Top Catal (2012) 55:1231–1246 DOI 10.1007/s11244-012-9891-2
REVIEW ARTICLE
Cellulase Immobilized Nanostructured Supports for Efficient Saccharification of Cellulosic Substrates Ankush A. Gokhale • Ilsoon Lee
Published online: 9 October 2012 Ó Springer Science+Business Media New York 2012
Abstract Functionalized nanomaterials are promising candidates for enzyme immobilization to develop efficient industrial biocatalysts with tailor-made catalytic properties. Cellulase, a saccharifying hydrolase, can be immobilized on various nanostructured supports using different types of binding chemistries. This review examines prior cellulase immobilization strategies and promising future techniques to integrate nanotechnology with biocatalysis. Keywords Nanobiotechnology Cellulase Immobilization Nanosupports Biocatalysis
1 Introduction In recent years, with the search for cleaner and greener routes for bulk synthesis of industrially relevant products intensifying, enzyme technology has emerged as a viable alternative. Enzymes are rightly called ‘biocatalysts’ due to their innate capability of aiding complex transformations in a more subtle way as nature originally conceived it to be [1]. Often regarded as the ‘third wave’ following the successful implementation of biotechnology in agricultural and pharmaceutical applications [2, 3], new developments in industrial enzyme technology are credited to developing highly substrate specific catalytic mechanisms to produce enantiomerically pure products. Examples include the conversion of b-tetralone to the desired amine by the S-selective transaminase with an enantiomeric excess (ee) of around 80–94 % [4] as well as bacterial lipase A. A. Gokhale I. Lee (&) Department of Chemical Engineering and Materials Science, Michigan State University, East Lansing, MI 48824-1226, USA e-mail:
[email protected]
developed from Pseudomonas aeruginosa with an ee of over 90 % as opposed to 2 % for wild-type lipase [5]. Biocatalytic processes have often been used to produce higher end products usually in the range of US $20–30 per kg [6]. However, enzyme technology has received a considerable boost in the past decade because of cheaper downstream processing, better understanding of genomics and a plethora of information provided by bioinformatics and high throughput screening studies [7]. This has enabled increased production of lower end goods at cost effective price. The use of nitrile hydratases as a biocatalyst to convert acrylonitrile to acrylamide, a well know commodity chemical shows that enzyme technology can be easily scaled up [8, 9]. Case studies have also shown that the use of enzyme technology helps cut the industrial operating costs by 10–50 % by bringing down the energy and raw material costs [10]. The key to improved biocatalysis is to develop engineered enzymes suitable for industrial processes by making suitable changes in the proposed biocatalyst design as well as the microenvironment in which they operate. Sitedirected mutagenesis and directed evolution are two commonly used tools in protein engineering to modify the catalytic potential of the enzymes. Site-directed mutagenesis provides invaluable insights regarding the relative importance of specific residues in the overall catalytic mechanism. Typically, site-directed mutagenesis involves substitution of specific amino acid residues followed by expression of the protein. Holloway et al. [11] studied the effect of site-directed mutagenesis techniques on improving the range of substrates that can be dechlorinated by X. autotrophicus haloalkane dehalogenase GJ10. Similarly, Igarashi et al. [12] have discussed the substitution of certain key amino acid residues near the active site to improve the substrate specificity of pyrroloquinoline quinone
123
1232
(PQQ)-harboring water-soluble glucose dehydrogenase. Directed evolution on the other hand aims to create multiple mutant libraries using recombining or non-recombining mutagenesis methods. This is followed by screening and selection based on desired parameters such as catalytic improvements or substrate specificity to accumulate a large number of beneficial mutant genes [13]. The usefulness of directed evolution techniques in enzyme biocatalysis has been demonstrated by several studies. Desantis et.al. [14] showed the preparation of highly enantioselective nitrilase using the gene site saturation mutagenesis (GSSM) evolution strategy. Bornscheuer et al. [15] studied the esterase catalyzed hydrolysis of the 3-hydroxy esters and showed a 25 % ee after using the directed evolution techniques. The emergence of nanobiotechnology as a new integrated discipline has helped diminish rigid boundaries between materials science, chemistry and biology. New techniques to synthesize size-controlled multifunctional nanoparticles developed by material scientists, novel physico-chemical binding mechanisms contributed by chemists and fundamental understanding of cellular systems provided by microbiologists have helped shape the contours of this multidisciplinary field. In the past decade, the use of nanostructured platforms to incorporate bioactive components has generated considerable interest especially in the fields of pharmacology, and biosensor industry [16–25]. In certain specialized pharmacological cases, nanobiotechnology has even made it possible to carry as well as release drug molecules to a specific organ of interest [26, 27]. Enzymes such as glucose oxidase have been successfully immobilized on graphene based nanomaterials to yield a fast response and improved sensitivity [28, 29]. The widespread use of nanobiotechnology in pharmaceutical and biosensor applications has led to an increased effort to replicate its success in other areas of interest such as bioenergy. The success of nanobiotechnology would however depend on the availability of mature technologies that can reasonably integrate and support miniaturization of biological functions. The use of functionalized nanosized materials as supports for enzyme immobilization has been gaining ground in recent years. Bare nanostructured supports are susceptible to aggregation. Hence these supports are usually modified by surface functionalization techniques. The type of surface chemistry used during modification is highly specific to the nanoparticle-enzyme system under consideration. Lee and co-workers [16, 17, 21–23] at MSU have utilized functionalized nanostructured lipid membranes tethered on electrodes to immobilize proteins such as the catalytically active esterase domains of neuropathy target esterase (NTE) by maintaining the activity of membrane proteins for the development of highly sensitive nanobiosensors. In their work, they utilized self-assembly, soft
123
Top Catal (2012) 55:1231–1246
lithography, layer-by-layer assembly, and molecular interactions at the interfaces [18, 30–45]. Nanosized materials usually have high surface area to volume ratio providing ample opportunities to anchor a number of bioactive materials [46, 47]. Bioconjugated nanostructured supports offer several advantages ranging from high loading, improved dispersability and reduced mass-transfer limitation [48–50]. Proteins undergo conformational changes as they bind to nanosized supports. Lundqvist et al. [51] linked the perturbations in the protein secondary structures to the size of the nanoparticles on which they are immobilized. Smaller nanoparticles with higher surface curvature allow proteins to retain their native structure. Nanoparticles because of their small sizes, often display Brownian motion. An interesting study by Jia et al. [52] used a simple model based on collision theory and Stokesequation to predict the enzymatic activity of a-Chymotrypsin on polymeric nanoparticles with different sizes. Particle size of the supports, enzyme mobility and protein conformational changes are some of the important factors that govern the ability of immobilized enzyme systems to deliver improved biocatalytic performance. There has been growing interest in cellulosic ethanol as the next generation fuel in recent years. Depleting fossil fuels, concerns about energy security as well as magnanimous government incentives in the form of subsidies and tax-breaks have positioned cellulosic ethanol to be a strong contender as an alternative energy source. With several governments around the world recommending the use of bioethanol blended fuel to meet the ever increasing energy demands, there has been considerable effort to find a practical and economically feasible solution to production of ethanol from renewable resources. Conversion of cellulosic biomass into reducing sugars has been historically accomplished by acid-based techniques and enzymatic processes. According to some estimates, besides being environmental friendly, enzyme based processes present roughly the same projected costs as compared to acidbased processes [53]. In a typical hydrolysis process, cellulose is hydrolyzed to reducing sugars by a bunch of catalytic and noncatalytic enzyme modules secreted by cellulolytic microorganisms. These modules are collectively referred to as cellulases. Cellulases comprise of a number of enzyme systems such as endo, exo glucanase and b-glucosidase. While endo-exo glucanases are responsible for disrupting the cellulosic matrix, b-glucosidase converts the cellobiose, an intermediate product generated during endo-exo synergism to reducing sugars [54]. The large scale use of cellulase as a biocatalyst for the production of cellulosic ethanol would entail careful optimization of the specific ratios in which the enzymes need to be expressed as well as the specific activities of individual enzymes. Moreover, it would also need developing
Top Catal (2012) 55:1231–1246
effective strategies to ensure the complete use of the biocatalytic potential of the added cellulase. According to some studies, cellulases account for as much as 50 % of the total hydrolysis costs [55, 56]. A truly cost efficient process would therefore demand evolving sustainable routes to ensure recovery of cellulase followed by its reuse. However, recovery and reuse is complicated due to the fact that during hydrolysis cellulase tends to get redistributed over two heterogeneous phases: the solid substrate and the liquid supernatant [56, 57]. An alternative strategy to prevent this redistribution and facilitate reuse is the technique of immobilization. The cellulase technology is still in the nascent stage of development and it is likely that newer breed of microbes producing highly specialized enzymes would be designed in the near future. However, the sheer robustness of the immobilization technique would allow ready assimilation of these new enzymes within the preexisting architecture with or without minor modifications. In addition, for any industrial process stability of catalyst over a relatively wide range of process parameters is necessary. Immobilization of enzymes on supports in general helps to improve the operational stability by increasing the resilience of enzymes to variation in pH and temperature [58, 59]. In this review, we discuss past and current developments in the field of cellulase immobilization on nanostructured supports for effective conversion of cellulosic substrates to reducing sugars. Before that, it is important to understand the structure of cellulase and the mechanism by which it operates in order to design effective immobilization strategies.
