Fluorescent Cellulose Microfibrils As Substrate for the Detection of

Station Biologique de Roscoff, UMR 1931 (CNRS and Laboratoires Goëmar), Place Georges Teissier, BP 74, 29682 Roscoff Cedex, France, Centre de ...
0 downloads 0 Views 194KB Size
Biomacromolecules 2003, 4, 481-487

481

Articles Fluorescent Cellulose Microfibrils As Substrate for the Detection of Cellulase Activity† William Helbert,‡,§ Henri Chanzy,§ Tommy Lykke Husum,| Martin Schu¨lein,| and Steffen Ernst*,| Station Biologique de Roscoff, UMR 1931 (CNRS and Laboratoires Goe¨mar), Place Georges Teissier, BP 74, 29682 Roscoff Cedex, France, Centre de Recherches sur les Macomole´ cules Ve´ ge´ tales (CERMAV-CNRS) (affiliated with the Joseph Fourier University, Grenoble, France), BP 53, 38041 Grenoble Cedex 9, France, and Novozymes A/S, Smoermosevej 11, DK-2880 Bagsvaerd, Denmark Received June 27, 2002

To devise a sensitive cellulase assay based on substrates having most of the physical characteristics of native cellulose, 5-(4,6-dichlorotriazinyl)aminofluorescein (DTAF) was used as a grafting agent to prepare suspensions of fluorescent microfibrils from bacterial cellulose. These suspensions were digested by a series of commercially relevant cellulases from Humicola insolens origin: cloned Cel6B and Cel 45A as well as crude H. insolens complex. The digestion induced the release of fluorescent cellodextrins as well as reducing sugars. After adequate centrifugation, these soluble products were analyzed as a function of grafting content, digestion time, and cellulase characteristics. The resulting data allowed the grafting conditions to be optimized in order to maximize the quantity of soluble products and therefore to increase the sensitivity of the detection. A comparison between the amount of released fluorescence and that of released reducing sugar allowed the differentiation between processive exo and endo cellulase activities. The casting of films of DTAF-grafted microfibrils at the bottom of the microwell titer plates also led to sensitive cellulase detection. As these films kept their integrity and remained firmly glued to the well bottom during the digestion time, they are tailored made for a full automation of the cellulases testing. Introduction Cellulases are used in laundry detergent and in the textile and in the paper industry; however, the assays that are available for their characterization are mostly tedious or based on cellulose substrate in a non-natural form. Cellulases catalyze the hydrolysis of the β(1f4) glycosidic bonds linking the glucosyl units of cellulose. They are functionally categorized by their mode of action as cellobiohydrolases (EC 3.2.1.91, which release the disaccharide cellobiose from the nonreducing ends of the cellulose chains) and endoglucanases (EC 3.2.1.4, which cleave endolytically). On the basis of amino acid sequence similarities, cellulases can also be classified in specific families, each of them containing a number of cellobiohydrolases and endoglucanases.1-4 Cellulose is the ubiquitous structural component of a plant cell wall, but it is also observed in some species of a wide spectrum of living organisms, namely, bacteria, fungi, algae, * To whom correspondence may be addressed. E-mail: sffe@ novozymes.com. † In memory of Martin Schu ¨ lein, who passed away August 2001. ‡ Station Biologique de Roscoff, UMR 1931 (CNRS and Laboratoires Goe¨mar). § Centre de Recherches sur les Macomole ´ cules Ve´ge´tales (CERMAVCNRS). | Novozymes A/S.

amoebae, and sea animals. In these, cellulose occurs naturally as insoluble crystalline microfibrils. The wide morphological and structural diversity of cellulose microfibrils, their hierarchical organization, and their low accessibility and reactivity have always presented some problem for the biochemical analysis of cellulase activity. Several strategies have been used to characterize the mode of action of cellulases, ranging from the use of soluble cellulosic derivatives, that of model compounds, or the use of amorphous or microcrystalline cellulose. Soluble substrates such as carboxymethyl cellulose (CMC) are examples of compounds that are very useful to identify endoglucanase activities by following the rapid loss of viscosity of CMC solutions. On the other hand, the cellobiohydrolase activity can be followed by the detection of reducing sugar after digestion of underivatized cellulose. Despite their simplicity, these methods have the disadvantage of using substrates that are far from native cellulose in terms of chemical and/or physical characteristics. Well-characterized dispersions of cellulose microfibrils from algae or bacteria have been described as more reliable substrates for the characterization of cellulases because they contain most of the structural and morphological character of “real” cellulose materials. In addition, their dispersed state

