Planta
Planta (1955)166:530-536
9 Springer-Verlag 1985
p-Glucanases in developing cotton (Gossypium hirsutum L.) fibres * P. Bucheli, M. Dfirr, A.J. Buchala** and H. Meier Institut de Biologic v6g6tale et de Phytochimie, Universit6 de Fribourg, CH-1700 Fribourg, Switzerland
Abstract. Cotton fibres possess several fl-glucanase activities which appear to be associated with the cell wall, but which can be partially solubilised in buffers. The main activity detected was that of an exo-(1-,3)-fl-D-glucanase (EC 3.2.1.58) but which also had the characteristics of a fl-glucosidase (EC 3.2.1.21). Endo-(l~3)-fl-D-glucanase activity (EC 3.2.1.39) and much lower levels of (1 ~4)-fl-D-glucanase activity were also detected. The exo-(l ~ 3 ) fl-glucanase showed a maximum late on (40 days post-anthesis) in the development of the fibres, whereas the endo-(l~3)-fl-glucanase activity remained constant throughout fibre development. The fl-glucanase complex associated with the cotton-fibre cell wall also functions as a transglucosylase introducing, inter alia, (1 ~6)-fl-glucosyl linkages into the disaccharide cellobiose to give the trisaccharide 4-O-fl-gentiobiosylglucose. Key words: Callose - Cellulose - f l - G l u c a n a s e fl-Glucosidase - Gossypium (fl-glucanase) - Transglucosylation.
or an obligatory intermediate in the formation of cellulose. Either of these possibilities imply the involvement of fl-glucanases or transglucosylases. Preliminary work showed the presence of several glycosidases (Huwyler 1978) and fl-D-glucanases bound to the cotton-fibre cell wall. Such enzymes have also been found associated with the cell walls of many other higher plants (Clarke and Stone 1962) but their physiological role is not clear. Some hydrolases can also behave as transglycosylases, catalysing the transfer of glycosyl residues to oligosaccharides or polysaccharides. The present work was undertaken in order to characterise the fl-glucanase or fl-glucosidase activity in cotton fibres and to study any eventual variations of these activities during fibre development. Material and methods Plant material. Cotton plants (Gossypium hirsutum L. var. Stone~ ville No. 406) were grown in a greenhouse at a temperature of 30 ~ C during the day and 20 ~ C at night. The fruits, at various stages of development, were all harvested at the same time and kept on ice until opened. Preparation of the fibre cell-wall homogenates. As soon as possi-
Introduction
Recent studies on the incorporation of radioactive precursors into the polysaccharides of the cell wall of intact cotton fibres have shown that one of the products, a (1 ~3)-fl-D-glucan (callose), undergoes a turnover during the stage of secondary cell-wall formation (Meier et al. 1981 ; Meier 1983). It was not clear, however, if the callose was a facultative * Presented at the Third Cell Wall Meeting held in Fribourg in 1984 ** To whom correspondence should be addressed
Abbreviations: CMC=carboxymethylcellulose; O N P G = o - n i trophenyl-fl-D-glucopyranoside; T L C = t h i n - l a y e r chromatography
ble, the fruit capsules were dissected and the individual seed clusters separated. The fibres were quickly detached from the seeds in acetate buffer (50 mM, pH 5.5), finely cut with scissors and quantitatively transferred to the chamber of a McCrone Micronising Mill (Lovibond, London, U K ) equipped with corundum mill elements. Homogenisation was carried out at 4 ~ C for 30 min in the above buffer and the homogenate was quantitatively transferred to a dialysis tube and exhaustively dialysed against distilled water at 4 ~ C. The final volume was measured and aliquots were taken for analysis.
Extraction of low-molecular-weight sugars. One seed cluster from each fruit was taken for ulterior analysis of the neutral sugars and sugar phosphates (results not given here) and for determination of the cell-wall dry weight and callose content. The extraction with hot aqueous 80% methanol (v/v) and the determination of the cell-wall dry weight were carried out as described previously (Jaquet et al. 1982).
P. Bucheli et al. : fl-Glucanases in cotton fibres
Determination of the callose content of the cotton fibres. The method involving extraction with hot dimethyl sulphoxide, followed by acid hydrolysis of the isolated product, was carried out as described by Jaquet et al. (1982). This method has been shown to be reproducible and to be quite quantitative (Pillonel et al. 1980). Enzyme assays. Preliminary experiments were carried out in order to determine the extent of autohydrolysis under the conditions to be used for the enzyme assays since the homogenate contains not only the activities to be determined, but also their endogeneous substrate, callose. It was found that for the relatively short period of incubation used no correction for autolysis was necessary. The total (1 ~3)-/~-glucanase activity (endo + exo) was measured as the release of reducing sugars, estimated by the colorimetric method of Nelson and Somogyi (Somogyi 1952). Digests contained laminaran (0.17 % ; United States Biochemical Corp., Cleveland, Ohio, USA), acetate buffer (50 mM, pH 5.5) and fibre homogenate (5 ml) to give a final volume of 6 ml. Incubation was carried out at 35~ in a gyratory shaker. Samples (0.5 ml) were withdrawn after 0, 1, and 2 h and added directly to the copper-containing reagent in order to stop the reaction. Before the final absorbance measurements, the reaction mixture was centrifuged (10 rain, 10000 g). The exo-(l~3)-fl-glucanase activity was measured as the release of glucose, determined by the glucose-oxidase Perid method (Boehringer, Mannheim, FRG), obtained under the condition described above for the determination of the total (1--+3)-/~-glucanase activity. Samples (0.6 ml) were withdrawn after 0, 1, and 2 h and added directly to a solution of trichloroacetic acid (10%, 0.4 ml) to stop the reaction. Aliquots (0.2 ml) were taken for analysis. Endo-(l~3)-~-glucanase activity was determined as described above for the total (1-,3)-p-glucanase activity, but the incubation mixture also contained 10-4M nojirimycin (Meiji Seika Kaisha Inc., Yokohama, Japan). An additional control was carried out - the glucose released under these conditions, determined by the glucose-oxidase method, was negligible. /~-Glucosidase activity was determined by incubating o-nitrophenyl-fl-D-glucoside (ONPG; 0.2%; Fluka AG, Buchs, Switzerland) or laminaribiose (10mM) in acetate buffer (50 mM, pH 5.5) with fibre homogenate (1 ml) in a total volume of 2 ml at 35 ~ C. The reactions were terminated by the addition of trichloroacetic acid (10%, 2 ml) after 0, 1 and 2 h. After centrifugation (10 min, 1000 g) the glucose released was measured using the glucose-oxidase Perid method, as above, but the fl-glucosidase known to be present in the commercial enzyme preparation was totally inhibited by adjusting the pH of the enzyme-buffer reagent to 8.5 with 1 M 2-amino-2-(hydroxymethyl)-l,3-propanediol immediately before analysis (Dahlqvist 1974). (l~4)-fl-glucanase (carboxymethylcellulase) activity was measured as the release of reducing sugars, determined by the Nelson-Somogyi method (Somogyi 1952), when carboxymethylcellulose (CMC; 4M6F; Hercules, Wilmington, USA; at a final concentration of 0.5% in 50 mM acetate buffer at pH 5.0) was incubated with fibre homogenate (1 ml) at 35 ~ C for 6 h. In each of the above methods, the change in absorbance compared with that of the sample taken at 0 h was measured. The results are expressed in gmol glucose (or glucose-equivalents) released per hour for the fibres of one seed, i.e. on a constant cell-number basis. The enzyme assays were carried out in duplicate for each fruit at a particular stage of development, as were the individual sugar determinations, with good reproducibility, and the mean values are reported.
