Purification and properties of the xylanases from the termite Macrotermes bellicosus and its symbiotic fungus Termitomyces sp.

~

Pergamon

Comp. Biochem. Physiol. Vol. I12B, No. 4, pp. 629-635, 1995 Copyright © 1995 Inc. Printed in Great Britain. All rights reserved 0305-0491/95 $9.50 + 0.00

ElsevierScience

0305-0491(95)00103-4

Purification and properties of the xylanases from the termite Macrotermes bellicosus and its symbiotic fungus Termitomyces sp. M. Matoub and C. Rouland Laboratoire d'Ecophysiologie des Invert6br6s, Universit6 Paris XII-Val de Marne, 94010 Cr6teil C6dex, France Four xylanases were purified, two from the termite Macrotermes bellicosus workers (X1T and X2T) and two from its symbiotic fungus Termitomyces sp. (X1Mc and X2Mc). The analysis of the step required for the purification of X1T and X1Mc and the comparison of their different properties suggested that xylanases X1T and X1Mc were the same enzyme, X1. The determination of the reducing sugars by TLC revealed that X1 was an endoxylanase (EC 3.2.1.8) and X2T and X2Mc were exoxylanases (EC 3.2.1.37). The apparent molecular weights of the three xylanases, determined by SDS-polyacrylamide gel electrophoresis, were 36 kDa for X1, 56 kDa for X2T and 22.5 kDa for X2Mc. The optimal pH of the three xylanases was - 5 . 5 , and Km values determined with birchwood xylan as substrate were 0.2% for X1, 0.1% for X2T and 0.3% for X2Mc, showing a high affinity for this substrate. The three enzymes differed also by their thermal stability. Key words: Xylanases; Purification; Characterization; Macrotermitinae; Fungus; Macrotermes bellicosus; Termitomyces sp.; Symbiosis.

Comp. Biochem. Physiol. I I2B, 629-635, 1995.

Introduction

Macrotermes bellicosus is a termite of the subfamily Macrotermitinae. These termites have a symbiotic relationship with a fungus from the genus Termitomyces. The fungus grows on structures built by termite workers, called fungus comb, that are enclosed into termite nests. Termitomyces is found as mycelium in the fungus comb and as white asexual nodules, called mycotetes, on the comb (Heim, 1977). The inferior part of fungus comb, degraded by fungus, is eaten up by termites for their nutrition. The fungus comb is essential for the termites' survival (Noirot, 1952; Sands, 1956; Ausat et al., 1960; Rouland et al., 1987), but the nature of symbiosis has not yet been established. Abo-Katwa (1978)

Correspondence to: Dr. M. Matoub. Received 17 February 1994; revised 20 April 1995; accepted 27 April 1995.

and Martin and Martin (1978, 1979) suggested, for Macrotermes natalensis and Macroterrnes subhyalinus, that the Termitomyces was a producer of cellulases that could be ingested by termites when they eat the fungus comb. Rouland et al. (1988-1989) demonstrated that in the digestive tract of the forest termite Macrotermes muelleri, enzymes produced by two different organisms co-existed: one endocellulase and one endoxylanase produced by the symbiotic fungus and ingested by termites and one exoceUulase synthesized by the termites salivary glands. The synergistic action of both cellulases significantly hydrolysed cellulose as a staple for termite nutrition. In opposite to these results, the correlation made between the cellulase fungus activity and the eating habits of M. michalseni and M. subhyalinus workers can supply no evidence to support the hypothesis of acquired enzymes in these termites (Veivers et al., 1991; Slaytor, 1992). These results clearly show that the symbiotic

629

630

M. Matoub and C. Rouland

role of the fungus garden must be regarded as not yet well known. It is established, as a law for the termites, that their first nutrient is the cellulose; then the degradation of xylan by termites and their symbiotic fungus was studied. In the savanna termite M. bellicosus, samples from both the worker's gut and from the mycotetes of the symbiotic fungus showed xylanase as the major hydrolytic activity (Rouland et al., 1991). To get new information on digestive symbiosis in fungus-growing termite, it seems to us more pertinent to study the xylanolytic complex in this species. We describe the procedure for the purification of the xylanases from M. bellicosus and its symbiotic fungus Termitomyces sp. and provides data about their properties.