1233
consumption is estimated to reach about US $400 million per year [8]. And this is assuming that cellulase would be used for hydrolyzing the corn-stover produced in the Midwest alone. The actual requirement would be considerably higher if the traditional applications of cellulase such as in detergent, textile, pulp and paper industries are all factored in. Certain studies were conducted in the past to examine the efficiency of cellulase from fungal sources by estimating the enzyme turn over number (kcat). Sinnot [65] showed that the kcat for cellulase produced from Trichoderma reesei acting on b-cellobiosyl fluoride was two orders of magnitude lower than that of Aspergillus glucoamylase acting on a-glucosyl fluoride. It is hypothesized that cellulase utilize the energy produced from the cleavage of glycosyl bonds for its own functions as opposed to improving the process of hydrolysis. Lower catalytic efficiency means higher cellulase requirement for cellulose to glucose conversion. This drives up the cost of the hydrolysis process making commercialization difficult. Recent development in cellulase technology through collaboration between industry and Department of Energy (DOE) has helped lower the enzyme cost from a high of US $5.40 per gallon of ethanol to approximately 20 cents per gallon of ethanol [66]. However, for the process to be truly competitive, the cost of cellulase needs to drop to less than 7 cents per gallon of ethanol [60, 67]. Further reduction in the cost of cellulase through genetic engineering or downstream processing could be very challenging [56]. Thus cellulase recovery and reuse facilitated through immobilization on supports can help offset the cost of the enzymes and make the process economically viable. 2.2 Structure
2 Cellulases 2.1 Background The use of the term ‘cellulase’ for the class of enzymes that degrade cellulosic materials goes back to the beginning of the 20th century [60, 61]. Major impetus to enzyme technology including cellulase biochemistry was provided during the Second World War when new protocols for effective protein separation started emerging [60]. Cellulase was initially conceived to be a consortium of several enzymes with one set of enzymes termed ‘C1’ exclusively meant for de-crystallization of the cellulose whereas another set called ‘Cx’ responsible for the hydrolytic action [62]. In recent years, there is a common agreement that the action of cellulase in degrading cellulose to reducing sugars is a result of three major classes of enzymes namely endoglucanase (EG), exoglucanase and b-glucosidase [63, 64]. With the bioethanol industry growing steadily over the past few years, the market potential for the cellulase
The modular structure of most cellulases (EG, exoglucanase and b-glucosidase) is conceived to exhibit at least two functional domains: a catalytic domain (CD) that participates in the cleavage of glycosidic bonds and the cellulose binding domain (CBD) which as the name suggests binds to the substrate [68, 69]. A peptide sequence links the two domains together. Figure 1 shows a typical schematic diagram of the various modules in a generic cellulolytic enzyme. Various experimental techniques such as X-ray crystallography, small angle X-ray scattering and nuclear magnetic resonance (NMR) have been used in the past to examine the 3D architecture of cellulase [71–73]. Studies have confirmed the existence of two domain proteins in case of T. reesei cellobiohydrolase (CBH) enzymes—the larger one corresponding to CD whereas the smaller one representing CBD [74, 75]. Kraulis et al. [71] showed through NMR studies that terminal CBD in case of CBH 1 has two sides one hydrophobic and the other hydrophilic.
123
1234
Top Catal (2012) 55:1231–1246
Fig. 1 Simplified scheme showing different functional domains of cellobiohydrolase, one of the exo-cellulolytic enzymes: The CBD anchors the enzyme to the substrate whereas the CD attacks the glycosidic bonds. The polypeptide linker joins the two domains. Reproduced with permission from [70]
The hydrophilic side is rich in tyrosine residues. Since tyrosine plays a significant role in cellulose binding, it has been hypothesized that the hydrophilic surface is primarily responsible for attachment of the CBD to the cellulosic substrate [76]. The exact role played by the CBD is relatively unclear. Some studies claim that the sole function of CBD is to anchor the enzyme on the cellulosic substrate [76] whereas others hypothesize that the CBDs are actually responsible for releasing individual cellulosic chains from the substrate of interest [77]. The role of the polypeptide linker joining the two domains is related to the function of the CBD. If CBD is assumed to play the mere role of binding on to the cellulose, the linker is probably just a spacer connecting the two domains [76]. On the other hand, if CBD plays a more prominent role in cellulose hydrolysis as claimed above, the linker can be responsible for the degree of penetration of the CBD into the substrate and in controlling the flow of cellulosic chains to the active sites [78]. Some studies have also concluded that the length of the linker peptide joining the CD and CBD domain is of great significance [79–81]. In case of endoglucanse A derived from Cellulomonas fimi, the deletion of the linker chain resulted in decrease in catalytic ability, though there was little effect on the ability to adsorb on microcrystalline cellulose [80]. On the other hand, in case of Cellobiohydrolase I (CBH I) derived from T. reesei, the removal of the first one-third part of the linker reduced the binding capacity but did not affect the catalytic ability [81]. Inspite of several studies, it is difficult to validate the exact structure of cellulase mainly because the architecture tends to vary depending on the family of cellulolytic microbes used for expressing the enzyme. Hence, cellulase is classified into different types depending on the type and nature of CDs. 2.3 Synergistic Mechanism of Cellulases The synergistic action of the various enzyme sub-systems within cellulase is well-known. While EG attacks the
123
internal b-1,4 bonds, exoglucanase cleave the terminal ends of the cellulose chain. Exoglucanase usually exists in two types: CBH I and cellobiohydrolase II (CBH II). CBH I and CBH II are processive in nature because they tend to attack the reducing and non-reducing ends as they progress along the cellulose chain. The main product from endo-exo catalysis is cellobiose. b-Glucosidase, another enzyme subsystem converts cellobiose to reducing sugar. Industrial strains of T. reesei can produce up to five types of EGs and two types of CBHs (CBH I and CBH II) [82]. EGs possess a higher catalytic ability when it comes to cleaving the internal glycosidic bonds. EGs have the active site situated on an open cleft unlike CBH whose catalytic center is located in a tunnel shaped region [83]. The mechanism of how EGs and CBHs act cooperatively is still a debatable question. Various synergistic mechanisms such as exo–exo or endo-exo have been suggested. The exo–exo synergism has been suggested on the basis of experimental proof shown by Fagerstam and Pettersson [84]. The more commonly used endo-exo synergism explains the formation of free cellulosic ends by EGs followed by action of CBHs on both reducing and non-reducing ends to produce cellobiose [83]. Figure 2 shows the cleavage of the amorphous cellulosic chains by EGs followed by the processive action of the CBHs. Glucose is produced from cellobiose units by b-glucosidase. The extent of synergism between EG and CBH also depends on the nature of substrate. Very little or no synergism was found to exist between the two enzymes when soluble substrates such as Carboxymethyl Cellulose (CMC) were used [87]. For crystalline substrates like commercial avicel, the effect of CBH is more pronounced because the internal b-1,4 bonds are relatively inaccessible to EGs [88]. However, most crystalline substrates have some amorphous regions which remain accessible to EGs. So the synergistic effect cannot be completely ruled out in this case. The degree of synergy, an important parameter used to evaluate the cooperative behavior among different cellulase sub-sets, is a ratio of the effect observed in presence
Top Catal (2012) 55:1231–1246
1235
Fig. 2 The endo-exo synergistic action of the cellulolytic enzymes: EGs cleave the internal b-1,4 bonds. CBHs attack the reducing/nonreducing ends. b-Glucosidase produces glucose from cellobiose units. Adapted with permission from [85, 86]