531
Chromatographic analysis of hydrolysis products. Paper chromatography was carried out on Schieicher and Schiill No. 2043b paper (Schleicher and Schiill, Dassel, FRG) using the following irrigants (a) ethyl acetate: pyridine: water (8 : 2 : 1, by vol.), (b) propan-l-ol:ethyl acetate :water (7 : 1 : 2, by vol.) and thin-layer chromatography (TLC) on Kieselgel 60 (Merck, Darmstadt, F R G with the irrigants (c) acetone:water (88:12, v/v), (d) acetone: water (85 : 15, v/v), (e) butan-l-ol : acetic acid : ether: water (9:6:3:1.5, by vol.) or (f) benzene:ethanol:acetic acid:water (200:47 : 15 : 1, by vol.). Detection was either with alkaline AgNO3 (Trevelyan et al. 1950) or with ~-naphthol/conc. H2SO4 (Stahl 1969) as appropriate; radioactive sugars on chromatograms were detected using a TLC Linear Analyser LB 282 (Berthold, Wildbad, FRG). Preparation of radioactive substrates, fl-Glucans (callose and cellulose) in cotton fibres were labelled with [14C]glucose by incubating isolated seed clusters with [U-l~C]sucrose as described by Pillonel et al. (1980). Cellobiose, uniformly 14C-labelled only in the reducing moiety, and cellobiose, uniformly ~4C-labelled only in the non-reducing glucose moieties were prepared enzymatically using cellobiose phosphorylase (EC 2.4.1.20) derived from Cellvibrio gilvus (Canevascini et al. 1979). Autolysis. Autolysis experiments were carried out with finely cut (razor blade) detached cotton fibres where the endogenous fl-glucans had been labelled (see above). Weighed aliquots of the fibres were thoroughly washed free of low-molecular-weight material by suspending them in Miracloth (Checopee Manufacturing Co., Cornelia, Ca., USA) bags in deionised water at 0 ~ overnight. Aliquots (0.5 g) were suspended in acetate buffer (50 mM, pH 5.5) and incubated at 30 ~ C. At various intervals the autolysis was terminated by heating the samples for 5 min in boiling water. The samples were centrifuged (10 min, 10000g), the radioactivity in the supernatant measured, and the glucose and reducing sugars liberated were determined by the glucose-oxidase and Nelson-Somogyi methods, respectively. For the inhibition experiments the incubation medium contained D-glucono-l,5-1actone (10-5-10-ZM) or nojirimycin (1 mM). In the control experiments the samples were heat-inactivated before incubation. Transglucosylation activity. (i) With detached cotton fibres. The fibres were prepared as described above for the autolysis experiments. Aliquots were incubated with cellobiose (290 mM) or cellotriose (100 raM) for 1 h at 35 ~ C in acetate buffer (50 raM, pH 5.5). The water-soluble material was examined by TLC (irrigant d). (ii) With an enzyme fraction solubilised from the cell wall. Fibres were homogenised as described for the/~-glucanase assays, but the simple acetate buffer was replaced with a buffer of high ionic strength containing, in addition, sodium azide (0.02%), glycerine (5%), ethane thiol (10 mM) and lithium chloride (3 M). The homogenate was then filtered through four layers of Miracloth, the filtrate centrifuged (10 min, 10000 g) and the supernatant was exhaustively dialysed against distilled water at 5~ C. The enzyme-containing solution (0.5 ml) was incubated with laminaran (0.5%) and cellobiose (100 mM; approx. 3 000 Bq), where the reducing moiety contained uniformly 14C-labelled glucose, for 6 h at 35 ~ C. The reaction mixture was examined by paper chromatography (irrigant b) and TLC (irrigant c) and then fractionated by chromatography on a column (50 cm long, 2.2 cm inner diameter) of Fractogel TSK HW-40(S) (Merck, Darmstadt, FRG). The major labelled product was re-examined by paper chromatography (irrigant b, Rg~c 0.18) and TLC (irrigant d, RgLc0.52) and found to be chromatographically homogeneous.
532
Analysis of the transglucosylation product. A sample of the product was reduced with NaBH4 and hydrolysed (2M trifluoroacetic acid; 1 h) as described by Buchala et al. (1972) for xylooligosaccharides. The products, glucose and glucitol, were separated by TLC (irrigant c) and the radioactivity measured - only glucitol was found to be labelled. The reduced oligosaccharide was treated with the fl-glucosidase preparation from almond emulsin (Fluka, Buchs, Switzerland) in acetate buffer (50 raM, pH 5.5) and the products examined by TLC (irrigant d) - glucose and labelled cellobiitol were found. The unreduced oligosaccharide, synthesised from eellohiose labelled in the non-reducing glucose moiety, was methylated and hydrolysed (Jansson et al. 1976). Thin-layer chromatography of the hydrolysate (irrigant e) gave 2,3,4,6-tetra-O-methylglucose, 2,3,6-tri-O-methylglucose and 2,3,4-tri-O-methylglucose, where only the latter tri-O-methylglucose was radioactive. No di-O-methylglucose derivatives were detected. The identities of these sugars were confirmed by capillary column gas chromatography (WCOT OV-225, Chrompack B.V., Middelburg, The Netherlands; 25 m, 170 ~ C) of the alditol acetate derivatives of the methyl sugars by comparison with authentic standards (Jansson et al. 1976).
Results
Qualitative examination of the fl-glucanase activity. Crude extracts of detached cotton fibres and fibre homogenates contain several fl-glucanase activities. Laminaran, periodate-oxidised laminaran and pachyman were effectively degraded to give glucose and a series of laminarioligosaccharides indicating the presence of endo-(l~3)-fl-D-glucanase, exo(l~3)-fl-D-glucanase or fl-glucosidase activities. The latter activity was also demonstrated by the degradation of glucose-containing fl-linked oligosaccharides and ONPG. The polysaccharides lichenan and oat caryopsis fl-glucan, which both contain (1--+3)-fl-, and (l~4)-fl-D-glucosidic linkages, and to a lesser extent CMC, could also be hydrolysed showing the presence of (1 ~4)-fl-D-glucanase activity. All of these activities were detected both at the primary and secondary cell-wall stages of fibre development. On the other hand, starch, pullulan, pustulan, and nigeran were not degraded (see Table 1). No systematic study of the solubility of the flglucanases was made. However, the activity solubilised depended on the ionic strength of the extracting buffer; with the low-ionic-strength buffer used for homogenisation of the fibres for the quantitative studies described below, the activity was essentially associated with the buffer-insoluble material, whilst about 60% of the fl-glucanase activity (determined either with O N P G or laminaran) could be solubilised if the homogenisation buffer contained 3 M LiC1.
Definition of the various fl-glucanase activities. Although the distinction between the exo-(l-~3)-fl-
P. Bueheli et al. : fl-Glucanases in cotton fibres Table 1. Action of the cell-wall-bound fl-glucanases on various substrates. Aliquots of finely cut fibres, free of low-molecularweight carbohydrates, were incubated with the suhstrate (10raM for low-molecular-weight substrates and 0.5% for polysaccharides) at 30~ for 1 h. The glucose released was measured by the glucose-oxidase method ~ (for the low-molecular-weight substrates) and reducing sugars (for polysaccharides) by the Nelson-Somogyi method b. The values are corrected for cell-wall autolysis Substrate
Rate of hydrolysis (gmol glc seed- 1 h - "
Sophorose Laminaribiose Cellobiose Gentiobiose ONPG Maltose p-Nitrophenyl-fl-Dglucoside Laminaran Pachyman Lichenan CMC Pustulan Pullulan Nigeran Starch
1.6 5.6 0.7 4.0 3.1 0.02 0.01 4.6 2.0 0.9 0.02 0.01 0.02 0
(gmol glc equiv. seed- 1 h - b m
m
5.1 2.5 2.5 0.04 0 0 0 0
glucanase and the fl-glucosidase activities is not clear, the former activity is referred to when laminaran was used as substrate and where the glucose released was measured. The latter activity is referred to when laminaribiose or O N P G was used as the substrate. Total (1--->3)-fl-glucanase activity (exo + endo) was measured using laminaran as substrate and determining the reducing sugars (expressed in glucose equivalents) released. Endo-(1--->3)-fl-glucanase activity could not be satisfactorily measured viscosimetrically using carboxymethylpachyman as substrate (Clarke and Stone 1962). Nojirimycin is a known inhibitor of fl-glucosidase, and to a lesser extent, exo-fl-glucanase activity. It was found that with nojirimycin at a concentration of 10-4M and laminaran as substrate, only laminarioligosaccharides, and no glucose, were produced. The reducing sugars thus released, could be taken as a measure of the endo(l~3)-fl-glucanase activity. Similar values were obtained by substracting the value for the exo-(l ~3)-fl-glucanase activity (glucose measured by the glucose-oxidase method) from the value for the total (1---,3)-fl-glucanase activity (measured by the Somogyi-Nelson method), thereby showing the validity of the inhibition method for short-term incubations.
