Regulation of Biosynthesis of Individual Cellulases in Thermomonospora fusca

Abstract

Regulation of the biosynthesis of the six cellulases comprising the cellulolytic system of the thermophilic soil bacterium Thermomonospora fusca ER1 was studied. The levels of the individual enzymes produced on different noninducing and inducing carbon sources were determined. The lowest level of cellulase synthesis (3 nM) was observed with xylose as a carbon source, and the highest level (247 to 1,670 nM for different enzymes) was found in cultures grown on microcrystalline cellulose. Endocellulases and exocellulases showed distinctly different regulation patterns. Differences in the regulation of individual enzymes appear to be determined by the specific structural organization of the upstream regulatory sequences of their genes.

Thermomonospora fusca (Actinomycetaceae) is a filamentous thermophilic soil bacterium and an important species degrading cellulose and hemicellulose in plant residues. The cellulolytic system of T. fusca is quite complex, consisting of six extracellular cellulases. There are three β-(1,4)-endoglucanases, two β-(1,4)-exoglucanases, and one processive endoglucanase (9, 10, 13, 28), which act cooperatively to convert insoluble cellulose into cellobiose and other soluble sugars. The activity, mechanisms of catalysis, and binding to substrate have been extensively studied (1, 3, 7, 8, 20, 23, 24).

Cellulase biosynthesis in T. fusca is regulated by induction by cellobiose and repression by readily metabolized carbon sources (14). Studies of the T. fusca celE gene suggested that it was transcriptionally regulated. A protein involved in cellulase induction and its binding site on the celE gene were identified (15, 16).

However, little is known about the levels of the individual enzymes produced under different physiological conditions. The aim of this study was to quantitate the individual cellulases in T. fusca grown on different carbon sources and to correlate individual enzyme levels with cellulolytic activity.

MATERIALS AND METHODS

Culture conditions and sample preparation.

T. fusca ER1 was grown on modified Hagerdahl medium (5, 14) containing 0.5% cellobiose. Stationary-phase cultures were used to inoculate the same medium, supplemented with either a 0.5% soluble or a 1% insoluble carbon source. Glucose, cellobiose, xylose, and medium-viscosity carboxymethyl cellulose sodium salt (CMC; Sigma Chemical Co.) were the soluble carbon sources, while Solka Floc (microcrystalline cellulose; Mallinckrodt) and dry ground grass were the insoluble carbon sources. The substrate concentrations used in this study were not growth limiting. Cells were cultured at 55°C and 200 rpm for 72 h in triplicate on each carbon source. For extracellular cellulase assays, culture samples were centrifuged at 5,000 × g for 10 min, and the supernatants were brought to 0.1 mM phenylmethylsulfonyl fluoride. Supernatants were stored at −20°C.

Protein assays.

Culture samples were centrifuged at 5,000 × g for 5 min, and the supernatants were assayed for extracellular protein by the Lowry method (17). Purified E5 cellulase from T. fusca was used as a standard. For cultures grown on ground grass, extracts which contained plant pigments that interfered with the protein assay were quantitated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (12) followed by staining of the gel with SYPRO orange protein stain as described by the supplier (Molecular Probes). The concentrations of E5 and other pure T. fusca cellulases used in this study were determined by absorbance at 280 nm based on their molar extinction coefficients calculated from their predicted amino acid compositions (8).

Enzyme activity assays.

Culture supernatants were assayed for cellulase activity by using filter paper or 1% low-viscosity CMC (Sigma) as the substrate. The amount of reducing sugar (primarily cellobiose) produced was measured spectrophotometrically at 600 nm after reaction with dinitrosalicylic acid as described previously (8). Xylanase activity was measured by the same method except that 1% xylan from birchwood (Sigma) was used as the substrate. Assays were carried out at 50°C for a fixed time for each substrate at several T. fusca supernatant concentrations above and below the target digestion (0.5 μmol of reducing sugar released in 16 h for filter paper, 60 min for CMC, and 15 min for xylan). The amount of extracellular protein required to give the target level for each substrate was determined graphically. If the target digestion could not be achieved in the filter paper assays, then the specific activity was calculated from the enzyme required to release 0.2 μmol of cellobiose in the same time period. Each data point was determined in triplicate.

Quantitation of soluble cellulases.

