which can dissolve cellulose and amorphous polysaccharides, did not affect cell viability. There were also no differences in cell viability due to variations in the degree of zwitterion modification per strand (Figure b). This low cytotoxicity is presumably due to the overall neutral charge of the side chains, which prevents interactions with cell membranes. , Thus, ZIPs can interact with cellulose or amorphous polysaccharides in the cell wall for cell wall denaturation without causing cytotoxicity to plant cells.

Possible Mechanism of Cellulose Hydrogen Bond Disruption by ZIPs

The present results demonstrate that highly side chain-modified ZIPs effectively disrupt the hydrogen bond network within cellulosic substrate filter paper and plant cell walls, providing new chemical methods to reduce cell wall recalcitrance without causing significant cytotoxicity.

The hydrogen bond disruption of cellulosic substrates, filter paper, and cell walls was confirmed by FT-IR analysis, showing a decrease in the peak area around 3,400 cm^–1 associated with hydrogen bonding among cellulose (Figure S7). As a result, the leaching of cellulose, particularly amorphous cellulose, from the substrates was confirmed (Figure a, a, b, c, Figure S7, S11). ZIPs exhibited hydrogen cleavage ability in aqueous environments at low concentrations (2 mg/mL), overcoming the water sensitivity limitation associated with ZIL-based approaches. In particular, the amount of leached cellulose was greater for highly modified ZIPs (Figure , Figure f, Figure S7 and S11). In a 2 mg/mL ZIP aqueous solution, P1 contains 3 × 10^–3 M of zwitterionic structures, whereas P2 and P3 contains 5 × 10^–3 M, representing approximately 1.6 times the amount of the effective zwitterionic functional groups. If the cellulose release were purely dependent on the effective concentration, the release amount of P2 and P3 would be expected to increase only about 1.6-fold relative to P1. However, the actual released amount was approximately 2.5 times higher, confirming a release enhancement beyond the concentration ratio. Furthermore, FT-IR results confirmed that P2 and P3 led to significantly greater disruption of hydrogen bonds and a larger amount of dissolved amorphous cellulose. These findings suggest that for effective cellulose elution, it is important not only to increase the concentration of zwitterionic structures but also to locally concentrate them. In addition, the morphology of the leached cellulose varied depending on the ZIP applied. The highly side chain-modified ZIPs, especially P3, resulted in the dispersal of shorter cellulose fibrils or collapsed crystalline structures (Figure S8d and Figure c). These differences were probably also due to variations in the hydrogen bond cleavability of the ZIPs originating from the differences in side chain density. The structural variation in the repeating units among P1, P2, and P3 may also affect the hydrogen bond cleavability.

The morphology and amount of released cellulose varied among the ZIP treatments; however, WAXS measurements revealed that the crystalline structure of the cellulose Iβ remained intact after treatment with the respective ZIP solutions (Figure S12). Considering that ZIP interacts with cellulose microfibrils in the cell wall, resulting in the formation of rod-like cellulose Iβ crystals, highly side chain-modified ZIPs are suggested to solubilize and interact with amorphous cellulose or polysaccharides rather than interacting with the crystalline region at low concentrations (Figure a, b, c). Treatment with highly side chain-modified ZIPs (P2 and P3) distorted the orientation of the cellulose microfibrils embedded in the cell wall, making their alignment more flexible (more isotropic), as clarified by SAXS (Figure f). The contrast between the preservation of crystallinity (Figure S12) and the significant perturbation of microfibril alignment (Figure f) physicochemically proves that ZIP action is strictly confined to these interfibrillar interfaces (the amorphous barrier). This structural change suggests that the hydrogen bond network in the cell wall uniformly collapsed after ZIP treatment because of the removal of tethered amorphous polysaccharides with cellulose microfibrils in the cell wall that act as a polysaccharide glue to fix the anisotropic alignment of the cellulose microfibrils. ,,, Generally, the cell wall recalcitrance is fundamentally rooted in two hierarchical factors: (i) the inherent structural rigidity of the cellulose crystalline core, and (ii) the physical shielding of cellulose surfaces by a dense, hydrogen-bonded network of amorphous polysaccharides (e.g., pectin and hemicellulose) that coat and tether the microfibrils. ZIPs are highly targeted hydrogen bond disruptors that reduce recalcitrance by addressing these amorphous barriers while preserving crystalline and overall structural integrity.