P. Bucheli et al. : fl-Glucanases in cotton fibres
0
533
(a)
,-
A
Control
start
Laminaribiose
glucose
~ A
0o
time(h)
Fig. l a , b. Autolysis of (l~3)-fl-glucan in detached cotton fibres: a seed clusters were incubated with 14C-sucrose in order to label the (1 ~ 3)-fl-glucan in situ. The fibres were detached, washed and incubated at 35 ~ C and at various intervals the radioactivity in the incubation medium was measured. In the control experiment the enzymes in fibres were deactivated by boiling for 10 rain in the incubation buffer; b radiochromatographic analysis (irrigant b) of the soluble products after 8 h of incubation in the absence and presence of nojirimycin
Characteristics of the various fl-glucanase activities. An optimum value of about pH 5.8 was obtained for the exo-(1--,3)-~-glucanase activity using laminaran as substrate. A slightly lower value of pH 5.5 was obtained for the fl-glucosidase activity with either laminaribiose or O N P G as substrate. All the enzymic determinations were, however, carried out at pH 5.5. The Km value for the fl-glucosidase was determined for the substrates O N P G (5 mM), laminaribiose (6.2 mM), gentiobiose ( l . 4 m M ) , cellobiose (0.6mM) and sophorose (0.4 mM) by the method of Lineweaver and Burk (1934) with fibre homogenates. Similarly, values of 0.04 m M and 0.12 m M were obtained for the exo-(l~3)-fl-glucanase with laminaran as substrate and determining the reducing sugars or the glucose released, respectively. These latter values are only apparent since clearly the presence of the endo-(l--,3)-/?-glucanase has an effect on the substrate concentration; effectively reducing its concentration for a true exo-enzyme or increasing its concentration for a true glucosidase. The value obtained for the endo-(l ~3)-/~-glucanase was 0.10 raM. The effect of various enzyme inhibitors was sought and it was found that D-glucono-l,5-1actone and nojirimycin inhibited the hydrolysis of ONPG, laminaribiose and laminaran. Attempts were made to distinguish the (1--*3)-fl-exoglucanase from the /?-glucosidase since it has been reported (Reese et al. 1971) that much higher concentrations of the inhibitors are necessary to inhibit the former activity. With glucono-l,5-1actone,
a concentration of 10-5M, normally sufficient to inhibit fl-glucosidase, had little effect when laminaribiose was the substrate. At higher concentrations (greater than 5.10-3M), both types of activity were inhibited. Similar results were obtained with nojirimycin except that total inhibition was achieved at 10- 4M.
Cell-wall autolysis. When the fibre homogenates were incubated in buffer at pH 5.5, there was in some cases a appreciable release of glucose caused by the autolysis of the cell wall which contains, in addition to the fl-glucanases, the endogenous substrates callose and cellulose. It was observed that the rate of autolysis depended on the length of time of dialysis of the fibre homogenate, when the endogenous callose was slowly degraded and the products eliminated. Fibre homogenates were prepared from intact seed clusters which had been incubated with 14C-sucrose (Pillonel et al. 1980) in order to label the endogenous/~-glucans. Autolysis experiments showed that 25% of the labelled callose could be degraded within 8 h. Figure 1 a shows the results obtained for a typical experiment. Chromatographic analysis of the products obtained after 8 h showed that glucose was the only labelled low-molecular-weight product (see Fig. 1 b). When nojirimycin (1 mM) was included in the incubation medium the rate of release of radioactivity into the incubation medium was reduced (not shown) and a appreciable amount of laminarioligosaccharides (principally laminaribiose) was detected.
534
P. Bucheli et al. : fl-Glucanases in cotton fibres
750
5
50 5O
500
,to ~
~ol
3
~
2
30 ~>,
..9o
o
4
I
250
t 20
10
1'5
2'1 2'4
2'8
3=6 410 4 4
Days post anthesis
Fig. 2. Changes in the cell-wall dry weight ( o - - o ) and in the
callose content (o--o) of the cotton fibres during development; the values were calculated for the fibres of one seed
5 p-,,
4 3 2 o
1 ll5
2'1 2'4
218
316 410 4'4
Days post anthesis
Fig. 3. Changes in the fl-glucanase activities of the cotton fibres during development; (e--o) total ~-glucanase, (zx--zx)exo-(l ~ 3)-~-glucanase, (i--i) endo-(1~3)-/?-glucanase
Relationship between the various fl-glucanase activities and fibre development. In order to relate the fl-glucanase activities to fibre development, several previously established sampling criteria (Jaquet et al. 1982) had to be satisfied for the series of fruits harvested. A regular increase in the cell-wall dry weight and the presence of a distinct maximum in the (1 ~3)-fl-glucan content at the stage corresponding to the beginning of massive secondary wall formation (Fig. 2) indicated the success in sampling. The enzyme activities are expressed in micromoles of glucose or glucose-equivalents released per hour for the fibres of one seed, i.e. on a constant cell number basis and are corrected where appropriate for autolysis.
ll5
2'1 214 " 218
3'6
4'0
4'4
Days post anthesis
Fig. 4. Comparison of the exo-(J~3)-fl-glucanase activity (~x--A) of cotton fibres during development with the glucosidase activity obtained using laminaribiose ( o - - o ) or O N P G ( o - - o ) as substrates
The changes in the various fl-glucanase activities, which occurred during fibre development, are shown in Fig. 3. It can be seen that during the formation of the primary cell wall, i.e. up to about 24 days post-anthesis, the absolute values for the activities were relatively low. The total (1--, 3)-flglucanase activity increased regularly throughout fibre development to reach a maximum at 40 days post-anthesis and then decreased. This same pattern was found for the exo-(] ~3)-fl-glucanase and fl-glucosidase activities; Fig. 4 compares the latter two activities. On the other hand, the endo-(l --*3)fl-glucanase activity was more or less constant after 15 days post-anthesis throughout fibre development (Fig. 3).
