Imaging Changes in Cell Walls of Engineered Poplar by Stimulated

Research Article pubs.acs.org/journal/ascecg

Cite This: ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Imaging Changes in Cell Walls of Engineered Poplar by Stimulated Raman Scattering and Atomic Force Microscopy Wei Shen,†,‡ Cynthia Collings,†,‡ Muyang Li,†,‡ Jake Markovicz,† John Ralph,‡,§ Shawn D. Mansfield,‡,∥ and Shi-You Ding*,†,‡ †

Department of Plant Biology, Michigan State University, 612 Wilson Road, East Lansing, Michigan 48824, United States DOE Great Lakes Bioenergy Research Center (GLBRC), East Lansing, Michigan 48824, United States § Department of Biochemistry, University of Wisconsin−Madison, 433 Babcock Drive, Madison, Wisconsin 53706, United States ∥ Department of Wood Science, University of British Columbia, 4030-2424 Main Mall, Vancouver, British Columbia V6T 1Z4, Canada

Downloaded via GUILFORD COLG on July 18, 2019 at 11:58:14 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

S Supporting Information *

ABSTRACT: Lignin is a major impediment in the deconstruction of plant biomass to its soluble monomeric constituents. Modification of the lignin biosynthetic pathway has proven to be an effective means of reducing biomass recalcitrance but can often result in impaired growth. As one of the successful examples of “designer” biomass, “Zip-lignin” has been generated in poplar to introduce ester linkages into the lignin backbone. The resulting plants grow normally and have improved biomass deconstructability, especially under mild alkaline conditions that can effectively cleave the uniquely introduced ester bonds. In order to further understand the structural and chemical features that may be associated with the observed improved processing efficiency, we used hyperspectral stimulated Raman scattering (hsSRS) to map the Zip-lignin in planta using lignin conjugated α−β carbon double bonds as the proxy. Such double bonds are presented in all types of lignin, but unique structures in Zip-lignins experience a ∼20 cm−1 shift from 1650 to 1630 cm−1 in their Raman spectra due to the conjugation between the double bond and an associated ester carbonyl group. Analysis of the hsSRS images has revealed that the Raman signal specifically representing Zip-lignin can be estimated. Moreover, it exhibits a distribution pattern in cell wall layers similar to that of the native lignin, but its intensity varies in different transgenic poplar lines. In addition, when the cell walls are imaged by atomic force microscopy and Simons’ staining, we found that the poplar containing the Zip-lignin had increased accessible areas, which may be beneficial to the chemical penetration during pretreatment and substrate accessibility during enzymatic hydrolysis. Our study suggests that both physical and chemical modification of cell wall in Zip-lignin poplar could contribute to the observed improvements in sugar yields. KEYWORDS: Hyperspectral imaging, Simons’ Stain, Poplar cell walls, Biomass pretreatment, Cell wall porosity

■

INTRODUCTION Lignin is a complex natural biopolymer that is derived from the oxidative coupling of various hydroxycinnamyl alcohol monomers, including p-coumaryl, coniferyl, and sinapyl alcohols that, respectively, give rise to H, G, and S units in lignin.1 The combinatorial coupling process of lignin polymerization produces a variety of structures in the polymer, with no fixed pattern or sequence. The resulting supermolecules are hydrophobic and may be cross-linked with polysaccharides forming complex matrix structures in the secondary cell walls. Such features contribute to the recalcitrant nature that impedes the chemical and enzymatic deconstruction of biomass.2 The relatively strong lignin backbone linkages make it difficult to efficiently and selectively remove lignin from the cell wall structure by chemical reactions; the weakest links are featured in the prominent β-aryl ether bonds which still require harsh conditions (e.g., 1 M alkali at typically 170 °C or 1 M acid at 190−200 °C, each for several hours) in order to be cleaved. © 2019 American Chemical Society

Typically, cleavage of at least some of the lignin is required to provide access to the polysaccharide constituents, cellulose and hemicelluloses.3 One of the recent breakthroughs in lignin modification has been the successful engineering of Zip-lignins.4 Such “designer lignins” were produced by introducing a ferulate monolignol transferase (FMT) gene into Arabidopsis and poplar. The resulting transgenics plants employ the ensuing monolignol ferulate conjugates as component lignification precursors (“monomers”) and incorporate at both the monolignol moiety, in the same manner as for the conventional unacylated monolignols and the ferulate moiety, by radical coupling reactions with other monomers and/or with the growing polymer. The resulting cell wall lignins were shown to possess Received: February 27, 2019 Revised: April 10, 2019 Published: May 27, 2019 10616

