INTRODUCTION

Top Previous Next References

Among the eukaryotic organisms, only a few species of fungi have the ability to thrive at temperatures between 45 and 55°C. Such fungi comprise thermophilic and thermotolerant forms, which are arbitrarily distinguished on the basis of their minimum and maximum temperature of growth (63): the thermophilic fungi have a growth temperature minimum at or above 20°C and a growth temperature maximum at or above 50°C, and the thermotolerant forms have a temperature range of growth from below 20 to ~55°C. Thermophily in fungi is not as extreme as in eubacteria or archaea, some species of which are able to grow near or above 100°C in thermal springs, solfatara fields, or hydrothermal vents (36, 45). Perhaps because of their moderate degree of thermophily and because their habitats are not exotic, thermophilic fungi have not received much publicity and attention. However, considering that the vast majority of eukaryotes cannot survive prolonged exposure to temperatures above 40 to 45°C (8), the ability of some 30 species, out of approximately 50,000 recorded fungal species, to breach the upper temperature limit of eukaryotes is a phenomenon that deserves elucidation. Moreover, this group of fungi provides scientists with valuable experimental material for investigations of the mechanisms which, although allowing their growth at moderately high temperatures, limit it beyond 60 to 62°C (243).

Thermophilic fungi are the chief components of the microflora that develops in heaped masses of plant material, piles of agricultural and forestry products, and other accumulations of organic matter wherein the warm, humid, and aerobic environment provides the basic conditions for their development (10, 172). They constitute a heterogeneous physiological group of various genera in the Phycomycetes, Ascomycetes, Fungi Imperfecti, and Mycelia Sterilia (182).

While reviewing the literature, we faced difficulties on account of the confusing nomenclature of thermophilic fungi. The confusion is due to several reasons. Since the early taxonomic literature is scattered and is often in languages other than English, it was difficult to ascertain the priority associated with the names of a species. As a result, some species have been described repeatedly under different names. As and when the earlier names were discovered, the fundamental rule of priority was applied and the names of the taxa were changed from time to time. For example, the ubiquitous fungus Thermomyces lanuginosus, which has been frequently used in experimental studies, has several synonyms (Table 1). Even in recent times, in several papers this fungus has been referred to by its earlier name, Humicola lanuginosa. Another source of confusion is the practice of interchangeably using the names of the asexual (anamorph) and the sexual (teleomorph) stages of the same fungus. For example, Sporotrichum (Chrysosporium) thermophile and Myceliophthora thermophila are, respectively, the anamorph and teleomorph stages of the same fungus. This fungus is reportedly a heterothallic ascomycete (182), but the heterothallic nature of an isolate cannot be demonstrated unless a compatible strain of the opposite mating type is available. There also are instances in the literature of misidentifications of thermophilic fungi. One example is Thermoascus aurantiacus, which has featured in early physiological investigations and in several recent reports dealing with enzymological studies. Some investigators have identified their isolates based on a description of T. aurantiacus given by Cooney and Emerson (63), who depicted this taxon as having an asexual stage, although in reality it lacks an asexual stage. Rather, the diagnosis of T. aurantiacus as given by Cooney and Emerson fits that of Dactylomyces crustaceus, which has a Paecilomyces asexual stage (16). Since both T. aurantiacus and D. crustaceus became a source of confusion, Mouchacca (182) proposed that the name T. aurantiacus be retained whereas D. crustaceus be renamed as Coonemeria crustacea. As is now known, T. aurantiacus is an ascomycetous fungus with a bright orange color, elliptical ascospores, and no asexual stage. Unfortunately, unless cultures used by different investigators under the name of T. aurantiacus are reexamined, it may not be possible to determine which fungus was actually used.

To nonmycologists, confusion has also resulted from the merging of what, for many years, had been regarded as different taxa. For example, several scientific papers deal with polysaccharide-degrading enzymes and trehalase of Humicola insolens, H. grisea var. thermoidea, or Torula thermophila, fungi which are commonly found in mushroom composts and in soil. All these are now thought to represent one single variable species, Scytalidium thermophilum (236). Supposing a biologist wishes to follow on the report of an interesting enzyme found in H. grisea var. thermoidea, he or she may be at a loss to reproduce the observations unless the original culture used by the author is available. Finally, in some cases, the specific epithet thermophilum (or variants thereof) has been used without adhering to the proposed definition of a thermophilic fungus. To name a few, these cases include Achaetomium thermophilum, Sordaria thermophila, or Gilmaniella thermophila, which are thermotolerant rather than thermophilic species (however, the dividing line between the two types of fungi is thin). Mouchacca (182) has attempted to remedy the confusion that had arisen by performing a critical analysis of the nomenclature and taxonomic status of thermophilic fungi. The current names of thermophilic fungi and their synonyms are given in Table 1. It will be some times before the proposed names of thermophilic fungi are stabilized. In this paper, when the work of earlier authors is reviewed, the names of the fungi as reported in the original publications have been retained, but these should be cross-checked by reference to Table 1.