of all enzymes to the sum of effects observed in presence of individual enzymes. Karlsson et al. [83] carried out a number of studies to study the effect of synergism on cellulase systems using steam treated willow as the substrate. One of the prominent observations made in this study is that presence of b-glucosidase plays a significant role in cellulolytic hydrolysis. Conversion was noticeably lower when no b-glucosidase was added. However this was not the case when EGs were used. This behavior can be explained on the basis of functions that each enzyme carries out. CBH is primarily responsible to convert the reducing and non-reducing ends of individual cellulosic chains into cellobiose. The products of cellulolytic hydrolysis can cause significant inhibition. The presence of bglucosidase helps eliminate accumulation of cellobiose by facilitating its conversion to glucose. As per Karlsson et al. [83], when plotted as a function of the adsorbed enzyme, EGs show almost similar conversions in presence or absence of b-glucosidase. These studies also conclude that if used independently, at lower conversions, the absolute quantity of sugars produced by EGs per unit of enzyme used is comparable to that produced by CBHs. At higher conversions, CBH is more efficient. EGs have a special preference for disorganized amorphous regions of the substrate. Once these regions are cleaved by EGs, further conversion of the remaining crystalline parts of the substrate proceeds at a slower pace. If EG and CBH are used together, results show that the combined effect of the two enzymes acting synergistically is quite high as compared to the sum of their individual effects. Also it was observed that the degree of synergy was quite high in the initial duration of hydrolysis. As the time of hydrolysis increased the value decreased and finally leveled off. The high synergistic effect could be because of EGs rapidly attacking the easily hydrolyzable areas of the substrate and generating several loose ends for the CBH to work with. As the concentration of these easily hydrolyzable areas goes down with passage of time, the EGs have to compete with the
CBH for finding appropriate substrate sites. Glucose is another product which has been extensively studied for its inhibitory effect [18, 89]. Accumulation of glucose can limit the b-glucosidase activity leading to build up of unutilized cellobiose. This indirectly affects the performance of EGs and CBHs. Some reports claim that the rate limiting step in the hydrolysis of crystalline cellulose is probably the binding of enzyme to the cellulosic substrate followed by channelization of individual cellulosic chains towards the active sites [90]. A number of mechanisms have been suggested to explain the conversion of cellulose to reducing sugars. Prominent mechanisms include retention or inversion of anomeric configurations. Identification of stereoisomers using p-NMR studies can help identify the possible mechanism. It has been hypothesized that cellulases having similar active sites tend to produce stereochemically similar isomers [91]. While is most cellulases, the active sites are located in two carboxylic acid residues [92], one of which acts as a proton donor and the other as a nucleophile. The structural topography of the active centers is such that in retaining enzymes, the distance between the ˚ which is proton donor and nucleophile is about 5.5 A ˚ in case of inversion type enzymes [93]. enlarged to 10 A While inversion is a single step process, retaining enzymes form a substrate-enzyme intermediate which is then transformed into the product [92]. 2.4 Immobilization of Cellulases on Nanostructured Supports Immobilization of cellulases on nanomaterials has been achieved using a diverse range of methodologies. These methods have been classified on the basis of the interaction between the cellulase and the support used for immobilization. Table 1 references prior work in this area by providing information about the binding chemistries used for cellulase immobilization, nanosupports used for binding and the types of substrates used for cellulosic hydrolysis.
123
1236
Top Catal (2012) 55:1231–1246
Table 1 Immobilization of cellulases on various types of nanostructured supports using different binding chemistries Immobilization technique
Nanosupport used for immobilization
Substrate used for hydrolysis
Enzyme or enzyme sub-type
Reference
Adsorption on surface of nanostructured materials
Nanoclay materials
p-Nitrophenyl-b-D-glucopyranoside
b-glucosidase
[94]
Covalent binding of cellulase on nanomaterials
Immobilization within nanopores
Soil colloidal particles
p-Nitrophenyl-b-D-glucoside
b-glucosidase
[95]
Hydrophobic teflon silica nanoparticles Activated carbon
Laminarin
Endo-b-1,3glucanase Cellulase from A. niger
[96]
Superparamagnetic iron-oxide nanoparticles
CMC
Cellulase from Trichoderma viride
[98]
Carbon nanotubes Latex colloidal particles
4-Nitrophenyl-b–D-glucopyranoside 2-Nitrophenyl-b-D-glucopyranoside
b-glucosidase b-glucosidase
[99] [100]
PMMA cores-shell particles
CMC
Cellulase from Aspergillus Sp.
[101]
Nanofibrous PAN membrane
CMC
Cellulases from A. niger
[102]
PVA membrane
CMC
Cellulase, Source: –
[103]
Non-porous silica nanoparticles
CMC
Cellulase from T. reesei
[104]
Gold nanoparticles
CMC
EG from Fusarium Sp.
[105]
Liposomes
CMC, Cellulose powder CC31
Cellulase from Trichoderma viride
[106]
Alginate and polyacrylamide gels, pore size: –
4-nitrophenyl-b–D-glucopyranoside
b-glucosidase isolated from A. niger
[107]
SBA-15 pore size: 5.4–11 nm
Microcrystalline cellulose
Cellulase from Trichoderma viride
[108]
FDU-12 (entrance about: 9–10.8 nm and cavity about 19–28 nm)
CMC
Cellulase from T. reesei
[109]
Small pore 2–4 nm mesoporous silica nanoparticles (SPMSN)
Cellulose power pretreated with ionic liquid 1-butyl-3-methyl-imidazolium chloride (BMIM)Cl
Cellulase from T. reesei
[110]
Large pore 20–40 nm mesoporous silica nanoparticles (LPMSN)
Methylcellulose solution
2.5 Adsorption of Cellulase on the Surface of Nanostructured Supports Physical adsorption of cellulase on solid supports is perhaps the most straightforward technique to achieve immobilization. Lower costs and relatively non-toxic mode of attachment are some of the major advantages of using this technique [111]. Vander–Waal forces of attraction, hydrogen bonding and hydrophobic interactions are some of the common modes of attachment of the protein molecules on the supports [112]. Some studies have also claimed that non-specific adsorption of enzymes in general reduces the chances of internal mass transfer diffusion, a
123
[97]
phenomenon seen in some other modes of immobilization [111]. A detailed study regarding sorption and desorption of cellulase on silicate clay materials has been shown elsewhere [113]. This study claims that organic matter with higher C/N ratio showed higher affinity for cellulase sorption. In another recent study, b-glucosidase, one of the three cellulolytic enzymes, was immobilized on nanoclay materials [94] as well as fine/coarse soil colloidal particles [95] and its effect on enzyme activity was monitored. In case of fine soil colloidal particles, greater enzyme loading was reported mainly due to higher surface area available for immobilization. The retention of enzyme activity on fine colloids was also reported to be higher. Soil colloids in
Top Catal (2012) 55:1231–1246
general contain a large number of metallic residues. The catalytic potential of the immobilized b-glucosidase improved because of the formation of enzyme-metal complexes which are reported to facilitate better substrate binding. Site-directed immobilization of enzymes can ensure that the CDs of the enzyme are always accessible to the substrate. However, despite several advantages, adsorption of enzymes on supports remains highly non-specific. Moreover, cellulolytic enzymes bound to solid supports are likely to desorb and instead re-adsorb on the cellulosic substrate. As explained in the previous section, cellulases share a strong affinity for cellulosic substrates because of the presence of CBDs. The forces responsible for physical adsorption of cellulase on solid supports are inherently weak and in the presence of cellulose, a substrate with which cellulases enjoy natural chemistry, these interactions may not be strong enough to stop enzyme leaching. With reference to the glucosidase adsorption on soil colloids as shown earlier [95], studies have shown that desorption of b-glucosidase from colloidal soil particles ranged from 17 to 20 % for fine colloids and around 28 % for coarse colloids. And this was even before the immobilized enzymes were subjected to hydrolysis. The extent of enzyme leaching during and after the hydrolysis cycle should be comparatively higher. In another study LamA, a type of endo-b-1,3-glucanase, was adsorbed on hydrophobic Teflon particles (215 nm) as well as hydrophilic silica nanoparticles (13 nm) [96]. Based on spectroscopic data and Km values, this study claims very little change occurred in the conformational entropy during the immobilization process making its contribution to the entropy of adsorption minimal. Surface coverage for Teflon and silica particles was reported to be 78 and 34 %, respectively. Hydrophobic interactions between the apolar Teflon and protein segments of LamA ensure significantly higher adsorption as compared to electrostatic forces acting at the interface of silica particles. Specific activity of immobilized enzymes was around 50 % for both Teflon and silica particles as compared to free enzymes in solution with the loss in catalytic potential being attributed to random orientation of the enzyme molecule on the surface of the support. A more quantitative study of the adsorption process is described elsewhere [97]. In this study, cellulase from Aspergillus niger was immobilized on activated carbon. Immobilization occurs both on the surface as well as within the micropores. Thermodynamic parameters were estimated to understand the interplay of temperature, equilibrium enzyme concentration, the distribution coefficient and their combined effect on cellulase adsorption. Negative value for change in Gibb’s free energy, positive change in enthalpy and entropy reinforce the view that adsorption of