Transglucosylase activity. When fibre homogenates, or material solubilised with 3 M LiC1 from fibre homogenates, was incubated with high concentrations (100 mM) offl-linked, glucose-containing, homooligosaccharides as acceptor molecules and laminaran (0.5%) as donor molecule, new oligosaccharides were synthesised. The transglucosylation appeared to introduce (l~6)-fl-glucosyl linkages into the acceptor molecules; this was demonstrated for cellobiose. Under the above conditions, cellobiose 14C-labelled in the reducing glucose moiety, gave at least two labelled products (Fig. 5). The major product had a chromatographic mobility corresponding to a tri-, or a tetrasaccharide and was not cellotriose, cellotetraose, laminaritriose nor laminaritetraose. The oligosaccharide was purified by molecular-size chromatography and partially characterised. Acid hydrolysis or treatment with a fl-glucosidase gave only glu-
P. Bucheli et al. : fl-Glucanases in cotton fibres
A start
transglucosylation i
i
C~
L4
cellobiose
product L3
Fig. 5. Radiochromatography (irrigant b) of the transglucosylation medium after 6 h of incubation with laminaran and labelled cellobiose; L3, L,~, and C3 show the locations of laminaritriose, laminaritetraose, and cellotriose, repectively
cose. Reduction of the oligosaccharide with sodium borohydride followed by hydrolysis gave unlabelled glucose and labelled glucitol, and treatment of the reduced oligosaccharide with fl-glucosidase gave unlabelled glucose and labelled cellobiitol, thus showing that glucosyl transfer to the reducing group of cellobiose had not occurred. The oligosaccharide, synthesised from cellobiose labelled in the non-reducing moiety was fully methylated (Jansson et al. 1976) and the hydrolysis products identified by TLC (irrigant e) and capillary-column gas chromatography as 2,3,4,6-tetra-O-methylglucose (unlabelled), 2,3,6-tri-O-methylglucose (unlabelled), and 2,3,4-tri-O-methylglucose (labelled) in roughly equal amounts. The oligosaccharide was tentatively identified as 4-O-fl-gentiobiosylglucose, i.e. the transglucosylation reaction introduced a (1 -*6)-fl-linked glucosyl residue at the C6-hydroxyl position of the non-reducing moiety of cellobiose. A similar, but less rigorous, characterisation showed that an analogous product was obtained from laminaribiose. Discussion
The fl-glucanase activities were determined using cotton-fibre homogenates since only part of the activity could be solubilised even in buffers of high ionic strength. No clear distinction could be made between the exo-(1 ~3)-fl-glucanase and the fl-glucosidase activities either on the basis of the rate of hydrolysis of the appropriate substrates, laminaran and laminaribiose, respectively (Barras et al. 1969), or on the basis of the susceptibility of the activities to the inhibitors glucono-l,5-1actone or nojirimycin (Reese et al. 1971). Preliminary fractionation studies of the buffer-soluble material, although resulting in the elimination of the endo-
535
(1--*3)-fl-glucanase activity, did not give any further indications of the nature of the two types of activity. A similar situation was encountered with an enzyme present in the walls of soybean cells cultured in vitro (Cline and Albersheim 1981). There is a clear difference between the endo(I ~ 3)-fl-glucanase activity, which remains more or less constant after 15 days post-anthesis throughout the fibre development, and the exo-(1--* 3)-/% glucanase activity. The former activity may be associated uniquely with the primary cell wall and may play a role in the elongation of the cotton fibre through hydrolytic, rather than transglycosylase activity (Cleland 1981), since there is good reason to believe that the observed transglucosylation may be attributed to the exo-(l~3)-fl-glucanase or fl-glucosidase activities. These two latter activities increase regularly with fibre development to reach a maximum late on (40 days post-anthesis). The increase may be simply related to the increase in cell-wall weight, but it is interesting to compare the maximum in the callose content at 28 days post-anthesis with the curve obtained for the total /%glucanase activity (Fig. 3); the activity also diminishes when most of the callose has been metabolised. The (1 ~3)-fl-glucan hydrolase activity may then be related to the reported turnover of the callose (Meier et al. 1981) since the glucose produced on hydrolysis could re-enter the low-molecular-weight sugar pools and be used inter alia for the synthesis of cellulose. F r o m the analyses of Jaquet et al. (1982) it can be calculated that the callose pool of a fibre should undergo turnover two (at the beginning of secondary wall formation) to about eight (at later stages) times within 24 h in order to deliver the glucose residues necessary for the synthesis of cellulose during the same period (Buchala and Meier 1985). If the rate of hydrolysis in vivo of the endogenous callose is anything similar to that in vitro of the model substrate, laminaran, then the amount of glucose that would be produced is more than sufficient to account for the synthesis of cellulose at any stage of secondary wall formation. The (1 -~4)-fl-glucanase (CMC-ase) activity was found to be almost negligible throughout fibre development. The possibility that the/~-glucanases could also function as transglucosylases was also examined. When fibre homogenates or solubilised crude pglucanase preparations were incubated with high concentrations of suitable donor molecules, e.g. laminaran, and of suitable acceptor molecules, e.g. cellobiose, transglucosylation occurred. With cellobiose a trisaccharide with (1 ~6)-fl-linked glucose was obtained. By enzymic hydrolysis and methyla-
536
tion analysis the trisaccharide was shown to be 4-O-fl-gentiobiosylglucose. This type of transfer corresponds to the introduction of branching into linear polysaccharide chains and it is interesting to note that the callose isolated from cotton fibres does contain some (l~6)-linkages (Meier et al. 1981). N o such linkages have been reported for cellulose. In addition to the trisaccharide a small amount of a glucose-containing polymer was obtained and further work is being carried out on this material. The role o f the fl-glucanases is still unclear. Although the endoglucanase may be involved in cellwall loosening at the primary cell-wall elongation phase, the absolute amount of callose present in the cell-wall at that stage is relatively low. The enzyme may also cooperate with the exoglucanases in a hydrolytic degradation of the cell-wall callose at the secondary cell-wall stage. It is difficult to assign a role to the transglucosylase activity until the enzyme has been purified and characterised or before a quantitative estimation can be made of its activity in the cotton fibre during development. The authors wish to thank Meiji Seika Kaisha Inc. for kindly supplying the nojirimycin. This work was supported by the Swiss National Science Foundation.
References Barras, D.R., Moore, A.E., Stone, B.A. (1966) Enzyme-substrate relationship among fl-glucan hydrolases. Adv. Chem. Series 95, 105-138 Buchala, A.J., Fraser, C.G., Wilkie, K.C.B. (1972) An acidic galactoarabinoxylan from the stem of Arena sativa. Phytochemistry 11, 2803-2814 Buchala, A.J., Meier, H. (1985) Biosynthesis of fl-glucans in growing cotton (Gossypiurn arboreum L. and Gossypium hirsuture L.) fibres. In: Biochemistry of plant cell walls, pp. 221-241. Brett, C., Hillman, J., eds. SEB Seminar 28, Cambridge University Press, Cambridge Canevascini, G., Coudray, M.R., Rey, J.P., Southgate, R.J.G., Meier, H. (1979) Induction and catabolite repression of cellulase synthesis in the thermophilic fungus Sporotrichum thermophile. J. Gen. Microbiol. 110, 291-303
P. Bucheli et al. : fl-Glucanases in cotton fibres Clarke, A.E., Stone, B.A. (1962) fl-l,3-Glucan hydrolases from the grape vine (Vitis vinifera) and other plants. Phytochemistry 1, 175-188 CMand, R.E. (1981) Wall extensibility: Hormones and wall extension. In: Encyclopedia of plant physiology, N.S., vol. 13B, pp. 255-2273 Tanner, W., Loewus, F.A., eds. Springer Berlin Heidelberg New York Cline, K., Albersheim, P. (1981) Purification and characterization of a beta glucosyl hydrolase/transferase in the walls of soy bean cells. Plant Physiol. 68, 221-228 Dahlqvist, A. (1974) Disaccharidasen, In: Methoden der enzymatischen Analyse, 3rd edn., pp. 95~957, Bergmeyer, U., ed. Verlag Chemic, Weinheim Huwyler, H.R. (1978) Die Zellwandkomponenten und Glycosidasenaktivitgten in Baumwollharen (Gossypium arboreum L.) verschiedener Entwicklungstadien. Doctoral Thesis, University of Fribourg, Switzerland Jansson, P.E., Kenne, L., Liedgren, H., Lindberg, B., L6nngren, J. (1976) A practical guide to the methylation analysis of carbohydrates. Univ. Stockholm Chem. Commun. 8, 1-75 Jaquet, J.P., Buchala, A.J., Meier, H. (1982) Changes in the nonstructural carbohydrate content of cotton (Gossypium spp.) fibres at different stages of development. Planta 156, 481-486 Lineweaver, H., Burk, D. (1934) The determination of enzyme dissociation constants. J. Am. Chem. Soc. 56, 658-666 Meier, H. (1983) Biosynthesis of (1--.3)- and (1 ~4)-fl-D-glucans in cotton fibers (Gossypium arboreum and Gossypium hirsuturn), J. Appl. Polymer Sci. 37, 123-130 Meier, H., Buchs, L., Buchala, A.J., Homewood, T. (1981) (1 --*3)-fl-D-Glucan (callose) is a probable precursor of cellulose of cotton fibres. Nature 289, 821-822 PilloneI, Ch., Buchala, A.J., Meier, H. (1980) Glucan synthesis by intact cotton fibres fed with different precursors at the stages of primary and secondary wall formation. Planta 149, 306,312 Reese, E.T., Parrish, F.W., Ettlinger, M. (1971) Nojirimycin and D-glucono-l,5-1actone as inhibitors of carbohydrases. Carbohydr. Res. 18, 381 388 Somogyi, M. (1952) Notes on sugar determination. J. Biol. Chem. 195, 139-147 Stahl, E. (1969) Thin-layer chromatography. Springer, New York Trevelyan, W.E., Proctor, D.P., Harrison, J.S. (1950) Detection of sugars on paper chromatograms. Nature 166, 444445
Received 13 May; accepted 5 July 1985
Planta (1955)166:530-536
9 Springer-Verlag 1985
p-Glucanases in developing cotton (Gossypium hirsutum L.) fibres * P. Bucheli, M. Dfirr, A.J. Buchala** and H. Meier Institut de Biologic v6g6tale et de Phytochimie, Universit6 de Fribourg, CH-1700 Fribourg, Switzerland
Abstract. Cotton fibres possess several fl-glucanase activities which appear to be associated with the cell wall, but which can be partially solubilised in buffers. The main activity detected was that of an exo-(1-,3)-fl-D-glucanase (EC 3.2.1.58) but which also had the characteristics of a fl-glucosidase (EC 3.2.1.21). Endo-(l~3)-fl-D-glucanase activity (EC 3.2.1.39) and much lower levels of (1 ~4)-fl-D-glucanase activity were also detected. The exo-(l ~ 3 ) fl-glucanase showed a maximum late on (40 days post-anthesis) in the development of the fibres, whereas the endo-(l~3)-fl-glucanase activity remained constant throughout fibre development. The fl-glucanase complex associated with the cotton-fibre cell wall also functions as a transglucosylase introducing, inter alia, (1 ~6)-fl-glucosyl linkages into the disaccharide cellobiose to give the trisaccharide 4-O-fl-gentiobiosylglucose. Key words: Callose - Cellulose - f l - G l u c a n a s e fl-Glucosidase - Gossypium (fl-glucanase) - Transglucosylation.