The first of the known thermophilic fungi, Mucor pusillus, was isolated from bread and described over a century ago by Lindt (148). A little later, Tsiklinskaya discovered another thermophilic fungus, Thermomyces lanuginosus, growing on potato which had been inoculated with garden soil (252). Both these molds were essentially discovered as chance contaminants. The natural habitats of thermophilic fungi and the biotic conditions which favored their growth remained unknown until Hugo Miehe investigated the causes of self-heating and spontaneous combustion of damp haystacks (172). In solving the puzzle of thermogenesis of stored agricultural products, Miehe was drawn to study the microflora present therein. He was the first person to work extensively on thermophilic microorganisms. He isolated four species of thermophilic fungi from self-heating hay: Mucor pusillus, Thermomyces lanuginosus, Thermoidium sulfureum, and Thermoascus aurantiacus. He compared the heating capacities of mesophilic and thermophilic fungi (173, 174). He inoculated sterilized hay and other substrates kept inside insulated flasks with pure cultures of individual fungi and observed that the final temperature of the material depended on the maximum temperature of growth of the fungus used. He demonstrated thereby that heating of packed plant material was caused by the microorganisms present therein. Miehe explained the self-heating of hay and other plant material as follows. Initially, because of the exothermic reactions of the saprophytic, mesophilic microflora present therein, the temperature of the material rises to ~40°C. The resulting warmed environment favors the germination of spores of the thermophilic microflora, and eventually the latter outgrows the mesophilic microflora; in the process, the temperature of the mass is raised further to 60°C or even higher.

By the beginning of the 20th century, Miehe's work had led to the discovery of a small group of thermophilic fungi and to their primary habitats. Their unique thermal adaptation attracted the attention of Kurt Noack (188), who isolated thermophilic fungi from several natural substrates. He was intrigued by the fact that in addition to self-heating masses of hay and compost heaps of leaves, these fungi were present in places where temperatures conducive to their growth occur only infrequently, for example in soils of the Temperate Zone. This puzzling aspect of the ecology of thermophilic fungi provided the foundation for Noack's pioneering investigations of their physiology. Using respiration as the probe, Noack sought to determine if thermophilic fungi had an unusually high rate of respiration whereby the released metabolic heat could warm their environment, allowing them to complete their life cycle rapidly. He found, however, that the respiration of thermophilic fungi does not confer any special advantage on them.

During the Second World War, the need for finding alternate sources of rubber led to studies of the rubber-producing guayule shrub, Parthenium argentatum. It had been observed that the extractability and physical properties of rubber from the shrub improved when the plant material was chopped and stored in a mass before being milled. Allen and Emerson (10) demonstrated that the observed improvement from the above treatment (retting) resulted primarily from utilization and reduction in the amount of resin in crude rubber by a thermophilic microflora. From the self-heating mass of chopped guayule, Allen and Emerson isolated several species of thermophilic fungi that had temperature limits extending up to 60°C. They demonstrated that for optimal development of thermophilic fungi in the mass of material, its moisture and nutrient content were crucial. In addition, although size of the mass was an important factor for reducing the outward dissipation of heat, the mass of material had to be sufficiently porous for air to diffuse inside and allow aerobic respiration of fungi. Based on the isolates of thermophilic fungi from the retting guayule shrub and on collections of cultures from other investigators, Cooney and Emerson (63) provided taxonomic descriptions of 13 species known at that time, an account of their habitats, and the general biology of thermophilic fungi. This monograph, in English, for the first time made mycologists generally aware of the existence of thermophilic fungi. It stimulated the search for new species in order to understand their taxonomic diversity as well as to investigate their potential use as sources of commercially important enzymes. The taxonomy (182) and ecology (153, 244) of thermophilic fungi have been reviewed previously. The other important areas that have been studied in this group of fungi include their physiology and the purification and study of the functional characteristics of their enzymes. This review therefore covers the studies in these two areas from the time when experimental work on thermophilic fungi began.