1237
cellulose on activated carbon is a spontaneous process. A small positive value for entropy represents relatively lower randomness in the system which indicates that very little conformational changes occurred during the immobilization process. Enzyme activity tests revealed higher glucose production when immobilized enzymes with higher enzyme loadings were used to hydrolyze the substrate. In addition, the immobilized enzymes display good biocatalytic potential (around 70 %) when subjected to five cycles of reuse. The nature of surface charge (whether positive or negative) and the extent of charge distribution on the surface of proteins are strong functions of solution pH [112]. Figure 3 illustrates how enzyme molecules can exhibit different types of charges depending on the charge of individual protein segments. Hence, in such cases, a detailed examination of ionization states of the enzyme and the charge on the support/carrier needs to be carried out. Electrostatic adsorption is a versatile approach that can be easily extended to a broad spectrum of nanomaterials such super paramagnetic particles [98] as well as carbon nanotubes [99]. The benign nature of iron-oxide nanoparticles towards biological applications allows these particles to be used in several cellular and drug delivery applications with or without functionalization. Besides this, magnetic nanoparticles provide a facile route to recycle and reuse the immobilized enzyme over multiple cycles. Higher pH stability was reported when cellulase was electrostatically immobilized on magnetic nanoparticles [98]. However, after the cellulase attachment the size of the nanoparticles recorded more than a ten-fold increase indicating some degree of aggregation. Another example of electrostatic interaction used for immobilization of cellulolytic enzymes is the attachment of b-glucosidase to carbon nanotubes [99]. Carbon nanotube with its high surface area is another nanoscale material that can serve as a very good candidate for anchoring a host of biomolecules. Negatively charged carbon nanotubes were prepared by pretreatment with nitric acid prior to enzyme immobilization [99]. Under acidic pH, b-glucosidase has positively charged protein segments which can be electrostatically assembled on the anionic carbon nanotubes. However, activity reported after immobilization was about 20 % as compared to free enzymes. The possibility of formation of amide bond between the amino groups of the proteins and the carboxylic groups on the acid pretreated carbon nanotubes leading to conformational changes in the immobilized enzyme structure cannot be completely ruled out. The usefulness of long-chained polyelectrolyte brushes to immobilize biomolecules on inorganic or polymeric supports has been demonstrated by Caruso et al. [114, 115]. The use of polyelectrolytes creates a benign
123
1238
Fig. 3 Selection of suitable support depending on the charge of the protein segments on the surface of the enzyme molecule. Reproduced with permission from [112]
microenvironment for the enzymes to operate on the substrate of interest by reducing support-protein steric hindrance. In a related study, b-glucosidase was immobilized on latex beads using quenched and annealed brushes [100]. The general procedure adopted by the authors for immobilization is depicted in Fig. 4. The robustness of this approach allows immobilization of enzymes with different shapes and conformations such as glucoamylase (dumbbell shaped with flexible linker) as well as b-glucosidase (rigid globular structure). The b-glucosidase activity after immobilization was well preserved because the spatial structure of the enzyme Fig. 4 Immobilization of enzymes (b-glucosidase and glucoamylase) on polyelectrolyte brush functionalized latex nanoparticles. Reproduced with permission from [100]
123
Top Catal (2012) 55:1231–1246
was largely retained in the brush matrix. The long-chained polymer segments of the brushes ensure minimum protein– protein interaction thus ensuring no aggregation within the brush layers. Another proof of the benign nature of polyelectrolytes is that the Michaelis–Menten parameters Km and kcat for the immobilized glucosidase are very close to those of free enzymes. No remarkable increase was observed in the value of Km after immobilization indicating that the active sites of b-glucosidase are still readily accessible to the substrate [116]. Another significant aspect that needs careful monitoring during cellulase adsorption is the isoelectric point (pI) of the enzyme. Proteins in general, show better binding capability when the pH is close to pI. Charged enzymes molecules experience mutual repulsion between different protein segments when the solution pH is far away from the pI. Near the pI, the net charge on the enzyme molecule is almost zero because of minimum intramolecular electrostatic repulsion. This facilitates higher protein binding. Previous reports have confirmed the veracity of this theory [104]. 2.6 Covalent Binding of Cellulase to Nanomaterials The covalent immobilization of enzymes on solid supports is often considered as a safe route to minimize protein desorption. In order to extract the complete biocatalytic potential of the immobilized enzymes, most industrially relevant processes require protein leaching to be as low as possible. Immobilization of cellulases on solid supports can be carried out using different types of covalent chemistries. Commonly used binding agents include glutaraldehyde and carbodiimide derivatives such as1-ethyl-3-[3-dimethylaminopropyl]carbodiimide hydrochloride (EDAC) [117]. However, certain reactive groups are used up during the
Top Catal (2012) 55:1231–1246
formation of the covalent bond. The identification and location of these groups is of primary importance. Groups which form a part of the active site of the enzyme should be generally avoided for bond formation. The general mechanism for formation of EDAC mediated covalent bond [118] and glutaraldehyde cross-linking [119] is depicted in Fig. 5a, b, respectively. Covalent binding is usually preceded by surface modification of nanomaterials. Ho et al. [101] used a unique approach to synthesize PMMA-cellulase core–shell nanoenzymes. In this study, well defined poly(methyl methacrylate) (PMMA) cores covalently bound to cellulase were synthesized from mixture of methyl methacrylate (MMA) and cellulase by direct graft polymerization. The formation of these novel structures was guided by initiation of amine groups on the enzyme. Highly uniform and thick enzyme shells, good pH and temperature stability, and single step process to synthesize nanometer scale particles with narrow size distribution are some attractive features of this approach. Certain issues such as lower activity, high temperature involved in synthesis (80 °C) and development of a facile method to recycle and reuse the immobilized enzymes once addressed would make these core–shell nanoenzymes truly attractive candidates for cellulase immobilization. Cellulase immobilized on electrospun nanofibrous polyacrylonitrile (PAN) membranes has been reported recently [102]. Previous studies focused on complicated surface functionalization of PAN membranes followed by enzyme immobilization. However, Hung et al. [102] suggested a more convenient direct activation of nitrile groups on PAN membranes by amidination reaction as per procedure described elsewhere [120–122]. The activation time can influence the number of active sites on the PAN membrane to which the cellulase can covalently bind to. An activation time of 7.5 min was found optimum for cellulase immobilization especially since the structural stability of the PAN membrane is doubtful at longer activation times [120]. Individual cellulase molecules tend to occupy the active sites of the PAN membrane rather than stack over each other, thus ensuring that the catalytic centers of the enzyme enjoy continued access to the cellulosic substrate [102]. Besides using PAN membranes, use of polyvinyl alcohol (PVA) nanofibers as scaffolds for cellulase immobilization has been also reported in literature [103, 123]. Glutaraldehyde can covalently cross-link the –OH groups of PVA fibers with the amino groups of the proteins. Just like the activation time plays an important role in opening up active sites on PAN membrane, a similar comparison can be drawn regarding the glutaraldehyde cross-linking. Longer cross-linking time can reduce the biocatalytic activity of immobilized cellulase because then the various protein segments within the enzyme undergo a
1239
