The in vitro synthesis of cellulose – A mini-review

The implementation of cellulose as a green alternative to classical polymers sparks research on the synthesis of defined derivatives of this biopolymer for various high-tech applications. Apart from the scientific challenge, the in vitro synthesis of cellulose using a bottom-up approach provides specimens with absolutely accurate substituent patterns and degrees of polymerization, not accessible from native cellulose.

Available online 7 February 2022

0144-8617/© 2022 The Authors Published by Elsevier Ltd This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

The in vitro synthesis of cellulose – A mini-review

Anna F Lehrhofera,1, Takaaki Gotoa,b,1, Toshinari Kawadac, Thomas Rosenaua,d,

Hubert Hetteggera,*

aUniversity of Natural Resources and Life Sciences, Vienna (BOKU), Department of Chemistry, Institute of Chemistry of Renewable Resources, Muthgasse 18, A-1190

Vienna, Austria

bWood K Plus – Competence Center for Wood Composites and Wood Chemistry, Altenberger Straße 69, A-4040 Linz, Austria

cKyoto Prefectural University, Graduate School of Life and Environmental Sciences, Nakaragi-cho 1-5, Shimogamo, Sakyo-ku, Kyoto 606-8522, Japan

dJohan Gadolin Process Chemistry Centre, Åbo Akademi University, Porthansgatan 3, FI-20500 Åbo/Turku, Finland

A R T I C L E I N F O

Keywords:

Anhydroglucose

Biopolymer

Cellulose

In vitro synthesis

Polysaccharide

Ring-opening polymerization

A B S T R A C T The implementation of cellulose as a green alternative to classical polymers sparks research on the synthesis of defined derivatives of this biopolymer for various high-tech applications Apart from the scientific challenge, the

in vitro synthesis of cellulose using a bottom-up approach provides specimens with absolutely accurate

substit-uent patterns and degrees of polymerization, not accessible from native cellulose Synthetic cellulose exhibiting a comparably high degree of polymerization (DP) was obtained starting from cellobiose by biocatalytic synthesis implementing cellulase Cationic ring-opening polymerization has been established in the last two decades, representing an excellent means of precise modification with regards to regio- and stereoselective substitution This method rendered isotopically enriched cellulose as well as enantiomers of native cellulose (“L-cellulose”, “D,

L-cellulose”) accessible In this review, techniques for in vitro cellulose synthesis are summarized and critically

compared – with a special focus on more recent developments This is complemented by a brief overview of alternative enzymatic approaches

1 Introduction and scope

Cellulose, as the essential building block of cell walls, is the most

abundant natural organic polymer Already in 1838, cellulose was

discovered, isolated, and named as such by Anselme Payen, who also

found the molecular formula of the biopolymer to be C6H10O5 (Klemm

et al., 2005) Its structure was first determined by Hermann Staudinger,

the “father” of polymer chemistry, in 1920, who identified cellulose as a

linear homopolysaccharide consisting of covalently linked β-(1 → 4)D-

glucose units (Staudinger, 1920) Since the glycosidic bond is formed by

condensation of two glucose units and is thus accompanied by the loss of

water, the repeating unit is often denoted as anhydroglucose unit (AGU)

For a long time, the dimer of glucose, cellobiose, was given as a

repeating unit and is still denoted as such in literature (Brown, 1996)

However, the elucidation of the three-dimensional structure of the

cellulose-synthesizing enzyme cellulose synthase (CesA) led to the

conclusion that natural synthesis starts from glucose (as the respective

UDP derivative), and thus D-anhydroglucopyranose/AGU is universally

accepted nowadays as the fundamental building block that makes up cellulose (French, 2017)

Cellulose – in contrast to starch and especially highly branched amylopectin – is arranged in a characteristic straight chain and can reach a length of up to several 1000 AGUs (French et al., 2018; Nunes,

2017) Because of the inherent chirality of the repeating unit D -anhy-droglucopyranose with its five chiral centers (C1-C5), the resulting polymer cellulose is also a chiral molecule Due to the high number of hydroxy groups present in cellulose – three hydroxy groups per AGU – an

exceptionally strong and complex inter- and intramolecular hydrogen

bond network is formed, which is the reason why the cellulose molecule

is insoluble both in water and many organic solvents According to its degree of order, these hydrogen bond interactions are responsible for the formation of higher structures consisting of crystalline (highly ordered)

as well as amorphous (less highly ordered) regions The H-bonds are also held accountable for the biopolymer's rigidity and stability, both phys-ically and chemphys-ically Cellulose is a polysaccharide exhibiting poly-morphism, with four main allomorphs being known The major one is

* Corresponding author

E-mail addresses: anna.lehrhofer@boku.ac.at (A.F Lehrhofer), takaaki.goto@boku.ac.at (T Goto), kawada@kpu.ac.jp (T Kawada), thomas.rosenau@boku.ac.at

(T Rosenau), hubert.hettegger@boku.ac.at (H Hettegger)

1 These authors contributed equally to this work

Contents lists available at ScienceDirect Carbohydrate Polymers journal homepage: www.elsevier.com/locate/carbpol

https://doi.org/10.1016/j.carbpol.2022.119222

Received 20 December 2021; Received in revised form 31 January 2022; Accepted 1 February 2022

cellulose I, which is the native form synthesized by the majority of living

organisms, whereas natural cellulose II is only synthesized by a very

limited number of species, most of which are bacteria (Brown, 1996;

Nunes, 2017) Native cellulose I found as a structural polymer in cell

walls is arranged in rod-shaped structures known as microfibrils These

microfibril rods make up the stiff, supporting part of the cell wall's

biocomposite, “glued” together by xyloglucans, pectins, proteins, and

lignin for providing additional strength and plasticity (Brown, 1996;

French et al., 2018)

Cellulose I consists of the allomorphs Iα/Iβ, which differ in their

crystal packing, molecular conformation as well as H-bonding, and the

ratio of the latter two forms can significantly vary depending on the

source of cellulose (Atalla & Vanderhart, 1984) In contrast to cellulose

I, which is packed in parallel strands, the chains in cellulose II are

ar-ranged in an antiparallel manner with extensive intersheet H-bonds

resulting from differences in the crystal structure (Brown, 1996; French

et al., 2018; Nunes, 2017) Cellulose II exhibits higher thermodynamic

stability than cellulose I and can be obtained by mercerization, swelling

in aqueous sodium hydroxide, or dissolution/regeneration The

modi-fication of cellulose I to attain cellulose II typically comes with a

sig-nificant reduction in its molecular weight (Brown, 1996) Cellulose III

can be accessed by reacting cellulose I or II with various amines (e.g.,

liquid ammonia) When cellulose is treated with glycerol at high

tem-peratures, cellulose IV is obtained (French et al., 2018; Nunes, 2017)

While the chemical structure of cellulose is rather simple (only one

monomer involved, strict linearity, strictly regioselective linkages), its

biosynthesis is by far more complex (Brown, 1996) Besides vascular

plants, which produce the majority of all naturally occurring cellulose,

also other organisms like algae, bacteria, tunicates, or oomycetes can

synthesize cellulose in a highly specific and orderly process (Bessueille &

Bulone, 2008; Brown, 1996; Saxena & Brown, 2005) Using a freeze-

fracture technique in 1976, Brown and Montezinos established that

the production of this biopolymer involves a linear, rod-like assembly

located at the growing end of a microfibril, which they called “terminal

complex” (TC) (Brown & Montezinos, 1976) The TCs of different

or-ganisms show various morphologies, with the most frequent one being a

hexagonal rosette-shaped structure found in higher plants and some

Chlorophyta (green algae) as well as various linear structures found in

other algae (Williamson et al., 2002) The dimensions of the synthesized

microfibrils mainly depend on the precise geometry of the TC, more

specifically the number of catalytic centers located there and the relative

arrangement of the latter to one another, which are responsible for

micro-sheet formation as the first stage of crystalline cellulose

con-struction (Brown, 1996; Brown & Montezinos, 1976)

Within the terminal complex, the plant enzyme cellulose synthase

(CesA) or bacterial cellulose synthase (BcsA), both examples of

glyco-syltransferases, catalyze the formation of cellulose by glycosidic bond

formation and chain extension at the non-reducing C4 terminal unit

(Koyama et al., 1997; Olek et al., 2014) Typically, the cellulose

syn-thases are arranged in hexameric clusters called rosettes, which are

located in the plasma membrane of the cell, with each rosette subunit

containing six CesA or BcsA and thus synthesizing a total of 36 cellulose

strands per TC, which then form an elementary fibril by spontaneous

lateral aggregation and bundling (Cosgrove, 2005; Heath, 1974; Olek

et al., 2014) More recent investigations on the microfibrils' exact

structure implementing X-ray diffraction techniques rather suggest an

arrangement containing three CesA enzymes per subunit and thus only

18 cellulose strands forming an elementary fibril, however, this is still a

matter of ongoing discussion (Jarvis, 2013) The formation of β-linkages

in the glucan chains starting from α-UDP-glucopyranose as the substrate

is catalyzed within the active site of CesA, corresponding to the so-called

inverting mechanism This inversion of the stereogenic center at the C1

position is coupled to the extension of the growing glucan chains from

their non-reducing ends (Bessueille & Bulone, 2008) The catalysis

oc-curs at the cytoplasmic side of the membrane, where the formed

cellu-lose chains are extruded through the plasma to reach the cell wall by a

so-far unidentified mechanism to fulfill their function as a structural biopolymer (Bessueille & Bulone, 2008) The biosynthesis of cellulose has been extensively described and reviewed (Bessueille & Bulone,