PHYSIOLOGY

Top Previous Next References

Growth Medium

Noack (188) grew Thermoascus aurantiacus, Anixia spadicea (Chaetomium thermophile?), Mucor pusillus, Thermomyces lanuginosus, and Thermoidium sulfureum in a glucose-salt liquid medium fortified with peptone and, often, with a decoction of hay. Until the 1980s, thermophilic fungi were thought to have complex or unusual nutritional requirements. For example, Miller et al. (176) remarked that "no defined medium could be produced in which the thermophilic fungi would grow ... ." Rosenberg (218) reported that nearly half of the species of thermophilic and thermotolerant fungi tested required 0.01% yeast extract for growth in a solid medium. Wali et al. (260) reported that for growth in a liquid medium containing glucose and ammonium sulfate, the thermophilic fungi required a supplementation of succinic acid, a tricarboxylic acid cycle intermediate. While this observation was confirmed in our laboratory (102), we additionally demonstrated that because of the low phosphate concentration in the culture medium, the pH of the medium in the absence of the organic acid dropped to ~3.4 after 12 to 24 h and the growth ceased. Moreover, any one of the several tricarboxylic acid cycle acids tested stimulated growth, which was due to their buffering action in the medium rather than to their nutritional role: thermophilic fungi grew satisfactorily in a minimal medium if the pH of the medium was controlled between 5.5 and 7.0 by increasing the phosphate concentration in the medium, readjusting the pH by addition of an alkali, including powdered calcium carbonate as a reserve alkali, or replacing the inorganic nitrogen source with an organic nitrogen source (L-asparagine). The low pH reduces the solubility of CO₂ in the growth medium and limits its availability for assimilation by the anaplerotic enzyme pyruvate carboxylase (102). Although CO₂ is not regarded as a nutritional requirement for fungi, growth of T. lanuginosus was severely affected if the gas phase in the culture flask was devoid of CO₂. The concentration of CO₂ inside composts can be as high as 10 to 15% (74); therefore, it is likely that its assimilation plays nutritional and morphogenetic roles in the development of thermophilic fungi, which are the primary components of the microflora of such habitats. It is interesting that this gas has been identified as essential for the axenic culture of a rust fungus, which had long been regarded as an obligate parasite on plants (37).

Thermophilic fungi have a widespread distribution in tropical as well as temperate regions (157). Tendler et al. (247) remarked that "the ubiquitous distribution of organisms, whose minimal temperature for growth exceeds the temperatures obtainable in the natural environment from whence they were isolated, still stands as a `perfect crime' story in the library of biological systems." They considered whether eukaryotic thermophily is an artifact of the nutritional environment. To test this, thermophilic isolates of Humicola, Thermoascus, and Aspergillus were incubated in a nutritionally rich liquid medium that included glucose, mannitol, starch, Casamino Acids, yeast extract, and peptone. After 10 days at 20°C, these fungi had generated good growth, although they had failed to grow below 30°C in a sucrose-salts medium that lacked complex supplements. It was suggested that the complex materials contained a factor(s) which the organisms could not synthesize at the lower temperature. This suggestion was supported by the observation that the growth of Talaromyces thermophilus at a suboptimal temperature (33°C) benefited from the supplementation of culture medium with 5 µg of ergosterol per ml (264).

Furthermore, the choice of inoculum, i.e., spores versus germinated spores or mycelium, may also influence the minimal temperature of growth of thermophilic fungi. Whereas Mucor miehei did not grow at 25°C in submerged cultures when spores were used as the inoculum, substantial growth occurred at this temperature (nearly 64% of that at 48°C in 72 h) in as much time when pregerminated spores were used as inocula (237). Similarly, we observed that a mycelial inoculum, but not a spore inoculum, resulted in a near maximal yield of Thermomyces lanuginosus at 25°C; growth occurred without a perceptible lag but at a lower rate than at 50°C (Fig. 1). In Thermoascus aurantiacus, the lowest temperature at which the ascospores germinated was 10 to 12°C higher than that for hyphal growth (71). Presumably this may be a means of minimizing competition from the mesophilic fungi and ensuring that warm conditions would be available for yet more time until the mycelium is established for resource capture. In light of the observations that the conditions for spore germination can be more exacting than those for hyphal growth, the reported minimal temperature of growth of thermophilic fungi should be redetermined, specifying the method of inoculation and the composition of the medium used. The above observations also suggest that once the spores have been induced to germinate at high temperature, the requirement of high temperature for sustaining growth may not be critical.

Many organisms vary the fatty acid composition of their membrane phospholipids as a function of growth temperature so that their membrane fluidity is kept constant for the optimal functioning of membrane-localized transporters and enzymes. For example, with an increase in temperature, there is an increase in the proportion of saturated fatty acids incorporated into phospholipids, whereas at lower temperature, a higher proportion of unsaturated fatty acids is incorporated. This phenomenon is called homeoviscous adaptation (229). Wright et al. (264) examined whether an inability to regulate membrane fluidity may be a reason for the high minimum temperature of growth of thermophilic fungi. They reported that when Talaromyces thermophilus was shifted from a high (50°C) to a low (33°C) growth temperature, the degree of unsaturation of fatty acids at the two stated temperatures remained virtually unchanged. This was thought to be the result of a metabolic limitation, presumably due to a nonfunctional fatty acid desaturase, which restricted the ability of the fungus to convert oleate to linoleate at low temperature. However, results with T. lanuginosus were different. In this fungus, the concentration of linoleic acid (18:2) was twofold higher at 30 than at 50°C. The degree of unsaturation of phospholipid fatty acids was 0.88 in mycelia grown at 50°C but 1.0 in the temperature-shifted cultures (from 50 to 30°C) and 1.06 in cultures grown at constant 30°C (209). A decrease in the degree of unsaturation was also observed in Chaetomium thermophile when it was subjected to heat shock (190). As mentioned above, some species of thermophilic fungi are capable of growth even at mesophilic temperatures. Therefore, it seems unlikely that the inability to adjust membrane fluidity is the general reason for their high minimum temperature of growth. As was reasoned for thermophilic bacteria (240), the loss of catalytic potential of one or more vital enzymes, caused by conformational changes and/or ribosomal assembly, may be an important determinant of the minimal growth temperature.