Materials and Methods

Biological material M. bellicosus and its symbiotic fungus Termitomyces sp. came from the savanna of Lamto (Ivory Coast). They were collected directly from the nest and then stored frozen at - 20oc. Biochemical products Carob galactomannan was kindly provided by the laboratory of Biochemistry (University of Paris V). All the other substrates (oat spelt and birchwood xylan, ONP/3 xyloside, laminarin, starch, arabinogalactan) were obtained from Sigma Chemical (St. Louis, MO, U.S.A.). Hydroxyapatite was prepared by a modification (Levin, 1962) of the method of Tisselius et al. (1956). DEAE Sepharose came from Pharmacia Fine Chemicals (Uppsala, Sweden). CM Econopack was purchased from Bio-Rad (Richmond, CA, U.S.A.). QA Trisacryl was from Sepracor/IBF S.A. (Villeneuve La Garenne, France). Assay of enzymes activity During the purification procedures, xylanolytic activities were assayed by incubation in 100 /zl of enzyme solution, 100 /A of birchwood xylan (2%, w/v) and 50/zl of McIlvaine buffer (1921), pH 5.2. After 15 min of incubation at 37°C, the release of reducing sugars was determined by the Somogyi-Nelson (Somogyi, 1945; Nelson, 1944) micromethod (Williams et al., 1978). Protein concentrations were determined according to the Bradford (1976) dye binding assay using bovine albumin (Sigma) as a standard. Specific activity was expressed as/.~mol of xylose liberated per min (= U) and per mg of protein. Other enzyme assays were performed ac-

cording to the methods described in a previous study (Rouland et al., 1986).

Thin-layer chromatography (TLC) Products of the enzymatic hydrolysis of xylan were detected by TLC (silicagel-coated F1500 plastic sheets, Schleicher & Schiill, Germany). The solvent system was 7 : 3 : 3 : 2 : 2 : 3 N-propanol-EtOH-EtOAc-AcOH-pyridine-water. Chromatograms were subsequently sprayed with an alcoholic solution of/3-naphthylamine (Sigma) followed by a treatment with 2% H2SO 4 in EtOH (Petek, 1962). Xylose, xylobiose and xylotriose used as standard were obtained from Sigma. Polyacrylamide gel electrophoresis (PAGE) Eiectrophoresis was usually performed on 7.5% polyacrylamide disc gels (100 x 5 mm) (Maizel, 1964) with a current intensity of 6 mA/gel. The gels were stained with Coomassie Brilliant Blue (Rouland et al., 1988c). Molecular weights of purified enzymes were determined by electrophoresis on polyacrylamide gels using the Hedrick and Smith (1968) method and by SDS-PAGE following the procedure of Weber and Osborn (1969). Reference proteins used were phosphorylase a monomer, bovine serum albumin, glutamic acid dehydrogenase, chicken ovalbumin, pepsin, and trypsin inhibitor (Sigma). The four purified xylanases were tested for carbohydrate content by periodic acid-Schiff staining (Chippendale and Beck, 1966). Results

Purification procedures Preparation of crude extracts. All operations were performed at + 4°C. Old and young workers were mixed, because the screening of their enzymatic activities showed no differences (Matoub, 1993). The collected termites (30 g) or mycotetes (12 g) were homogenized with 180 ml (termites) or 80 ml (fungus) of 0.9% NaC1 (w/v) solution in an Ultra-turrax (Junkel Kunkel) and then sonicated as previously described (Rouland et al., 1988a). The homogenates were centrifuged at 20,000 g for 20 rain (Sigma 2 K 15). Solid ammonium sulfate was slowly added to the collected supernatants to a final concentration of 4.2 M (80% saturation). These precipitates, after being centrifuged as above, were suspended in I00 ml (termite homogenate) or 40 ml (mycotetes homogenate) of 2 mM potassium phosphate buffer, pH 5.3, and were dialysed overnight against the same buffer in a pig gut membrane (Ets Sousanna,

Purification and properties of termite xylanases

631

Thiais, France). The dialysed solution constituted the crude extract.