high degree of cross-linking blocking the active sites from accessing the substrate. Also, prolonged cross-linking can cause –OH groups of the PVA fibers to cross-link among themselves thus reducing the protein-PVA interaction. The use of gold nanoparticles for bioconjugation is quite common. Reliable techniques to synthesize gold nanoparticles starting from 2 to 250 nm, fairly high degree of colloidal stability over long times and a wide range of surface functionalization approaches available in literature have made these particles promising candidates for protein immobilization [124]. Gole et al. [105] showed the covalent immobilization of endoglucanse on gold nanoparticles with mean size of around 5 nm. A combination of thiolate linkages through cysteine residues as well as amine binding to gold nanoparticles is predicted to be the possible cause for cellulase attachment. The catalytic behavior of EG conjugated gold nanoparticles is very much similar to free enzymes with the optimum of almost 100 % recorded at a pH of 7 and a temperature of 60 °C. Cellulase has been covalently bound to nonporous silica nanoparticles [104] as well as iron-oxide nanoparticles [125, 126]. Using both physical adsorption as well as covalent cross-linking, Afsahi et al. [104] showed a quantitative comparison of both the methods by immobilizing cellulase on nonporous silica particles with an average particle size of 14.8 nm and surface area of 25 m2/gm. These results show for the given system under consideration, covalent cross-linking with glutaraldehyde provided better results as compared to physical adsorption. However enzyme activity in general was subdued for both the cases (not more than 35–40 %) indicating that the immobilization strategies used for cellulase attachment on silica particles may be responsible for constraining the mobility of the attached enzyme. To overcome the problem of enzyme mobility, Garcia et al. [126] suggested the use of long-chained spacer molecules to serve as a bridge between the cellulase enzyme and the support. Use of high molecular weight ligands such PVA or polyethylene glycol (PEG) helped to improve the biocatalytic activity of the immobilized cellulase. Another interesting approach suggested by Yoshimoto et al. [106] is using liposomes of various diameters to act as spacers between the chitosan beads and cellulase. Liposomes with mean diameters of 50, 100 and 200 nm with aldehyde terminal groups were used in this study. Protein loading was size dependent, with 50 nm liposomes showing least protein loading while 200 nm liposomes showing highest loading. However, 50 nm liposome showed the highest specific activity thus indicating that the size of the spacer and the orientation of the active sites of the enzyme during immobilization via a spacer can play a significant role in the biocatalytic activity. The same study also concluded that cellulase immobilized using 50 nm liposomes were best suited for multiple cycles of reuse since these
123
1240
Top Catal (2012) 55:1231–1246
Fig. 5 Mechanisms for formation of covalent bonds a EDAC mediated amide bond formation proceeds through formation of an unstable o-acylisourea ester (adapted with permission from [118], b Glutaraldehyde crosslinking of proteins occurs through formation of intermediate imine. Adapted with permission from [119]
liposomes were least susceptible to disruption during the course of the reaction. 2.7 Immobilization of Enzymes Within the Nanopores Another immobilization technique suggested in literature is entrapment or encapsulation of the enzyme within an inorganic or polymeric matrix. Enzyme immobilization
123
within porous materials can provide the same degree of stability to the immobilized enzymes as compared to covalent binding or physical adsorption [127]. Entrapment exploits the size difference between the substrate or the product molecule and the enzyme. The entrapping medium is so selected that it allows free flow of substrate from the bulk medium to the active site of the enzyme [128]. However, enzyme leaching is a common problem with this
Top Catal (2012) 55:1231–1246
Fig. 6 Enzyme immobilization within nanopores: hybrid bimodal mesoporous silica spheres used for enzyme encapsulation. Reproduced with permission from [129]
process since the mode of attachment is more physical than covalent. Ortega et al. [107] immobilized b-glucosidase from A. niger into alginate and polyacrylamide gels. The size of the pores as well as the charge on the gel matrix plays an important role in deciding the extent of immobilization followed by retention of activity. Wang and Caruso [129] reported the use of polyelectrolyte modified hybrid bimodal mesoporous silica spheres to immobilize a host of proteins and enzymes. Though not specially used for cellulase immobilization, the versatility of this approach as illustrated in Fig. 6 clearly shows that this method can be easily extended to cellulase immobilization. In order to study the effect of pore size on cellulase immobilization, Takimoto et al. [108] used mesoporous Santa Barbara amorphous silica-15 (SBA-15) to immobilize cellulase. The pore size of the silica was varied from 5.4 to 11 nm. The authors conclude that the protein loading on silica particles not only depends on the relative surface area available for immobilization but also the pore size of the particles. The authors claim that the enzyme activity for silica materials with pore size 8.9 nm is highest because this size closely matches the dimensions of the cellulase enzyme used in this study. At this pore size the enzyme is immobilized at the very opening of the pore allowing better accessibility to the substrate molecules. Nitrogen adsorption studies confirm this hypothesis. In another study, both functionalized and unfunctionalized porous face centered cubic mesoporous silica materials called FDU-12 produced using a technique developed at the Fudan University, China
1241
were used to immobilize cellulase [109]. The immobilized cellulase activity was not just the function of the pore dimensions but also the type of organosilane used for the synthesis of the mesoporous silica. Organo-functionalized silica showed lower enzyme leaching as compared to nonfunctionalized ones thus enhancing the operational stability. It has been noted that the extent of cellulase adsorption on SBA-15 is limited because of its 2D hexagonal shape [110]. Amine functionalized FDU-12, on the other hand, can immobilize a significantly higher quantity of cellulase among the FDU class supports yet showed lower catalytic activity [110]. This is because cellulases typically have their active centers located in aspartic and glutamate acid residues [130]. Hence, the formation of a covalent bond between the amine groups of the aminopropyltriethoxysilane (APTES) functionalized silica materials with the carboxylic groups of the catalytic centers can lead to subdued enzyme activity. Vinyl functionalized FDU is another type of support that has been investigated for cellulase immobilization. The highest activity achieved using SBA supports tends to be closer to 70 % [108] whereas when using the vinyl functionalized FDU supports, it is 71–90 % [109]. However, the two results are not comparable because of the difference in the nature of substrates used. The activity of cellulase immobilized on SBA-15 supports was tested using microcrystalline cellulose which is an insoluble substrate. On the other hand, cellulase immobilized on FDU-12 supports was used to hydrolyze CMC, a semi-soluble form of cellulose. Insoluble substrates encounter greater difficulty in diffusing through the porous structure thus severely limiting enzyme-substrate interactions. Soluble substrates on the other hand, can freely access the enzyme entrapped in the porous silica support. In addition to the diffusion effects which are inherent in these systems, cellulase immobilization within a porous matrix also depends on the nature of interaction between the enzyme and the immobilizing support. The result is a synergistic effect where a strong interplay of diffusion, mode of attachment, resultant stability and enzyme activity is commonly observed. A recent study tries to understand this combinational effect by immobilizing cellulase on porous silica nanoparticles with varying pore dimensions and by using different types of bonding chemistries [110]. Cellulase was physically adsorbed on both small pore mesoporous silica nanoparticles (SPMSN) and large pore mesoporous silica nanoparticles (LPMSN). The different methods used in the synthesis of these materials give rise to different functional groups on the surface. As a result, SPMSN carry a net negative charge whereas LPMSN carry zero to mildly positive charge. The net negative charge on cellulase at the immobilization pH along with the small size of the pores (2–4 nm) inhibits the electrostatic adsorption of proteins. LPMSN, however, performs better
123
1242
