A Cold-Active Glucanase from the Ruminal Bacterium Fibrobacter succinogenes S85

Abstract

We previously characterized two endoglucanases, CelG and EGD, from the mesophilic ruminal anaerobe Fibrobacter succinogenes S85. Further comparative experiments have shown that CelG is a cold-active enzyme whose catalytic properties are superior to those of several other intensively studied cold-active enzymes. It has a lower temperature optimum, of 25°C, and retains about 70% of its maximum activity at 0°C, while EGD has a temperature optimum of 35°C and retains only about 18% of its maximal activity at 0°C. When assayed at 4°C, CelG exhibits a 33-fold-higher k_cat value and a 73-fold-higher physiological efficiency (k_cat/K_m) than EGD. CelG has a low thermal stability, as indicated by the effect of temperature on its activity and secondary structure. The presence of small amino acids around the putative catalytic residues may add to the flexibility of the enzyme, thereby increasing its activity at cold temperatures. Its activity is modulated by sodium chloride, with an increase of over 1.8-fold at an ionic strength of 0.03. Possible explanations for the presence of a cold-active enzyme in a mesophile are that cold-active enzymes are more broadly distributed than previously expected, that lateral transfer of the gene from a psychrophile occurred, or that F. succinogenes originated from the marine environment.

Fibrobacter succinogenes S85, a gram-negative anaerobe, is one of the major cellulolytic organisms in the rumen (10). Although it produces multiple cellulases, so far none has been characterized as being cold active. Cold-active cellulases are uncommon even among psychrophiles, but there is a report on a thermolabile amylase from the Antarctic bacterium Alteromonas haloplanctis A23 (8). This may not be surprising since the organism itself is a psychrotroph. With the increasing sophistication of molecular cloning methods, previously unculturable organisms are now identifiable (1), increasing the possibility of discovering new strains or even identifying previously characterized strains in much different environments. For instance, there are indications that Fibrobacter species are present in oceans (12) and tundra soil (29), where temperatures are very low. These discoveries may have important impacts on our knowledge of microorganisms.

While enzymes from thermophiles are stabilized by a combination of noncovalent interactions making use of a small number of amino acid replacements (19), less is known about residues important for adaptation of their psychrophilic counterparts. It is thought that the thermal instability in this group is a result of their highly flexible and loose structure (7, 17). As a rule, cold-active enzymes exhibit less temperature dependence at low temperature and possess higher k_cat values, higher physiological efficiencies (k_cat/K_m values), and lower activation energies than their thermophilic counterparts but also exhibit an increased heat lability (7, 26).

In this report, we present data on the CelG enzyme from F. succinogenes S85 with respect to the effect of temperature on its enzymatic activity, thermal stability, and structure.

MATERIALS AND METHODS

Bacterial strains, plasmids, and enzyme assays.

Escherichia coli DH5α (14) was used for routine propagation of plasmids, while E. coli BL21(DE3) (13, 25) was used for protein expression. The plasmids containing the two genes encoding the CelG and EGD enzymes have been previously described (18, 23). Growth of plasmid-containing strains as well as expression and purification of enzymes were done by previously described methods (18).

The standard assay for endoglucanase activity was carried out as described previously (18), using low-viscosity carboxymethyl cellulose (CMC) at a final concentration of 1% (wt/vol). Assays for CelG and EGD were carried out at 25 and 35°C, respectively. To determine the effect of temperature on enzyme activity, purified enzymes were incubated with substrate in a buffer of the optimum pH (18, 23) at different temperatures. The thermal stabilities of CelG and EGD were determined by incubating the enzymes at various temperatures for 2 h and assaying residual activity at timed intervals. Kinetic parameters were determined directly from Lineweaver-Burk plots (22), using low-viscosity CMC at concentrations from 5 to 20 mg ml⁻¹. Determinations were done at 4 and 25°C for both enzymes. The amount of reducing sugar produced at the end of the incubation period in all cases was determined by the method of Lever (21). Briefly, enzyme assay reactions (200-μl volumes) were stopped by addition of 1.5 ml of p-hydroxybenzoic acid hydrazide containing 1 mM bismuth nitrate and 500 mM sodium hydroxide, which inactivates the enzyme completely. After the reactions were stopped, the tubes were incubated at 70°C (10 min) for color development and the amount of reducing sugar was estimated by measuring the absorbance of the solution at 410 nm. The protein concentration was estimated by the dye binding method of Bradford (2), using a Bio-Rad (Richmond, Calif.) protein assay kit, with bovine serum albumin fraction V (Sigma Biochemicals) as the standard.

Ea determination.

Activation energy (Ea) was determined from the slope (−Ea/R) of Arrhenius plots of log V (in units per milligram per minute) versus 1/T, while thermodynamic activation parameters were calculated by using the following equations: ΔG* = ΔH* − TΔS*, ΔH* = Ea − RT, and ΔS* = 2.303R(log k_cat − 10.753 − log T + Ea/2.303 RT), where ΔG represents change in free energy, ΔH is change in euthalpy, ΔS is change in entropy, R is the universal gas constant, and T is the absolute temperature (in Kelvin).