Purification of the xylanases from the termite M. bellicosus. The crude extract was applied to a column (2.5 x 7 cm) of hydroxyapatire equilibrated with 2 mM potassium phosphate buffer, pH 5.3. Elution with the same phosphate buffer provided a peak with xylanase activity. After dialysis, the active fraction was applied to a column (1.8 x 12 cm) of DEAE-Sepharose equilibrated with a 10 mM Tris-HC1 buffer, pH 7.6. The elution with a Tris-HCl buffer 10 mM + 10 mM NaC1 provided a peak with xylanase activity (X1T) (fraction 1). A second xylanase activity (X2T) was recovered with 10 mM + 100 mM NaCI Tris-HC1 buffer (fraction 2). Fraction 1 was electrophoretically pure (Fig. 1), and the single proteic band was associated with the xylanasic activity. The final X1T was 33-fold purified and its specific activity towards birchwood xylan was 54 U/mg of protein (Table 1). Xylanase XIT represented 15.5% of the overall activity recovered after DEAE-Sepharose column, whereas the second fraction represented 84.5%. After dialysis, fraction 2 was adsorbed onto a QA Tris Acryl column (1.8 x 5 cm) equilibrated with a 10 mM Tris-HC1 buffer, pH 8.2. The xylanase activity was eluted with the same buffer containing 10 mM NaCI. Purification was achieved by chromatography onto a CM Econopack column equilibrated by a 10 mM sodium-acetate buffer, pH 3. This column was eluted with a linear gradient of NaCI; the xylanase was eluted with 2% NaCI. At this stage, PAGE showed only one proteic band possessing xylanase activity (Fig. 1). The xylanase X2T is purified 264-fold and presents a specific activity of 427 U/mg of protein (Table I). Purification of the xylanases from Termitomyces sp. Mycotetes crude extract was loaded onto a column of hydroxyapatite equilibrated as described above. The active fraction on xylan was unbound to the column and therefore eluted by the equilibrating buffer. This fraction was then applied to a column of DEAE-Sepharose equilibrated with 10 mM Tris-HC1 buffer, pH 7.6. Separation on this ion exchange chromatography gave two xylanasic activity peaks. One activity peak (X1Mc) was eluted from the column with a buffer containing 10 mM NaCI. This first fraction appeared as a single proteic band on gel electrophoresis (Fig. 1) on which xylanasic activity was associated. This electrophoretically pure enzyme, 37-fold purified, had a specific activity of 54 U/mg of protein (Table 1) with a 3.8% overall yield of activity.

Fig. 1. Ten percent SDS-PAGE of xylanases XIT and X2T purified from the digestive tract of the termite Macrotermes bellicosus and xylanases XIMc and X2Mc purified from the mycotetes of its symbiotic fungus Termitornyces sp. at successive stages of purification. Lane 1, mycotetes crude extract; lane 2, hydroxyapatite; lane 3, purified xylanase X2Mc; lane 4, purified xylanase X1Mc; lane 5, purified xylanase XIT; lane 6, purified xylanase X2T; lane 7, Hydroxyapatite; and lane 8, termite crude extract.

The second xylanasic peak activity, called X2Mc, not pure after the DEAE-Sepharose column, was applied to a QA Tris Acryl column equilibrated with a 10 mM Tris-HCl buffer, pH 8.2. The active fraction, eluted by 10 mM NaC1, appeared electrophoretically pure (Fig. 1) with associated xylanasic activity. The specific activity of pure X2Mc was 1240 U/mg. The overall yield of activity was 40% and a 868-fold purification was achieved (Table 1).

Properties of the four purified xylanases Enzyme specificity. Xylanases X1T, X2T, X1Mc and X2Mc hydrolysed both types of xylan (Table 2) but had no activity on the other hemicelluloses tested. Xylanases X2Mc and X2T also showed a slight activity on ONP/3xyloside. TLC revealed that xylanases X2T and X2Mc hydrolysed xylan releasing only xylobiose, whereas xylanases X1T and X1Mc action on xylan led to the formation of xylobiose, xylotriose and higher homologues in equal amounts. These results suggested that XIT and X1Mc were endoxylanases (EC 3.2.1.8) and X2T and X2Mc were exoxylanases (EC 3.2.1.37). Effects of pH on xylan hydrolysis. The effect of pH on the activity of all the xylanases was studied by using Mcllvaine buffers of pH 2.6-8. The pH activity profiles were very similar for the four enzymes (Figs. 2 and 3) showing a maximum at pH 5.2-5.6. Thermal denaturation. The four xylanases remained stable at +4°C for several months

M. Matoub and C. Rouland

632

Table 1. Purification of the xylanases of the termite Macrotermes bellicosus and its symbiotic fungus Termitomyces sp.