or an obligatory intermediate in the formation of cellulose. Either of these possibilities imply the involvement of fl-glucanases or transglucosylases. Preliminary work showed the presence of several glycosidases (Huwyler 1978) and fl-D-glucanases bound to the cotton-fibre cell wall. Such enzymes have also been found associated with the cell walls of many other higher plants (Clarke and Stone 1962) but their physiological role is not clear. Some hydrolases can also behave as transglycosylases, catalysing the transfer of glycosyl residues to oligosaccharides or polysaccharides. The present work was undertaken in order to characterise the fl-glucanase or fl-glucosidase activity in cotton fibres and to study any eventual variations of these activities during fibre development. Material and methods Plant material. Cotton plants (Gossypium hirsutum L. var. Stone~ ville No. 406) were grown in a greenhouse at a temperature of 30 ~ C during the day and 20 ~ C at night. The fruits, at various stages of development, were all harvested at the same time and kept on ice until opened. Preparation of the fibre cell-wall homogenates. As soon as possi-
Introduction
Recent studies on the incorporation of radioactive precursors into the polysaccharides of the cell wall of intact cotton fibres have shown that one of the products, a (1 ~3)-fl-D-glucan (callose), undergoes a turnover during the stage of secondary cell-wall formation (Meier et al. 1981 ; Meier 1983). It was not clear, however, if the callose was a facultative * Presented at the Third Cell Wall Meeting held in Fribourg in 1984 ** To whom correspondence should be addressed
Abbreviations: CMC=carboxymethylcellulose; O N P G = o - n i trophenyl-fl-D-glucopyranoside; T L C = t h i n - l a y e r chromatography
ble, the fruit capsules were dissected and the individual seed clusters separated. The fibres were quickly detached from the seeds in acetate buffer (50 mM, pH 5.5), finely cut with scissors and quantitatively transferred to the chamber of a McCrone Micronising Mill (Lovibond, London, U K ) equipped with corundum mill elements. Homogenisation was carried out at 4 ~ C for 30 min in the above buffer and the homogenate was quantitatively transferred to a dialysis tube and exhaustively dialysed against distilled water at 4 ~ C. The final volume was measured and aliquots were taken for analysis.
Extraction of low-molecular-weight sugars. One seed cluster from each fruit was taken for ulterior analysis of the neutral sugars and sugar phosphates (results not given here) and for determination of the cell-wall dry weight and callose content. The extraction with hot aqueous 80% methanol (v/v) and the determination of the cell-wall dry weight were carried out as described previously (Jaquet et al. 1982).
P. Bucheli et al. : fl-Glucanases in cotton fibres
Determination of the callose content of the cotton fibres. The method involving extraction with hot dimethyl sulphoxide, followed by acid hydrolysis of the isolated product, was carried out as described by Jaquet et al. (1982). This method has been shown to be reproducible and to be quite quantitative (Pillonel et al. 1980). Enzyme assays. Preliminary experiments were carried out in order to determine the extent of autohydrolysis under the conditions to be used for the enzyme assays since the homogenate contains not only the activities to be determined, but also their endogeneous substrate, callose. It was found that for the relatively short period of incubation used no correction for autolysis was necessary. The total (1 ~3)-/~-glucanase activity (endo + exo) was measured as the release of reducing sugars, estimated by the colorimetric method of Nelson and Somogyi (Somogyi 1952). Digests contained laminaran (0.17 % ; United States Biochemical Corp., Cleveland, Ohio, USA), acetate buffer (50 mM, pH 5.5) and fibre homogenate (5 ml) to give a final volume of 6 ml. Incubation was carried out at 35~ in a gyratory shaker. Samples (0.5 ml) were withdrawn after 0, 1, and 2 h and added directly to the copper-containing reagent in order to stop the reaction. Before the final absorbance measurements, the reaction mixture was centrifuged (10 rain, 10000 g). The exo-(l~3)-fl-glucanase activity was measured as the release of glucose, determined by the glucose-oxidase Perid method (Boehringer, Mannheim, FRG), obtained under the condition described above for the determination of the total (1--+3)-/~-glucanase activity. Samples (0.6 ml) were withdrawn after 0, 1, and 2 h and added directly to a solution of trichloroacetic acid (10%, 0.4 ml) to stop the reaction. Aliquots (0.2 ml) were taken for analysis. Endo-(l~3)-~-glucanase activity was determined as described above for the total (1-,3)-p-glucanase activity, but the incubation mixture also contained 10-4M nojirimycin (Meiji Seika Kaisha Inc., Yokohama, Japan). An additional control was carried out - the glucose released under these conditions, determined by the glucose-oxidase method, was negligible. /~-Glucosidase activity was determined by incubating o-nitrophenyl-fl-D-glucoside (ONPG; 0.2%; Fluka AG, Buchs, Switzerland) or laminaribiose (10mM) in acetate buffer (50 mM, pH 5.5) with fibre homogenate (1 ml) in a total volume of 2 ml at 35 ~ C. The reactions were terminated by the addition of trichloroacetic acid (10%, 2 ml) after 0, 1 and 2 h. After centrifugation (10 min, 1000 g) the glucose released was measured using the glucose-oxidase Perid method, as above, but the fl-glucosidase known to be present in the commercial enzyme preparation was totally inhibited by adjusting the pH of the enzyme-buffer reagent to 8.5 with 1 M 2-amino-2-(hydroxymethyl)-l,3-propanediol immediately before analysis (Dahlqvist 1974). (l~4)-fl-glucanase (carboxymethylcellulase) activity was measured as the release of reducing sugars, determined by the Nelson-Somogyi method (Somogyi 1952), when carboxymethylcellulose (CMC; 4M6F; Hercules, Wilmington, USA; at a final concentration of 0.5% in 50 mM acetate buffer at pH 5.0) was incubated with fibre homogenate (1 ml) at 35 ~ C for 6 h. In each of the above methods, the change in absorbance compared with that of the sample taken at 0 h was measured. The results are expressed in gmol glucose (or glucose-equivalents) released per hour for the fibres of one seed, i.e. on a constant cell-number basis. The enzyme assays were carried out in duplicate for each fruit at a particular stage of development, as were the individual sugar determinations, with good reproducibility, and the mean values are reported.