in terms of adsorption because of lower level of mutual repulsion between the support and the adsorbed proteins. The study also reveals that covalent functionalized LPMSN yielded activities upward of 80 % even though the same type of chemistry described elsewhere showed subdued cellulase [109]. Better characterization techniques are therefore needed to identify whether the functional groups of cellulase participating in the immobilization process belong to the CD or the CBD region of the enzyme in order to explain the deviation in the results. Similarly, ensuring substrate compatibility would also help in a more robust comparison. 2.8 New Advances in Cellulase Technology and Its Implication on Immobilization 2.8.1 Cellulosome Unlike their aerobic counterparts that secrete the regular cellulase or hemicellulase, anaerobic microbes (cellulosomal microbes) during the course of their evolution developed novel enzyme architecture integrating multiple enzymatic sub-units to degrade highly crystalline cellulosic plant cells in a more efficient way [131]. The presence of dockerin modules and attachment of CBDs to the scaffoldin sub-unit are some of the key features of cellulosomal enzymes. The scaffoldin sub-units act as vehicles for efficient assembly of various enzyme modules with different sub-units joined by flexible linkers. Thus, while in regular noncellulosomal enzymes, each enzyme sub-unit usually comes with a CBD to bind to the substrate, cellulosomes frequently incorporate multiple CBDs within one scaffoldin sub-unit [131, 132]. Instead of individual enzymes binding to the substrate, all the enzyme modules on the same scaffoldin sub-unit are bound to the cellulosic substrate through the common scaffoldin sub-unit. In addition few enzyme modules may have separate CBDs as well. The difference between the non-cellulosomal enzymes and cellulosomal enzymes can be understood from Figs. 1 and 6. Figure 1 shows how each enzyme unit (CBH, EG and b-glucosidase) participate in the cellulosic hydrolysis by binding and attacking substrates/intermediate products whereas Fig. 7 shows the modular structure of cellulosomes with multiple CDs attached to the scaffoldin subunit through cohesion-dockerin interactions. The binding of CBD on cellulosic substrate has been an area of intense research. Studies have identified the mechanism and the aromatic residues responsible for the CBD binding on the cellulosic substrate [134, 135]. Plant biomass displays a heterogeneous structure with some portions being highly crystalline while some others showing with high degree of soluble glycan chains [131, 136]. Likewise, the CBDs also display preferential binding
123
Top Catal (2012) 55:1231–1246
ability in response to the difference in the nature of plant cell-wall. After the scaffoldin is bound to the plant biomass depending on the preferential affinity of the CBDs, it is hypothesized that cellulosome undergoes conformational changes, which includes reorientation of its CDs to accommodate for the substrate specific degradation. The highly specialized cohesion-dockerin interaction, one of the strongest protein–protein interlocking forces, is another key feature of cellulosomes [137, 138]. The unique plugsocket design of the cohesion-dockerin assembly ensures that the CDs of the enzyme remain fused to the main scaffoldin sub-unit. Cellulosomes are nature’s answer to improved cellulose degradation. The ability to blend several enzyme-sub units together using the same scaffoldin is a powerful tool to achieve synergy at the nanometer scale. Looking in terms of enzyme immobilization, each scaffolding sub-unit represents a ‘nanostructured support’ on which several enzyme units are ‘immobilized’ using the cohesin-dockerin arrangement. In the previous sections, the significance of the use of spacer arms such as polyelectrolyte brushes and high molecular weight polymer chains to reduce steric hindrance and improve enzyme-substrate accessibility has been highlighted. The cohesin-dockerin assembly also serves the same purpose. Moreover, most of the scaffoldin units come with their CBDs which enable efficient adsorption of the cellulosome on to the cellulosic substrate. 2.8.2 Artificial or ‘Designer’ Cellulosomes The use of cellulosomes as nanostructured scaffolds has found support within the scientific community in recent years. Efforts are being made to engineer ‘designer cellulosomes’ to accommodate the various types of dockerincohesin modules as well as CDs and CBDs belonging to different families on the scaffoldin units [139, 140]. In yet another case, the CDs and the CBDs were assembled on streptavidin and streptavidin functionalized inorganic nanoparticles [141]. Enzymes immobilized on curved supports are observed to have a higher biocatalytic activity [142, 143]. Thus by attaching CDs and CBDs belonging to different protein families on cadmium selenide (CdSe) nanoparticles, Kim et al. [141] effectively used the curved surface area of the CdSe nanoparticle for bioconjuagation. The illustration of the same is provided in Fig. 8. Highly clustered CBDs on CdSe nanoparticles helped achieve more than seven fold increase in the biocatalytic activity as compared to free native enzymes. In case of CD and CBD assembled on streptavidin, improved results were obtained when an insoluble amorphous substrate was used in place of soluble substrate. Biotin-avidin interactions, one of the most commonly used noncovalent forces in binding biomolecules were utilized for effective clustering of CDs
Top Catal (2012) 55:1231–1246
1243
significant improvement in the enzyme activity. Heteroclustering of both CDs as well as CBDs helped improve the catalytic efficiency for amorphous and microcrystalline substrates. From the Lineweaver–Burke plot, the authors observed that the clustering of CDs and CBDs on inorganic nanoparticles is responsible for enhanced substrate-enzyme interactions. On the other hand, the immobilization of CDs and CBDs on streptavidin improves the kinetic rate of the process. Thus the support used for immobilization of enzymatic modules plays a significant role in the mechanism of activity enhancement.
3 Conclusion
Fig. 7 Simplified scheme showing different components of a cellulosome: scaffoldin sub-unit includes multiple cohesion modules and a C-terminal divergent dockerin for increased interactions between the plant cell-wall and bacterial cell envelope. CDs and CBDs are anchored on scaffoldin through cohesion-dockerin interactions. Reproduced with permission from [133]
Fig. 8 Artificial/‘Designer’ cellulosomes: catalytic/CBD modules expressed by different families of cellulolytic microbes assembled on streptavidin and functionalized nanoparticles to form hybrid cellulosomes. Reproduced with permission from [141]
and CBD modules on streptavidin and CdSe nanoparticles. An important observation made in this study was that the mere homoclustering of CDs may not always provide a
Nanobiotechnology has a unique potential to engineer a plethora of bioconjugated nanoscale materials with novel functional properties. In order to make optimum use of available resources in the most compact manner, the world is moving rapidly towards developing hybrid technologies that support miniaturization in a big way. Nanobiotechnology is well positioned to deliver these technologies and is already making significant inroads in certain specialized industries such as in production of fine chemicals, drug delivery and many more. The application of cellulase immobilization on nanostructured supports is yet another example. Here, in this review various traditional aspects of cellulase immobilization were examined including the nature, shape and size of support used the different types of interactions, cross-linking agents, incubation time and their collective effect on the biocatalytic activity of the enzyme. Inspite of attractive features such as remarkable dispersability, high aspect ratio, relative ease of functionalization, high degree of robustness, easy application to various forms of reactor configuration and opportunities to recycle, the use of nanomaterials as scaffolds for cellulase immobilization is not commonly seen in the bioethanol industry. Two typical reasons are preventing the ascent of this technology. Firstly, concerns regarding the toxicological effects of nanomaterials and a perception that large scale, cost effective production of these materials may not be feasible has led to a considerable delay. Secondly, with the focus shifting to engineering improved varieties of cellulolytic microbes using the newly developed techniques of proteomics, investment in immobilization techniques has relatively gone down. While concerns regarding handling of nanomaterials in an industrial setting are not unfounded, it is hard to ignore the obvious advantages. Similarly, immobilization should not be treated as a rival technology to the developing field of cellulase engineering. Instead both technologies are complementary. An excellent example of assimilation of both immobilization and proteomics presented in this review is the development of
123
1244
artificial cellulosomes, more specifically the assembly of various enzymatic modules on streptavidin and CdSe nanoparticles. With cellulases projected to soon become the second largest industrially used enzymes, the prospect of such examples incorporating fusion of these technologies appear bright. Acknowledgments The funding from the Michigan University Research Corridor and the Michigan Initiative for Innovation and Entrepreneurship and in part from the National Science Foundation (0609164, 0832730) to support this research is greatly appreciated.