2008; Brown, 1996; Delmer, 1999; Saxena & Brown, 2005)

The isolation of cellulose from wood or other plant material is essential for the use of this biopolymer and is conducted by pulping the raw material This technical process has been established especially in the paper industry and has already been described in the literature in detail (Biermann, 1996; Mboowa, 2021; Ragnar et al., 2014)

Here the question arises why a chemist would synthesize cellulose when nature provides enormous amounts of this material every year Methods for the utilization and purification of natural cellulose are available on a large scale After all, the classic paper and pulp industry – one of the most important monetary pillars in many national economies – is based on the isolation and utilization of cellulose as a basic material for paper, textile fibers, and cellulose derivatives So why prepare cel-lulose in a complex multi-step synthesis, in small quantities probably below the gram scale?

The first, general answer is scientific curiosity Will it be possible to copy the process that nature has invented and optimized so perfectly for the production of the – in terms of mass – most prominent and important natural substance? The second answer is based on the fact that biosyn-thesis naturally always generates cellulose This is done with great

perfection, but on the other hand, the possibilities for alterations, e.g., in

isotope content or substituents/functionalities, are almost non-existent

If such modifications are desired, one must always start from the natural polymer Reactions on cellulose are always “polymer analogous”, in the best cases they are almost regioselective, but just not completely This would be different for chemically synthesized cellulose and cellulose derivatives By using selectively modified monomers, which, for example, have different isotope or substitution patterns, celluloses or cellulose derivatives could be built up with 100% selectivity, which would otherwise not be accessible in this perfection starting from nat-ural cellulose A third question is hoped to be answered with the syn-thesis of cellulose: which allomorph is obtained in such “non-natural” syntheses, cellulose I or II, or continuous amorphous material? Despite the structure of cellulose being determined already in 1920 and numerous attempts of chemists to either enzymatically or chemi-cally synthesize cellulose, the published examples of actually successful

chemical in vitro synthesis of this biopolymer are rather limited This can

be attributed to the fact that bottom-up synthesis approaches are all quite challenging, mostly starting from glucose, the logical and obvious monomer choice In 2001, Kobayashi and co-workers stated that a variable and convenient method for the synthesis of different poly-saccharides had not yet been established (Kobayashi et al., 2001) Cellooligomers with a defined DP can be obtained employing a top-

down approach For example, Isogai et al reported the hydrolysis of

alkali-treated native and regenerated cellulose with dilute acid The obtained products exhibited a DP of 35–101 for the major, high- molecular-mass fraction and around 20 for the minor low-molecular- mass fraction (Isogai et al., 2008) In 2016, Zweckmair et al reported

a method for the preparation and isolation of well-defined cellooligo-saccharides from cellulose The authors acetolyzed a native cellulose sample and isolated monodisperse peracetylated cellooligomers with a

DP of up to 20 using preparative normal-phase HPLC (Zweckmair et al.,

2016) However, these top-down approaches do not present an in vitro

cellulose synthesis and are not discussed further here

In this review, a brief overview of existing methods for the in vitro

synthesis of cellulose using chemical as well as enzymatic methods is provided Kobayashi and co-workers already discussed some of the chemical and especially enzymatic approaches for obtaining synthetic cellulose and related polysaccharides in 2001, including some mecha-nistic aspects (Kobayashi, 2005; Kobayashi et al., 2001) Further reviews also discussing methods for the synthesis of polysaccharides (Kobayashi

& Makino, 2009; Yoshida, 2001) as well as generally cellulose as a biopolymer (Klemm et al., 2005) have been published in the last

decades The present article focuses on more recent developments

within the last 20 years, with a special emphasis on the accessibility of

cellulose via chemical synthesis

2 Chemical synthesis of cellulose

To achieve the chemical synthesis of cellulose, three conditions need

to be fulfilled: a) regioselectivity of the linkage (only 1 → 4 linkages and

no disubstitution, which would cause branching), b) stereoselectivity

(strictly β-configured anomeric carbon), and c) DP control (to allow

precise regulation of the chain length or at least a low dispersity) While

regiocontrol can be achieved by selective protection of the

corre-sponding monosaccharide glucose prior to the glycosylation reaction,

the choice of the reaction type itself is crucial for accomplishing precise

stereocontrol The formation of the O-glycosidic linkage is usually

conducted by nucleophilic attack of the free hydroxy group of the

glycosyl donor at the anomeric carbon of the glycosyl acceptor carrying

a leaving group Complete stereocontrol of the conformation at the

anomeric carbon (C1) can thus be achieved employing an SN2 reaction

at an α-configured acceptor, in which the anomeric carbon undergoes

complete inversion On the contrary, an SN1 pathway via an

oxocarbe-nium cation intermediate (Hosoya et al., 2014; Xiao & Grinstaff, 2017)

permits nucleophilic attack from both sides of the ring and consequently

gives a racemic mixture of both the α- and β-anomers Even though the

latter one is preferred because of the often-discussed “anomeric effect”

(in short, the adjacent oxygen atom in the pyranose ring exerts a

sta-bilizing effect on the β-anomer due to orbital interactions) (Graczyk &

Mikolajczyk, 1994; Juaristi & Cuevas, 1992), the SN1 pathway is

disadvantageous with regard to anomeric stereocontrol To ensure

ste-reoregularity, precise selection of the type of leaving group, catalyst,

protecting groups, and the reaction conditions is required to enforce

selective progression of the glycosylation via an SN2 pathway (Fügedi,

2006; Xiao & Grinstaff, 2017; Zhu & Schmidt, 2009)

There are mainly two approaches to chemically synthesize cellulose:

1) the stepwise addition method and 2) the polymerization and

poly-condensation method

2.1 Stepwise addition method

The stepwise method is classically employed in oligosaccharide

synthesis, where it is used to obtain a substance with defined

composi-tion, linkage pattern, and DP, or at least a very narrow dispersity In this

method, the glycosylation reaction is repeated until the desired DP is

achieved using a derivative of the repeating unit The glycosylation

re-action is formally a dehydrative condensation rere-action between the

hydroxy group of a glycosyl donor carrying the anomeric carbon of the

new glycosidic bond and the hydroxy group of a glycosyl acceptor An

important example, directly illustrating the principle, is the acid-

catalyzed Fischer glycosylation, which is depicted in Fig 1

To control the configuration at the anomeric carbon, several

glyco-sylation methods utilizing glycosyl donors with specific leaving groups

and reactivities and special catalysts have been developed, inter alia the

K¨onigs-Knorr method (Koenigs & Knorr, 1901), the Schmidt imidate

method (Schmidt & Michel, 1980), the thioglycosylation method

(Fügedi et al., 1987) and the pentenyl method (Mootoo et al., 1988) The

imidate method, mostly relying on trichloroacetimidate, is the most

widely used one because the imidoyl group can be easily introduced, the imidate is generally stable enough to outlast purification steps and at the same time reactive enough to allow high yields in the glycosylation reactions (Schmidt & Jung, 2000) The chemical structures of frequently used glycosyl donors are shown in Fig 2

It is worth mentioning that the first regiocontrolled glycosylation

was reported by Shapiro et al in 1969 for the synthesis of lactose from 2,3-di-O-acetyl-1,6-anhydro-β-D-glucopyranose with tetra-O-acetyl-α-D- galactopyranosyl bromide (acetobromogalactose) (see Fig 3) (Shapiro

et al., 1969)