The need to explain the occurrence of thermophilic fungi in soil, which may warm to favorable temperatures because of solar radiation, but only for a transitory period, prompted Noack (188) to investigate the effects of subminimal temperatures and rewarming. He noticed that when an actively growing culture of Thermoascus aurantiacus was cooled to 31°C (4°C below the lower temperature limit of growth), its respiration after 2 days declined only minimally but that cultures kept at 21°C for 24 h stopped respiring. When the culture medium was rewarmed to 46°C, practically no respiration was observed. Whether the low resistance to lower temperature displayed by T. aurantiacus is also applicable to other thermophilic fungi is not known. However, until more information is available, it should not necessarily be assumed that mycelial cultures of thermophilic fungi can be stored under refrigeration or at subminimal temperatures without loss of viability.

Noack (188) recognized that during the self-heating process the environment in composts would become oxygen deficient. He therefore studied the behavior of Thermoascus aurantiacus subjected to anaerobiosis and found that the withdrawal of oxygen severely affected its respiration and growth. Although thermophilic fungi do not have the ability to undergo anaerobic growth (136), Humicola insolens was reported to grow better under anaerobic or microaerobic conditions than under aerobic conditions at elevated temperatures (116; also see reference 78). Cooney and Emerson (63) reported an interesting morphogenetic effect of anaerobiosis in Talaromyces (Penicillium) duponti. This thermophilic fungus formed only a conidial stage (Penicillium) in aerobic cultures; the sexual stage (Talaromyces) was initiated in agar cultures only when they were flushed with nitrogen. It would be worthwhile to study the effect of anaerobic conditions in species in which the sexual stage has not been discovered so far or is observed only infrequently.

We referred earlier to the work of Kurt Noack (188). He sought to determine whether thermophilic fungi have an exceptionally high rate of metabolism accompanied by a high rate of substrate conversion. He observed that the economic yield (grams of sugar consumed per gram of mycelial dry weight formed) of Thermoascus aurantiacus grown in a minimal medium at 45°C (1.89) was the same as that of the mesophilic fungus Aspergillus niger grown at 25°C. From this, he inferred that the overall metabolisms of the two types of fungi must be quite similar. Moreover, he estimated that on average both types of fungi converted 55% of sugar for the synthesis of fungal biomass and 45% for metabolism. Since few values of economic coefficients have been reported, we compared the molar growth yield (grams of biomass produced per mole of glucose utilized) (Y_G) of fungi. From the data available (Table 2), the average Y_G values of mesophilic and thermophilic fungi at their respective temperature optima are quite comparable (86 to 88 g/mol), suggesting that similar proportions of carbohydrate are used by both types of fungi for macromolecular synthesis. This value is close to that found by Noack for the fungal species studied by him.

Under the culture conditions used by us, some thermophilic fungi (Thermomyces lanuginosus, Penicillium duponti, Sporotrichum thermophile, and Malbranchea pulchella var. sulfurea) produced exceptionally homogeneous mycelial suspensions when grown in a glucose-asparagine medium in shake cultures at 50°C (102, 154, 200). This allowed the sampling of mycelia by using pipettes for quantitative measurements during their growth and facilitated the determination of growth rates, growth yields (see above), the effect of temperature-shift, and other aspects of physiology. The ranges of growth rates of thermophilic and meophilic fungi were similar (Table 2). Interestingly, although the growth of T. lanuginosus at a suboptimal temperature was slowed, biomass production was not affected (Fig. 1). Using CO₂ produced as an index of development of fungal biomass, Wiegant (262) observed that the exponential growth rate of the thermophilic fungus Scytalidium thermophilum (a common thermophilic fungus in mushroom compost) at 45°C in a liquid medium supplemented with malt and yeast extracts was 0.41 h¹. Thus, contrary to a common belief, thermophilic fungi do not, in general, grow faster than mesophilic fungi. The situation is similar to that in thermophilic bacteria (44, 240).

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Growth of Thermomyces lanuginous in submerged cultures at different thermal regimens (semilogarithmic plot). Reprinted from reference 208 with permission of the publisher.