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering

modulated by an electro-optical modulator (EOM) at 20 MHz, serving as a Stokes beam. The combined beam was sent into an inverted laser-scanning microscope (Olympus IX83, with Fluoview 1200 scanning-head). A 60× water-immersion objective lens (Olympus UPLSAPO 60XW/IR) was used to focus the beams onto the sample, which was mounted between a 1 mm microscopy glass slide and a 150 μm coverslip. A plastic sticky spacer was employed to reduce the distortion of the cell wall structure between the slide and coverslip. The output beam, after passing through the microscope, was filtered with a Chroma short-pass filter (ET890/220m) to block the Stokes beam. The pump beam was detected with a photodiode and lock-in amp (APE GmbH) at 20 MHz which corresponds to the stimulated Raman loss through frequency transfer. The temporal delay between two lasers controls the Raman shift that was detected at each frame. Each frame was scanned in 1.12 s, providing images of 512 × 512 pixels. In total, a stack of hyperspectral images across a wavelength range of 300 cm−1 took ∼1 min. Simons’ Stain. Direct Blue 1 (DB, Pontamine Fast Sky Blue 6BX) and Direct Orange 15 (DO, Pontamine Fast Orange 6RN) dyes were obtained from Sigma-Aldrich. Biomass samples (∼10 mg) of each Zip-lignin line and wild-type trees were weighed into Eppendorf conical tubes. DB and DO dye solutions (100 μL of 10% solutions in water) were each added to the tubes with distilled water to make up the final volume to 1 mL. The final concentration for each dye was 1%. All tubes were placed on a rotator at minimal speed (10 rpm) overnight to permit complete staining. The absorbance of the supernatant solution was then obtained on a Thermo Nanodrop 2000C spectrophotometer at 455 and 628 nm, the wavelengths representing the maximal absorbance for DO and DB, respectively.15 The amounts of dyes absorbed by individual biomass samples were represented by the difference in absorbances before and after the reaction. Atomic Force Microscopy. Wild-type and Zip-5 samples in the AFM experiments were imaged at room temperature on a Dimension AFM with Nanoscope controller V (Fastscan, Bruker Nano, Santa Barbara, CA, U.S.A.) fit with an acoustic and vibration isolation system. SCANASYST-air (Bruker, Camarillo, CA, U.S.A.) probes were used for all imaging, as they are suitable for imaging in an air environment. The AFM operation software (Nanoscope V9.1) was employed to control the scan size, set point, and gain values. Before AFM imaging, the scanner was carefully calibrated using a calibration kit (Bruker, Camarillo, CA, U.S.A.) to ensure that all the measurements were authentic. The scan rate was normally set at 1−3 Hz. All images were taken at 512 × 512 pixels. At least five images with different scan sizes (i.e., 0.5, 1, 2, and 5 μm) were taken in the same scan area and on the same piece of cell wall, and at least five different areas were measured.

ester bonds in the lignin backbone, as designed. Subsequently, Zip-lignin poplar demonstrated improved cell wall digestibility after mild alkaline pretreatment,4 improved chemical pulping,5 as well as beneficial processing with ionic liquid treatment6 and under copper-catalyzed alkaline hydrogen peroxide (Cu-AHP) oxidation.7 However, the complex chemical and structural features in the cell walls affecting the digestibility of Zip-lignin poplar remain poorly understood and, in particular, we do not fully understand the efficiency or distribution of the monomer conjugates and their integration throughout the secondary cell wall. Previously, we have reported on the applications of stimulated Raman scattering (SRS) to image lignin in plant cell walls with high spatial resolution and fast image acquisition based on the inherent aromatic Raman vibration band at 1600 cm−1 for lignin and the C−O stretching band at 1100 cm−1 ascribed to cellulose.1,8−11 We have also shown that in a typical poplar cell wall transverse section, without treatment to minimize autofluorescence background, the distribution of cellulose appears to be relatively uniform, whereas lignin concentration varies in different wall layers, as has been long known.12 For example, it has been shown that lignin is inherently highest in concentration in the cell corner, followed by the compound middle lamella.13 The lignified secondary wall contains three sublayers from the outside to the inside denoted as S1, S2, and S3; S1 and S3 are thin, and S2 constitutes the bulk of the secondary wall and contains the majority of the lignin, although the compound middle lamellae contain the highest concentration. Using SRS microscopy, we have also recently demonstrated real-time imaging of lignins and polysaccharides in the plant cell walls during chemical pretreatment and hydrolysis by extracellular fungal cellulases.9 We show that SRS microscopy offers diffraction-limited spatial resolution and pixel dwell times of down to 100 ns, allowing video-rate imaging (2−100 μs per pixel are sufficient and routinely used) and quantitative microscopy at the subcellular scale. Here, we demonstrate a new imaging modality to enable in situ visualization and quantification of the Zip-lignin in plant cell walls leveraging hyperspectral SRS microscopy and the unique features inherent in Zip-lignin’s molecular structure. We were able to achieve high Raman intensity comparable to previous studies and maintain fast imaging speed, while unambiguously identifying the conjugated lignin esters that comprise a fraction of the lignin complex. In addition, Simons’ staining and atomic force microscopy (AFM) were used to investigate the molecular scale accessibility and structure of these unique cell walls.