Tackles towards the chemical synthesis of cellulose from a bottom-up approach employing the stepwise addition method had begun with the synthesis of cellooligosaccharides To achieve cellulose synthesis, the glycosylation reaction must regioselectively occur between C1 of a glycosyl donor and C4-OH of a glycosyl acceptor Such regiocontrol is possible by selective protection of all other hydroxy groups except the glycosyl acceptor's C4-OH group To obtain cellooligomers or cellulose and not just the dimer cellobiose, repetitive glycosylation is required to elongate the chain After each glycosylation step, the obtained in-termediates need to be isolated, purified and converted into the corre-sponding glycosyl donor or acceptor once again The first arrow using this stepwise addition method was fired by Freudenberg and Nagai to hit cellobiose synthesis (DP = 2) already in 1933 Starting from levoglu-cosan (1,6-anhydro-β-D-glucopyranose) and acetobromoglucose (1- bromo-α-D-glucose tetraacetate), they reported the synthesis of octaa-cetyl cellobiose, however, the yield was still very low (Freudenberg & Nagai, 1933) Another example of this stepwise addition method is the synthesis of α-cellotriose (DP = 3) by Hall and Lawler in 1971 by

condensation of selectively protected 2,3,6-tri-O-substituted methyl β-D-

glucopyranosides with tetra-O-acetyl-α-D-glucopyranosyl bromide in the

first and hepta-O-acetyl-α-D-cellobiosyl bromide in the second step, respectively After substitution of the methyl- by acetyl-groups, α-D- cellotriose hendecaacetate was obtained, which could be efficiently transformed into the β-anomer (Hall & Lawler, 1971) Ten years later, Takeo and co-workers reported a similar synthesis of these α- and

β-cellotriose hendecaacetates and an additional pathway to obtain 6,6′,6′′-tri-substituted derivatives of methyl β-cellotrioside (Takeo et al.,

1981) Cellotetraose (DP = 4) was first reported by Schmidt and Michel

in 1982 (Schmidt & Michel, 1982) As a starting material, they used the

α-trichloroacetimidate derivative of

2,3,4,6-tetra-O-acetyl-glucopyr-anose, which they coupled stepwise with high β-selectivity to obtain both linear and branched cellooligomers In this case, three steps were necessary to convert the obtained intermediates – the suitably protected mono-, di- and trisaccharide – after successful glycosylation to the next glycosyl donor (imidate) again before being able to couple it once more Takeo and co-workers also reported the synthesis of cellotetraose by stepwise elongation in 1983 The authors described the selective

Fig 1 Acid-catalyzed Fischer glycosylation of glucose yielding a glucoside (R

=alkyl or aryl, typically methyl) Fig 2 Chemical structures of glycosyl donors comprising a) halide, b) tri-chloroacetimidate, c) thioalkyl, and d) O-pentenyl leaving groups

condensation of hepta-O-acetyl-α-D-cellobiosyl bromide with benzyl

2,3,6,2′,3′,6′-hexa-O-benzyl-β-D-cellobioside and subsequent

depro-tection to yield the desired product (Takeo et al., 1983)

An additional challenge the stepwise synthesis faced was the high

number of steps involved, such as those that entail byproduct formation

or require purification of the products, which significantly decrease the

overall yield of the synthesis To overcome this problem, Nakatsubo

et al proposed a convergent synthetic method (Nakatsubo et al., 1985)

They used a precursor with active X and Y groups at the C1-OH and C4-

OH, respectively, with the other hydroxy groups being protected (see

Fig 4) The X and Y groups can be cleaved selectively, and thus the

glycosyl donor can be prepared in only two steps (removal of X group

and introduction of the leaving group Z, see Fig 4, left), and the glycosyl

acceptor in only one step (removal of Y group, see Fig 4, right) The

product has the same combination of terminal protecting groups as the

respective precursors (X and Y) and can thus be directly used as a

re-agent for subsequent repetitive coupling using the same chemistry and

conditions

Using this strategy, cellopentaose to cellooctaose derivatives were

synthesized by Kawada and co-workers in the form of perbenzylated

allyl 4n’-O-p-methoxybenzyl cellooligosaccharides starting from

2,3,6,2′,3′,6′-hexa-O-benzyl-4′-O-(p-methoxybenzyl)-β-D-cellobioside by

repetitive alternating removal of the p-methoxybenzyl group on the one

hand and stereoselective β-glycosylation using the imidate method on

the other hand (Kawada et al., 1990) Further, the authors reported the

first synthesis of cellooctaose acetate starting from allyl 2,3,6-tri-O- benzyl-4-O-(p-methoxybenzyl)-β-D-glucopyranoside (Kawada et al.,

1994)

In this synthesis approach, the “1 + n” convergent method or linear synthetic method, they employed benzyl protective groups at O-2, O-3, and O-6, which significantly decreased the reactivity of both the glycon

(i.e., glycosyl donor) and the aglycon (i.e., glycosyl acceptor) with

increasing DP Thus, an “n + n” convergent method could not be realized

employing per-benzyl protection and the synthesis of

cellooligo-saccharides bearing a DP higher than eight was found to be challenging Therefore, Takano and co-workers thoroughly studied the substituent effects on the glycosylation reactions, not only regarding their reactivity but also focusing on the stereoselectivity They investigated various protective groups and found that an electron-withdrawing pivaloyl group at the O-2 and O-6 position of the glycon and an electron-donating benzyl group at the O-3 of the aglycon were the most convenient choice This finely tuned protective system made an “n + n” convergent syn-thesis of cellooligomers possible (Takano et al., 1990, 1994)

Combining the new findings by implementing this convergent method and the ideal choice of protective groups as suggested by Takano

and co-workers (vide supra), Nishimura and Nakatsubo synthesized a

series of regioselectively substituted oligomers up to celloeicosaose (DP

=20) starting from cellooctaose by stepwise chain extension using a monomer derivative (Nishimura & Nakatsubo, 1997) This protective system proved the possibility of synthesizing cellooligosaccharides with

Fig 3 Regio- and stereo-controlled synthesis of a lactose precursor as described by Shapiro and co-workers (Shapiro et al., 1969)

Fig 4 Schematic of the glycosylation procedure in the convergent synthetic method reported by Nakatsubo and co-workers (Nakatsubo et al., 1985)

a higher DP, and thus opened up opportunities for later ring-opening

polymerization methods using a similar arrangement

This stepwise method, of course, still required multiple stages to

prepare the precursors, and individual glycosylation steps are required

In addition to these time-consuming, laborious chemical steps, also

purification is usually required after each step Moreover, solubility and

reactivity usually get lower and lower as the DP of this “homologous

series” of the cellooligomers increases This is especially true in the case

of a linear oligo- or polysaccharide with extensive inter- and

intra-molecular hydrogen bonds, such as cellulose, rendering the practical

aspects of the synthesis even more challenging However, despite its

challenges, the stepwise method remains the method of choice for the

synthesis of cellooligosaccharides or cellulose with a single, strictly set,

and well-defined DP

2.2 Polycondensation method

In contrast to the stepwise addition method, no intermediate

pro-tection or depropro-tection steps are required in the polycondensation

method for the synthesis of cellulose Only one single type of monomer

precursor containing reactive groups at corresponding carbon atoms is

used To enable synthesis of cellulose implementing polycondensation, a

monomeric glucose derivative exhibiting an anomeric leaving group at

C1 and a free, unprotected hydroxy group at C4, which can

simulta-neously act as a glycosyl donor as well as an acceptor, is needed (Xiao &

Grinstaff, 2017)

Already in 1941, Schlubach and Lührs reported first attempts to

polycondensate D-glucose by treatment with dry gaseous HCl The

au-thors stated that these polyglucosans exhibited a spherical or highly

branched structure and a high molecular weight of 12 000 g/mol, which

had not been previously described in the literature for a synthetic

molecule at that time However, the authors did not state the

confor-mation of the anomeric carbon in the polymer or whether their product

exhibited stereoregularity (Schlubach & Lührs, 1941) The first reported

attempt to synthesize a stereoregular, linear polysaccharide based on

glucose using a polycondensation approach in a controlled manner was

conducted by Haq and Whelan in 1956, which was inspired by the

K¨onigs-Knorr method for stepwise addition (vide supra) (Haq & Whelan,

1956) Implementing a regioselective polymerization, the authors

star-ted from selectively protecstar-ted 2,3,4-tri-O-acetyl-α-D-glucopyranosyl

bromide, which polycondensed upon treatment with silver oxide in

chloroform However, their attempt caused also intramolecular side

reactions with the formation of O-acetylated levoglucosan (which is also

known as a chemical biomass tracer of carbohydrate pyrolysis) and

gentiodextrins, i.e., (1 → 6)-linked β-glucopyranose derivatives (see

Fig 5) with a low DP of up to only nine repeating units

A few years later, Husemann and Müller as well as Hirano attempted

polycondensation of a glucose derivative inspired by a non-

regioselective polymerization approach of different monosaccharides

previously proposed by Micheel and co-workers (Micheel et al., 1961)