Using the volume of carbon dioxide evolved over time as a measure of metabolism, Noack (188) compared a thermophilic fungus (Thermoascus aurantiacus) with a mesophilic fungus (Penicillium glaucum) grown in identical medium. He observed that the volume of carbon dioxide released by P. glaucum in 24 h was equivalent to 67% of its dry weight at 15°C and 133% at 25°C. He argued that if this fungus could grow at 45°C, the extrapolated value of carbon dioxide, according to van't Hoff rule, would be 532%. However, the actual value for T. aurantiacus at 45°C was 310%. From this, Noack inferred that at a given temperature the metabolism of a thermophilic fungus is actually slower than that of a mesophilic fungus. Subsequent studies have shown that the two types of fungi have nearly comparable respiratory rates at their respective temperature optima (203, 210).

Noack (188) observed that with increases in temperature, the increase in the respiration of Thermoascus aurantiacus was lower than that of mesophilic fungi. However, we found that the rates of oxygen uptake of homogeneous mycelial suspensions of thermophilic fungi (Thermomyces lanuginosus and Penicillium duponti), measured by Warburg manometry, were markedly responsive to changes in temperature between their minimal (30°C) and optimal (50°C) temperatures of growth (203, 210). In contrast, the oxygen uptake rates of the mesophilic fungi tested (Aspergillus niger, A. phoenicis, and Trichoderma viride) were either independent of temperature changes between 15 and 40°C or affected to a lesser degree, at least during the period of measurement (Fig. 2). The biochemical basis of this difference is not known, but this behavior may be significant in relation to their growth in nature. If mycelia in the nutritionally poor environment of soil respond similarly, then fungi which maintain an optimal metabolic rate over a broad range of temperatures may be expected to have a competitive advantage over those which lack this ability. Mesophilic fungi, rather than thermophilic fungi, would be better adjusted to soil, where temperatures vary both spatially and temporally. Although a Warburg apparatus or an oxygraph is no longer standard equipment in laboratories, the investigations of respiratory metabolism still are very likely to lead to insights to physiological adaptations. In this context, another interesting observation was that whereas in the mesophilic fungus Agaricus bisporus both cytochrome and alternative respiratory pathways are present, in the thermophilic fungus Scytalidium thermophilum only the cytochrome-mediated respiratory pathway was present (74). Furthermore, high concentrations of CO and HCN severely inhibited the growth of S. thermophilum but not of A. bisporus.

View larger version (19K): [in this window] [in a new window]   FIG. 2.   Arrhenius plots of QO2 (microliters of oxygen taken up per mg [dry weight] of mycelium per hour) of shaker-grown mycelia of Thermomyces lanuginosus and Aspergillus niger. Reprinted from reference 203 with permission of the publisher.

Of the possible factors which restrict the growth of thermophilic fungi at ordinary temperatures, the reduction in the rate of uptake of nutrients appears to be particularly important. However, little information is available on the effect of growth temperature on the synthesis and activity of transport systems in fungi. Palanivelu et al. (196) identified a specific sucrose transport activity in T. lanuginosus, which was coinduced with invertase activity in mycelia exposed to -fructofuranosides (sucrose or raffinose). Both activities appeared in sucrose-grown mycelia at about the same time, and both declined simultaneously following the exhaustion of sucrose in the medium. The uptake of sucrose was inhibited by ionophores that dissipate the proton gradient, suggesting that transport of sucrose is H⁺ coupled. Furthermore, the uptake measured at 50°C followed Michaelis-Menten kinetics, with an apparent K_m of 250 µM (154). Transport of glucose in T. lanuginosus was a constitutive, specific, carrier-mediated process (K_m = 290 µM) that functioned as a proton-driven symport (A. K. Rajasekaran and R. Maheshwari, unpublished data). In preliminary experiments, an unexpected effect of the assay temperature on the rate of glucose uptake by T. lanuginosus was observed. Regardless of the growth temperature of the mycelia (50 or 30°C), glucose uptake by mycelia was saturated when assayed at 50°C. In contrast, it increased linearly with increasing concentrations of 2-deoxyglucose (tested up to 2 mM) at 30°C. The temperature-dependent sensitivity in the kinetics of glucose uptake may be due to a temperature-induced conformational change in glucose transporter.

The characteristics of sulfate permease in Penicillium duponti were similar to those in the mesophilic fungus P. chrysogenum (an Na⁺/H⁺/SO₄² symport system derepressed by sulfur starvation, activated by divalent cations and inhibited by molybdate, vanadate, and tungstate, K_m of 57 µM) (263). They differed in that the permease of the thermophilic fungus was optimally active at 45°C and in its sensitivity to the ionic strength of the solution: mycelium that had been washed with deionized water lost transport activity.