Furthermore, studies of the mechanical properties of ZIP-treated plant cell walls (primary cell wall of the onion epidermis) support the removal of amorphous polysaccharide by highly side chain-modified ZIPs. The indentation modulus calculated from the highly side chain-modified ZIP-treated cell walls increased (Figure b). The indentation modulus is generally influenced by the elastic compliance of the pectin layer composed of homogalacturonan and the bendability of the cellulose microfibrils on the cell wall surface. , Considering that ZIPs did not deteriorate the structure of crystalline cellulose, highly side chain-modified ZIPs have the potential to affect amorphous polysaccharides (pectin and xyloglucan) on the cell wall surface, resulting in a change in the apparent elastic compliance calculated from the indentation modulus. The increase in the indentation modulus might occur because the cellulose surface might be exposed by the removal of amorphous polysaccharides by ZIPs, promoting direct contacts between cellulose microfibrils. The initial elastic properties, which can be calculated from the initial elastic region (λ = 0–0.1) measured by tensile tests, did not differ between ZIP-treated cell walls and 20 mM HEPES buffer-treated cell walls (Figure ). Theoretically and experimentally, cellulose microfibrils, rather than tethering hemicelluloses or abundant pectin (not from viscoelastic amorphous polysaccharides), are considered to be the load-bearing polymers in the elastic region (λ = 0–0.08). ,, In particular, the elastic stiffness is governed primarily by chain packing proximity within the crystalline structure of cellulose, as indicated by the d-spacing of the (200) plane. , The preservation of the elastic properties of the cell wall after ZIP treatment aligns with the results of WAXS measurements, which revealed that the distance between cellulose chains in cellulose Iβ (d-spacing of the (200) plane) remained unaffected by ZIP treatment. Indeed, crystalline cellulose remained intact; however, given the high hydrogen bond cleavability of ZIPs, it is likely that nonload bearing amorphous cellulose and polysaccharides were affected, resulting in the minimal impact on the elastic modulus. The discrepancies between the moduli obtained from indentation and tensile tests are typical, as the indentation modulus is influenced by the compliance of amorphous polysaccharides, whereas the tensile properties reflect the compliance of cellulose microfibrils alone. Considering that hydrogen bond cleavage and solubilization by ZIP occurred mainly with amorphous cellulose and polysaccharides, the results obtained from indentation and tensile tests can also be understood, and the discrepancy in the calculated modulus is considered acceptable.

In the cell wall, expansin and hydrogen bond cleaving proteins modulate the extensibility of the cell wall during plant growth. During cell wall expansion via “expansin”, the sliding of cellulose microfibrils, which can be detected as a rheological creep activity of the cell wall, is induced by transient disruption of noncovalent bonding between contacting surfaces of cellulose microfibrils. Therefore, hydrogen bond-cleaving ZIPs were also expected to induce creep activity of the cell wall and act as synthetic wall-loosening proteins similar to expansin. Contrary to expectations, negligible changes of creep activity was observed after the highly side chain-modified P3 treatment (Figure S13). This suggested that the collapsed anisotropic network of cellulose and the solvation of the amorphous polysaccharides between those fibrils induced by ZIPs has little effect toward sliding of the cellulose microfibrils. The lack of macroscopic creep despite clear nanoscale structural perturbation underscores a fundamental difference between mechanical wall loosening and chemical reduction of recalcitrance. Expansin activity relies on hotspot-specific disruption of load-bearing junctions formed by cellulose–hemicellulose assemblies, a process that requires precise molecular recognition enabled by planar aromatic motifs. In contrast, ZIPs do not form planar binding motifs in aqueous solution (Figure S4), and have no targeting ability to such “biomechanical hotspot”. Therefore, ZIPs act through spatially distributed hydrogen-bond disruption. Such nanoscale disorder is sufficient to alter microfibril alignment and increase molecular accessibility, as evidenced by SAXS, yet insufficient to compromise long-range mechanical connectivity of the wall. In summary, the ZIP-modified cell wall obtained here had no significant cell wall mechanical depression, which is supported by the high selectivity of the interaction between ZIP and hydrogen bonding. Additionally, the ZIPs exhibited no cytotoxicity toward plant cells but achieved effective hydrogen bond cleavage and cell wall denaturation at low concentrations, suggesting that they successfully addressed the trade-off between hydrogen bond cleavability and cytotoxicity.