10.1021/bm020076i CCC: $25.00 © 2003 American Chemical Society Published on Web 03/27/2003

482

Biomacromolecules, Vol. 4, No. 3, 2003

Helbert et al.

Materials and Methods

Figure 1. Chemical structure of 5-(4,6-dichlorotriazinyl)aminofluorescein (DTAF).

makes them quite effective as it eliminates the problem of substrate accessibility, which creates difficulties for the understanding of the various actions of cellulases. In this respect, the direct visualization of digested cellulose microfibrils by transmission electron microscopy (TEM) has allowed an accurate accounting for the mode of action of the enzymes.5-9 The resulting images have indicated that endoglucanases preferentially degrade amorphous or less organized parts of the microfibrils, which is consistent with their endo character. On the other hand, crystalline cellulose microfibrils were thinned down by the action of cellobiohydrolases: this is interpreted as the result of the processive character of these enzymes. Despite the advantage of giving clear-cut fundamental description of the cellulase actions, the TEM technique not only is labor intensive but also requires fairly large amounts of purified enzymes to reveal the effect of a given cellulase on model cellulose substrates. More practical approaches have therefore to be developed if one wants to screen quickly and efficiently the presence of cellulases in a large number of purified or crude enzyme extracts. With this goal in mind, we have implemented the preparation of fluorescent cellulose microfibrils in view of detecting minute amounts of endoglucanase and cellobiohydrolase. In principle, endoglucanase activity should result in the solubilization of fluorescent cellodextrins in the medium. Also, if the derivatization is not too important, cellobiohydrolase should be detectable by measuring the reducing end of the released soluble sugars. Fluorescent dyes bearing mono- or dichlorotriazinyl group such as Procion dyes are known to react with the hydroxyl groups of polysaccharides.10,11 A less known fluorescent dye is DTAF (5-(4,6-dichlorotriazinyl)aminofluorescein) (Figure 1) which combines the reactivity of dichlorotriazinyl group with the fluorescent properties of fluorescein. To our knowledge, this commercial dye has not been used to label any polysaccharide, with the exception of dextran for which a fluorescent DTAF derivative has been reported.12 DTAF was therefore a good candidate for the direct grafting of a fluorescent probe at the surface of cellulose microfibrils. The aim of this paper is to describe the development of a sensitive assay for determining cellulase activity toward slightly modified cellulose that mimics “natural” cellulose. At first, we delineate conditions of labeling on native cellulose with DTAF that do not modify the physical integrity of the microfibrils. Then, these fluorescent substrates proved to be very useful for the detection and the characterization of minute amount of cellulases.