Thermophilic fungi develop in composts during the high-temperature phase, succeeding a mesophilic microflora (55, 115). Since most of the initially available soluble carbon sources (sugars, amino acids, and organic acids) would have been depleted, the carbon source available for the growth of thermophilic fungi would be mainly the polysaccharide constituents of the biomass, of which cellulose is the chief constituent. Interestingly, some compost fungi are unable to utilize cellulose, for example, Thermomyces lanuginosus (55, 115, 204), Talaromyces duponti, Malbranchea pulchella var. sulfurea, Mucor pusillus (55), and Melanocarpus albomyces (156). The noncellulolytic species in compost can grow commensally by utilizing sugars released during the hydrolysis of hemicellulose and cellulose by the cellulolytic partner. For example, T. lanuginosus showed profuse growth in mixed cultures with a cellulolytic fungus, Chaetomium thermophile (115). Moreover, several noncellulolytic species readily utilize xylan, which is external to cellulose in the plant cell wall and is apparently a more accessible carbon source (199). Indeed, some fungi (C. thermophile and Humicola insolens) grow even better on xylan than on simple sugars (55). The secretion of thermostable extracellular polysaccharide-degrading enzymes and the simultaneous uptake of sugars would be important attributes of thermophilic fungi in self-heating masses of plant material. In T. lanuginosus, a single transporter was identified for glucose, xylose and mannose, the hydrolytic products of cellulose and hemicellulose (Rajasekaran and Maheshwari, unpublished).

Because the measurement of biomass of fungi growing on insoluble polysaccharides is indirect, it has rarely been done. We measured the growth of Sporotrichum thermophile on cellulose in terms of insoluble nitrogen or as an increase in mycelial dry weight after selectively estimating the amount of cellulose and subtracting its weight from that of the samples (32). Interestingly, the exponential growth rate of the fungus on cellulose (0.09 to 0.16 h¹) was similar to that on glucose (0.1 h¹), revealing the remarkable ability of this fungus to utilize cellulose as efficiently as glucose. The visual characteristics of the fungus were strikingly different in submerged cultures grown with cellobiose (repeating unit of cellulose) or cellulose (91). The mycelia in cellobiose-grown cultures retained a prolonged filamentous and healthy appearance, whereas in cellulose medium they rapidly autolysed and sporulated. Perhaps oligosaccharides derived from the hydrolysis of cellulose regulate the gene expression and metabolic process differently from when the fungus is growing on soluble sugars. Another interesting observation was the influence of culture temperature on fungal morphology when grown with cellulose. At a suboptimal temperature (30°C), the conidia of S. thermophile formed a very limited mycelium that precociously developed asexual reproductive structures (microcycle conidiation). Although the mechanism of this cellular response is not understood, microcycle conidiation (Fig. 3) may be a survival strategy of producing propagules in the shortest possible time under suboptimal conditions.

View larger version (137K): [in this window] [in a new window]   FIG. 3.   Microcycle conidiation in Sporotrichum thermophile. The fungus was grown in shake cultures with shredded Whatman filter paper as the carbon source. (A) Phase-contrast micrograph of a 24-h-old germling that has produced oval asexual spores. The insoluble particle is a piece of cellulose fiber. (B) Phase-contrast micrograph showing precocious differentiation of asexual spores in a 72-h-old germling grown at 30°C. The germinated conidium is indicated by an arrow. Bars, 50 µm.

In composting plant material, the hydrolysis of polysaccharide constituents by the secreted enzymes would be expected to produce a mixture of sugars. We determined if thermophilic fungi utilize one sugar at a time or a mixture of sugars simultaneously (154). In the only study so far with fungi, a combination of glucose and sucrose was chosen, because the concentrations of these sugars in the medium can easily be determined using commercially available enzymes. The fungi studied, Thermomyces lanuginosus and Penicillium duponti, concurrently utilized glucose and sucrose at 50°C, with sucrose being utilized faster than glucose. The phenomenon was studied further with T. lanuginosus. Its growth rates on single- or mixed-carbon sources were comparable. The rate of utilization of glucose and sucrose in the mixture was lowered unequally compared to when the sugars were provided singly, indicating that the two sugars reciprocally influenced their utilization in the mixture. The simultaneous utilization of sucrose in the presence of glucose occurred because (i) invertase was insensitive to catabolite repression by glucose and (ii) the activity of the glucose uptake system was repressed by glucose itself as well as by sucrose. Both sugars were also utilized concurrently at 30°C but at nearly identical rates. This observation indicates that the activity of nutrient transporters and the sensitivity of catabolic enzymes to glucose repression can be influenced differently at different temperatures.

One of the early hypotheses put forward to explain thermophily in bacteria was that rapid breakdown of proteins at elevated growth temperatures is compensated by their fast resynthesis (9). This rapid-turnover hypothesis came to be known as the dynamic hypothesis of thermophily. To examine its applicability to thermophilic fungi, Miller et al. (176) compared protein breakdown rates in thermophilic (Penicillium duponti, Malbranchea pulchella, and Mucor miehei) and mesophilic (Penicillium notatum and P. chrysogenum) fungi by monitoring the breakdown of pulse-labeled protein. The growing cells in both types of fungi had a negligible rate of breakdown of bulk protein. In the nongrowing cells of both types, the breakdown rate was similar (5.2 to 6.7% per h). However, the breakdown rate of the soluble-protein fraction in thermophilic fungi was twice that in the mesophilic fungi. The authors suggested that the increased turnover rate of soluble protein is important in the survival of thermophilic fungi at the high temperatures.