Husemann and Müller reported polycondensation of selectively

pro-tected 2,3,6-tri-O-(N-phenylcarbamoyl)-D-glucopyranose (“2,3,6-

glucose tricarbanilate”) in a mixture of chloroform and dimethyl sulf-oxide (DMSO) in the presence of phosphorus pentsulf-oxide for the removal

of water that is generated during polycondensation After removal of the protecting groups, a predominantly β-(1 → 4)-linked, slightly branched glucan with a DP of 48–63 was obtained (Husemann & Müller, 1966) Shortly after, Hirano reported a similar synthesis procedure using the same glucose derivative, 2,3,6-glucose tricarbanilate, DMSO, and phosphorus pentoxide without any additional solvents to give a similar polysaccharide The predominant formation of the β-linkage was attributed to the steric hindrance of the protective groups and thus the disfavoring of α-linkages (Hirano, 1973) Since then, no conceptually different attempts to synthesize cellulose by a polycondensation approach have been conducted and, to the best of our knowledge to date, no successful stereo- and regioselective synthesis route towards cellulose with a sufficiently high DP has been reported in the literature

2.3 Ring-opening polymerization method

This polymerization method requires a monomer containing a strained ring system readily available for ring-opening polymerization When synthesizing polysaccharides in a bottom-up approach, ring- opening polymerization of anhydrosugars has become the method of choice An anhydroglucose derivative that either forms a strained ring system itself or carries a cyclic protecting group can be used as a pre-cursor These bi- or tricyclic ring systems are prone to undergo poly-merization upon release of the ring strain as the respective driving force (Xiao & Grinstaff, 2017)

The precursors are typically prepared by vacuum pyrolysis of monosaccharides and subsequent protection of the free hydroxy groups

After that, polymerization is started by adding a Lewis acid catalyst at a

low temperature under the exclusion of water (Xiao & Grinstaff, 2017) This general approach was established by Ruckel and Schuerch in 1966 The reactivity of the monomers is based on the release of the strain in the monomers They reported successful preparation of a stereoregular

linear polysaccharide starting from 1,6-anhydro-2,3,4-tri-O-substituted

β-D-glucopyranose monomers, yielding polymeric compounds with a DP

of up to 300 (Ruckel and Schuerch 1966a and 1966b) Although this approach – employing levoglucosan (1,6-anhydrosugar) derivatives – is the most established and thoroughly studied one, obviously 1,4-anhy-droglucose derivatives are needed as precursors for cellulose synthesis Their synthesis is made more challenging, on the one hand, by much higher ring strain and, on the other hand, by the fact that they can be regarded as both 1,4-anhydropyranoses as well as 1,5-anhydrofurano-ses Thus, the precursors can polymerize according to two different ring-opening pathways to give either (1 → 4)-pyranosidic or (1 → 5)- furanosidic repeating units, both possibly with α- or β-linkages (see Fig 6)

Synthesis of cellulose-type polymeric compounds requires selective

1,4-cleavage of a 1,4-anhydropyranose, in most cases via a

tri-alkyloxonium ion intermediate As the 1,5-anhydro ring oxygen pos-sesses higher basicity than the 1,4-anhydro ring oxygen (Xiao & Grinstaff, 2017), the formation of C1-O+-C5 oxonium ions as in-termediates is preferred Consequently, (1 → 5)-linked glucofuranosides are predominantly formed during polymerization Thus, the precise se-lection of catalyst, choice of protecting groups, and careful control of the reaction conditions are required to direct the preferred pathway and accomplish selective 1,4-scission for the successful synthesis of cellulose (Uryu, Yamanouchi, Hayashi, et al., 1983; Uryu, Yamanouchi, Kato,

et al., 1983; Xiao & Grinstaff, 2017) The catalyst must selectively complex the 1,4-anhydro ring oxygen, and at the same time, the 1,4- anhydro ring oxygen of the co-reacting monomer must nucleophili-cally attack C1 from the reverse side of the C1-O4 bond Due to the involvement of the oxonium species, the reaction type is specified as

“cationic”, and CROP is often used as an abbreviation for the “cationic ring-opening polymerization” sequence

The cationic ring-opening polymerization of 1,4-anhydro-2,3,6-tri-

Fig 5 Structures of levoglucosan (1,6-anhydro-β-D-glucopyranose) and

gen-tiobiose (1-β-D-glucopyranosyl-6-D-glucopyranose)

O-benzyl-α-D-glucopyranose was reported by Micheel and Brodde as

well as Uryu et al., but mainly (1 → 5)-glucofuranose repeating units

were obtained due to the preferred glucopyranose-scission pathway

However, they also attributed this outcome to the steric hindrance of the

bulky benzyloxy-groups, suspecting inhibition of complexation by the

catalyst (Micheel & Brodde, 1974; Uryu et al., 1985)

In 1994, Kochetkov reported the synthesis of a completely

stereo-regular β-(1 → 4)-D-glucopyranoside under high pressure by

poly-condensation of 4,6-di-O-acetyl-3-O-trityl-1‚2-O-cyanoethylidene-α-D-

glucose, however, the author did not reveal detailed information about

the conditions or the product (Kochetkov, 1994) Kamitakahara and co-

workers described a synthetic approach using 1,4-anhydro derivatives,

such as 1,4-anhydro-3,6-di-O-benzyl-2-O-pivaloyl-α-D-glucopyranose,

which led to a regular 3,6-di-O-benzyl-2-O-pivaloyl-β-(1 → 5)-D

-gluco-furanan and the respective non-natural polysaccharide after

depro-tection (Fig 7, top) (Kamitakahara et al., 1994) The authors stated that

only the 1,4-anhydro-3,6-di-O-benzyl-2-O-pivaloyl-α-D-glucopyranose,

but not the 1,4-anhydro-2,3-di-O-benzyl-6-O-pivaloyl-α-D

-glucopyr-anose, led to a stereoregular polymer, which they attributed to a

participation of the pivaloyl as the neighboring group at the O-2 during

the polymerization Zachoval and Schuerch already previously

dis-cussed this phenomenon of neighboring group participation of ester

groups during polymerization of sugars in detail (Zachoval & Schuerch,

1969) Additionally, coordination of the catalyst PF5 is facilitated, as the

benzyl group at the O-6 position increases the electron density of the

1,5-anhydro ring oxygen (Kamitakahara et al., 1994) Shortly after that,

in 1996, Nakatsubo et al achieved the first successful synthesis of

cellulose by CROP using a derivative of the precursor implemented by Kamitakahara Instead of incorporating simple 1,4-anhydro derivatives, they used a different intermediate in which O-1 and O-4 were separated

by a C1-spacer This way, sufficient strain – and thus reactivity – was maintained and the competitive glucopyranose ring scission was

pre-vented The OH groups 1, 2, and 4 were integrated into an orthopivalate

structure, a system with high ring strain, but different geometry from the above 1,4-anhydroglucopyranoses Starting from D-glucose and

involving several protection and deprotection steps, 3,6-di-O-benzyl-

α-D-glucose 1,2,4-orthopivalate was synthesized as the reactive

precur-sor, which underwent highly selective ring-opening catalyzed by trity-lium tetrafluoroborate (Ph3CBF4) to give a highly stereoregular, β-(1 → 4)-linked cellulose derivative with a DP of about 20 after cleavage of the protecting groups (Fig 7, bottom) The pivaloyl group remained at O-2, while O-6 and O-3 were protected by benzyl groups from the beginning

A careful selection of the protecting groups was crucial to direct the

mechanism towards selective ring-opening (the desired 1,4-cleavage vs

the 1,2- and 2,4-alternatives) and thus the formation of selectively linked β-(1 → 4)-glucopyranose repeating units (Nakatsubo et al., 1996)

Especially the 3-O-benzyl group was found to be essential for

cellu-lose formation when studying the substituent effect of the protecting group in the O-3 position According to Kamitakahara and co-workers –

due to the rather sterically demanding axial 3-O-benzyl group – the polymerization of this orthoester might proceed via the so-called

diox-alenium ion mechanism rather than the trialkyloxonium ion mechanism proposed for other stereoregular ring-opening polymerizations of anhydrosugars (Kamitakahara et al., 1996) Hori and co-workers further

studied the influence of the orthoester group, which was implemented during this synthesis of cellulose Studying orthopropionate-, orthoace-tate- as well as orthobenzoate-derivatives and comparing their poly-merization behavior to the respective orthopivalate derivative, they

found that the latter was crucial for regioregularity, as the other de-rivatives gave a mixture of β-(1 → 4)-linked and β-(1 → 2)-linked glucopyranose-derivatives (Hori et al., 1997; Yoshida, 2001)

The authors further adapted their polymerization method and suc-cessfully synthesized 13C-labeled cellulose with 99% isotopic

enrich-ment, which was otherwise not accessible employing in vivo synthesis

(Adelw¨ohrer et al., 2009) and was used to study dynamic changes of the hydrogen bond patterns during swelling and dissolution (Rosenau et al.,

2019) Recently, also the enantiomeric mirror image of native cellulose,

i.e., “L-cellulose”, was obtained by this approach (see Fig 8 for the chemical structure) The successful synthesis of β-(1 → 4)-L -glucopyr-anan starting from L-glucose exhibiting an average DP of 32.8 and an