Microfibrils of Bacterial Cellulose. Cubes of bacterial cellulose from Nata de cocosa food grade commercial cellulose (Fujico, Kobe, Japan)swere used throughout. The cubes were extensively washed with tap water in order to remove the flavored sweet syrup before being homogenized in a Waring blender operated at full speed for 10 min. The resulting cellulose suspension was resuspended twice in 1% NaOH by centrifugation and kept overnight in the alkali solution under mild stirring at room temperature. The cellulose suspension was neutralized by at least three centrifugations and redispersions in distilled water. The resulting suspension was deproteinized by treating it at 70 °C for 2 h with an excess of bleaching solution consisting of 1 volume of 1.7% aqueous NaClO2 and 1 volume of acetate buffer (pH 4.9) completed with 3 volumes of distilled water. Finally, the bacterial cellulose microfibrils were washed from the bleaching solution by several centrifugations with distilled water. The purified cellulose suspension was homogenized again with the Waring blender for 20 min at full speed and stored at 4 °C with 0.01% (w/v) NaN3. Transformation of Cellulose I into Cellulose IIII. Cellulose I was converted into cellulose IIII according to a procedure described previously.13,14 The microfibril suspensions were transferred to anhydrous methanol and then to anhydrous ethylenediamine by successive centrifugations. The mixture was kept overnight at room temperature in ethylenediamine before being resuspended for a few hours in pure methanol. The successive methanol/ethylenediamine treatment was repeated six times until the complete conversion of cellulose I into the cellulose IIII was observed. This transformation was monitored by X-ray diffractometry using a Warhus flat film camera mounted on Philips PW1720 X-ray generator, emitting Ni-filtered Cu KR radiation and operating at 30 kV and 20 mA. The full conversion of native cellulose into cellulose IIII was estimated when the cellulose I diffraction diagram exhibiting the characteristic diffraction lines at d-spacings 0.39, 0.53, and 0.6 nm were completely replaced by a cellulose IIII diagram showing diffraction lines at d-spacings 0.40, 0.52, and 0.75 nm. Grafting of DTAF on Cellulose Microfibrils. One-Step Grafting on Cellulose. DTAF (Figure 1), purchased from Sigma, was used without further purification. A first set of DTAF-cellulose was prepared by dissolving 10-70 mg of DTAF into 10 mL of a suspension containing 100 mg of cellulose microfibils in 0.1 N NaOH. These mixtures were stirred at room temperature for 24 h. The cellulose specimens were then washed free of unreacted DTAF by at least six centrifugations with distilled water. A second set of grafted cellulose was prepared as for the first set, but the amount of DTAF was in the range of 70-115 mg and the cellulose suspension was made in 0.2 N NaOH. Multistep Grafting on Cellulose. Cellulose grafting assays were conducted by dissolving 60 mg of DTAF in 10 mL of a suspension containing 100 mg of cellulose microfibils in 0.2 N NaOH. The mixture was stirred for 24 h at room temperature. The specimens were then washed extensively with distilled water by successive centrifugations. The

Fluorescent Cellulose Microfibrils

procedure was repeated several times, and the final grafted cellulose suspensions were stored at 4 °C. Measurement of the Degree of Substitution Ds. The mean degree of substitution (Ds) was calculated from elementary analysis composition in carbon and nitrogen. It was expressed as a fraction describing the number of moles of DTAF bound per mole of anhydroglucose (AGU) residues. These analyses were performed at the De´partement d’Analyze Ele´mentaireCNRS (Vernaison, France). Casting of Cellulose Films in Microwell Titer Plates. Preliminary tests of deposition of unlabeled cellulose at the bottom of 96 microwell titer plates (Nunc-immuno PlateMaxsorp, Nunc, Denmark) were achieved by drying at 37 °C specific volumes of cellulose suspension (50, 100, and 200 µL) of various concentrations (0.1-2 mg/mL). It appeared rapidly that the films did not stick at the surface of the wells when the total amount of cellulose dried was above 150 µg. Also, when the volume of suspension was larger than 200 µL, the cellulose dried onto the wall of the well in a nonreproducible fashion. The best films were obtained by drying 100 µL of suspension having a concentration of about 1 mg of cellulose/mL of water or below. The reproducibility of the film constitution and reactivity was tested toward their susceptibility to enzymatic degradation. Typically, the films were incubated at 37 °C with 200 µL of 50 mM phosphate buffer at pH 6.5 and with 20 µL of Humicola insolens complex (1 mg/mL). At various times of incubation, eight samplings of 100 µL were collected, and the amount of solubilized reducing sugars produced was measured by the ferricyanide method. 15 The measurements were quite reproducible as a standard of only 15% resulted from the digestion of the eight identical samples. Cellulases. Cellulases from the filamentous fungus H. insolens form the basis for several commercial products, such as Carezyme, Endolase, and Celluzyme (all available from Novozymes A/S). H. insolens harbors at least two cellobiohydrolases of families 6 and 7, and five endoglucanases of families 5, 6, 7, 12, and 45, all of which have been cloned and expressed recombinantly in yeast and/or Aspergillus.16,17 Here, we chose to use (i) a complex of enzymes from H. insolens recovered by filtration from the fermentation supernatant, (ii) samples of isolated and purified recombinant endoglucanases Cel 45 (formerly called endoglucanase V) and Cel6B (formerly called endoglucanase VI) expressed in Aspergillus oryzae as described earlier,16,18 and (iii) Cel6B and Cel6B-D316N expressed in Saccharomyces cereVisiae (YNG318) as described earlier 19 and analyzed directly in the cell culture medium without purification. Cel6B-D316N was a Cel6B mutant constructed by site-directed mutagenesis. It was devoid of any cellulolytic activity, as in its active site the catalytically active aspartate 316 had been replaced by an asparagine. Digestion of DTAF-Cellulose Suspensions. Enzymatic digestion of labeled cellulose was performed as follows: 600 µL of grafted cellulose suspension (100 µg/100 µL) in 50 mM phosphate buffer at pH 6.5 was mixed with 20 µL of the H. insolens complex (1 mg/mL), or 20 µL of endoglucanase Cel6B (1 mg/ml), or 20 µL of water as standard. Assays were conducted for 4 h at 37 °C without agitation.