The energy expended in increased protein turnover in thermophilic fungi would be expected to affect their growth yield compared to mesophilic fungi. However, the growth yields of thermophilic and mesophilic fungi examined were in a similar range (Table 2). Therefore, we reexamined protein breakdown in fungi (208) by selecting fungi which produced homogeneous mycelia, thereby obviating the sonication treatment used by Miller et al. (176) for rendering mycelia homogeneous for sampling and measuring radioactivity. Moreover, we attempted a general labeling of cellular proteins by adding radioactive amino acids at two different times and avoided the use of toxic concentrations of amino acids during the chase. Furthermore, we used the radioactivity of the whole cells as an index of the radioactivity of total protein rather than of the protein that was extractable by ultrasonic disruption of mycelia. The results pointed to a definite protein turnover in the growing cells of fungi, although the rate of protein turnover varied among the different species (Table 3). Under different conditions of incubation, the protein breakdown rate was somewhat lower in thermophilic fungi than in mesophilic fungi. Although we measured only protein breakdown, it is likely that the results would be similar had protein turnover been measured. As in bacteria (240), the rapid-turnover hypothesis in thermophilic fungi is not supported by experimental results. Nonetheless, specific enzymes may well have a high turnover rate.

Acquired thermotolerance refers to the enhanced survival of organisms at lethal temperature after a brief exposure to sublethal temperatures. A common phenomenon in mesophilic species is the synthesis of a set of proteins, called heat shock proteins (HSPs), following a sudden exposure of organisms to elevated temperature. The synthesis of HSPs is thought to be an adaptive response to increased thermotolerance and survival in the face of stressful conditions. Trent et al. (251) demonstrated that in common with thermophilic bacteria and archaea, thermophilic fungi also synthesize HSPs and acquire thermotolerance. They observed that conidia of T. lanuginosus, germinated at 50°C and heat shocked at 55°C for 60 min prior to exposure to 58°C, showed enhanced survival compared to non-heat-shocked conidia. Thermotolerance was eliminated if protein synthesis during the heat shock period was inhibited by cycloheximide. Pulse-labeling of proteins during the heat shock period, followed by their separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), showed increased synthesis of eight HSPs. Of these, three small HSPs ranging from 31 to 33 kDa dominated the heat shock response. Since T. lanuginosus is capable of growth up to 60 to 62°C, the role of HSPs induced at 55°C is not clear. A transient HSP synthesis was also observed in Chaetomium thermophile var. thermophile (190). In particular, a constitutive, abundantly expressed protein, HSP60, belonging to the intercompartmental transport proteins was thought to be important in thermophily.

In closing this section, we should mention that the impression gained is that only cursory explorations of the physiology of thermophilic fungi have been made, prompted primarily by a desire to explain their widespread distribution in soila habitat characterized by changing temperatures, uncertain supplies of nutrients and water, and the coexistence with a numerically high mesophilic microflora and fauna consisting of competitors and predators. Notwithstanding the reported latent abilities of some species to grow at ordinary temperatures, the controversy concerning the growth and reproduction of thermophilic fungi in soils is far from settled. Nonetheless, the attempts so far have dispelled certain early notions concerning their rates of growth, their nutritional needs, and their overall metabolism.

Proteases have long been used in the food, dairy, and detergent industries and for leather processing. The need to overcome the limitation of obtaining chymosin, the milk-curdling enzyme from the stomach contents of milk-feeding calves, which is used in the industrial preparation of cheese, led to a search for substitutes. Arima et al. (17) screened about 800 microorganisms and obtained a soil isolate of Mucor pusillus that produced an enzyme with a high ratio of milk-clotting to proteolytic activity, enabling the production of high yields of curds. Subsequently, a strong milk-clotting activity was also observed in M. miehei (195). The milk-clotting activity of the enzyme was due to its selective attack on the k-casein fraction, which stabilizes the casein micelle in milk. The split k-casein loses its stabilizing activity, and the micelles of casein coagulate in the presence of calcium (233). The finding of a milk-clotting activity of utility resulted in serious efforts to isolate other thermophilic fungi in order to exploit them for industrially useful enzymes.

The Mucor rennins were produced by growing the fungus on wheat bran, from which they were extracted with water. The crude extract was then purified and crystallized (18, 20, 127, 195, 230, 234). Since the rennins hydrolyzed both casein and hemoglobin optimally at pH 3.7, they were classified as acid protease (EC 3.4.23.6). Both M. pusillus and M. miehei rennins hydrolyzed peptide bonds in synthetic peptides with an aromatic amino acid as the carboxyl donor (20, 195, 233). The M. pusillus enzyme was stable from pH 3.0 to 6.0 and showed maximum activity at 55°C (273). The enzyme was inhibited by the aspartic protease inhibitors diazoacetyl-DL-norleucine methyl ester and pepstatin (241). Sequence comparison with the other well-characterized aspartic proteases confirmed the presence of aspartic acid at their active site (25). M. pusillus and M. miehei rennins had similar molecular masses (38.5 and 42 kDa) and isoelectric points (3.9 and 4.1), respectively (84). Although the Mucor rennins (aspartic proteases) are structurally homologous, they differ from other fungal aspartic proteases and from mammalian proteases (25, 187, 249, 269). The specificity of Mucor rennins is similar to that of pepsin and calf rennin.