Mw/Mn ratio of 1.97 was reported by Yagura et al in 2020, and its

enantiomeric identity was confirmed by optical rotation of the respec-tive peracetate derivarespec-tive The synthesized L-cellulose triacetate had a positive specific optical rotation of +8.3◦, whereas authentic D-cellulose triacetate had a negative specific optical rotation of − 23.4◦ (Yagura

et al., 2020)

Fig 6 Ring-opening pathways of 1,4-anhydrosugar derivatives (redrawn from (Xiao & Grinstaff, 2017))

Fig 7 First successful chemical synthesis of cellulose by CROP of orthopivalate

precursors as reported by Nakatsubo and co-workers Reprinted with

permis-sion from Nakatsubo et al., 1996 Copyright © 1996 American

Chemi-cal Society

Shortly after, the authors also reported the synthesis of optically

inactive cellulose using the same method, starting from a “racemic

mixture” of D- and L-glucose The equimolar mixture of the

corre-sponding orthopivalate precursors yielded “racemic” 3-O-benzyl-2,6-di-

O-pivaloyl-β-(1 → 4)-D,L-glucopyranan, which was deprotected and

subsequently derivatized to give an acetylated derivative with an

average DP of 18.6 and a specific optical rotation of +0.01◦, suggesting

that the underivatized biopolymer consisted of a nearly racemic mixture

of D- and L-anhydroglucose units, namely “D,L-cellulose” (Ikegami et al.,

2021)

Further, CROP was used to synthesize highly regioselectively

alky-lated cellulose derivatives Kamitakahara and co-workers prepared 2,6-

di-O-methylcellulose by CROP of the previously reported precursor 3,6-

di-O-benzyl-α-D-glucose 1,2,4-orthopivalate, removal of the pivaloyl

protecting groups, methylation using MeI in the O-2 and O-6 position,

and subsequent three-step deprotection of O-3 (Kamitakahara et al.,

2008) Moreover, Kamitakahara et al studied the influence of

regiose-lectively substituted synthetic cellulose bearing methyl and ethyl groups

in the O-2 and O-3 positions on the biopolymer's solubility The authors

synthesized the cellulose derivatives by polymerization of

1,2,4-ortho-pivalate-type precursors, partly already carrying ethyl or methyl

sub-stituents in the O-6 position of the monomer and thus demonstrating

derivatization before CROP (Kamitakahara et al., 2009) In 2010,

Kamitakahara and co-workers reported the successful synthesis of

regioselectively substituted 2-O-, 3-O- and 6-O-ethyl celluloses Starting

from a 1,2,4-orthopivalate cellulose precursor carrying different

sub-stituents in O-2, O-3, and O-6 positions, they prepared the polymeric

compound by CROP and subsequent (multiple) deprotection steps to

study the structure/property relationship of the cellulose derivatives

(Kamitakahara et al., 2010)

3 Enzymatic synthesis of cellulose

As nature is synthesizing cellulose through enzymes far more

spe-cifically and efficiently than any chemist in the laboratory, it is logical

that several enzymatic in vitro approaches towards cellulose have been

proposed Regioselective and stereoregular formation of the β-(1 → 4)-

glycosidic linkage is especially challenging by chemical synthesis, while

native enzymes can do this with ease Enzymatic catalysis for the in vitro

synthesis of cellulose and similar polysaccharides, including

mecha-nistic aspects, has already been reviewed in detail by Kobayashi

(Kobayashi et al., 2001; Kobayashi & Makino, 2009), Kadokawa

(Kadokawa, 2011), and Shoda et al (Shoda et al., 2016) Thus, only a

brief overview is provided in this review for the sake of completeness

Using a UDP-forming cellulose synthase enzyme isolated from

Ace-tobacter xylinum, Aloni et al and Lin et al synthesized a β-(1 → 4)-D-

glucan (Aloni et al., 1982; Lin et al., 1985) Aloni et al used a PEG-

supported enzyme to reach a rate of almost 40% of in vivo cellulose

production However, they only concluded that the reaction product was

cellulose by using derivatization or digestion with cellulose hydrolyzing

enzymes, but did not provide detailed analytical data or information

about the crystalline and microfibrillar structure (Aloni et al., 1982)

Shortly after, Lin et al reported an in vitro synthesis of cellulose

micro-fibrils which, however, accumulated in clusters, according to electron

microscopy, and were much shorter compared to the typical fibrils

produced by Acetobacter xylinum cellulose synthase in vivo, which might

be attributed to the absence of the TCs in the solubilized, isolated enzyme (Lin et al., 1985)

A major drawback of cellulose synthesis by the biosynthetic pathway using glycosyltransferases is the comparatively high cost of the glycosyl donor in the case of UDP-forming cellulose synthase Further, these enzymes are prone to product-inhibition, which significantly diminishes the yield (Kobayashi et al., 2001) Thus, also non-biosynthetic pathways are frequently employed In 1991, the first confirmed enzymatic syn-thesis of cellulose was accomplished by Kobayashi and co-workers The authors reported an entirely non-biosynthetic pathway for cellulose synthesis implementing an enzyme catalyst, a glycosylase or glycosyl hydrolase of which the biocatalytic activity was reversed Glycosylases are a convenient choice for biocatalysis since they usually have high glycosylation activity and are comparatively tolerant against organic solvents They are, furthermore, readily available and relatively well- studied (Kobayashi et al., 2001) Cellulase, an example of glyco-sylases, is originally a degrading (hydrolyzing) enzyme As using cellu-lase for cellulose synthesis is the reverse reaction, various ideas, mainly

an equilibrium control and kinetic control, have been developed Shoda and co-workers developed the glycosyl fluoride method as a new tech-nology for glycosylation reactions (Shoda et al., 2016) Kobayashi and co-workers implemented this technique and used β-D-cellobiosyl fluo-ride as a starting material, which was polymerized with cellulase as the enzymatic catalyst in acetonitrile/acetate buffer (pH 5) to obtain the

first-ever example of in vitro synthesized cellulose To promote the

ste-reoregular formation of the β-(1 → 4)-glycosidic linkage, the β-config-uration of the starting substrate was crucial The authors confirmed the structure of the isolated polymeric compound by comparing the 13C solid-state NMR and IR spectra to the respective spectra of native cel-lulose and by conducting a hydrolysis experiment Further, they acety-lated the synthetic compound and measured a DP of at least 22 using gel permeation chromatography, which is still relatively low compared to native cellulose (Kobayashi et al., 1991) Using an extensively purified

enzyme, Lee and co-workers achieved the first and highly remarkable in

vitro synthesis of the native cellulose I allomorph (Lee et al., 1994) This technique was further extended to the synthesis of cellulose derivatives

exhibiting, inter alia, alternating methyl groups at C6-OH (Okamoto

et al., 1997) or glucose and N-acetylglucosamine units (Kobayashi et al.,

2006) using fluorinated disaccharide monomers and cellulase catalysis

A major drawback of this method, however, is the rather complex, multistep synthesis of the glycosyl fluorides, making the use of different monomers desirable (Shoda et al., 2016)

Samain et al successfully synthesized cellooligosaccharides

employing cellodextrin phosphorylase (CDP) for catalysis, together with cellobiosyl derivatives and α-D-glucopyranosyl 1-phosphate as glycosyl donors However, the rather costly glycosyl donor was also a major drawback in this case and the achieved DP of around eight was comparatively low (Samain et al., 1995) Starting from glucose and α-D- glucopyranosyl 1-phosphate, Hiraishi and co-workers synthesized highly ordered cellulose II cellooligosaccharides also using CDP as a catalyst (Hiraishi et al., 2009) Serizawa et al were the first ones to

implement C1-modified β-D-glucose derivatives as end-group substrates

Fig 8 Chemical structure of native “D-cellulose compared to its artificial enantiomer “L-cellulose” as synthesized by Yagura et al (Yagura et al., 2020)

(“primers”) in CDP-catalyzed polycondensation of α-D-glucopyranosyl 1-

phosphate to allow precise end-group modification of the cellulose chain

at its reducing end Due to the poor substrate specificity of CDP, the

modified glycosyl acceptors were readily accepted by the enzyme For

example, it was used as a catalyst in a related synthesis of alkyl

β-cel-lulosides, i.e., O-alkyl reducing-end-modified cellooligomers By varying

the length of this single alkyl group, the crystallization behavior of the

oligomers could be influenced While the use of n-butyl or shorter-chain

alkyls promoted the formation of antiparallel cellulose II, n-hexyl or

longer chains led to the parallel cellulose I allomorph, which also

affected the self-assembly of the strands into different tertiary structures

(Serizawa et al., 2021; Yataka et al., 2016) The authors also synthesized

azide-containing cellulose oligomers by the same path Implementing a

β-glucosyl azide primer as the glycosyl acceptor, Yataka et al

synthe-sized a cellulose II crystalline allomorph with reactive N3-groups

situ-ated at the reducing ends on the surface prone to post-functionalization

by Cu(I)-catalyzed Huisgen alkyne-azide cycloaddition reactions (Yataka

et al., 2015) The same group recently reported the synthesis of block

copolymers of cellooligosaccharides and oligo(ethylene glycol) using

bifunctional oligomeric primers and CDP catalysis, in which a cellulose

II-like crystalline structure was observed for the products (Sugiura et al.,

2021)