Biomacromolecules, Vol. 4, No. 3, 2003 483

The extent of cellulose digestion was deduced by the ferricyanide method15 that measures the amount of soluble reducing sugars in the supernatant after centrifugation of the digestion mixtures. In parallel, the release of the fluorescent probes in the digestion medium was followed with a PerkinElmer spectrofluorometer (488 nm excitation, 515 nm emission). The fluorescence was expressed in terms of relative intensity (R.I.), corresponding to the measurement of fluorescence released when the enzymes used were subtracted from that of a control experiment where only water replaced the enzyme solution. Digestion of DTAF-Cellulose Films in Microwell Titer Plates. The reactivities of grafted and ungrafted control films toward Cel45A and Cel6B were assayed by adding 200 µL of 50 mM phosphate buffer pH 6.5 and 10 µL of enzymes (0.1 mg/mL) in the microwells. The plates were kept at 37 °C, and eight samplings were collected at several times during digestion. The samples were diluted eight times, and 200 µL of these diluted samples was analyzed by spectrofluorometry. Digestion of the grafted cellulose films by yeast extracts was achieved by mixing 200 µL of 50 mM phosphate buffer pH 6.5 with 50 µL of yeast culture broth. The solubilized cellulose was also analyzed by spectrofluorometry, but the samples were diluted only twice because of the strong quenching effect resulting from the cultures. Transmission Electron Microscopy (TEM). All TEM was achieved with a Philips CM 200 CRYO electron microscope operated at an accelerating voltage of 80 kV. For the visualization of individual cellulose microfibrils, drops of their dispersions were deposited on carbon-coated copper grids, which had been pretreated by glow discharge. After adequate drying of the specimens, images were recorded under low dose conditions. Surface replicas of the cellulose films were achieved by shadowing the film with carbon/platinum at a shadow angle of 30°. Cellulose was dissolved in 72% (w/v) H2SO4, and the carbon/platinum replica were floated off, rinsed in distilled water, and mounted on copper grids prior to observation under normal beam conditions.

Results and Discussion Optimizing the Grafting of DTAF on Cellulose. Among the number of parameters used in this study, the quantity of DTAF and the alkalinity of the reaction media appeared to be important for optimum grafting in the one-step grafting experiments. Figure 2 illustrates the influence of these parameters by following the release of fluorescence when the DTAF-treated microfibrils in suspension were digested by H. insolens complex, assuming that the fluorescence released was proportional to the amount of grafted DTAF. The two curves shown in Figure 2 indicate that a plateau was reached when the amount of DTAF was increased beyond a certain value in the initial cellulose suspension. This plateau appears to depend on the NaOH concentration in the grafting medium, a more alkaline suspensions leading to more grafting. As seen in Figure 2, when the grafting was

484

Biomacromolecules, Vol. 4, No. 3, 2003

Figure 2. Effect of the amount of DTAF reacted with cellulose microfibrils on the release of the fluorescent probe after 4 h of digestion by the H. insolens complex. In this experiment, the supernatant was diluted eight times before fluorescence measurements.