References 1. Altreuter DH, Clark DS (1999) Curr Opin Biotechnol 10: 130–136 2. Marrs B, Delagrave S, Murphy D (1999) Curr Opin Microbiol 2:241–245 3. Organisation de coope´ration et de de´veloppement e´conomiques. La biotechnologie au service de produits et de proce´de´s industriels propres vers un de´veloppement industriel durable. Organisation de coope´ration et de de´veloppement e´conomiques, Paris, 1998 4. Kuchner O, Arnold FH (1997) Trends Biotechnol 15:523–530 5. Jaeger KE, Reetz MT (2000) Curr Opin Chem Biol 4:68–73 6. Schmid A, Dordick JS, Hauer B, Kiener A, Wubbolts M, Witholt B (2001) Nature 409:258–268 7. Schoemaker HE, Mink D, Wubbolts MG (2003) Science 299:1694–1697 8. van Beilen JB, Li Z (2002) Curr Opin Biotechnol 13:338–344 9. Schulze B, Wubbolts MG (1999) Curr Opin Biotechnol 10:609–615 10. Organisation for Economic Co-operation and Development (2001) The application of biotechnology to industrial sustainability. OECD Publishing 11. Holloway P, Knoke KL, Trevors JT, Lee H (1998) Biotechnol Bioeng 59:520–523 12. Igarashi S, Hirokawa T, Sode K (2004) Biomol Eng 21:81–89 13. Bornscheuer UT (2005) Trends and challenges in enzyme technology, In: Biotechnology for the future, pp 181–203 14. DeSantis G, Wong K, Farwell B, Chatman K, Zhu Z, Tomlinson G, Huang H, Tan X, Bibbs L, Chen P, Kretz K, Burk MJ (2003) J Am Chem Soc 125:11476–11477 15. Bornscheuer UT, Altenbuchner J, Meyer HH (1998) Biotechnol Bioeng 58:554–559 16. Hassler BL, Amundsen TJ, Zeikus JG, Lee I, Worden RM (2008) Biosens Bioelectron 23:1481–1487 17. Hassler BL, Kohli N, Zeikus JG, Lee I, Worden RM (2007) Langmuir 23:7127–7133 18. Holtzapple M, Cognata M, Shu Y, Hendrickson C (1990) Biotechnol Bioeng 36:275–287 19. Kohli N, Dvornic PR, Kaganove SN, Worden RM, Lee I (2004) Macromol Rapid Commun 25:935–941 20. Kohli N, Hassler BL, Parthasarathy L, Richardson RJ, Ofoli RY, Worden RM, Lee I (2006) Biomacromolecules 7:3327–3335 21. Kohli N, Lee I, Richardson RJ, Worden RM (2010) J Electroanal Chem 641:104–110 22. Kohli N, Srivastava D, Sun J, Richardson RJ, Lee I, Worden RM (2007) Anal Chem 79:5196–5203 23. Kohli N, Vaidya S, Ofoli RY, Worden RM, Lee I (2006) J Colloid Interface Sci 301:461–469 24. Kohli N, Worden RM, Lee I (2005) Chem Commun 316–318
123
Top Catal (2012) 55:1231–1246 25. Kohli N, Worden RM, Lee I (2007) Macromol Biosci 7:789–797 26. Zhang LM, Xia JG, Zhao QH, Liu LW, Zhang ZJ (2010) Small 6:537–544 27. Ferrari M (2005) Nat Rev Cancer 5:161–171 28. Lu J, Do I, Drzal LT, Worden RM, Lee I (2008) ACS Nano 2:1825–1832 29. Lu J, Drzal LT, Worden RM, Lee I (2007) Chem Mater 19:6240–6246 30. Ahn JS, Hammond PT, Rubner MF, Lee I (2005) Colloids and Surf A Physicochem Eng Asp 259:45–53 31. Ahn JS, Hendricks TR, Lee I (2007) Adv Funct Mater 17:3619–3625 32. Kidambi S, Chan C, Lee I (2008) Langmuir 24:224–230 33. Kidambi S, Chan C, Lee I (2004) J Am Chem Soc 126: 4697–4703 34. Kidambi S, Lee I, Chan C (2004) J Am Chem Soc 126: 16286–16287 35. Kidambi S, Lee I, Chan C (2008) Adv Funct Mater 18:294–301 36. Kidambi S, Sheng LF, Yarmush ML, Toner M, Lee I, Chan C (2007) Macromol Biosci 7:344–353 37. Kidambi S, Udpa N, Schroeder SA, Findlan R, Lee I, Chan C (2007) Tissue Eng 13:2105–2117 38. Mehrotra S, Hunley SC, Pawelec KM, Zhang LX, Lee I, Baek S, Chan C (2010) Langmuir 26:12794–12802 39. Mehrotra S, Lee I, Chan C (2009) Acta Biomater 5:1474–1488 40. Mehrotra S, Lee I, Liu C, Chan C (2011) Ind Eng Chem Res 8851–8858 41. Mehrotra S, Lynam D, Maloney R, Pawelec KM, Tuszynski MH, Lee I, Chan C, Sakamoto J (2010) Adv Funct Mater 20:247–258 42. Lee I, Wool RP (2000) Thin Solid Films 379:94–100 43. Lee I, Wool RP (2000) Macromolecules 33:2680–2687 44. Lee I, Wool RP (2001) J Adhes 75:299–324 45. Lee I, Wool RP (2002) J Polym Sci B Polym Phys 40:2343–2353 46. Srivastava D, Hendricks TR, Lee I (2007) Nanotechnology 18:245305–245310 47. Srivastava D, Lee I (2006) Adv Mater 18:2471 48. Dandavate V, Keharia H, Madamwar D (2009) Process Biochem 44:349–352 49. Nagayama K, Yamasaki N, Imai M (2002) Biochem Eng J 12:231–236 50. Kim J, Grate JW, Wang P (2006) Chem Eng Sci 61:1017–1026 51. Lundqvist M, Sethson I, Jonsson B-H (2004) Langmuir 20:10639–10647 52. Jia HF, Zhu GY, Wang P (2003) Biotechnol Bioeng 84:406–414 53. Lynd LR (1996) Annu Rev Energy Env 21:403–465 54. Medve J, Karlsson J, Lee D, Tjerneld F (1998) Biotechnol Bioeng 59:621–634 55. Galbe M, Zacchi G (2002) Appl Microbiol Biotechnol 59: 618–628 56. Tu MB, Chandra RP, Saddler JN (2007) Biotechnol Prog 23: 398–406 57. Lee D, Yu AHC, Saddler JN (1995) Biotechnol Bioeng 45: 328–336 58. Cao LQ, van Langen L, Sheldon RA (2003) Curr Opin Biotechnol 14:387–394 59. Sheldon RA (2007) Adv Synth Catal 349:1289–1307 60. Himmel ME, Ruth MF, Wyman CE (1999) Curr Opin Biotechnol 10:358–364 61. Pringsheim H (1912) Hoppe-Seyler’s Z Physiol Chem 78:266–291 62. Reese ET, Siu RGH, Levinson HS (1950) J Bacteriol 59: 485–497 63. Dashtban M, Schraft H, Qin WS (2009) Int J Biol Sci 5:578–595 64. Maki M, Leung KT, Qin WS (2009) Int J Biol Sci 5:500–516
Top Catal (2012) 55:1231–1246 65. Sinnott ML (1998) Biochem Soc Trans 26:160–164 66. Moreira N (2005) Sci News 168:218 67. Finkelstein M, Davison BH (2000) Twenty-first symposium on biotechnology for fuels and chemicals. Humana, Totowa, 2–6 May 1999 68. Henrissat B (1994) Cellulose 1:169–196 69. Gilkes NR, Warren RAJ, Miller RC, Kilburn DG (1988) J Biol Chem 263:10401–10407 70. Mosier N, Hall P, Ladisch C, Ladisch M, Tsao G, Brainard A, Bungay H, Cao N, Cen P, Chen Z, Du J, Foody B, Gong C, Hall P, Ho N, Irwin D, Iyer P, Jeffries T, Ladisch C, Ladisch M, Lee Y, Mosier N, Mu¨hlemann H, Sedlak M, Shi N, Tsao G, Tolan J, Torget R, Wilson D, Xia L (1999) Reaction kinetics, molecular action, and mechanisms of cellulolytic proteins recent progress in bioconversion of lignocellulosics. Springer, Heidelberg, pp 23–40 71. Kraulis PJ, Clore GM, Nilges M, Jones TA, Pettersson G, Knowles J, Gronenborn AM (1989) Biochemistry 28:7241–7257 72. Rouvinen J, Bergfors T, Teeri T, Knowles JKC, Jones TA (1990) Science 249:380–386 73. Schmuck M, Pilz I, Hayn M, Esterbauer H (1986) Biotechnol Lett 8:397–402 74. Abuja PM, Schmuck M, Pilz I, Tomme P, Claeyssens M, Esterbauer H (1988) Eur Biophys J Biophys Lett 15:339–342 75. Abuja PM, Pilz I, Claeyssens M, Tomme P (1988) Biochem Biophys Res Commun 156:180–185 76. Teeri T, Reinikainen T, Ruohonen L, Jones TA, Knowles JKC (1992) J Biotechnol 24:169–176 77. Knowles J, Lehtovaara P, Teeri T (1987) Trends Biotechnol 5:255–261 78. Wyman C (1996) Handbook on bioethanol: production and utilization. Taylor & Francis, Washington, DC 79. Black GW, Rixon JE, Clarke JH, Hazlewood GP, Ferreira LMA, Bolam DN, Gilbert HJ (1997) J Biotechnol 57:59–69 80. Shen H, Schmuck M, Pilz I, Gilkes NR, Kilburn DG, Miller RC, Warren RAJ (1991) J Biol Chem 266:11335–11340 81. Srisodsuk M, Reinikainen T, Penttila M, Teeri TT (1993) J Biol Chem 268:20756–20761 82. Saloheimo M, NakariSetala T, Tenkanen M, Penttila M (1997) Eur J Biochem 249:584–591 83. Karlsson J, Medve J, Tjerneld F (1999) Appl Biochem Biotechnol 82:243–258 84. Fagerstam LG, Pettersson LG (1980) FEBS Lett 119:97–100 85. Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS (2002) Microbiol Mol Biol Rev 66:506 86. Tomme P, Warren RAJ, Gilkes NR, Poole RK (1995) Cellulose hydrolysis by bacteria and fungi. Advances in microbial physiology. Academic Press, London, pp 1–81 87. Woodward J (1991) Bioresour Technol 36:67–75 88. Dashtban M, Maki M, Leung KT, Mao CQ, Qin WS (2010) Crit Rev Biotechnol 30:302–309 89. Andric P, Meyer AS, Jensen PA, Dam-Johansen K (2010) Biotechnol Adv 28:308–324 90. Wilson DB (2009) Curr Opin Biotechnol 20:295–299 91. Gebler J, Gilkes NR, Claeyssens M, Wilson DB, Beguin P, Wakarchuk WW, Kilburn DG, Miller RC, Warren RAJ, Withers SG (1992) J Biol Chem 267:12559–12561 92. Bayer EA, Chanzy H, Lamed R, Shoham Y (1998) Curr Opin Struct Biol 8:548–557 93. Davies G, Henrissat B (1995) Structure 3:853–859 94. Serefoglou E, Litina K, Gournis D, Kalogeris E, Tzialla AA, Pavlidis IV, Stamatis H, Maccallini E, Lubomska M, Rudolf P (2008) Chem Mater 20:4106–4115 95. Yan JL, Pan GX, Li LQ, Quan GX, Ding C, Luo AL (2010) J Colloid Interface Sci 348:565–570 96. Koutsopoulos S, van der Oost J, Norde W (2005) Biophys J 88:467–474
1245 97. Daoud FBO, Kaddour S, Sadoun T (2010) Colloids Surf B Biointerfaces 75:93–99 98. Khoshnevisan K, Bordbar AK, Zare D, Davoodi D, Noruzi M, Barkhi M, Tabatabaei M (2011) Chem Eng J 171:669–673 99. Gomez JM, Romero MD, Fernandez TM (2005) Catal Lett 101:275–278 100. Haupt B, Neumann T, Wittemann A, Ballauff M (2005) Biomacromolecules 6:948–955 101. Ho KM, Mao XP, Gu LQ, Li P (1042) Langmuir 24: 11036–11042 102. Hung TC, Fu CC, Su CH, Chen JY, Wu WT, Lin YS (2011) Enzym Microb Technol 49:30–37 103. Wu LL, Yuan XY, Sheng J (2005) J Membr Sci 250:167–173 104. Afsahi B, Kazemi A, Kheirolomoom A, Nejati S (2007) Sci Iran 14:379–383 105. Gole A, Vyas S, Phadtare S, Lachke A, Sastry M (2002) Colloids Surf B Biointerfaces 25:129–138 106. Yoshimoto M, Li CZ, Matsunaga T, Nakagawa H, Fukunaga K, Nakao K (2006) Biotechnol Prog 22:459–464 107. Ortega N, Busto MD, Perez-Mateos M (1998) Bioresour Technol 64:105–111 108. Takimoto A, Shiomi T, Ino K, Tsunoda T, Kawai A, Mizukami F, Sakaguchi K (2008) Microporous Mesoporous Mater 116:601–606 109. Hartono SB, Qiao SZ, Liu J, Jack K, Ladewig BP, Hao ZP, Lu GQM (2010) J Phys Chem C 114:8353–8362 110. Chang RHY, Jang J, Wu KCW (2011) Green Chem 13:2844–2850 111. Jegannathan KR, Abang S, Poncelet D, Chan ES, Ravindra P (2008) Crit Rev Biotechnol 28:253–264 112. Hanefeld U, Gardossi L, Magner E (2009) Chem Soc Rev 38:453–468 113. Sinegani AAS, Emtiazi G, Shariatmadari H (2005) J Colloid Interface Sci 290:39–44 114. Caruso F, Trau D, Mohwald H, Renneberg R (2000) Langmuir 16:1485–1488 115. Caruso F, Niikura K, Furlong DN, Okahata Y (1997) Langmuir 13:3427–3433 116. Xia JL, Mattison K, Romano V, Dubin PL, Muhoberac BB (1997) Biopolymers 41:359–365 117. Makhongela H, Glowacka A, Agarkar V, Sewell B, Weber B, Cameron R, Cowan D, Burton S (2007) Appl Microbiol Biotechnol 75:801–811 118. Nakajima N, Ikada Y (1995) Bioconjug Chem 6:123–130 119. Walt DR, Agayn VI (1994) Trends Anal Chem 13:425–430 120. Li SF, Chen JP, Wu WT (2007) J Mol Catal B Enzym 47:117–124 121. Li SF, Wu WT (2009) Biochem Eng J 45:48–53 122. Yajima H, Hirose A, Ishii T, Ohsawa T, Endo R (1989) Biotechnol Bioeng 33:795–798 123. Kim Byoung C, Nair S, Kim Seong H, Sang Byoung I, Kim J, Gu Man B (2008) Enzyme-nanofiber composites for biocatalysis applications In: Biomolecular catalysis. American Chemical Society, pp 254–262 124. Wilson R (2008) Chem Soc Rev 37:2028–2045 125. Jordan J, Kumar CSSR, Theegala C (2011) J Mol Catal B Enzym 68:139–146 126. Garcia A, Oh S, Engler CR (1989) Biotechnol Bioeng 33:321–326 127. Kennedy JF, Melo EHM, Jumel K (1990) Chem Eng Prog 86:81–89 128. Brash JL (1996) Interfacial phenomena and bioproducts. Dekker, New York 129. Wang YJ, Caruso F (2005) Chem Mater 17:953–961 130. Bhat MK, Bhat S (1997) Biotechnol Adv 15:583–620 131. Bayer EA, Belaich J-P, Shoham Y, Lamed R (2004) Annu Rev Microbiol 58:521–554
123
1246 132. Bayer EA, Morag E, Lamed R (1994) Trends Biotechnol 12:379–386 133. C.M.G.A. Fontes, H.J. Gilbert, Annual Review of Biochemistry, 79 655-681 134. Lehtio J, Sugiyama J, Gustavsson M, Fransson L, Linder M, Teeri TT (2003) Proc Nat Acad Sci USA 100:484–489 135. Shimon LJW, Pages S, Belaich A, Belaich JP, Bayer EA, Lamed R, Shoham Y, Frolow F (2000) Acta Crystallogr Sect D Biol Crystallogr 56:1560–1568 136. Gilbert HJ, Davies GJ, Henrissat B, Svensson BHE, Royal Society of Chemistry (GB) (2000) Recent advances in carbohydrate bioengineering [Texte imprime´]: Proceedings of the 3rd carbohydrate bioengineering meeting held at the University of Newcastle upon Tyne on 11–14 April 1999. Royal Society of Chemistry, London
123
Top Catal (2012) 55:1231–1246 137. Mechaly A, Fierobe HP, Belaich A, Belaich JP, Lamed R, Shoham Y, Bayer EA (2001) J Biol Chem 276:9883–9888 138. Fierobe HP, Mechaly A, Tardif C, Belaich A, Lamed R, Shoham Y, Belaich JP, Bayer EA (2001) J Biol Chem 276:21257–21261 139. Mingardon F, Chanal A, Tardif C, Bayer EA, Fierobe HP (2007) Appl Environ Microbiol 73:7138–7149 140. Caspi J, Irwin D, Lamed R, Shoham Y, Fierobe H-P, Wilson DB, Bayer EA (2006) Biocatal Biotransform 24:3–12 141. Kim DM, Umetsu M, Takai K, Matsuyama T, Ishida N, Takahashi H, Asano R, Kumagai I (2011) Small 7:656–664 142. Kim J, Jia HF, Wang P (2006) Biotechnol Adv 24:296–308 143. Wang P, Dai S, Waezsada SD, Tsao AY, Davison BH (2001) Biotechnol Bioeng 74:249–255