Protein (rag)

43 119 115

3683 2724 139

2576 100 2.6

1.43 27.24 53.5

1 19 37

43 119 87 95

3683 2724 1898 1490

2576 100 3 1.2

1.43 27.24 633 1240

1 19 443 868

100 155 112

413 240 28

254 3.84 0.52

1.62 62.5 54

I 39 33

100 58 7

100 155 84 75 27

413 240 153 104 76.8

254 3.84 0.42 0.27 0.18

1.62 62.5 364 385 427

1 39 225 237 264

100 58 37 25 19

Volume (ml)

Steps Xylanase X1Mc Crude extract Hydroxyapatite DEAE-Sepharose Xylanase X2Mc

Crude extract Hydroxyapatite DEAE-Sepharose QA Tris-Acryl Xylanase X 1T

Crude extract Hydroxyapatite DEAE-Sepharose

Xylanase X2T Crude extract Hydroxyapatite DEAE-Sepharose QA Tris Acryl CM Econopack

Specific activity (units/mg protein)

Total activity (units)

Purification

Yield

(fold)

(%)

100 74 3.8 100 74 51.5 40.5

Xylanasic activities were tested using a 2% (w/v) birchwood xylan solution as substrate. One unit is a/xmole of xylose equivalent per minute.

without apparent loss of activity. The four xylanases were heat sensitive, but X2T and X2Mc were the most thermolabile. Preincubation at + 50°C for 5 min caused, respectively, 60% and 80% loss of activity, whereas under the same conditions, X1T and X1Mc had no loss of activity (Figs. 4 and 5). Effect of substrate concentrations. The effect of varying the substrate concentration on the reaction rate was studied with birchwood xylan. Values of Km, calculated with Lineweaver-Burk plots (1934), were 0.2 _+ 0.03% for xylanases X1Mc and XIT, 0.1 _+ 0.02% for xylanase X2T and 0.3 _+ 0.02% for xylanase X2Mc. Molecular weights. The molecular weights determined by the method of Hedrick and Smith (1968) were 36 _ 4 kDa for xylanase X1Mc, 22.5 _ 3 kDa for xylanase X2Mc, 36 ___ 4 kDa for xylanases XIT and 56 __+ 6 kDa for xylanase X2T. They were confirmed by SDS-PAGE, suggesting a monomeric structure for these enzymes.

Glycoproteic nature. The four purified xylanases were tested for the presence of carbohydrate by periodic acid-Schiff staining on PAGE. Only xylanases X1T and X1Mc presented a positive response, suggesting a glycoproteic nature for these two enzymes. Discussion Comparison of the different properties of xylanases X1T and X1Mc and analysis of the steps required for their purification led to the following observations. Xylanase XIT detected in the gut of M. bellicosus and xylanase XIMc isolated from Termitomyces sp. mycotetes showed the same elution profile on all tested chromatographic matrices. Their specific activities on xylan were almost identical. The K m values calculated for both enzymes and the indistinguishable pattern of the xylooligosaccharides released during xylan hydrolysis suggested a common mechanism of action. Furthermore, both enzymes had the

Table 2. Substrate specificity of the purified xylanases XIMc, X2Mc, X1T and X2T Substrates Xylobiose ONP/3 xyloside Oat spelt xylan Birchwood xylan

Hydrolysis products

Xylanase X1Mc

Xylanase X2Mc

Xylanase X1T

Xylanase X2T

0 0 53 _+ 3.5 54 _+ 3 Oligoxyloside, PD -> 3

0 18 _+ 0.8 1040 _+ 95 1240 _+ 110 Xylobiose

0 0 51 -+ 2.5 53 _+ 4 Oligoxyloside, PD -> 3

0 8.5 -+ 0.4 334 +- 15 427 _+ 18

Specific activity is expressed in units per mg protein. Values are means _+ SD.