In 2007, Egusa et al successfully synthesized cellulose exhibiting a

high DP from non-substituted cellobiose, implementing catalysis with a

commercially available cellulase They employed the well-known N,N-

dimethylacetamide (DMAc)/LiCl solvent system: first, to overcome the

solubility issue of cellulose, and second, to prevent partial inactivation

of the cellulase by acetonitrile, which were the main reasons for the low

DP of the product previously obtained by Kobayashi and co-workers

The authors introduced the enzyme in the form of a

cellulase/surfac-tant complex Using this method, synthetic cellulose exhibiting a DP

>100 was obtained, whose structure was confirmed by NMR

spectros-copy as well as X-ray diffraction However, the yield of higher molecular

weight products (DP ≥6) was less than 5%, and more than half of the

starting cellobiose was not consumed, which was attributed to a possible

decreasing effect of aprotic polar solvents such as DMAc or DMSO on

enzymatic activity (Egusa et al., 2007) Later, Egusa et al further

opti-mized the conditions and were able to achieve higher conversion rates A

surfactant-enveloped enzyme (SEE) and a protic component (acid) as an

additive in the same aprotic organic solvent system enabled the

syn-thesis of cellulose with a DP of more than 120 at approximately 26%

conversion (see Fig 9) Furthermore, the protective surfactant in the

SEE, i.e., dioleyl-N-D-glucona-L-glutamate, largely prevented

deactiva-tion of the cellulase in the strongly polar, aprotic organic solvent system

DMAc/LiCl and thus allowed for a more efficient biocatalytic synthesis

(Egusa et al., 2012)

4 Concluding remarks and outlook

Only limited options are available to efficiently synthesize cellulose

in vitro, even though much effort has been put into establishing efficient,

convenient, and practical protocols The complete regio- and stereo-control during the synthesis of the polysaccharide is especially chal-lenging and requires accurate fine-tuning of protecting groups and reaction conditions The method of choice strongly depends on the purpose of the synthesis If isotopically labeled and completely regio-selectively substituted cellulose or cellulose with a discrete DP (defined chain length, no molar mass distribution) are needed, chemical synthesis might be a suitable choice Especially the stepwise addition method can provide cellulose (cellooligosaccharides) with exactly defined DP (or at least a very low dispersity) However, the high number of synthetic steps required and the low amounts accessible are significant drawbacks, as is the limited accessibility of higher DPs Chemical polymerization, espe-cially the ring-opening polymerization approach, is also able to provide pure cellulose Yet, this method is prone to various quenching mecha-nisms, which again significantly limits the accessible DP range With this method, however, D-cellulose, isotopically labeled celluloses, cellulose derivatives with complete substitution selectivity as well as the enan-tiomers “L-cellulose” and optically-inactive “D,L-cellulose” have been synthesized In general, celluloses chemically synthesized are identical

to genuine celluloses according to all analytical techniques, and usually

belong to the cellulose II allomorph In vitro chemical synthesis of

cel-lulose exhibiting a type I crystalline structure may thus also be desirable and was already achieved under special conditions (Lee et al., 1994) If a higher molecular weight is required than accessible by chemical syn-thesis, enzyme-mediated synthesis is a convenient choice, although the conversion yield of the literature protocols is still very low due to the deactivation of the implemented cellulases by the required organic (co-) solvent, and further purification of the product might be more challenging

A common challenge in all the discussed methods is the insolubility

of cellulose in water as well as in almost all organic solvents Especially with increasing DP, the biopolymer precipitates and renders efficient elongation of the chain challenging, both by chemical and enzymatic means In terms of enzymatic catalysis, the available solvent systems, such as DMAc/LiCl, are critical for the enzyme's activity For efficient polymerization, the careful exclusion of any type of quenching agent is crucial for obtaining a high DP Even a solvent-free method might be accessible in the future, also rendering the synthesis more convenient in terms of workup and cellulose isolation Soon, novel bottom-up ap-proaches to cellulose and cellulose derivative synthesis can be expected, which make a completely regioselective substitution of the biopolymer

or well-defined isotopomers easier accessible and thus more widely available for specialized studies

Fig 9 Biocatalytic cellulose synthesis using a surfactant-enveloped enzyme (SEE) and a protic component (acid) in an organic medium Reprinted with permission

from Egusa et al., 2012, Copyright © 2012 American Chemical Society

CRediT authorship contribution statement

Anna F Lehrhofer: Writing – original draft, Visualization Takaaki

Goto: Writing – original draft, Visualization Toshinari Kawada:

Writing – review & editing, Supervision Thomas Rosenau:

Conceptu-alization, Writing – review & editing, Supervision, Project

administra-tion, Funding acquisition Hubert Hettegger: Conceptualizaadministra-tion,

Writing – review & editing, Visualization, Supervision, Project

admin-istration, Funding acquisition

Declaration of competing interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence

the work reported in this paper

Acknowledgements

We would like to thank the University of Natural Resources and Life

Sciences, Vienna (BOKU), the County of Lower Austria, and Lenzing AG

for their financial support in the framework of the “Austrian Biorefinery

Center Tulln” (ABCT) and the BOKU doctoral school “Advanced

Bio-refineries: Chemistry & Materials” (ABC&M) The financial support by

Wood K plus (T.G.) and the GFF Gesellschaft für Forschungsf¨orderung

Nieder¨osterreich m.b.H (A.F.L and H.H., project LSC20-002) is

grate-fully acknowledged

References

Adelw¨ohrer, C., Takano, T., Nakatsubo, F., & Rosenau, T (2009) Synthesis of 13 C-

perlabeled cellulose with more than 99% isotopic enrichment by a cationic ring-

opening polymerization approach Biomacromolecules, 10, 2817–2822, 10/brw7jb

Aloni, Y., Delmer, D P., & Benziman, M (1982) Achievement of high rates of in vitro

synthesis of 1,4-β-D-glucan: Activation by cooperative interaction of the acetobacter

xylinum enzyme system with GTP, polyethylene glycol, and a protein factor

Proceedings of the National Academy of Sciences of the United States of America,

6448–6452, 10/cswkgb

Atalla, R H., & Vanderhart, D L (1984) Native cellulose: A composite of two distinct

crystalline forms Science, 223, 283–285, 10/d7kjp4

Bessueille, L., & Bulone, V (2008) A survey of cellulose biosynthesis in higher plants

Plant Biotechnology, 25, 315–322, 10/bkh69r

Biermann, C J (1996) Handbook of pulping and papermaking (2nd ed.) Corvallis,

Oregon: Academic Press - An Imprint of Elsevier

Brown, R M (1996) The biosynthesis of cellulose Journal of Macromolecular Science,

Part A, 33, 1345–1373, 10/bwbd7h

Brown, R M., & Montezinos, D (1976) Cellulose microfibrils: Visualization of

biosynthetic and orienting complexes in association with the plasma membrane

Proceedings of the National Academy of Sciences, 73, 143–147, 10/fw3kgf

Cosgrove, D J (2005) Growth of the plant cell wall Nature Reviews Molecular Cell

Biology, 6, 850–861, 10/bn26z3

Delmer, D P P (1999) Cellulose biosynthesis: Exciting times for a difficult field of

study Annual Review of Plant Physiology and Plant Molecular Biology, 50, 245–276,

10/fv4r3q

Egusa, S., Goto, M., & Kitaoka, T (2012) One-step synthesis of cellulose from cellobiose

via protic acid-assisted enzymatic dehydration in aprotic organic media

Biomacromolecules, 13, 2716–2722, 10/f379b4

Egusa, S., Kitaoka, T., Goto, M., & Wariishi, H (2007) Synthesis of cellulose in vitro by

using a cellulase/surfactant complex in a nonaqueous medium Angewandte Chemie,

International Edition, 46, 2063–2065, 10/c5fhw6

French, A D (2017) Glucose, not cellobiose, is the repeating unit of cellulose and why

that is important Cellulose, 24, 4605–4609, 10/gch6nf

French, A D., P´erez, S., & Bulone, V (2018) Cellulose In Encyclopedia of polymer science

and technology (1st ed., pp 1–69) John Wiley & Sons, Inc

Freudenberg, K., & Nagai, W (1933) Die Synthese der Cellobiose Berichte der Deutschen