The resolution of this trade-off, which allows ZIPs to exhibit high solubility in cellulose and amorphous polysaccharides without significant cytotoxicity, is attributed to the “synergistic effect” of the highly side chain-modified ZIPs. A single hydrogen bond in an aqueous solution is relatively weak, with a bond energy of approximately 6.5 kJ/mol. However, in filter paper or cell walls, extensive and multiple hydrogen bonds are formed between cellulose–cellulose or cellulose–amorphous polysaccharides, and these bonds are cooperatively stabilized by electron delocalization, leading to the formation of a strong hydrogen bond network. ,− In this network, when a single hydrogen bond donor or a zwitterionic moiety approaches, it can easily cleave one hydrogen bond such as opening a zipper. The multiple hydrogen bonds formed between the polysaccharides prevent the donor from cleaving all of the bonds formed among the polysaccharides. However, in ZIPs, the zwitterion moieties are not diffuse but are confined to a specific point by covalent bonds; hence, ZIPs possess a localized hydrogen cleavage site that is inserted between the polysaccharides and weakens multiple hydrogen bonds formed among the polysaccharides. The inserted ZIP starts to “unzip” the multiple points of hydrogen bonds between the polysaccharides and may allow water molecules to come between the polysaccharides, resulting in complete dissociation (Figure ).

8.

Localization of zwitterionic moieties (the so-called “synergistic effect”) enhances the defibrillation of cellulose fibrils.

Conclusions

In summary, the ZIPs developed in this study could be alternative chemical reagents for efficient cell wall modification in water-rich environments without cytotoxicity. Localization of the ionic moieties by densely linking the zwitterionic structures with covalent bonds significantly solubilized and dispersed the amorphous cellulose or polysaccharides by cleaving densely linked hydrogen bonds. These findings indicate that treating cell walls with a highly side chain-modified ZIP (P2 or P3) is a promising possibility for reducing the recalcitrance of biomass and offers a new way to handle cellulosic biomass. This approach not only results in low cytotoxicity but also preserves mechanical strength during cell wall modification, making it an attractive pretreatment for reducing cell wall recalcitrance. More detailed studies are needed to fully understand how ZIPs interact with polysaccharides at the molecular level, especially regarding the role of water molecules in the ZIP solution. Considering that ZIPs do not affect the crystalline region of cellulose, relatively bulky ZIPs might directly cleave the hydrogen bonds in amorphous polysaccharides to dissociate them. This research could provide new insights into improving the solubilization of other biopolymers, such as chitin, silk fibroin, and starch, which also feature strong hydrogen bond networks. By leveraging these insights, ZIPs could become more valuable reagents for a broader range of applications in biorefineries and biopolymer engineering.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.biomac.5c02456.

• Experimental details; synthesis of ZIPs, ¹H NMR spectra, and DLS profiles of ZIPs; SEM images and FT-IR spectra of ZIP-treated filter paper surfaces; DLS profiles of cellulose released from filter paper under ZIP solutions; FT-IR spectra and X-ray scattering profiles of ZIP-treated cell walls; and mechanical properties of ZIP-treated cell walls (PDF)

R.N., D.J.C., K. Tsuchiya, and K.N. conceived and designed the research. R.N. and K. Terada wrote the manuscript. R.N. and L.W. conducted the AFM observations. R.N. and H. M conducted the SAXS measurements. R.N., K. Terada, and K. Tsuchiya performed all the other experiments and analyzed the data. D.J.C., K. Tsuchiya, and K.N. edited the manuscript.

The authors declare no competing financial interest.