Far-UV CD spectroscopy.

Circular dichroic (CD) spectra of proteins (0.2 mg/ml) in 10 mM phosphate buffer, pH 7.5, were recorded with a Jasco model J600 spectropolarimeter (Japan Spectroscopic Co. Ltd., Tokyo, Japan) in a 0.1-cm-light-path cell under a constant nitrogen purge. Samples were scanned an average of five times between the wavelengths of 190 and 250 nm. Secondary-structure fractions were calculated by using the Jasco secondary-structure program based on the algorithm of Chang et al. (4) and the database of Hennessey and Johnson (15). Temperature was controlled thermostatically by the use of water-jacketed curvettes.

Nucleotide sequence accession numbers.

The nucleotide sequences for the celD and celG genes were previously assigned GenBank accession no. U05897 and U33887, respectively.

RESULTS AND DISCUSSION

The CelG enzyme had a pH optimum of 5.5 and a temperature optimum of 25°C. The thermodependence curves of CelG and EGD endoglucanase activities are shown in Fig. 1. While it is true that CelG shares a lot of properties with other Fibrobacter endoglucanases and exhibits structural similarities to the members of the family 5 cellulases to which it belongs (16), its thermodependence makes it unique. Both CelG and EGD show a rapid decrease in activity at 45°C, indicating the induction of thermal alteration of the catalytic mechanisms of the two enzymes. However, at 0°C there is a clear difference between the two enzymes. While CelG maintains about 70% of its maximum activity at this temperature, EGD retains only 18% of its maximum activity. In terms of thermal stability, EGD is more stable than CelG, maintaining about 55% of its maximum activity after 2 h of exposure at 42.5°C. At the same temperature and exposure time, CelG maintains about 30% of its maximum activity. An interesting difference is seen when both enzymes are exposed at 40°C and assayed under optimal conditions: EGD maintains about 90% of its maximum activity, while CelG maintains 70%. This difference between the behaviors of the enzymes at 40 and 42.5°C alludes to the ability of proteins to renature under favorable conditions and also shows that for the two enzymes, induction of irreversible denaturation occurs from about 42.5°C and that once this temperature is reached, complete denaturation sets in very quickly (Fig. 2). Enzyme renaturation is a feature that is consistent with several endoglucanases. The optimum temperature for EGD activity is 35°C, which is much higher than that for CelG. The low temperature optimum reported for CelG has never been reported for any other F. succinogenes enzymes, and the high catalytic activity at 0°C documents that it satisfies the criteria for being a cold-active enzyme (7, 28).

FIG. 1.

Effects of temperature on the activities of the CelG (■) and EGD (●) endoglucanases. Assays were performed in 50 mM sodium acetate buffer, pH 5.5, for CelG and in 50 mM sodium phosphate, pH 6.0, for EGD. The maximum specific activity of EGD, which is 12 U (mg of protein)⁻¹, was set as 100%; the specific activity of CelG at various temperatures was related to the 100% value for EGD at 35°C.

FIG. 2.

Effect of temperature on the stability of CelG (A) and EGD (B). Assays were carried out under optimal conditions for each enzyme after exposure to different temperatures for 2 h. The numbers next to the individual plots represent temperature in degrees Celsius.

The values for the kinetic parameters k_cat and k_cat/K_m exhibited by CelG are much higher than those for EGD (Table 1). The k_cat of CelG toward CMC at 4°C is 34 times higher than that for EGD, while the k_cat/K_m ratio, which determines the physiological efficiency of an enzyme, is 75-fold higher for CelG than for EGD. High k_cat and physiological efficiency, including a decreased K_m at lower temperatures, are features consistent with cold-active enzymes (5). The large degrees of variation in these important kinetic properties of psychrophilic enzymes are thought to compensate for the ordinarily low reaction rates that normally occur at low temperatures (6). The fact that the same scenario is true for CelG is evident upon comparison of its kinetic and thermodynamic activation parameters with those of EGD, which is considered by all standards to be a mesophilic enzyme from the same organism.

TABLE 1.