531
Chromatographic analysis of hydrolysis products. Paper chromatography was carried out on Schieicher and Schiill No. 2043b paper (Schleicher and Schiill, Dassel, FRG) using the following irrigants (a) ethyl acetate: pyridine: water (8 : 2 : 1, by vol.), (b) propan-l-ol:ethyl acetate :water (7 : 1 : 2, by vol.) and thin-layer chromatography (TLC) on Kieselgel 60 (Merck, Darmstadt, F R G with the irrigants (c) acetone:water (88:12, v/v), (d) acetone: water (85 : 15, v/v), (e) butan-l-ol : acetic acid : ether: water (9:6:3:1.5, by vol.) or (f) benzene:ethanol:acetic acid:water (200:47 : 15 : 1, by vol.). Detection was either with alkaline AgNO3 (Trevelyan et al. 1950) or with ~-naphthol/conc. H2SO4 (Stahl 1969) as appropriate; radioactive sugars on chromatograms were detected using a TLC Linear Analyser LB 282 (Berthold, Wildbad, FRG). Preparation of radioactive substrates, fl-Glucans (callose and cellulose) in cotton fibres were labelled with [14C]glucose by incubating isolated seed clusters with [U-l~C]sucrose as described by Pillonel et al. (1980). Cellobiose, uniformly 14C-labelled only in the reducing moiety, and cellobiose, uniformly ~4C-labelled only in the non-reducing glucose moieties were prepared enzymatically using cellobiose phosphorylase (EC 2.4.1.20) derived from Cellvibrio gilvus (Canevascini et al. 1979). Autolysis. Autolysis experiments were carried out with finely cut (razor blade) detached cotton fibres where the endogenous fl-glucans had been labelled (see above). Weighed aliquots of the fibres were thoroughly washed free of low-molecular-weight material by suspending them in Miracloth (Checopee Manufacturing Co., Cornelia, Ca., USA) bags in deionised water at 0 ~ overnight. Aliquots (0.5 g) were suspended in acetate buffer (50 mM, pH 5.5) and incubated at 30 ~ C. At various intervals the autolysis was terminated by heating the samples for 5 min in boiling water. The samples were centrifuged (10 min, 10000g), the radioactivity in the supernatant measured, and the glucose and reducing sugars liberated were determined by the glucose-oxidase and Nelson-Somogyi methods, respectively. For the inhibition experiments the incubation medium contained D-glucono-l,5-1actone (10-5-10-ZM) or nojirimycin (1 mM). In the control experiments the samples were heat-inactivated before incubation. Transglucosylation activity. (i) With detached cotton fibres. The fibres were prepared as described above for the autolysis experiments. Aliquots were incubated with cellobiose (290 mM) or cellotriose (100 raM) for 1 h at 35 ~ C in acetate buffer (50 raM, pH 5.5). The water-soluble material was examined by TLC (irrigant d). (ii) With an enzyme fraction solubilised from the cell wall. Fibres were homogenised as described for the/~-glucanase assays, but the simple acetate buffer was replaced with a buffer of high ionic strength containing, in addition, sodium azide (0.02%), glycerine (5%), ethane thiol (10 mM) and lithium chloride (3 M). The homogenate was then filtered through four layers of Miracloth, the filtrate centrifuged (10 min, 10000 g) and the supernatant was exhaustively dialysed against distilled water at 5~ C. The enzyme-containing solution (0.5 ml) was incubated with laminaran (0.5%) and cellobiose (100 mM; approx. 3 000 Bq), where the reducing moiety contained uniformly 14C-labelled glucose, for 6 h at 35 ~ C. The reaction mixture was examined by paper chromatography (irrigant b) and TLC (irrigant c) and then fractionated by chromatography on a column (50 cm long, 2.2 cm inner diameter) of Fractogel TSK HW-40(S) (Merck, Darmstadt, FRG). The major labelled product was re-examined by paper chromatography (irrigant b, Rg~c 0.18) and TLC (irrigant d, RgLc0.52) and found to be chromatographically homogeneous.
532
Analysis of the transglucosylation product. A sample of the product was reduced with NaBH4 and hydrolysed (2M trifluoroacetic acid; 1 h) as described by Buchala et al. (1972) for xylooligosaccharides. The products, glucose and glucitol, were separated by TLC (irrigant c) and the radioactivity measured - only glucitol was found to be labelled. The reduced oligosaccharide was treated with the fl-glucosidase preparation from almond emulsin (Fluka, Buchs, Switzerland) in acetate buffer (50 raM, pH 5.5) and the products examined by TLC (irrigant d) - glucose and labelled cellobiitol were found. The unreduced oligosaccharide, synthesised from eellohiose labelled in the non-reducing glucose moiety, was methylated and hydrolysed (Jansson et al. 1976). Thin-layer chromatography of the hydrolysate (irrigant e) gave 2,3,4,6-tetra-O-methylglucose, 2,3,6-tri-O-methylglucose and 2,3,4-tri-O-methylglucose, where only the latter tri-O-methylglucose was radioactive. No di-O-methylglucose derivatives were detected. The identities of these sugars were confirmed by capillary column gas chromatography (WCOT OV-225, Chrompack B.V., Middelburg, The Netherlands; 25 m, 170 ~ C) of the alditol acetate derivatives of the methyl sugars by comparison with authentic standards (Jansson et al. 1976).
Results
Qualitative examination of the fl-glucanase activity. Crude extracts of detached cotton fibres and fibre homogenates contain several fl-glucanase activities. Laminaran, periodate-oxidised laminaran and pachyman were effectively degraded to give glucose and a series of laminarioligosaccharides indicating the presence of endo-(l~3)-fl-D-glucanase, exo(l~3)-fl-D-glucanase or fl-glucosidase activities. The latter activity was also demonstrated by the degradation of glucose-containing fl-linked oligosaccharides and ONPG. The polysaccharides lichenan and oat caryopsis fl-glucan, which both contain (1--+3)-fl-, and (l~4)-fl-D-glucosidic linkages, and to a lesser extent CMC, could also be hydrolysed showing the presence of (1 ~4)-fl-D-glucanase activity. All of these activities were detected both at the primary and secondary cell-wall stages of fibre development. On the other hand, starch, pullulan, pustulan, and nigeran were not degraded (see Table 1). No systematic study of the solubility of the flglucanases was made. However, the activity solubilised depended on the ionic strength of the extracting buffer; with the low-ionic-strength buffer used for homogenisation of the fibres for the quantitative studies described below, the activity was essentially associated with the buffer-insoluble material, whilst about 60% of the fl-glucanase activity (determined either with O N P G or laminaran) could be solubilised if the homogenisation buffer contained 3 M LiC1.
Definition of the various fl-glucanase activities. Although the distinction between the exo-(l-~3)-fl-
P. Bueheli et al. : fl-Glucanases in cotton fibres Table 1. Action of the cell-wall-bound fl-glucanases on various substrates. Aliquots of finely cut fibres, free of low-molecularweight carbohydrates, were incubated with the suhstrate (10raM for low-molecular-weight substrates and 0.5% for polysaccharides) at 30~ for 1 h. The glucose released was measured by the glucose-oxidase method ~ (for the low-molecular-weight substrates) and reducing sugars (for polysaccharides) by the Nelson-Somogyi method b. The values are corrected for cell-wall autolysis Substrate
Rate of hydrolysis (gmol glc seed- 1 h - "
Sophorose Laminaribiose Cellobiose Gentiobiose ONPG Maltose p-Nitrophenyl-fl-Dglucoside Laminaran Pachyman Lichenan CMC Pustulan Pullulan Nigeran Starch
1.6 5.6 0.7 4.0 3.1 0.02 0.01 4.6 2.0 0.9 0.02 0.01 0.02 0
(gmol glc equiv. seed- 1 h - b m
m
5.1 2.5 2.5 0.04 0 0 0 0
glucanase and the fl-glucosidase activities is not clear, the former activity is referred to when laminaran was used as substrate and where the glucose released was measured. The latter activity is referred to when laminaribiose or O N P G was used as the substrate. Total (1--->3)-fl-glucanase activity (exo + endo) was measured using laminaran as substrate and determining the reducing sugars (expressed in glucose equivalents) released. Endo-(1--->3)-fl-glucanase activity could not be satisfactorily measured viscosimetrically using carboxymethylpachyman as substrate (Clarke and Stone 1962). Nojirimycin is a known inhibitor of fl-glucosidase, and to a lesser extent, exo-fl-glucanase activity. It was found that with nojirimycin at a concentration of 10-4M and laminaran as substrate, only laminarioligosaccharides, and no glucose, were produced. The reducing sugars thus released, could be taken as a measure of the endo(l~3)-fl-glucanase activity. Similar values were obtained by substracting the value for the exo-(l ~3)-fl-glucanase activity (glucose measured by the glucose-oxidase method) from the value for the total (1---,3)-fl-glucanase activity (measured by the Somogyi-Nelson method), thereby showing the validity of the inhibition method for short-term incubations.