Figure 3. Effect of the number of labeling steps on the Ds of DTAFtreated cellulose and on the amount of fluorescence released after digestion with Cel6B from A. oryzae. In the Ds measurements, errors bars have been added.

done in 0.1 N NaOH, the plateau was reached when 100 mg was treated by 40 mg or more of DTAF. If the alkali concentration was doubled, a 25% increase in fluorescence was released when 70 mg or more of DTAF was reacted with 10 mg of cellulose. We supposed that activation of cellulose increased with the amount of NaOH that enhances the reactivity of cellulose toward DTAF molecules, leading to a larger amount of chemically bonded fluorescent molecules onto a cellulose microfibril surface. Another way to increase the amount of grafting was to achieve a number of successive grafting operations, taking care of doing extensive washing between each operation. The effect of the multistep grafting is illustrated in Figure 3, which shows the result of applying seven successive grafting steps. In that case, the cellulose Ds increased steadily by steps of 0.0035, to reach a value of 0.025 at the seventh step. For the intermediate Ds, the DTAF molecules grafted at the cellulose surface did not seem to influence, inhibit, or activate, the labeling reaction. In terms of amount of labeling, no distinction could be made between the first and the last reaction, it seems that the accessibility of bacterial cellulose is uniform and that the DTAF molecules become attached randomly along the microfibrils.

Helbert et al.

Regarding the fluorescence released when the substrate was digested by the H. insolens complex, a steady increase was observed until the fourth grafting step. Beyond that, a decrease in the released fluorescence was observed. This decrease appears to be associated with an inhibition of the enzymatic system by the excess of DTAF at the surface of cellulose. The use of microfibrils converted into cellulose IIII and subsequently grafted with DTAF led to a release of fluorescence and reducing sugars (results not shown) that were qualitatively much higher than in the cellulose I counterpart. Thus, even if we did not measure precisely the Ds of the DTAF-treated cellulose IIII, it is inferred that it is markedly higher than that achieved in a similar experiment with cellulose I. Indeed, the conversion of bacterial cellulose I to IIII has induced substantial morphological modifications of the microfibrils. Despite the fact that the microfibrillar structure of bacterial cellulose is still conserved, the original rectangular cross-sectional shape of the initial microfibrils has become distorted.14 These altered microfibrils display a rough surface, with a consequence of an increase in their surface area. Because there is more cellulose surface accessible to DTAF, the reactivity of the microfibrils has raised and the overall amount of DTAF that can be grafted is potentially higher. In the same manner, we can argue that the alteration of microfibrils morphology has increased the accessibility and reactivity of cellulose toward cellulases. Finally, the rise of DTAF observed after the enzymatic digestion of labeled cellulose IIII could be attributed to a combination of increased Ds and higher accessibility for cellulases. Other experiments not shown here involved the use of Cel6B instead of the H. insolens complex. In that case, the release of fluorescence followed the same trend as in Figure 3, but the amount of released fluorescence was somewhat higher than that obtained in the corresponding test with the H. insolens complex. It is interesting to note that the grafting experiments did not affect the overall appearance of the cellulose microfibrils which displayed the same classical ribbonlike morphology before and after grafting as seen in Figure 4, where panel A corresponds to suspension of microfibrils and panel B is a surface replica of a film cast in a microwell titer plate. X-ray diffraction analysis (results not shown) indicated that there was no modification in the structural parameters of the derivatized samples. Enzymatic Assays on Fluorescent Cellulose Microfibrils in Suspension. Figure 5 illustrates the release of fluorescence and that of soluble sugars when DTAF-grafted microfibrils were incubated for 4 h with either the H. insolens complex (Figure 5A) or Cel 6B (Figure 5B). In both cases, the effects were studied as a function of the Ds of the fluorescent microfibrils. With both enzymatic systems, one sees that the released fluorescence increases steadily as a function of Ds to reach a maximum for a Ds value of 0.015 beyond which a significant drop was observed. For this Ds value, assuming that 25% of cellulose chains of one bacterial cellulose microfibrils are located at the surface, one can estimate that

Fluorescent Cellulose Microfibrils

Biomacromolecules, Vol. 4, No. 3, 2003 485

Figure 6. Mechanism of degradation of fluorescent cellulose microfibrils. (A) Only unlabeled products are released by the highly processive exo enzymes. (B) Random degradation of cellulose by endo-glucanases, leading to the solubilization of fluorescent soluble cellodextrins.