Chemischen Gesellschaft A/B, 66, 27–29, 10/d2pwqw

Fügedi, P (2006) Glycosylation methods In The Organic Chemistry of Sugars (pp

89–180) Boca Raton: CRC Press

Fügedi, P., Garegg, P J., L¨onn, H., & Norberg, T (1987) Thioglycosides as glycosylating

agents in oligosaccharide synthesis Glycoconjugate Journal, 4, 97–108, 10/cs5pdb

Graczyk, P P., & Mikolajczyk, M (1994) Anomeric effect: Origin and consequences In

Topics in stereochemistry (pp 159–350) New York, New York: John Wiley & Sons,

Inc

Hall, D M., & Lawler, T E (1971) New routes to the synthesis of 2,3,6-tri-O-substituted

methyl β-D-glucopyranosides an improved synthesis of α-cellotriose hendecaacetate

Carbohydrate Research, 1–7 https://doi.org/10.1016/S0008-6215(00)86092-9

Haq, S., & Whelan, W J (1956) 873 The chemical synthesis of polysaccharides Part I

Synthesis of gentiodextrins Journal of the Chemical Society, 4543–4549, 10/dwk4d7

Heath, I B (1974) A unified hypothesis for the role of membrane bound enzyme

complexes and mierotubules in plant Cell Wall synthesis Journal of Theoretical Biology, 48, 445–449 https://doi.org/10.1016/S0022-5193(74)80011-1 Hiraishi, M., Igarashi, K., Kimura, S., et al (2009) Synthesis of highly ordered cellulose

II in vitro using cellodextrin phosphorylase Carbohydrate Research, 344, 2468–2473,

10/fc35dj

Hirano, S (1973) The preparation of a cellulose-like polymer from 2,3,6-tri-O-(N -phenylcarbamyl)-D-glucopyranose by the action of phosphorus pentoxide in

dimethyl sulfoxide Agricultural and Biological Chemistry, 37, 187–189, 10/gmxs5r Hori, M., Kamitakahara, H., & Nakatsubo, F (1997) Substituent effects of the orthoester group on ring-opening polymerization of α-D-glucopyranose 1,2,4-orthoester

derivatives Macromolecules, 30, 6 https://doi.org/10.1021/ma961883a Hosoya, T., Takano, T., Kosma, P., & Rosenau, T (2014) Theoretical foundation for the

presence of oxacarbenium ions in chemical glycoside synthesis The Journal of

Organic Chemistry, 79, 7889–7894, 10/f6gx5h

Husemann, V E., & Müller, G J M (1966) Über die Synthese unverzweigter

Polysaccharide Makromolekulare Chemie, 91, 212–230, 10/bsbcc9

Ikegami, W., Kamitakahara, H., Teramoto, Y., & Takano, T (2021) Synthesis of optically

inactive cellulose via cationic ring-opening polymerization Cellulose, 28,

6125–6132, 10/gj6hq5

Isogai, T., Yanagisawa, M., & Isogai, A (2008) Degrees of polymerization (DP) and DP distribution of dilute acid-hydrolyzed products of alkali-treated native and

regenerated celluloses Cellulose, 15, 815–823, 10/fbn7jd

Jarvis, M C (2013) Cellulose biosynthesis: Counting the chains Plant Physiology, 163,

1485–1486, 10/gncwbf

Juaristi, E., & Cuevas, G (1992) Recent studies of the anomeric effect Tetrahedron, 48,

5019–5087, 10/dxtgt6

Kadokawa, J (2011) Precision polysaccharide synthesis catalyzed by enzymes Chemical Reviews, 111, 4308–4345 https://doi.org/10.1021/cr100285v

Kamitakahara, H., Funakoshi, T., Nakai, S., et al (2010) Synthesis and structure/ property relationships of regioselective 2-O-, 3-O- and 6-O-ethyl celluloses

Macromolecular Bioscience, 10, 638–647, 10/czz73n

Kamitakahara, H., Funakoshi, T., Takano, T., & Nakatsubo, F (2009) Syntheses of 2,6-O- alkyl celluloses: Influence of methyl and ethyl groups regioselectively introduced at

O-2 and O-6 positions on their solubility Cellulose, 16, 1167–1178, 10/fvdfd9

Kamitakahara, H., Hori, M., & Nakatsubo, F (1996) In , 29 Substituent effect on ring-

opening polymerization of regioselectively acylated β-D-glucopyranose 1,2,4-orthopivalate derivatives (p 6)

Kamitakahara, H., Koschella, A., Mikawa, Y., et al (2008) Syntheses and comparison of

2,6-di-O-methyl celluloses from natural and synthetic celluloses Macromolecular

Bioscience, 8, 690–700, 10/cwtms4

Kamitakahara, H., Nakatsubo, F., & Murakami, K (1994) In , 27 Ring-opening

polymerization of 1,4-anhydro- α -D-glucopyranose derivatives having acyl groups and synthesis of (1→5)-β-D-glucofuranan (p 6)

Kawada, T., Nakatsubo, F., & Murakami, K (1990) Synthetic studies of cellulose VII: Synthesis of a series of perbenzylated allyl 4n’-O-p-methoxybenzyl cello-

oligosaccharides up to octamer Cellulose Chemistry and Technology, 343–350

Kawada, T., Nakatsubo, F., Umezawa, T., et al (1994) Synthetic studies of cellulose XII:

First chemical synthesis of cellooctaose acetate Mokuzai Gakkaishi, 7, 738–743

Klemm, D., Heublein, B., Fink, H.-P., & Bohn, A (2005) Cellulose: Fascinating

biopolymer and sustainable raw material Angewandte Chemie, International Edition,

44, 3358–3393, 10/bs2fw2

Kobayashi, S (2005) Challenge of synthetic cellulose Journal of Polymer Science, Polymer

Chemistry Edition, 43, 693–710, 10/d2cdrt

Kobayashi, S., Kashiwa, K., Kawasaki, T., & Shoda, S (1991) Novel method for polysaccharide synthesis using an enzyme: The first in vitro synthesis of cellulose via

a nonbiosynthetic path utilizing cellulase as catalyst Journal of the American

Chemical Society, 113, 3079–3084, 10/cvrs7k

Kobayashi, S., & Makino, A (2009) Enzymatic polymer synthesis: An opportunity for

green polymer chemistry Chemical Reviews, 109, 5288–5353, 10/df2w7c

Kobayashi, S., Makino, A., Matsumoto, H., et al (2006) Enzymatic polymerization to novel polysaccharides having a glucose-N-acetylglucosamine repeating unit, a

cellulose− chitin hybrid polysaccharide Biomacromolecules, 7, 1644–1656, 10/

cmzn45

Kobayashi, S., Sakamoto, J., & Kimura, S (2001) In vitro synthesis of cellulose and

related polysaccharides Progress in Polymer Science, 26, 1525–1560, 10/fp7 ppf

Kochetkov, N K (1994) Recent developments in the synthesis of polysaccharides and

stereospecificity of glycosylation reactions In Studies in natural products chemistry

(pp 201–266) Elsevier

Koenigs, W., & Knorr, E (1901) Ueber einige Derivate des Traubenzuckers und der

Galactose Berichte der Deutschen Chemischen Gesellschaft, 34, 957–981, 10/cf2wv7

Koyama, M., Helbert, W., Imai, T., et al (1997) Parallel-up structure evidences the

molecular directionality during biosynthesis of bacterial cellulose Proceedings of the

National Academy of Sciences, 94, 9091–9095, 10/d894q2 Lee, J H., Brown, R M., Kuga, S., et al (1994) Assembly of synthetic cellulose I

Proceedings of the National Academy of Sciences, 91, 7425–7429 https://doi.org/ 10.1073/pnas.91.16.7425

Lin, F C., Malcolm Brown, R., Cooper, J B., & Delmer, D P P (1985) Synthesis of fibrils

in vitro by a solubilized cellulose synthase from azobacter xylinum Science, 230,

822–825 https://doi.org/10.1126/science.230.4727.822

Mboowa, D (2021) A review of the traditional pulping methods and the recent

improvements in the pulping processes Biomass Conversion and Biorefinery https:// doi.org/10.1007/s13399-020-01243-6

Micheel, F., & Brodde, O.-E (1974) Polymerisation von 1,4-Anhydro-2,3,6-tri-O-benzyl-

α-D-glucopyranose Justus Liebigs Annalen der Chemie, 1974, 702–708, 10/fnv7dh

Micheel, V F., B¨ockmann, A., & Meckstroth, W (1961) Darstellung und Struktur

synthetischer Polysaccharide Makromolekulare Chemie, 48, 1–16, 10/cbm8rc

Mootoo, D R., Vandana, D., & Bert, F.-R (1988) N-pentenyl glycosides permit the

chemospecific liberation of the anomeric center Journal of the American Chemical

Society, 110, 2662–2663, 10/dtb2rz

Nakatsubo, F., Kamitakahara, H., & Hori, M (1996) Cationic ring-opening

polymerization of 3,6-Di-O-benzyl-α-D-glucose 1,2,4-orthopivalate and the first

chemical synthesis of cellulose Journal of the American Chemical Society, 118,

1677–1681, 10/fd6vqw

Nakatsubo, F., Takano, T., & Kawada, T (1985) Synthetic studies of cellulose, 1:

Synthetic design and selection of the protective groups Memoirs of the College of

Agriculture, Kyoto University, 127, 37–47

Nishimura, T., & Nakatsubo, F (1997) Chemical synthesis of cellulose derivatives by a

convergent synthetic method and several of their properties Cellulose, 4, 109–130

https://doi.org/10.1023/A:1018423503762

Nunes, R C R (2017) Rubber nanocomposites with nanocellulose In Progress in rubber

nanocomposites (pp 463–494) Elsevier

Okamoto, E., Kiyosada, T., Shoda, S.-I., & Kobayashi, S (1997) Synthesis of alternatingly

6-O-methylated cellulose via enzymatic polymerization of a substituted cellobiosyl

fluoride monomer catalyzed by cellulase Cellulose, 4, 161–172 https://doi.org/

10.1023/A:1018479621509

Olek, A T., Rayon, C., Makowski, L., et al (2014) The structure of the catalytic domain

of a plant cellulose synthase and its assembly into dimers Plant Cell, 26, 2996–3009,

10/f6hj52

Ragnar, M., Henriksson, G., & Lindstr¨om, M E (2014) Pulp In Ullmann’s encyclopedia of

industrial chemistry (4th ed., pp 1–92) Weinheim: Wiley-VCH

Rosenau, T., Potthast, A., & Hofinger, A (2019) Toward a better understanding of

cellulose swelling, dissolution, and regeneration on the molecular level In Cellulose

science and technology Chemistry, analysis and applications (pp 99–125) Hoboken,

USA: John Wiley & Sons Inc

Ruckel, E R., & Schuerch, C (1966a) Chemical synthesis of a stereoregular linear

polysaccharide Journal of the American Chemical Society, 88, 2605–2606, 10/brkj3j

Ruckel, E R., & Schuerch, C (1966b) Preparation of high polymers from 1,6-Anhydro-

2,3,4-tri-O-substituted β-D-glucopyranose The Journal of Organic Chemistry, 31,

2233–2239 https://doi.org/10.1021/jo01345a035

Samain, E., Lancelon-Pin, C., F´erigo, F., et al (1995) Phosphorolytic synthesis of

cellodextrins Carbohydrate Research, 271, 217–226, 10/bc733f

Saxena, I M., & Brown, R M (2005) Cellulose biosynthesis: Current views and evolving

concepts Annals of Botany, 96, 9–21, 10/b9rx3k

Schlubach, H H., & Lührs, E (1941) Einwirkung von Chlorwasserstoff auf Glucose

Synthese eines Polyglucosans Justus Liebigs Annalen der Chemie, 547, 73–85, 10/

d423c7

Schmidt, R R., & Jung, K.-H (2000) Trichloroacetimidates In B Ernst, G W Hart, &

P Sina (Eds.), Carbohydrates in chemistry and biology (pp 5–59) Weinheim,

Schmidt, R R., & Michel, J (1980) Facile synthesis of α- and β-O-glycosyl imidates;

preparation of glycosides and disaccharides Angewandte Chemie (International Ed in

English), 19, 731–732, 10/b88hfh

Schmidt, R R., & Michel, J (1982) Synthesis of linear and branched cellotetraoses

Angewandte Chemie (International Ed.), 21, 72–73, 10/c93tdr

Serizawa, T., Tanaka, S., & Sawada, T (2021) Control of parallel versus antiparallel

molecular arrangements in crystalline assemblies of alkyl β-cellulosides Journal of

Colloid and Interface Science, 601, 505–516 https://doi.org/10.1016/j

jcis.2021.05.117

Shapiro, D., Rabinsohn, Y., & Diver-Haber, A (1969) A new micro-scale synthesis of

lactose Biochemical and Biophysical Research Communications, 37, 28–30, 10/chmpxj

Shoda, S., Uyama, H., Kadokawa, J., et al (2016) Enzymes as green catalysts for

precision macromolecular synthesis Chemical Reviews, 116, 2307–2413 https://doi org/10.1021/acs.chemrev.5b00472

Staudinger, H (1920) Über Polymerisation Berichte der Deutschen Chemischen

Gesellschaft A/B, 53, 1073–1085, 10/fq7gp3

Sugiura, K., Sawada, T., Tanaka, H., & Serizawa, T (2021) Enzyme-catalyzed propagation of cello-oligosaccharide chains from bifunctional oligomeric primers for

the preparation of block co-oligomers and their crystalline assemblies Polymer

Journal, 53, 1133–1143, 10/gkq3wt

Takano, T., Nakatsubo, F., & Murakami, K (1990) Synthetic studies of cellulose VI:

Effect of the substituent groups of glycon on β-glycosylation Cellulose Chemistry and

Technology, 333–341

Takano, T., Nakatsubo, F., & Murakami, K (1994) Synthetic studies of cellulose V: Effect

of the 3-O- and 6-O-substituent groups of the aglycon on β-glycosylation Cellulose

Chemistry and Technology, 135–145 Takeo, K., Okushio, K., Fukuyama, K., & Kuge, T (1983) Synthesis of cellobiose,

cellotriose, cellotetraose, and lactose Carbohydrate Research, 163–173 https://doi org/10.1016/0008-6215(83)84014-2

Takeo, K., Yasato, T., & Kuge, T (1981) Synthesis of α- and β-cellotriose hendecaacetates and of several 6,6 ′ ,6 ′′ -tri-substituted derivatives of methyl β-cellotrioside

Carbohydrate Research, 148–156 https://doi.org/10.1016/S0008-6215(00)80761-2 Uryu, T., Yamaguchi, C., Morikawa, K., et al (1985) Ring-opening polymerization of 1,4-anhydro-2,3,6-tri-O-benzyl-α-D-glucopyranose and 1,4-anhydro-2,3,6-tri-O-

benzyl-β-D-galactopyranose Macromolecules, 18, 599–605, 10/c8jzgb

Uryu, T., Yamanouchi, J., Hayashi, S., et al (1983a) Selective ring-opening polymerization of 1,4-Anhydro-2,3-di-O-benzyl-α-D-xylopyranose and synthesis of stereoregular (1→5)-α-D-xylofuranan Macromolecules, 16, 320–326

Uryu, T., Yamanouchi, J., Kato, T., et al (1983b) Selective ring-opening polymerization

of di-O-methylated and di-O-benzylated 1,4-anhydro-α-D-ribopyranoses and structure proof of synthetic cellulose-type polysaccharide (1→4)-β-D-ribopyranan and (1→5)-α-D-ribofuranan Journal of the American Chemical Society, 105,

6865–6871, 10/dtmf37

Williamson, R E., Burn, J E., & Hocart, C H (2002) Towards the mechanism of

cellulose synthesis Trends in Plant Science, 7, 461–467, 10/fpvkwr

Xiao, R., & Grinstaff, M W (2017) Chemical synthesis of polysaccharides and

polysaccharide mimetics Progress in Polymer Science, 74, 78–116, 10/gch58t

Yagura, T., Ikegami, W., Kamitakahara, H., & Takano, T (2020) Synthesis of an

enantiomer of cellulose via cationic ring-opening polymerization Cellulose, 27,

9755–9766, 10/gmxtb4

Yataka, Y., Sawada, T., & Serizawa, T (2015) Enzymatic synthesis and post- functionalization of two-dimensional crystalline cellulose oligomers with surface-

reactive groups Chemical Communications, 51, 12525–12528, 10/gn4h4q

Yataka, Y., Sawada, T., & Serizawa, T (2016) Multidimensional self-assembled structures of alkylated cellulose oligomers synthesized via in vitro enzymatic

reactions Langmuir, 32, 10120–10125, 10/f86zfs Yoshida, T (2001) Synthesis of polysaccharides having specifc biological activities

Progress in Polymer Science, 26, 379–441 https://doi.org/10.1016/S0079-6700(00) 00045-9

Zachoval, J., & Schuerch, C (1969) Steric control in the polymerization of 1,6-anhydro- β-D-glucopyranose derivatives Journal of the American Chemical Society, 91, 1165–1169 https://doi.org/10.1021/MA951454X

Zhu, X., & Schmidt, R R (2009) New principles for glycoside-bond formation

Angewandte Chemie (International Ed.), 48, 1900–1934, 10/c5fk7w

Zweckmair, T., Oberlerchner, J T., B¨ohmdorfer, S., et al (2016) Preparation and

analytical characterisation of pure fractions of cellooligosaccharides Journal of

Chromatography A, 1431, 47–54, 10/gmxtb5