T. fusca extracellular proteins were separated electrophoretically on SDS-polyacrylamide gels (12) and transferred (26) to Immobilon-P polyvinylidene difluoride membranes (Millipore). Rabbit polyclonal antisera raised against the appropriate T. fusca cellulases which had been purified from Streptomyces lividans clones were used for cellulase detection by Western blotting. Electrophoretically pure T. fusca cellulases were used as standards (8, 28). Goat anti-rabbit immunoglobulin-alkaline phosphatase conjugate was used as the second antibody. Cellulases were visualized with Vistra ECF substrate as described by the supplier (Amersham Life Science). Blots were scanned with a STORM 840 scanner, and images were quantitated with the ImageQuant program (Molecular Dynamics). Total soluble cellulases were calculated as the sum of the individual enzyme concentrations.

Quantitation of bound cellulases.

Samples of T. fusca cultures grown on 1% Solka Floc (1 ml) were centrifuged at 10,000 × g for 5 min. The pellets were resuspended in medium, sedimented by centrifugation, and dissolved in 200 μl of 50 mM Tris-HCl buffer (pH 6.8) containing 2% SDS, 0.1 M dithiothreitol, and 50% glycerol. Samples were boiled in a water bath for 4 to 5 min and centrifuged, and the solubilized cellulases were separated by electrophoresis and quantitated as described above. The residual concentration of Solka Floc was not measured.

Computer-assisted analysis.

Computer-assisted analysis of upstream regions of T. fusca cellulase genes was performed with the DARWIN program (19).

RESULTS

Enzyme activity in T. fusca cultures.

T. fusca ER1 was grown in minimal medium to stationary phase, and the supernatants were assayed for extracellular protein and the activities hydrolyzing three substrates: filter paper, CMC, and xylan (Table 1). Enzymatic activities in cultures grown on different carbon sources differed significantly. There was a positive correlation between total cellulase content in the cultures and their activities on CMC and filter paper. Cultures grown on Solka Floc (microcrystalline cellulose) exhibited the highest activity on all substrates, while cultures grown on ground grass had about one-third of these activities. The activities in cultures grown on noninducing substrates (glucose and xylose) were 12 to 30 times lower than those in cultures grown on Solka Floc. Although CMC gave a low yield of enzymes, it produced the highest specific activity on filter paper and CMC.

TABLE 1.

Extracellular protein and enzyme activities in T. fusca ER1 cultures grown on different carbon sources

Carbon source	Protein in supernatant (μg/ml)	Enzyme activity in supernatanta			Sp act (μmol of CB/min/mg of protein)
		CMC	Filter paper	Xylan	CMC	Filter paper	Xylan
Solka Floc	711	1.84	0.0767	19.1	2.59	0.108	26.9
Ground grass	385	0.765	0.0203	5.75	1.99	0.0527	14.9
CMC	44	0.138	0.0099	0.34	3.14	0.225	7.73
Cellobiose	436	0.595	0.0112	0.61	1.36	0.0257	1.40
Glucose	436	0.056	0.0055	0.74	0.128	0.0126	1.70
Xylose	187	0.107	0.0064	1.49	0.572	0.0342	7.97

^aExpressed as micromoles of cellobiose (CB) per minute per milliliter of supernatant for each substrate. 

Cultures grown on insoluble carbon sources exhibited considerable xylanase activity. Other authors observed simultaneous induction of both cellulase and xylanase activity also in Pseudomonas fluorescens (6) and Cellulomonas fimi (11). This appears to be due to the induction of xylanase genes by the products of cellulose digestion. Xylanase activity in cultures grown on xylose was lower than xylanase activity in cultures grown on Solka Floc and ground grass (14).

Individual cellulases in T. fusca cultures.

All six cellulases were found in T. fusca cultures grown on every carbon source tested, although their quantities and relative levels differed significantly (Table 2). The lowest levels of cellulases were found in the cultures grown on xylose. It is interesting that all cellulases in xylose-grown cultures except E2 were produced in nearly equal molar amounts (about 3 nM). This may represent the level of constitutive cellulase synthesis in T. fusca grown on noninducing carbon sources. The basal level of E2 synthesis was 10 times higher than that of the other cellulases. Similar data were obtained for cultures grown on glucose, although the levels of E3, E4, E5, and E6 produced on glucose were about two times higher.

TABLE 2.