Figure 4. Transmission electron micrographs of the bacterial cellulose microfibrils that were used here. (A) Low dose image of a series of dispersed microfibrils. (B) Surface replica after shadow casting with C/Pt of a film of bacterial cellulose microfibrils cast at the bottom of a microwell titer plate.

Figure 5. Effect of the degree of substitution on the release of fluorescence and reducing sugar. (A) Incubation by the H. insolens complex. (B) Incubation by recombinant Cel 6B from A. oryzae. In these experiments, the supernatant was diluted 16 times.

the Ds at the surface should be around 0.06, which corresponds to the grafting of 1 DTAF molecule every 17 glucose

units. In Figure 5A, the amount of reducing sugar solubilized by the H. insolens complex as function of Ds followed a different pattern as it decreased rapidly as a function of Ds, thus denoting a strong inhibition. Indeed, a Ds of 0.010 cuts the amount of reducing sugar released by a factor of 10 as opposed to the value corresponding to an underivatized sample. In the case of the sample digested with Cel6B, no inhibition could be observed, as the amount of reducing sugar released appeared to be insensitive to the presence or not of grafted DTAF. At first glance, the results shown in Figure 5 appear somewhat contradictory as it seems difficult to understand why the release in reducing sugar is so dependent on the grafting in the case of digestion with H. insolens complex and not when Cel6B alone was used. Also the difference between the variation of the amount of reducing sugar released and that of the fluorescence is another point that needs further thinking. To account for these phenomena, one needs to consider the enzyme composition of the H. insolens complex. According to literature data, this complex contains at least seven different cellulases, classified into five endoglucanases and two cellobiohydrolases.19 As schematized in Figure 6, these two classes of enzyme act in different way on grafted cellulose. The cellobiohydrolases are the main providers of reducing sugars during the digestion. Indeed these enzymes and in particular Cel 7A are not only active on crystalline cellulose but also display a highly processive character. 8 Thus, once a given Cel7A enzyme starts from the reducing end of a cellulose chain, it will keep on going releasing cellobiose until it reaches either the other chain end or a DTAF-labeled glucosyl moiety (Figure 6A). Thus one enzymatic attack will generate a large amount of reducing sugars in an underivatized cellulose chain and only a few in one holding DTAF grafts. In the case of Cel 6B, the attack is random along the chain and one expects no or only moderate processivity (Figure 6B). Thus the amount of reducing sugar released is relatively constant within the experimental Ds conditions corresponding to the case in Figure 5B. Experiments (not shown) where DTAF grafted cellulose IIII were incubated with Cel 6B indicated a small inhibition at high DTAF content. As stated above, the DTAF grafting of cellulose IIII samples was much higher than the

486

Biomacromolecules, Vol. 4, No. 3, 2003

Figure 7. Expected patterns of release of soluble fluorescent probe and reducing sugar according to the type of enzymes: cellulases mixtures containing endo and exo enzymes as opposed to endo cellulases only.