■

■

RESULTS AND DISCUSSION In addition to the lower energy required for cleaving the uniquely incorporated backbone ester bonds compared to that for the native ether bonds in lignin,16 the incorporation of monolignol ferulate conjugates into the lignin polymer may temporally and structurally affect the lignification process, which in turn may also contribute to the observed benefits in cell wall digestion.4,7 The level of ferulate conjugates incorporated into the secondary walls of Zip-lignin poplar has been estimated via derivatization followed by reductive cleavage (DFRC).4,17 However, although the method is entirely diagnostic, the quantification remains a best-estimate projection based on model systems as only a small fraction of lignin units are released during the degradation reactions. Independent Zip-lignin lines (i.e., Zip-5, -6, -7, and -8 lines) were generated, containing different levels of transcripts, proteins, and consequently lignified ferulate conjugates. Among these, Zip-5 and Zip-7 have consistently been shown to have higher sugar yields than the other transgenic lines and

EXPERIMENTAL SECTION

Plant Materials. Wild-type (WT) line P39 poplar (Populus alba x grandidentata) and Zip-lignin poplar lines (Zip 5, 6, 7, and 8) were produced and grown in a greenhouse at the University of British Columbia, Vancouver, British Columbia as previously described.4 For all imaging, 1-year-old poplar stems were cut in cross-sections at a thickness of 50 μm. Hyperspectral SRS Imaging Setup. The spectral focusing hsSRS imaging method was used for fast hyperspectral scanning.14 An ultrafast laser system with dual output (InSight DeepSee, SpectraPhysics) provided the excitation sources. The two broadband lasers which were centered on 1040 and 889 nm were chirped to approximately 1.5 ps using SF57 glass rods. Both lasers emitted pulses at 80 MHz. The 889 nm laser was routed through a motorized delay stage, serving as a pump beam. The 1040 nm laser was 10617

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering the wild-type trees.4 Although most of the lines have zip levels that align well with the digestibility data, there is one inconsistency; Zip-7 line had the highest detectable Zip-lignin level, whereas the Zip-5 line produced the highest sugar yields.4 There is therefore a pressing need for an alternative quantification method for ferulate conjugate (relative) levels, ideally from plant materials in their native state, to provide a deeper understanding of the current data, as well as to identify other associated changes that manifest from the conjugates’ introduction into plant cell walls. We hypothesize that both chemical and potential indirect structural changes in Zip-lignin poplar cell walls contribute to the variation in improved digestibility, as indicated by the improvement across a broad range of pretreatment methods.4,7 In order to identify the factors in Zip-lignin that may affect the deconstruction processes, we compared the wild-type poplar and different transgenic lines that have demonstrated differences in response to various pretreatments.4 SRS, AFM, and Simons’ staining methods were used to access the chemical and structural features. Hyperspectral SRS Imaging to Identify Zip-Lignin. SRS has been developed to examine the vibrational fingerprints of selected chemical species in situ, providing a fast and labelfree chemical contrast that may otherwise be difficult to obtain. Raman spectra between 1600 and 1650 cm−1 contain signature bands from lignin, originating from both aromatic rings and α−β carbon double bond. SRS microscopy creates chemicalspecific images based on signature bond vibration modes from target species. By scanning through a range of wavenumbers, we acquired hyperspectral images that could be used to isolate Raman responses from individual bands. In spectroscopic imaging, each pixel contains spectral information on interest. The total accumulation time is essentially limited by the number of pixels and the time required to spectrally sample data for each pixel. A requirement is for the Raman mode to have a relatively large cross-section, i.e., strong signal relating to the selected molecular bond. Another consideration in SRS imaging is that the Raman band representing the chemical species needs to be specific for the target of interest. For example, lignin is the major cell wall polymer that consists of aromatic molecules so aromatic ring stretching at 1600 cm−1 provides a good representation for lignin in general. When examining the distinction between native lignin and Zip-lignin, the additional moiety in Zip-lignin, i.e., the ester bond resulting from the ferulate conjugate, produces a Raman band at ∼1700 cm−1 due to the carbonyl group’s C=O bond. However, this is far from unique, as there are also many other groups in plant cells walls that can generate Raman band at 1700 cm−1, e.g., many carboxylic acids and esters which also contain carbonyl groups (see Figure S2). As such, the carbonyl stretch is simply insufficiently discrete to permit its specific use for the imaging of Zip-lignin. The π electrons on the Zip-lignin carbonyl, however, conjugate with the α−β carbon (ethylenic) double bond in some structures resulting from ferulate incorporation into the lignin (Figure S2). These are unique and therefore should be diagnostic. There are nonconjugated double bonds in end-group units in “normal” lignins, but those that are conjugated, in esters, shift the Raman band from 1650 to 1630 cm−1.18,19 Such a shift was predicted by DFT computation and was observed in lignin monomer derivatives modeling the structures involved.20 The α−β double bond conjugates, with both the carbonyl on Zip-lignin ester bond and the aromatic rings in lignin (Figure 1), provide a unique signature to