Molar levels of extracellular cellulases in T. fusca ER1 cultures grown on different carbon sources

Carbon source	Mean individual cellulasea concn (nM) ± SD						Total cellulase (μg/ml)
	E1 endo	E2 endo	E3 exo	E4 endo/exo	E5 endo	E6 exo
Solka Flocb	420 ± 138	507 ± 49	1,669 ± 290	277 ± 98	372 ± 136	1,719 ± 30	389
Ground grassc	14 ± 4.2	121 ± 32	273 ± 138	36 ± 11	31 ± 5	315 ± 50	61
CMC	3.4 ± 0.9	35 ± 4	21 ± 2	11.4 ± 4.1	10 ± 0.9	25 ± 6	7.3
Cellobiose	35 ± 6	104 ± 32	149 ± 39	59 ± 10	33 ± 3	177 ± 31	43
Glucose	3.9 ± 1.3	31 ± 5	8.4 ± 1.2	5.4 ± 1.2	6 ± 0.7	8.2 ± 1.3	3.9
Xylose	3.1 ± 1.2	35 ± 13	3.7 ± 0.5	3.2 ± 1.0	3.7 ± 0.4	3.3 ± 0.5	2.8

^aendo, endocellulase; exo, exocellulase. 

^bSoluble cellulases and cellulases bound to Solka Floc. 

^cSoluble cellulases only. 

There was a dramatic increase in cellulase synthesis in cultures grown on inducing carbon sources. The highest levels of cellulase production were observed in cultures grown on Solka Floc, where total soluble cellulases constituted one-half of the total extracellular protein. Strikingly, the level of the exocellulases E3 and E6 in Solka Floc-grown cultures was 4 times higher than that of the endocellulases E1, E2, and E5 and 6.6 times higher than that of E4. Most of the cellulases in cultures grown on Solka Floc were soluble; only about 4% of the enzymes were associated with Solka Floc (Table 3). Similar quantities of E1, E3, E5, and E6 were bound, while more E2 and E4 were bound.

TABLE 3.

T. fusca cellulases bound to Solka Floc

Bound cellulasea	Mean concn (nM) in culture medium ± SD	Bound cellulases/total cellulases (%)
E1 endo	14.3 ± 1.9	3.4
E2 endo	78.5 ± 3	15.5
E3 exo	48.9 ± 7.9	2.9
E4 endo/exo	30.3 ± 3.9	10.9
E5 endo	10.6 ± 1.9	2.8
E6 exo	49 ± 12.7	2.9

^aendo, endocellulase; exo, exocellulase. 

When cellobiose was the carbon source, cellulases were produced in quantities intermediate between those found in cultures grown on cellulose or noninducing sugars. The most marked decrease (about 10-fold), compared to cultures grown on cellulose, was observed for exocellulases E3 and E6 and endocellulases E1 and E5, while levels of endocellulases E2 and E4 decreased only 4-fold.

Cell yields in T. fusca cultures were measured in a separate set of experiments (data not shown). Intracellular protein concentrations at the stationary phase were similar for cultures grown on Solka Floc, cellobiose, glucose, and xylose (0.44 to 0.57 mg/ml), while CMC gave poor cell growth. We could not measure cell yields in cultures grown on ground grass due to the presence of plant pigments that interfered with the assay. It is clear that the minor variations in cell yields cannot account for the significant differences in enzyme concentrations and their relative levels in cultures grown on Solka Floc, cellobiose, glucose, and xylose. However, poor cell growth on CMC was evidently the reason for lower enzyme production on this carbon source.

DISCUSSION

The results (Table 2) show that the T. fusca cellulases are synthesized constitutively on noninducing carbon sources at the level of 3 nM, except for E2 (35 nM). The higher level for this enzyme may indicate a special role for E2 in releasing cellobiose from insoluble carbon sources which can induce and provide the cells with energy during the early stages of growth. E2 comprised 54 and 34% of the total cellulase in cultures grown on xylose and glucose. The proportion of E2 decreased in cultures grown on inducing carbon sources, 10.5% on cellobiose, 8.6% on ground grass, and only 4.9% on Solka Floc. Cellobiose gave lower levels of all enzymes than other inducing carbon sources because it can repress the synthesis at high concentrations (14).

There were distinct differences in the molar levels of the endocellulases. The level of E2 was always highest, while the level of E1 was lowest on all carbon sources except Solka Floc. The molar levels of E4 and E5 were similar in cultures grown on different carbon sources except cellobiose. The relative levels of E5 in T. fusca grown on Solka Floc, cellobiose, and glucose found in our experiments (1, 0.092, and 0.017) were very similar to the relative levels of the celE gene transcripts (1, 0.11, and 0.015) measured earlier by Lin and Wilson (15). This agreement between celE mRNA and E5 levels supports the hypothesis that cellulase synthesis in T. fusca is regulated at the level of transcription.