Kinetic parameters for the activities of the endoglucanases CelG and EGD from F. succinogenes S85

Endoglucanase	Assay temp (°C)	Kinetic parameter
		kcatS−1	Km (mg/ml)	kcat/Km
CelG	4	20.1	6.8	3.0
	25	93	55.0	1.7
EGD	4	0.6	16.1	0.04
	25	4	23.8	0.17

Naturally, enzyme activity increases with temperature in a predictable manner until a maximum activity is reached, at which point most enzymes become inactivated as temperature is further increased. Of course, there are enzymes which do not follow this trend, resulting in a much slower and nonexponential rise in velocity; examples include phosphoglycerate kinases from both mesophilic and thermophilic microorganisms (27). A nonexponential rise in velocity could indicate that thermal denaturation sets in very early in the reaction process, before a true exponential increase is seen. This may be the case for CelG, indicating a deviation from the Arrhenius equation. This has been corrected in the derivation of the activation energy for CelG for the basis of comparison, since it was obvious that the value for CelG would be lower than that of EGD. The value obtained for CelG was 6.5 kJ mol⁻¹, while that calculated for EGD within the same temperature range (0 to 25°C) was 35 kJ mol⁻¹ (Table 2). EGD, on the other hand, conforms to the expected trend. The activation energy for CelG is constant over the above-described temperature range (results not shown), indicating that the enzyme does not undergo major structural changes within this range, a fact which is supported by the CD spectra (Fig. 3) at both 4 and 20°C. At both temperatures the percentages of randomness and β-turns are maintained at 18 and 20%, respectively, while at 55°C randomness alone is almost 50%, showing the protein structure disorganization associated with reduced catalytic activity. The contribution of the thermodynamic activation parameters enthalpy and entropy is seen in the value of ΔG* for CelG (Table 2), which is much lower than that for EGD. This low value indicates that CelG will require less heat content and, hence, minimal entropy to achieve activation.

TABLE 2.

Thermodynamic activation parameters for CelG and EGD at 25°C

Endoglucanase	Parameter
	kcatS−1	Ea (kJ mol−1)	ΔG* (kJ mol−1)	ΔH* (kJ mol−1)	ΔS* (J mol−1K−1)
CelG	93	6.5	61	4	−190
EGD	4	35	90	33	−127

FIG. 3.

CD spectra of CelG recorded in 10 mM phosphate buffer (pH 7.5) at 4, 20, and 55°C. The protein concentration in the solution was 0.2 mg ml⁻¹.

While a low temperature optimum or high activity at low temperatures is a prerequisite for being categorized as cold active, there are presently no formal rules for identifying cold-active enzymes because of the paucity of these enzymes in the databases. In spite of these anomalies, certain trends are beginning to emerge. One of them is the stacking of small amino acids around catalytic residues (6, 8). Although this stacking is not very evident in CelG, it is interesting that the putative catalytic glutamic acid residue at position 166 is flanked by three small amino acids on the left and a glycine on the right. As yet, the number of these small amino acids required around catalytic residues has not been determined. It is thought that the presence of small amino acids reduces steric hindrance at the entrance of the active site, inducing active-site flexibility and providing an energetically favorable environment. This feature is essential for high activity at low temperatures and is consistent with heat lability (6).

Another feature of cold-active enzymes, according to Hochachka and Somero (17), is the possession of folding flexibility. For activity at a low temperature, an enzyme must have a more flexible structure to enable rapid and reversible catalytic cycles (6). The predicted CelG structure shows extensive loop formation (56%) (24), while CD analysis indicates that there is about 40%. This value, coupled with the presence of small amino acids at the entrance of the catalytic residue, may add to the active-site flexibility of CelG.

The celG gene, encoding the cold-active glucanase, probably originated in one of the following three ways. First, the cold-active nature of enzymes may be a more common phenomenon in mesophilic organisms than has been theorized. Reaction rates in this case may be temperature independent, as in perfectly evolved enzymes (7), whose reactions occur almost as soon as the enzyme and substrate come in contact. Second, the gene may have originated by lateral transfer from a cellulolytic psychrophilic organism; however, it seems that the opportunity for such a transfer would be rare. Third, ruminal F. succinogenes may have a marine origin, with the celG gene representing a vestigial gene not yet fully evolved, or, since it has low activity in a purely mesophilic environment, it may have been lost from most species of Fibrobacter, since we have previously shown that it is present only in the type species F. succinogenes S85 (18).

Support for the last hypothesis comes from the recent detection of Fibrobacter species in ocean water (12), where temperatures can be low. Gordon and Giovannoni (12) discovered the presence of a Fibrobacter species-related gene lineage in 16S rRNA clone libraries prepared from water samples collected by filtration from a depth of 80 m at a site in the western Sargasso Sea and from a depth of 120 m at a site in the Pacific Ocean. This hypothesis is further strengthened by the fact that increasing salt concentrations were found to be stimulatory to the activity of CelG, with a maximum stimulation of 1.8-fold at an ionic strength of 0.03. This is not surprising, since Bryant et al. (3) have previously shown that F. succinogenes S85 requires high salt concentrations for optimal growth, a characteristic exhibited by many marine bacteria (20). Also, procedures previously developed for the isolation of cell envelope fractions from the halophilic marine bacterium Alteromonas haloplancktis (9) have been successfully applied to F. succinogenes S85 (11).

Resolution of these hypotheses will be provided with the isolation of F. succinogenes species from marine and other cold environments.