Figure 1. Chemical structure of native lignin (a) and a selected example of a Zip-lignin structure (b). The addition of π electron conjugation from carbonyl group in Zip-lignin shifted the Raman peak of α−β carbon double bond. The difference in π electron conjugation is highlighted in red.

measure Zip-lignin. It should be noted that the native lignin with an aldehyde end group could still show up at 1630 cm−1 due to the similar π conjugation. The aldehyde lignin may therefore contribute part of the 1630 cm−1 in WT sample, but the gains in Zip-lignin lines are likely derived from the introduced ferulate conjugate. Additional studies will be carried out to further discriminate Zip-lignin from aldehyde lignin by selectively reducing aldehydes. We measured the Raman spectra of several synthetic samples, as well as lignin from wide-type and Zip-lignin line5 poplars (Figure 2). Both coniferyl alcohol and sinapyl alcohol clearly showed Raman peaks at 1600 and 1650 cm−1, representing aromatic ring stretch and ethylenic bond stretch, respectively.21−23 Coniferyl ferulate and ethyl cinnamate, both containing carbonyl-conjugated ethylenic bonds similar to those in Zip-lignin, showed peaks at 1600 and 1630 cm−1. The 20 cm−1 blue shift on the ethylenic bond stretch was caused by the conjugation with the carbonyl group. The Raman spectra of poplar samples showed broader bands in this region, due to broadening caused by the heterogeneity of plant cell walls and the variety of structures produced by ferulate coupling into lignin (Figure S2). Despite the broad spectra, the Zip-lignin poplar spectrum clearly had a higher contribution from the 1630 cm−1 component than wild-type poplar. The Raman bands measured from the poplar samples consist of overlapping contributions from all three Raman peaks. However, we can mathematically deconvolute the Raman spectra into individual Gaussian peaks, centered at 1600 cm−1, 1630 cm−1, and 1650 cm−1 which can subsequently be used to represent total lignin, Zip-lignin, and native lignin levels, respectively. (Figure 3) Raman Spectral Deconvolution and Zip-Lignin Quantity. Multivariate Curve Resolution (MCR) has been used to decompose hyperspectral images including hyperspectral SRS images.24−27 Briefly, the experimental data matrix consisting of Raman intensity at each pixel across a range of wavenumbers (hsSRS images) was decomposed into two smaller matrices that represent concentration maps for each component and their corresponding spectral features (i.e., Raman spectra). The experimental data matrix is therefore the scalar product of the two smaller matrices plus any errors. The major components are estimated through PCA, and the final results were determined through the MCR method when convergence was reached. At that point, we obtained the calculated concentration map at the three major Raman bands (1600, 1630, and 1650 cm−1, Figure 3c). We recreated the images with new concentration maps for each of the Raman bands (Figure 3a,b) and found that the distribution of Ziplignin and native lignin largely colocalized with each other. The 10618

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering

Figure 2. Raman spectra of selected standards and poplar samples measured by hyperspectral SRS imaging. The spectra clearly show the 1650 and 1630 cm−1 Raman peaks from the native and Zip-lignin like standards, whereas both poplar samples contain overlapping peaks from multiple Raman modes. CA: coniferyl alcohol, CF: coniferyl ferulate, EC: ethyl cinnamate, SA: sinapyl alcohol, WT: wild-type poplar, Zip5: Zip-lignin line 5.