P. Bucheli et al. : fl-Glucanases in cotton fibres
0
533
(a)
,-
A
Control
start
Laminaribiose
glucose
~ A
0o
time(h)
Fig. l a , b. Autolysis of (l~3)-fl-glucan in detached cotton fibres: a seed clusters were incubated with 14C-sucrose in order to label the (1 ~ 3)-fl-glucan in situ. The fibres were detached, washed and incubated at 35 ~ C and at various intervals the radioactivity in the incubation medium was measured. In the control experiment the enzymes in fibres were deactivated by boiling for 10 rain in the incubation buffer; b radiochromatographic analysis (irrigant b) of the soluble products after 8 h of incubation in the absence and presence of nojirimycin
Characteristics of the various fl-glucanase activities. An optimum value of about pH 5.8 was obtained for the exo-(1--,3)-~-glucanase activity using laminaran as substrate. A slightly lower value of pH 5.5 was obtained for the fl-glucosidase activity with either laminaribiose or O N P G as substrate. All the enzymic determinations were, however, carried out at pH 5.5. The Km value for the fl-glucosidase was determined for the substrates O N P G (5 mM), laminaribiose (6.2 mM), gentiobiose ( l . 4 m M ) , cellobiose (0.6mM) and sophorose (0.4 mM) by the method of Lineweaver and Burk (1934) with fibre homogenates. Similarly, values of 0.04 m M and 0.12 m M were obtained for the exo-(l~3)-fl-glucanase with laminaran as substrate and determining the reducing sugars or the glucose released, respectively. These latter values are only apparent since clearly the presence of the endo-(l--,3)-/?-glucanase has an effect on the substrate concentration; effectively reducing its concentration for a true exo-enzyme or increasing its concentration for a true glucosidase. The value obtained for the endo-(l ~3)-/~-glucanase was 0.10 raM. The effect of various enzyme inhibitors was sought and it was found that D-glucono-l,5-1actone and nojirimycin inhibited the hydrolysis of ONPG, laminaribiose and laminaran. Attempts were made to distinguish the (1--*3)-fl-exoglucanase from the /?-glucosidase since it has been reported (Reese et al. 1971) that much higher concentrations of the inhibitors are necessary to inhibit the former activity. With glucono-l,5-1actone,
a concentration of 10-5M, normally sufficient to inhibit fl-glucosidase, had little effect when laminaribiose was the substrate. At higher concentrations (greater than 5.10-3M), both types of activity were inhibited. Similar results were obtained with nojirimycin except that total inhibition was achieved at 10- 4M.
Cell-wall autolysis. When the fibre homogenates were incubated in buffer at pH 5.5, there was in some cases a appreciable release of glucose caused by the autolysis of the cell wall which contains, in addition to the fl-glucanases, the endogenous substrates callose and cellulose. It was observed that the rate of autolysis depended on the length of time of dialysis of the fibre homogenate, when the endogenous callose was slowly degraded and the products eliminated. Fibre homogenates were prepared from intact seed clusters which had been incubated with 14C-sucrose (Pillonel et al. 1980) in order to label the endogenous/~-glucans. Autolysis experiments showed that 25% of the labelled callose could be degraded within 8 h. Figure 1 a shows the results obtained for a typical experiment. Chromatographic analysis of the products obtained after 8 h showed that glucose was the only labelled low-molecular-weight product (see Fig. 1 b). When nojirimycin (1 mM) was included in the incubation medium the rate of release of radioactivity into the incubation medium was reduced (not shown) and a appreciable amount of laminarioligosaccharides (principally laminaribiose) was detected.
534
P. Bucheli et al. : fl-Glucanases in cotton fibres
750
5
50 5O
500
,to ~
~ol
3
~
2
30 ~>,
..9o
o
4
I
250
t 20
10
1'5
2'1 2'4
2'8
3=6 410 4 4
Days post anthesis
Fig. 2. Changes in the cell-wall dry weight ( o - - o ) and in the
callose content (o--o) of the cotton fibres during development; the values were calculated for the fibres of one seed
5 p-,,
4 3 2 o
1 ll5
2'1 2'4
218
316 410 4'4
Days post anthesis
Fig. 3. Changes in the fl-glucanase activities of the cotton fibres during development; (e--o) total ~-glucanase, (zx--zx)exo-(l ~ 3)-~-glucanase, (i--i) endo-(1~3)-/?-glucanase
Relationship between the various fl-glucanase activities and fibre development. In order to relate the fl-glucanase activities to fibre development, several previously established sampling criteria (Jaquet et al. 1982) had to be satisfied for the series of fruits harvested. A regular increase in the cell-wall dry weight and the presence of a distinct maximum in the (1 ~3)-fl-glucan content at the stage corresponding to the beginning of massive secondary wall formation (Fig. 2) indicated the success in sampling. The enzyme activities are expressed in micromoles of glucose or glucose-equivalents released per hour for the fibres of one seed, i.e. on a constant cell number basis and are corrected where appropriate for autolysis.
ll5
2'1 214 " 218
3'6
4'0
4'4
Days post anthesis
Fig. 4. Comparison of the exo-(J~3)-fl-glucanase activity (~x--A) of cotton fibres during development with the glucosidase activity obtained using laminaribiose ( o - - o ) or O N P G ( o - - o ) as substrates
The changes in the various fl-glucanase activities, which occurred during fibre development, are shown in Fig. 3. It can be seen that during the formation of the primary cell wall, i.e. up to about 24 days post-anthesis, the absolute values for the activities were relatively low. The total (1--, 3)-flglucanase activity increased regularly throughout fibre development to reach a maximum at 40 days post-anthesis and then decreased. This same pattern was found for the exo-(] ~3)-fl-glucanase and fl-glucosidase activities; Fig. 4 compares the latter two activities. On the other hand, the endo-(l --*3)fl-glucanase activity was more or less constant after 15 days post-anthesis throughout fibre development (Fig. 3).