Nearly equal amounts of E3 and E6 were found in cultures grown on all carbon sources. This result suggests a common mechanism of regulation for these enzymes, a possibility consistent with the 5′ upstream regions of the E3 and E6 genes, which have similar arrangements of their regulatory sequences (Table 4).

TABLE 4.

Functional elements in the upstream sequences of T. fusca cellulase genes and their distances from the start codon

Gene (product)	RBS	14-bp inverted repeat sequence	Homologous sequences in cel promoter regions
celA (E1)	12 ggaagg 7	74 TGGGAGCGCTCCCA 61	100 gtttttatg 92	127 taactattgac 117
celB (E2)	10 ggag 7	76 TGGGAGCGCTCCCA 63	194 gcttttacg 186	215 tagacattcac 205
		177 TGGGAGTTCTCC 166
celC (E3)	13 ggaag 9	205  GGGAGAACTCCCA 193	336 tactctcttt 327	359 tttacaaccg 350
		225 TGGGAGCGCTCCCA 212
		326 TGGGAGCGCTCCC 314
celD (E4)	9 ggag 6	49 TGGGAGCGCTCCCA 36	107 tcggtcaccac 97	137 gactactta 129
celE (E5)	9 ggagg 5	49 TGGGAGCGCTCCCA 36	122 tacggtctcac 112	149 gatcactta 141
celFa (E6)	17 ggagg 13	250 TGCCAGTGCTCCC 238	369 taccctccta 360	392 tttacacggc 383
		264 TGGGAGCGCTCCCA 251
		358 TGGGAGCGCTCCC 346

^aD. Irwin, unpublished data. 

The levels of synthesis of all the individual cellulases correlate with the structural organization of the 5′ regulatory regions of their genes (Table 4). All of the cellulase genes have a 14-bp inverted repeat sequence TGGGAGCGCTCCCA (9, 10, 13, 28), which is a binding site for a regulatory protein (16). DNase I footprinting studies showed that a regulatory protein specifically protects the 14-bp inverted repeat in the celE and celB genes. Gel retardation experiments also showed that the regulatory protein binds to both the 14- and 13-bp inverted repeats in the celC gene (12a, 16). The 14-bp inverted repeat is located 35 bp upstream from the start codon in the celD and celE genes, whose protein products (E4 and E5) are produced at similar molar levels on all carbon sources. The regulatory sequence is further upstream in the celA gene encoding E1, whose level is the lowest on most of the carbon sources. E2, displaying the highest level of synthesis of all endocellulases, possesses an additional imperfect copy of this sequence. Each of the exocellulases, all of which are produced at highest levels on inducing carbon sources, has two additional imperfect copies of this site located upstream from the start codon. These correlations suggest that cooperative binding of the regulatory protein may affect transcription of the celC and celF genes. Cellulase genes showing similar patterns of regulation also exhibited homologies in their putative promoter sequences (Table 4). In particular, local homologies were found between celC (bp 327 to 336 and 350 to 359 upstream of the start codon) and celF (bp 360 to 369 and 383 to 392) and between celD (bp 97 to 107 and 129 to 137) and celE (bp 112 to 122 and 141 to 149). Homologous elements were also found in the celA and celB genes. There also are differences in the location of the potential ribosome-binding sites (RBS) in the cellulase genes. In the genes encoding E1, E2, E4, and E5, the RBS is 5 to 7 nucleotides upstream from the start codon, whereas in the genes encoding E3 and E6, the distances are 9 and 13 nucleotides. These data suggest possible differences in the regulation of endo- and exocellulases at the level of translation.

The 14-bp inverted repeat is also found in the upstream regions of the cenC gene from Streptomyces strain KSM9 (18), the cel1 gene from S. reticuli (22), and the celA genes from S. halstedii (4) and S. lividans (25), as well as in the xylA gene from T. alba (2). Recently Walter and Schrempf (27) proposed that this sequence serves as the operator for a repressor protein in the cel1 gene from S. reticuli. Apparently, the sequence may be a universal regulatory site involved in transcriptional control of cel genes in high-GC gram-positive bacteria.

In conclusion, our results indicate that the cellulase genes of T. fusca share a common mechanism for regulation which is mediated through a protein interacting with a 14-bp inverted repeat sequence. The differences in the regulation of the individual enzymes appear to be due to the peculiar design of the 5′ regulatory regions of their genes. Apparently, cellulase genes in T. fusca are subject to coordinate control and may form a regulon.