Figure 3. Recreated Raman images of ferulate conjugates (a) and total lignin (b) show similar distribution patterns but differ in intensities. (c) Deconvolution of a Zip-lignin line-5 Raman spectrum into component Raman peaks. Blue dotted line: Zip-lignin line-5 spectrum. Solid lines in gray, green, and orange: deconvoluted Gaussian peaks at 1600, 1630, and 1650 cm−1, respectively. A.U.: arbitrary unit. (d) Raman intensities at 1630 cm−1 normalized to wild-type poplar as 1, representing the relative abundance of ester bonds. Higher bars indicate higher ester concentration. Scale bar = 5 μm (a, b); color bar shows the relative intensity.

describing these materials).4 It is difficult to estimate the association between the increase in ferulate conjugates and plant cell wall digestibility as many other factors such as the spatial localization of the ester linkages and overall degree of lignin polymerization can dramatically skew its impact. However, it is reasonable to assume that the increase in cell wall digestibility is largely determined by the incorporation of the ferulate conjugate and that the gains in pretreatment efficiency (∼25% increase on average) are largely derived from ester bond chemistry.4,7 Simons’ Stain and Increased Plant Cell Wall Accessibility. In order to investigate the overall physical accessibility to polymeric molecules, the Simons’ staining method was applied to both WT and Zip-lignin poplar lines. Direct dyes are

implication is that the Zip-lignin monomers were incorporated into the lignin in much the same way as the canonical monomers, as has been shown via DFRC etherification analysis, i.e., that they are fully compatible with the lignification process producing integrated copolymers rather than simply independent polymers that, admittedly, might not be readily distinguished here if they resided in the vicinity of native lignin across the plant cell walls. When we compared the Raman intensity at 1630 cm−1, which we previously ascribed to the conjugated esters, in the Zip-lignin and wild-type poplar samples we estimate that on average, the transgenic line has 43% greater ferulate conjugate levels (Figure 3d). By comparison, the DFRC levels of incorporation were estimated to be 10-fold higher (Table S4 and Figure S9 of the reference 10619

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering

Figure 4. Simons’ stain of poplar cell walls. Wild-type (a, b) and Zip-5 poplar (c, d) dyed with direct orange (a, c) and direct blue (b, d), respectively. Scale bar = 5 μm (a−d). (e) Dye solutions after adsorption by wild-type and Zip-lignin biomass samples. The control (no biomass) is on the left. (f) Reduction of absorbance of the liquid phase as a result of adsorption of Simons’ stain dyes on poplar samples. Orange and blue bars are for direct orange and direct blue, respectively. Error bars represent standard deviation of three replicates.

Figure 5. SRS images of poplar cell walls at 1600 cm−1 showing the overall lignin distribution. (a) Wild-type. (b−e) Zip-lignin lines 5, 6, 7, and 8, respectively. (f) Profile plot of orange line in part e (top) and green line in part a. Scale bar = 5 μm; color bar shows the relative intensity.

larger pores (>5 nm) and the surface. We hypothesize that part of the gains in sugar yield manifesting from the Zip-lignin poplar lines may be associated with an increase in accessibility due to disturbance of the lignification process. The Simon’s stain is based on the adsorption of the dyes which are a complex function of the surface conditions. It may not be sensitive to the new ester bonds in the Zip-lignin poplars or respond the same way as alkaline hydrolysis would. The difference in adsorption is therefore only derived from the

sensitive probes for the characterization of cellulose structure because of their linear structures and outstanding adherence specifically toward this polymer.15 Two dyes, direct blue (molecular weight 992.82 g/mol, molecular diameter ∼1 nm) and direct orange (molecular diameter ∼5 to 36 nm, variable molecular weight) were used to access the porosity of the cell walls. When cell walls are treated with a mixed solution of these two dyes, the blue dye enters all the pores with diameters larger than ∼1 nm, whereas the orange dye only populates the 10620