one observed with cellulose I. This could explain why no inhibition could be detected with cellulose I samples whereas a small one was observed with cellulose IIII at high grafting content. On the basis of the results presented in Figure 5 and the scheme in Figure 6, one can propose a general pattern for the variation of the reducing sugar and the fluorescence released during the digestion of DTAF-grafted cellulose. This pattern, presented in Figure 7, indicates that if the digesting enzymes are essentially of the endo type, a slow decrease of reducing sugar is likely to occur when the grafted content is increased. On the other end, pure exo enzymes or a mixture of endo and exo will lead to a rapid and spectacular decrease of the quantity of released reducing sugar. In both cases, however, the fluorescence released will maximize at given grafting amount beyond which it will drop toward smaller values. Thus, with the present fluorescent cellulose, a comparison between the evolution of the fluorescence and that of the reducing sugar release can account for the enzymatic composition in a given complex cellulase system. Enzymatic Assays on Microwell Titer Plates Coated with Fluorescent Microfibils. Films of fluorescent microfibrils cast at the bottom of microwell titer plates proved to be strongly held at the well bottom as they kept glued to the well at all times. Thanks to this property, aliquots devoid of any suspended cellulose could be pipeted out from the supernatant and there was no need for any centrifigation. These films presented efficient response upon enzymatic digestion. This is illustrated in Figures 8 and 9. In Figure 8, films prepared from cellulose I and cellulose IIII were assayed for Cel 45A activity. In these experiments, the released fluorescence was measured as a function of the digestion time. In both samples, there was a fast increase in released fluorescence for the first 2 h. From then the enzymatic digestion seemed to occur much more slowly as both curves in Figure 8 tapered off after 2 h of digestion. As expected, the fluorescence released was systematically close to three times higher in the case of cellulose IIII films as opposed to those of cellulose I. Figure 9 is another example demonstrating the use and sensitivity of cellulose films for cellulase assays. In this experiment, yeast broths containing recombinantly expressed Cel 6B were added to the microwell titer plate coated with fluorescent cellulose. As in the case of Cel45A, extensive

Helbert et al.

Figure 8. Kinetics of the release of the fluorescence from DTAFgrafted cellulose films cast at the bottom of microwell titer plates incubated with recombinant Cel 45A from A. oryzae: top curve, films of cellulose I; bottom curve, film of cellulose IIII.

Figure 9. Kinetics of the release of the fluorescence from DTAFgrafted cellulose IIII films cast at the bottom of microwell titer plates: top curve, release of fluorescence after incubation with S. cerevisae extracts containing the “wild-type” Cel 6B; bottom curve, release of the fluorescence after incubation with yeast extract containing the inactive mutant (Cel6B-D316N).

fluorescence was released during the first 2 h and not much after that. The substitution of Cel 6B by its inactive mutant Ce6B-D316N had a dramatic effect as only a constant and low background fluorescence could be observed at any time during the 5 h incubation. The digestion experiments achieved with microwell titer plates coated with films of fluorescent microfibrils are quite interesting as they open a way to automation of the cellulase assays.20 Indeed, one can conceive that robots could perform all the operations ranging from the film casting into the microwells of the titer plates, the various washing steps, the inoculation with enzyme broths, and subsequent pipeting for fluorescence and reducing sugar detection. Such operations are possible because the fluorescent cellulose films not only are firmly anchored at the bottom of the microwells but also keep their integrity throughout the washing and the moderate digestion steps that are required for the detection. The automation of the above cellulases assay opens the way to the screening of a very large number of different enzymatic systems where the detection of new cellulases is of importance. So far, screening for useful cellulases has always been labor intensive and time-consuming. The automation of the method of the microwell coated with fluorescent cellulose and the automatic detection of the released soluble products

Fluorescent Cellulose Microfibrils

could drastically cut down the cost of cellulase screening as it is formatted for high-throughput operation. The discovery of new cellulase systems is still timely as many industrial applications either do benefit or could benefit from optimized cellulase treatments. Examples of such applications include the treatment of cellulosic textiles or fabrics using cellulases, as ingredients either in laundry detergent or fabric softener formulations or in textile finishing such as stone washing of denim jeans or biopolishing. Other important applications of cellulolytic enzymes are in the treatment of pulp furnish for improving the drainage in the paper machine or for deinking of recycled paper. The cellulose films that we have described here are made of assemblies of cellulose microfibrils. We believe that they are a good representative of natural fibers such as cotton or wood, which are the main targets for the industrial applications of cellulase. The preparation of films from cotton microfibrils could also be achieved. 20 For this, we used never-dried unopened cotton bolls where the cellulose microfibrils are not locked into intermicrofibrillar hydrogen bond netwok.21 Preliminary tests showed that these samples could be microfibrillated and grafted easily. As in the case of the films of fluorescent bacterial cellulose microfibrils, the fluorescent films of cotton microfibrils were strongly glued to the bottom of the microwell titer plates. Therefore these cotton-based films proved also to be good candidates for the automatic detection of cellulases. The preparation of films of fluorescent wood pulp cellulose was not attempted, but with the current knowledge on cellulose microfibril suspensions, we do not see any problem in preparing homogeneous suspensions of these microfibrils, grafting them with DTAF and casting them in microwell titer plates. As in the case of bacterial and cotton cellulose, automation of the cellulase testing could also be envisaged for wood pulp samples. The assay that we have described here could be extended to selections of other polysaccharides. In addition to native cellulose, a number of microfibrillar polysaccharides exist