Transglucosylase activity. When fibre homogenates, or material solubilised with 3 M LiC1 from fibre homogenates, was incubated with high concentrations (100 mM) offl-linked, glucose-containing, homooligosaccharides as acceptor molecules and laminaran (0.5%) as donor molecule, new oligosaccharides were synthesised. The transglucosylation appeared to introduce (l~6)-fl-glucosyl linkages into the acceptor molecules; this was demonstrated for cellobiose. Under the above conditions, cellobiose 14C-labelled in the reducing glucose moiety, gave at least two labelled products (Fig. 5). The major product had a chromatographic mobility corresponding to a tri-, or a tetrasaccharide and was not cellotriose, cellotetraose, laminaritriose nor laminaritetraose. The oligosaccharide was purified by molecular-size chromatography and partially characterised. Acid hydrolysis or treatment with a fl-glucosidase gave only glu-
P. Bucheli et al. : fl-Glucanases in cotton fibres
A start
transglucosylation i
i
C~
L4
cellobiose
product L3
Fig. 5. Radiochromatography (irrigant b) of the transglucosylation medium after 6 h of incubation with laminaran and labelled cellobiose; L3, L,~, and C3 show the locations of laminaritriose, laminaritetraose, and cellotriose, repectively
cose. Reduction of the oligosaccharide with sodium borohydride followed by hydrolysis gave unlabelled glucose and labelled glucitol, and treatment of the reduced oligosaccharide with fl-glucosidase gave unlabelled glucose and labelled cellobiitol, thus showing that glucosyl transfer to the reducing group of cellobiose had not occurred. The oligosaccharide, synthesised from cellobiose labelled in the non-reducing moiety was fully methylated (Jansson et al. 1976) and the hydrolysis products identified by TLC (irrigant e) and capillary-column gas chromatography as 2,3,4,6-tetra-O-methylglucose (unlabelled), 2,3,6-tri-O-methylglucose (unlabelled), and 2,3,4-tri-O-methylglucose (labelled) in roughly equal amounts. The oligosaccharide was tentatively identified as 4-O-fl-gentiobiosylglucose, i.e. the transglucosylation reaction introduced a (1 -*6)-fl-linked glucosyl residue at the C6-hydroxyl position of the non-reducing moiety of cellobiose. A similar, but less rigorous, characterisation showed that an analogous product was obtained from laminaribiose. Discussion
The fl-glucanase activities were determined using cotton-fibre homogenates since only part of the activity could be solubilised even in buffers of high ionic strength. No clear distinction could be made between the exo-(1 ~3)-fl-glucanase and the fl-glucosidase activities either on the basis of the rate of hydrolysis of the appropriate substrates, laminaran and laminaribiose, respectively (Barras et al. 1969), or on the basis of the susceptibility of the activities to the inhibitors glucono-l,5-1actone or nojirimycin (Reese et al. 1971). Preliminary fractionation studies of the buffer-soluble material, although resulting in the elimination of the endo-
535
(1--*3)-fl-glucanase activity, did not give any further indications of the nature of the two types of activity. A similar situation was encountered with an enzyme present in the walls of soybean cells cultured in vitro (Cline and Albersheim 1981). There is a clear difference between the endo(I ~ 3)-fl-glucanase activity, which remains more or less constant after 15 days post-anthesis throughout the fibre development, and the exo-(1--* 3)-/% glucanase activity. The former activity may be associated uniquely with the primary cell wall and may play a role in the elongation of the cotton fibre through hydrolytic, rather than transglycosylase activity (Cleland 1981), since there is good reason to believe that the observed transglucosylation may be attributed to the exo-(l~3)-fl-glucanase or fl-glucosidase activities. These two latter activities increase regularly with fibre development to reach a maximum late on (40 days post-anthesis). The increase may be simply related to the increase in cell-wall weight, but it is interesting to compare the maximum in the callose content at 28 days post-anthesis with the curve obtained for the total /%glucanase activity (Fig. 3); the activity also diminishes when most of the callose has been metabolised. The (1 ~3)-fl-glucan hydrolase activity may then be related to the reported turnover of the callose (Meier et al. 1981) since the glucose produced on hydrolysis could re-enter the low-molecular-weight sugar pools and be used inter alia for the synthesis of cellulose. F r o m the analyses of Jaquet et al. (1982) it can be calculated that the callose pool of a fibre should undergo turnover two (at the beginning of secondary wall formation) to about eight (at later stages) times within 24 h in order to deliver the glucose residues necessary for the synthesis of cellulose during the same period (Buchala and Meier 1985). If the rate of hydrolysis in vivo of the endogenous callose is anything similar to that in vitro of the model substrate, laminaran, then the amount of glucose that would be produced is more than sufficient to account for the synthesis of cellulose at any stage of secondary wall formation. The (1 -~4)-fl-glucanase (CMC-ase) activity was found to be almost negligible throughout fibre development. The possibility that the/~-glucanases could also function as transglucosylases was also examined. When fibre homogenates or solubilised crude pglucanase preparations were incubated with high concentrations of suitable donor molecules, e.g. laminaran, and of suitable acceptor molecules, e.g. cellobiose, transglucosylation occurred. With cellobiose a trisaccharide with (1 ~6)-fl-linked glucose was obtained. By enzymic hydrolysis and methyla-
536
tion analysis the trisaccharide was shown to be 4-O-fl-gentiobiosylglucose. This type of transfer corresponds to the introduction of branching into linear polysaccharide chains and it is interesting to note that the callose isolated from cotton fibres does contain some (l~6)-linkages (Meier et al. 1981). N o such linkages have been reported for cellulose. In addition to the trisaccharide a small amount of a glucose-containing polymer was obtained and further work is being carried out on this material. The role o f the fl-glucanases is still unclear. Although the endoglucanase may be involved in cellwall loosening at the primary cell-wall elongation phase, the absolute amount of callose present in the cell-wall at that stage is relatively low. The enzyme may also cooperate with the exoglucanases in a hydrolytic degradation of the cell-wall callose at the secondary cell-wall stage. It is difficult to assign a role to the transglucosylase activity until the enzyme has been purified and characterised or before a quantitative estimation can be made of its activity in the cotton fibre during development. The authors wish to thank Meiji Seika Kaisha Inc. for kindly supplying the nojirimycin. This work was supported by the Swiss National Science Foundation.
References Barras, D.R., Moore, A.E., Stone, B.A. (1966) Enzyme-substrate relationship among fl-glucan hydrolases. Adv. Chem. Series 95, 105-138 Buchala, A.J., Fraser, C.G., Wilkie, K.C.B. (1972) An acidic galactoarabinoxylan from the stem of Arena sativa. Phytochemistry 11, 2803-2814 Buchala, A.J., Meier, H. (1985) Biosynthesis of fl-glucans in growing cotton (Gossypiurn arboreum L. and Gossypium hirsuture L.) fibres. In: Biochemistry of plant cell walls, pp. 221-241. Brett, C., Hillman, J., eds. SEB Seminar 28, Cambridge University Press, Cambridge Canevascini, G., Coudray, M.R., Rey, J.P., Southgate, R.J.G., Meier, H. (1979) Induction and catabolite repression of cellulase synthesis in the thermophilic fungus Sporotrichum thermophile. J. Gen. Microbiol. 110, 291-303
P. Bucheli et al. : fl-Glucanases in cotton fibres Clarke, A.E., Stone, B.A. (1962) fl-l,3-Glucan hydrolases from the grape vine (Vitis vinifera) and other plants. Phytochemistry 1, 175-188 CMand, R.E. (1981) Wall extensibility: Hormones and wall extension. In: Encyclopedia of plant physiology, N.S., vol. 13B, pp. 255-2273 Tanner, W., Loewus, F.A., eds. Springer Berlin Heidelberg New York Cline, K., Albersheim, P. (1981) Purification and characterization of a beta glucosyl hydrolase/transferase in the walls of soy bean cells. Plant Physiol. 68, 221-228 Dahlqvist, A. (1974) Disaccharidasen, In: Methoden der enzymatischen Analyse, 3rd edn., pp. 95~957, Bergmeyer, U., ed. Verlag Chemic, Weinheim Huwyler, H.R. (1978) Die Zellwandkomponenten und Glycosidasenaktivitgten in Baumwollharen (Gossypium arboreum L.) verschiedener Entwicklungstadien. Doctoral Thesis, University of Fribourg, Switzerland Jansson, P.E., Kenne, L., Liedgren, H., Lindberg, B., L6nngren, J. (1976) A practical guide to the methylation analysis of carbohydrates. Univ. Stockholm Chem. Commun. 8, 1-75 Jaquet, J.P., Buchala, A.J., Meier, H. (1982) Changes in the nonstructural carbohydrate content of cotton (Gossypium spp.) fibres at different stages of development. Planta 156, 481-486 Lineweaver, H., Burk, D. (1934) The determination of enzyme dissociation constants. J. Am. Chem. Soc. 56, 658-666 Meier, H. (1983) Biosynthesis of (1--.3)- and (1 ~4)-fl-D-glucans in cotton fibers (Gossypium arboreum and Gossypium hirsuturn), J. Appl. Polymer Sci. 37, 123-130 Meier, H., Buchs, L., Buchala, A.J., Homewood, T. (1981) (1 --*3)-fl-D-Glucan (callose) is a probable precursor of cellulose of cotton fibres. Nature 289, 821-822 PilloneI, Ch., Buchala, A.J., Meier, H. (1980) Glucan synthesis by intact cotton fibres fed with different precursors at the stages of primary and secondary wall formation. Planta 149, 306,312 Reese, E.T., Parrish, F.W., Ettlinger, M. (1971) Nojirimycin and D-glucono-l,5-1actone as inhibitors of carbohydrases. Carbohydr. Res. 18, 381 388 Somogyi, M. (1952) Notes on sugar determination. J. Biol. Chem. 195, 139-147 Stahl, E. (1969) Thin-layer chromatography. Springer, New York Trevelyan, W.E., Proctor, D.P., Harrison, J.S. (1950) Detection of sugars on paper chromatograms. Nature 166, 444445
Received 13 May; accepted 5 July 1985