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering accessibility of the corresponding biomass samples. We found significant increases in both dye adsorptions in the Zip-lignin samples (Figure 4), revealing an increase in cell wall accessibility as a result of an incorporation of Zip-lignin conjugates. Furthermore, the variation in dye adsorption follows the pattern of sugar yield when comparing the different Zip lines. In our results, Zip-lignin line-5 and line-7 have the highest dye adsorptions and also have the highest sugar yield after pretreatment.4 It was also reported that the Zip-lignin poplar showed improved glucose yield when treated with ionic liquid.6 Part of the improvement could be due to the increased cell wall accessibility. We calculated the correlation between dye adsorption and glucose yield after ionic liquid pretreatment across the wild-type/Zip-lignin lines; the R2 value for direct blue and direct orange were 0.797 and 0.842, respectively (Figure S1), suggesting that the cell wall accessibility is strongly correlated to the increases in final sugar yield at least for the ionic-liquid-pretreated biomass. Structural Changes in Zip-Lignin Poplar Cell Walls. SRS and AFM imaging were used to investigate the structural changes in plant cells and further understand the impact on chemical and physical accessibility. As discussed, there is no noticeable difference in lignin distribution between the ferulate conjugates in the transgenics vs the native lignins. This is consistent with reports that monolignol ferulate conjugates were incorporated into lignin backbones, as evidenced by the etherification status of both “ends” of the conjugates (Figure S8 in the original reference describing these materials).4 However, the Zip-lignin transgenic lines showed some abnormalities in the secondary wall lignin distribution, as shown in SRS images (Figures 5b−e), compared to the wildtype trees. We observed an increased accumulation of lignin in the secondary walls which coincided with a later stage of lignification. This can be clearly viewed from profile plots as shoulders in Figure 5f. It appears that the overexpression of the FMT gene did not change the lignin pathway when producing the monolignols including the ferulate conjugates as evidenced by their colocalization. The lignin patterning in the secondary wall indicates that the presence of ferulate conjugates might slow down lignification which leads to elevated lignin levels toward the end of the cell’s development, which may be due to a slowing of the diffusion/transport of the monolignols into the secondary wall. It is not immediately clear how this could be related to the increase in cell wall digestibility. It does show that the introduction of monolignol ferulate conjugates can impact the structure of cell walls to a certain extent. It may be even more informative to examine the ultrastructure on a nanometer scale as it is directly related to molecular and enzymatic accessibilities. AFM provides detailed surface topographic imaging by raster-scanning using a nanometer sharp probe. As the majority of biomass is in the S2 layer of the secondary cell wall, we cut the poplar samples longitudinally to expose the inner structure of the cell wall and make it accessible to an AFM tip. The overall structures of WT and Zip-5 appear similar in large-scale images, such as 5 μm × 5 μm scan areas and above (data not shown). The zoomed-in images (Figure 6) show that the majority of the cellulose microfibrils are arranged in a parallel manner, and matrix polymers, presumably lignin and hemicelluloses, are deposited between the cellulose microfibrils. At the cellulose microfibrillar level, we found that the Zip-lignin poplar appeared to be more disorganized compared to the wild-type, potentially creating higher porosities that promote cell wall accessibility,

Figure 6. Representative AFM images of poplar secondary cell walls showing the inner structure S2 layer of WT (a) and Zip 5 (b). Scale bar = 200 nm; color bar = 200 nm.

consistent with the results of the Simons’ stain. The individual cellulose microfibrils remained the same as the Zip-lignin modification should have minimal impact on cellulose synthesis itself, particularly because the cellulose is laid down in the cell wall before lignification is initiated. However, as we observed from lignin Raman images, the Zip-lignin poplar appeared to have slightly altered lignin distribution, especially in secondary cells walls. The spatial and temporal changes in lignification may be manifested in a rearrangement of the cellulose microfibrils, creating higher porosity environments and ultimately leading to increased pretreatment efficiency. It is not obvious how large an effect this provides, and it cannot be the sole reason for the observed improvement in the digestibility of the isolated cell walls that have been uniquely lignified using the conjugates, as the cellulose fibrils in these walls should have been deposited before lignification (as they would be in the wild-type plant). It is hard to imagine how the lignification, in either case, could seriously disrupt the cellulose network. Given that the Zip-lignin modification was mainly designed to target the alteration of the lignin structure, it is not yet clear if the conditions that cause increases in cell wall porosity are advantageous or detrimental to plant growth and development (even though no obvious agronomic differences have been noted). A targeted effort in the engineering of the plant cell wall accessibility through lignification manipulation may lead to further improvement in deconstructability and sugar yield.