Biomacromolecules, Vol. 4, No. 3, 2003 487

also in nature.22 Native microfibrils are indeed found for R and β chitin, (1f 4) mannan, (1f 3) glucan, or (1f 3) xylan. It is therefore likely that the process of fluorescent film casting can also be extended to these systems. The automation of the screening of the degrading enzymes specific to any of these samples could therefore be envisaged easily. References and Notes (1) Henrissat, B. Biochem. J. 1991, 280, 309. (2) Henrissat, B.; Bairoch, A. Biochem. J. 1993, 293, 781. (3) Henrissat, B.; Teeri, T. T.; Warren, R. A. J. FEBS Lett. 1998, 425, 352. (4) http://afmb.cnrs-mrs.fr/∼cazy/CAZY/. (5) Chanzy, H.; Henrissat, B. Carbohydr. Polym. 1983, 3, 161. (6) Chanzy, H.; Henrissat, B. FEBS Lett. 1985, 184, 285. (7) Boisset, C.; Chanzy, H.; Henrissat, B.; Lamed, R.; Soham, Y.; Bayer, E. A. Biochem J. 1999, 340, 829. (8) Boisset, C.; Fraschini, C.; Schu¨lein, M.; Henrissat, B.; Chanzy, H. Appl. EnViron. Microbiol. 2000, 66, 1444. (9) Boisset, C.; Pe´trequin, C.; Chanzy, H.; Henrissat, B.; Schu¨lein, M. Biotechnol. Bioeng. 2001, 72, 339. (10) Dudman, W. F.; Bishop, C. T. Can. J. Chem. 1968, 46, 3079. (11) Hackman, R. H.; Goldberg, M. Anal. Biochem. 1964, 8, 397. (12) de Belder, A. N.; Granath, K. Carbohydr. Res. 1973, 30, 375. (13) Roche, E.; Chanzy, H. Int. J. Biol. Macromol. 1981, 3, 201. (14) Chanzy, H.; Henrissat, B.; Vuong, R.; Revol, J.-F. Holzforschung 1986, 40 Suppl, 25. (15) Kidby, D. K.; Davidson, D. J. Anal. Biochem. 1973, 55, 321. (16) Schu¨lein, M. J. Biotechnol. 1997, 57, 71. (17) Schou, C.; Rasmussen, G.; Kaltoft M. B.; Henrissat, B.; Schulein, M. Eur. J. Biochem. 1993, 217, 947. (18) Schu¨lein, M.; Tikhomirov, D. F.; Schou, C. In Trichoderma reesei Cellulases and other Hydrolases; Suominen, P., Reinikainen, T., Eds.; Foundation for Biochemical and Industrial Fermentation Research: Helsinki, Finland, 1993; Vol. 8, p 109 (19) Okkels, J. S. 1997, WO 97/07205. (20) Helbert, W.; Chanzy, H.; Ernst, S.; Schu¨lein, M.; Husum, T. L. 2001, WO 01/25470. (21) Nelson, M. L.; Rousselle, M.-A.; Ramey, H. H. Jr; Barker, G. L. 1980, Text. Res. J. 50, 491. (22) Chanzy, H.; Vuong, R. In Polysaccharides. Topics in Structure and Morphology; Atkins, E. D. T., Ed.; Macmillan: London, 1985; p 41.

BM020076I