Xylobiose

633

Purification and properties of termite xylanases 1500'

.... O"-

Xylanas¢

.... o-'-

XlMc

1250

J,

Xylalm._sc XIT

/

60

Xylanase X2Mc

Xylanase X2T

'[

1000

<

\

750 '

"d

<

gl,

40

== 500'

20' 250

0 2

3

4

5

6

7

8

9

pH

Fig. 2. Effect of pH on rates of birchwood xylan hydrolysis by the purified xylanases X1T and X1Mc. Specific activities were expressed in U/mg protein. Each point was the mean value of five tests.

same molecular weight, 36 kDa, calculated from their relative migration rate on PAGE. This value was confirmed by SDS-PAGE in reducing conditions, which also demonstrated the monomeric nature of the two xylanases. They also showed the same pH activity profile and the same thermal stability. We could therefore conclude that xylanase X1T and xylanase X1Mc were indeed the same enzyme, called X1, produced by the fungus Termitomyces sp. and ingested by the termite M. belli-

pH

Fig. 3. Effect of pH on rates of birchwood xylan hydrolysis by the purified xylanases X2T and X2Mc. Specific activities were expressed in U/mg protein. Each point was calculated from the average value of five tests.

120 .

.

.

.

.

o. -

Xylan~ XIM¢ X~lan~¢ XIT

100"

~

8o

I 40

COStIS.

Xylanases X2T and X2Mc, purified from the termite and from its symbiotic fungus, respectively, hydrolysed xylan releasing only xylobiose and then X2T and X2Mc can be considered as exoxylanases (EC 3.2.1.37). The specific activity of xylanase X2Mc of 1240 U/mg protein is very important in corn-

2o

°35

45

SS Temperat=res (~CI

65

7~5

Fig. 4. Thermal stability of the purified xylanases X1T and X1Mc after 5 min of pre-incubation at different temperatures. Each point was the mean value of five tests.

M. Matoub and C. Rouland

634 . .o-

.

Xyiana~ X2M~ X ylanaw X21

i c~)

},

?.

J35

45

SS

65

75

Fig. 5. Thermal stability of the purified xylanases X2T and X2Mc after 5 min of pre-incubation at different temperatures. Each point was calculated from the average value of five tests.

parison with other fungus xylanases, like those of Aspergillus nidulans (FernandezEspinar et al., 1994) or Humicola sp. (Da Silva et al., 1994), but is rather similar to those of Piromyces sp. (Teunissen et al., 1993). The specific activity of X2T (427 U/mg protein) is particularly notable in comparison with other xylanases from insects (Chipoulet and Chararas, 1985). Xylanase X1, produced by Termitomyces sp., presented a very low specific activity that was also found for the xylanases produced by Termitomyces clipeatus (Ghosh et al., 1980; 1987). The molecular weight values obtained for the four purified xylanases are in accordance to the xylanase purified from another Termitomyces (Rouland et al., 1988c) and comparable with purified xylanases from other organisms (Tan et al., 1985; Biswas et al., 1990; Huang et al., 1991; Fern~mdez-Espinar et al., 1994; Da Silva et al., 1994). The three xylanases showed maximum activity at pH 5.2-5.6, a value very similar to optimal pHs observed for xylanases purified from other Termitomyces (Ghosh et al., 1980; Ghosh and Sengupta, 1987; Rouland et al., 1988c). The K m values calculated with birchwood xylan as substrate were 0.2% for X1, 0. I% for X2T and 0.3% for X2Mc. These values are lower than those obtained for purified xylanases from other Termitomyces (Ghosh et at., 1980; Ghosh and Sengupta, 1987; Rouland et al., 1988), showing a higher affinity of X1, X2T and X2Mc for the substrate. On the other hand, these values are comparable with the K m of xylanases purified from other fungus than Termitomyces (Teunissen et al., 1993;

Elegir et al., 1994; Fern~ndez-Espinar et al., 1994). Our studies show that xylanases X2T and X2Mc were sensitive to temperature, whereas xylanase X1 remained stable up to 55°C. The thermostability of X1 could be explained by its glycoproteinic nature (Rouland et al., 1988a and c) as it is suggested by its positive reaction with periodic acid-Schiff reagent. The presence of the fungal endoxylanase XI in termite gut results from the consumption by workers of fungus comb (containing Termitomyces mycelium) and/or nodules as it has been shown for Macrotermes muelleri (Rouland et al., 1988a). The fungal endoxylanase X1 is therefore an acquired enzyme. The fact that only one of the fungal endoxylanase is acquired remained to study. Kukor and Martin (1986) defined the conditions for an enzyme to be ingested: the enzyme must remain stable and active in the gut milieu, termite usual food must contain the specific substrate (xylan) of the acquired enzyme and termite must not be the same enzyme (endoxylanase) producer. Xylanase XI presented these four conditions, whereas X2Mc is not thermostable and has the same specific action on xylan (exocellulase) than X2T. On the other hand, experiments with Termitomyces sp. growing in liquid medium showed that only xylanase X1 was secreted in the culture medium (Ikhouane, unpublished data). We conclude therefore that this enzyme, produced and secreted by fungus cells in the comb, could be simultaneously ingested by M. bellicosus workers. The secretion of this enzyme seems to be another essential factor for its "appropriation" by the termite. Synergistic activities could exist between the different xylanases as has been observed for cellulases (Rouland et al., 1988b). Despite the little information we have on the division of labour in the M. bellicosus workers used for this study, it is difficult to estimate, as Veivers et al. (1991) have made for M. subhyalinus and M. michaelseni, the real importance of the acquired xylanase X I in the digestive metabolism of this termite.