■

CONCLUSIONS SRS imaging provides new insight into the identification of the presence and distribution of cell components. Here, we reported the unique Raman band associated with Zip-lignins which stems from its conjugated ester moieties into the polymer. We can qualitatively determine the enhanced levels of Zip-lignin in the transgenic poplar lines. We also examined the changes in cell wall porosity through imaging and staining and found a positive association between increased pretreatment efficiency and cell wall accessibility. Our results suggested that both lignin chemistry and plant cell wall accessibility, stemming from the Zip-lignin modification, were responsible for the improvement in pretreatment efficiency.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssuschemeng.9b01166. Correlation of dye adsorption and saccharification yield after [C2C1Im][OAc] pretreatment of WT and transgenic poplars, Zip-lignin molecular structure and coupling (PDF) 10621

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622

Research Article

ACS Sustainable Chemistry & Engineering

■

(14) Fu, D.; Holtom, G.; Freudiger, C.; Zhang, X.; Xie, X. S. Hyperspectral imaging with stimulated Raman scattering by chirped femtosecond lasers. J. Phys. Chem. B 2013, 117 (16), 4634−40. (15) Meng, X.; Foston, M.; Leisen, J.; DeMartini, J.; Wyman, C. E.; Ragauskas, A. J. Determination of porosity of lignocellulosic biomass before and after pretreatment by using Simons’ stain and NMR techniques. Bioresour. Technol. 2013, 144, 467−476. (16) Bhaskar, T.; Pandey, A.; Mohan, S. V.; Lee, D.-J.; Khanal, S. K. Synthesis of bioethanol from invasive weeds: Process design, optimization, and intensification with ultrasound. p - 445. (17) Petrik, D. L.; Karlen, S. D.; Cass, C. L.; Padmakshan, D.; Lu, F.; Liu, S.; Le Bris, P.; Antelme, S.; Santoro, N.; Wilkerson, C. G.; Sibout, R.; Lapierre, C.; Ralph, J.; Sedbrook, J. C. p-Coumaroyl-CoA:monolignol transferase (PMT) acts specifically in the lignin biosynthetic pathway in Brachypodium distachyon. Plant J. 2014, 77 (5), 713− 726. (18) Meyer, M. W.; Lupoi, J. S.; Smith, E. A. 1064 nm dispersive multichannel Raman spectroscopy for the analysis of plant lignin. Anal. Chim. Acta 2011, 706 (1), 164−70. (19) Ram, M. S.; Dowell, F. E.; Seitz, L. M. FT-Raman spectra of unsoaked and NaOH-soaked wheat kernels, bran, and ferulic acid. Cereal Chem. 2003, 80 (2), 188−192. (20) Sebastian, S.; Sundaraganesan, N.; Manoharan, S. Molecular structure, spectroscopic studies and first-order molecular hyperpolarizabilities of ferulic acid by density functional study. Spectrochim. Acta, Part A 2009, 74 (2), 312−323. (21) Agarwal, U. P.; Ralph, S. A. FT-Raman spectroscopy of wood: Identifying contributions of lignin and carbohydrate polymers in the spectrum of black spruce (Picea mariana). Appl. Spectrosc. 1997, 51, 1648−1655. (22) Agarwal, U. P. Raman imaging to investigate ultrastructure and composition of plant cell walls: distribution of lignin and cellulose in black spruce wood (Picea mariana). Planta 2006, 224, 1141−1153. (23) Horvath, L.; Peszlen, I.; Gierlinger, N.; Peralta, P.; Kelley, S.; Csoka, L. Distribution of wood polymers within the cell wall of transgenic aspen imaged by Raman microscopy. Holzforschung 2012, 66, 717−725. (24) Zhang, D.; Wang, P.; Slipchenko, M. N.; Ben-Amotz, D.; Weiner, A. M.; Cheng, J.-X. Quantitative vibrational imaging by hyperspectral stimulated Raman scattering microscopy and multivariate curve resolution analysis. Anal. Chem. 2013, 85 (1), 98−106. (25) Liu, B.; Wang, P.; Kim, J. I.; Zhang, D.; Xia, Y.; Chapple, C.; Cheng, J.-X. Vibrational fingerprint mapping reveals spatial distribution of functional groups of lignin in plant cell wall. Anal. Chem. 2015, 87, 9436−9442. (26) Felten, J.; Hall, H.; Jaumot, J.; Tauler, R.; de Juan, A.; Gorzsás, A. Vibrational spectroscopic image analysis of biological material using multivariate curve resolution-alternating least squares (MCRALS). Nat. Protoc. 2015, 10 (2), 217−40. (27) Jones, H. D. T.; Haaland, D. M.; Sinclair, M. B.; Melgaard, D. K.; Collins, A. M.; Timlin, J. a. Preprocessing strategies to improve MCR analyses of hyperspectral images. Chemom. Intell. Lab. Syst. 2012, 117, 149−158.

AUTHOR INFORMATION

Corresponding Author

*Phone: 1-517-432-4880. E-mail: [email protected]. ORCID

John Ralph: 0000-0002-6093-4521 Shi-You Ding: 0000-0002-1102-1507 Author Contributions

All authors were involved in data collection and analysis, and approved the final version of the paper. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Great Lakes Bioenergy Research Center, U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under Award Number DE-SC0018409 and DE-SC0019072.

■

REFERENCES

(1) Ralph, J. Hydroxycinnamates in lignification. Phytochem. Rev. 2010, 9 (1), 65−83. (2) Himmel, M. E.; Ding, S. Y.; Johnson, D. K.; Adney, W. S.; Nimlos, M. R.; Brady, J. W.; Foust, T. D. Biomass recalcitrance: Engineering plants and enzymes for biofuels production. Science 2007, 315 (5813), 804−807. (3) Zeng, Y.; Zhao, S.; Yang, S.; Ding, S.-Y. Lignin plays a negative role in the biochemical process for producing lignocellulosic biofuels. Curr. Opin. Biotechnol. 2014, 27, 38−45. (4) Wilkerson, C. G.; Mansfield, S. D.; Lu, F.; Withers, S.; Park, J.-Y. J.-Y.; Karlen, S. D.; Gonzales-Vigil, E.; Padmakshan, D.; Unda, F.; Rencoret, J.; Ralph, J. Monolignol ferulate transferase introduces chemically labile linkages into the lignin backbone. Science 2014, 344, 90−93. (5) Chen, K.; Wu, T.; Chen, T.; Zhou, T.; Wei, H.; Li, Y. Spectral focusing dual-comb coherent anti-Stokes Raman spectroscopic imaging. Opt. Lett. 2017, 42, 3634−3637. (6) Kim, K. H.; Dutta, T.; Ralph, J.; Mansfield, S. D.; Simmons, B. A.; Singh, S. Impact of lignin polymer backbone esters on ionic liquid pretreatment of poplar. Biotechnol. Biofuels 2017, 10 (1), 1−10. (7) Bhalla, A.; Bansal, N.; Pattathil, S.; Li, M.; Shen, W.; Particka, C. A.; Karlen, S. D.; Phongpreecha, T.; Semaan, R. R.; Gonzales-Vigil, E.; Ralph, J.; Mansfield, S. D.; Ding, S.-Y.; Hodge, D. B.; Hegg, E. L. Engineered lignin in poplar biomass facilitates Cu-catalyzed alkalineoxidative pretreatment. ACS Sustainable Chem. Eng. 2018, 6 (3), 2932−2941. (8) Ding, S.; Liu, Y.; Zeng, Y.; Himmel, M.; Baker, J.; Bayer, E. How does plant cell wall nanoscale architecture correlate with enzymatic digestibility? Science 2012, 338, 1055−1059. (9) Saar, B. G.; Zeng, Y.; Freudiger, C. W.; Liu, Y.-S.; Himmel, M. E.; Xie, X. S.; Ding, S.-Y. Label-free, real-time monitoring of biomass processing with stimulated Raman scattering microscopy. Angew. Chem. 2010, 49, 5608−5611. (10) Zeng, Y.; Saar, B. G.; Friedrich, M. G.; Chen, F.; Liu, Y.-S.; Dixon, R. a.; Himmel, M. E.; Xie, X. S.; Ding, S.-Y. Imaging lignindownregulated alfalfa using coherent anti-Stokes Raman scattering microscopy. BioEnergy Res. 2010, 3, 272−277. (11) Zeng, Y.; Himmel, M. E.; Ding, S. Y. Coherent Raman microscopy analysis of plant cell walls. Methods Mol. Biol. 2012, 908, 49−60. (12) Zeng, Y.; Himmel, M. E.; Ding, S. Y. Visualizing chemical functionality in plant cell walls. Biotechnol. Biofuels 2017, 10, 263. (13) Tobimatsu, Y.; Wagner, A.; Donaldson, L.; Mitra, P.; Niculaes, C.; Dima, O.; Kim, J. I.; Anderson, N.; Loque, D.; Boerjan, W.; Chapple, C.; Ralph, J. Visualization of plant cell wall lignification using fluorescence-tagged monolignols. Plant J. 2013, 76 (3), 357−66. 10622

DOI: 10.1021/acssuschemeng.9b01166 ACS Sustainable Chem. Eng. 2019, 7, 10616−10622