References Abo-Khatwa N. (1978) Cellulase of fungus growing termites: a new hypothesis on its origin. Specialia 34, 559-560. Ausat A., Cheema P. S., Koshi T., Perti S. L. and Ranganathan S. K. (1960) Laboratory culturing of termites. In Termites in the Humid Tropics, Proc. New Delhi, Symp., UNESCO Paris, pp. 121-125. Biswas S. R., Jana S. C., Mishra A. K. and Nanda G. (1990). Production, purification and characterization of xylanase from hyperxylanolytic mutant Aspergillus ochraceus. Biotechnol. Bioeng. 35, 244-251.

Purification and properties of termite xylanases Bradford M, M. (1976) A rapid and sensitive method for the quantification of/ag quantities of protein utilizing the principle of protein dye binding. Analy. Chem. 72, 248-254. Chipoulet J. M. and Chararas C. (1985) Survey and electrophoretical separation of the glycosidases of Rhagium inquisitor larvae. Comp. Biochem. Physiol. 80B, 241-246. Chippendale G. M. and Beck S. D. (1966) Staining of glycoproteins. J. Insect. Physiol. 12, 1628-1638. Da Silva R., Yim K. D. and Park K. Y. (1994) Purification and characterization of thermostable xylanases from thermophilic Humicola sp. and their application in pulp improvement. Rev. Microbiol. 25, 112-118. Elegir G., Szak~tcs G. and Jeffries T. W. (1994) Purification, characterization and substrates specificities of multiple xylanases from Streptomyces sp. strain B12-2. Appl. environ. Microbiol. 60, 2609-2615. Fern~tndez-Espinar M., Pi~aga F., De Graaff L., Visser J., Ram6n D. and Vallrs S. (1994) Purification, characterization and regulation of the synthesis of an Aspergillus nidulans acidic xylanase. Appl. microbiol. Biotechnol. 42, 555-562. Ghosh A. K. and Sengupta S. (1987) Multisubstrate specific amylase from mushroom Termitomyces clypeatus. J. Biosci. 4, 275-285. Ghosh A. K., Banerjee P. C. and Sengupta S. (1980) Purification and properties of xylan hydrolase from mushroom Termitornyces clypeatus. Biochem. biophys. Acta 612, 143-152. Hedrick J. L. and Smith A. (1968) Size and charge isomer separation of molecular weight of proteins by disk gel electrophoresis. Archs Biochem. Biophys. 126, 155-165. Heim R. (1977) Termites et champignons. Paris. (Ed by Boubre) Huang L., Hseu T. H. and Wey T. T. (1991) Purification and characterization of an endoxylanase from Trichoderma konigii G-39. Biochem. J. 278, 329-333. Kukor J. J. and Martin M. M. (1986) The effect of acquired microbial enzymes on assimilation efficiency in the common woodlouse, Tracheoniscus rathket. Oecologia 69, 360-366. Levin O. (1962) Protein chromatography on calcium phosphate columns. Methods Enzymol. 5, 658-666. Lineweaver H. and Burk K. (1936) The determination of enzyme dissociation constant. J. Am. Chem. Soc. 56, 658-666. Macllvain T. C. (1921) A buffer solution for colorimetric comparison. J. biol. Chem. 49, 183-188. Maizel J. V. Jr. (1964) Preparative electrophoresis in acrylamide gels. Ann. N.Y. Acad. Sci. 121, 381-390. Martin M. M. and Martin J. S. (1978) Cellulose digestion in the midgut of the fungus growing termite Macrotermes natalensis: the role of acquired digestive enzymes. Science 199, 11-21. Martin M. M. and Martin J. S. (1979) The distribution and origins of the cellulolytic enzymes of the higher termite Macrotermes natalensis. Physiol. Zool. 52, 11-21. Matoub M. (1993) La symbiose termite-champignon chez Macrotermes bellicosus (Termitidae, Macrotermitinae)--rrle des enzymes acquises dans la xylanolyse. Thesis d'Universitr, Paris XII-Val de Marne, Paris. Nelson N. (1944) Photometric adaptation of Somogyi method for determination of glucose. J. biol. Chem. 153, 375-380. Noirot Ch. (1952) Le soin et l'alimentation des jeunes chez les termites. Ann. Sci. Nat. Zool. XIV, 405-414.

635

Petek F. (1962) Recherches sur les activitrs hydrolysantes et transfrrantes des a-galactosidases. Ph.D. Th~se d'Universitr, Pharmacie, Paris V. Rouland C., Chararas C. and Renoux J. (1986) Etude comparre des osidases de trois esp~ces de termites africains ~t rrgime alimentaire diffrrent. C. R. Acad. Sci. Paris 9, 341-345. Rouland C., Mora Ph. and Renoux J. (1987) Essai d'interprrtation de la symbiose digestive chez Macrotermes muelleri (Termitidae, Macrotermitinae). Act. Coll. U.I.E.IS. 4, 111-118. Rouland C., Civas A., Renoux J. and Petek F. (1988a) Purification and properties of cellulases from the termite Macrotermes mi~lleri (Termitidae Macrotermitinae) and its symbiotic fungus Termitomyces sp. Comp. Biochem. Physiol. 91B, 449-458. Rouland C., Civas A., Renoux J. and Petek F. (1988b) Synergistic activity of the enzymes involved in cellulose degradation, purified from Macrotermes bellicosus and its symbiotic fungus Termitomyces sp. Comp. Biochem. Physiol. 91B, 459-465. Rouland C., Renoux J. and Petek F. (1988c) Purification and properties of two xylanases from Maerotermes muelleri (Termitidae Macrotermitinae) and its symbiotic fungus Termitomyces sp. Insect. Biochem. 18, 709-715. Rouland C., Brauman A., Keleke S., Labat M., Mora Ph. and Renoux J. (1989) Endosymbiosis and ectosymbiosis in the fungus growing termites. In Microbiology in Poecilotherms (Edited by Lesel R.), pp. 79-82. Elsevier Science, Amsterdam. Rouland C., Lenoir F. and Lepage M. (1991) The role of the symbiotic fungus in the digestive metabolism of several species of fungus growing termites. Comp. Biochem. Physiol. 99A, 657-663. Sands W. A. (1956) Some factors affecting the survival of Odontoterrnes badius. Insect. Soc. III, 531-536. Slaytor M. (1992) Cellulose digestion in termites and cockroaches: what role do symbionts play? Comp. Biochem. Physiol. 103B, 775-784. Somogyi M. (1945) Determination of blood sugar. J. biol. Chem. 160, 61-68. Tan L., Wong E. K. C. and Saddler J. N. (1985) Purification and characterization of two D-xylanases from Trichoderma harzanium. Enzyme Microbiol. Technol. 7, 425-430. Teunissen M. J., Hermans J. M. H., Huis In't Veld J. H. J. and Vogels G. D. (1993) Purification and characterization of a complex-bound and a free /3-1-4endoxylanase from the culture fluid of the anaerobic fungus Piromyces sp. strain E2. Arch. Microbiol. 159, 265-271. Tisselius A., Hjerten S. and Levin O. (1956) Protein chromatography on calcium phosphate columns. Ann. Biochem. Biophys. 65, 132-155. Veivers P. C., Miihlemann R., Slaytor M., Leuthold R. H. and Bignell D. E. (1991) Digestion, diet and polyethism in two fungus growing termites Macrotermes subhyalinus Rambur and Macrotermes michaelseni Sjostedt. J. Insect Physiol. 37, 675-682. Weber K. and Osborn M. (1969) The reliability of molecular weight determinations by Dodecyl-SulfatePolyacrylamide gel electrophoresis. J. biol. Chem. 244, 16, 4406-4412. Williams J., Villaroya H. and Petek F. (1968) aGalactosidase II, III and IV from seeds of Trifolium repens. Biochem. J. 175, 1069-1077.