The Mucor rennins are being used as model systems in investigations of the heterologous expression of fungal protein, the mechanism of zymogen processing, the refolding of recombinant protein from inclusion bodies, the effects of glycosylation on secretion, and the activity and stability of mutant enzymes with amino acid insertions (6, 38, 100, 118). The Mucor rennin genes were cloned in Escherichia coli and sequenced (38, 100, 249). Their deduced amino acid sequence showed that the enzyme was synthesized as a zymogen that contained an N-terminal leader region constituting a typical signal peptide of 22 amino acids and a propeptide of 44 to 47 amino acids. This leader sequence was not present in the mature protein (361 amino acids). The expression of the Mucor protease gene in E. coli resulted in the accumulation of unsecreted, inactive polypeptide (38, 100), but when the gene was expressed in yeast, a form of the zymogen that was more glycosylated than the native enzyme was secreted at a concentration exceeding 150 mg/liter (267). The prosequence of the heterologous secreted protein was removed by autocatalytic processing at acidic pH, yielding an active rennin with properties almost identical to those of the native enzyme (118, 267). A mutation of two of the three glycosylated asparagine sites in the recombinant M. pusillus rennin significantly reduced the amount of secreted rennin by the yeast cells and decreased its milk-clotting activity and thermal stability compared to those of nonmutated M. pusillus rennin (6). It has been hypothesized that the flexible carbohydrate structures act as "heat reservoirs" and stabilize the conformation of rennins (269). The recombinantly produced M. miehei protease in Aspergillus nidulans was similar in specific activity to that produced by M. miehei (100). Subsequently, using an -amylase promoter of Aspergillus oryzae, a recombinant A. oryzae strain was constructed that produced heterologous M. miehei protease in excess of 3 g/liter (61).

The high thermal stability of Mucor rennins turned out to be an undesirable property, since the residual enzyme activity, after cooking, can spoil the flavor of cheese during the long maturation process (268). Efforts are therefore being made to engineer Mucor rennins with lower thermostability but with the same milk-clotting potential. Nevertheless, the basic studies on Mucor enzymes have had interesting spinoffs. For example, the leader peptide of M. pusillus rennin was found to be useful for the secretion of a heterologous protein by yeast cells: when the human growth hormone gene was fused to the whole presequence and a part of the prosequence of M. pusillus rennin and expressed in yeast under the control of the yeast GAL7 promoter, the level of the secreted hormone reached approximately 10 mg/liter (119).

A source of one of the most thermostable fungal acid proteases was a strain of Penicillium duponti isolated from compost (105, 106). The enzyme was produced in submerged cultures, containing rice bran, at 50°C with vigorous aeration and agitation. It was purified by alcohol precipitation, ion-exchange chromatography, and gel filtration (107). Subsequently, it was crystallized with an yield of 3.3 g from 200 g of crude protease preparation (79). The enzyme was most active at pH 2.5 on casein and at pH 3.0 to 3.5 on hemoglobin. It had a molecular mass of 41 kDa and contained 4.3% carbohydrate. P. duponti protease retained its full activity after 1 h at 60°C and pH 4.5. By comparison, the acid protease of M. pusillus was irreversibly destroyed in 15 min at 65°C (230).

Two fungal sources of thermostable alkaline proteases were identified based on the zone of clearing of casein agar by culture filtrates as a semiquantitative assay of proteolytic activity: Malbranchea pulchella var. sulfurea and Humicola lanuginosa (193, 194). Both produced proteases during active growth in the presence of 2 and 8% (wt/vol) casein, respectively, suggesting that the enzyme was induced by external protein substrate (235). The production of protease by M. pulchella var. sulfurea was repressed by glucose, peptides, amino acids, or yeast extract (194). The M. pulchella protease, named thermomycolase, could be concentrated from the culture medium simply by vacuum evaporation at 45°C without a loss of activity. After dialysis and removal of pigments by adsorption on a cation-exchange column, a homogeneous preparation of the enzyme was obtained by affinity chromatography on a hydrophobic adsorbent with 78% yield. Its substrate specificity was investigated by analyzing the peptides formed in the reaction digests of glucagon and the A and B chains of oxidized insulin. At 45°C and pH 7.0, thermomycolase exhibited a general proteolytic activity rather than a well-defined specificity for any particular amino acid residue (235). The enzyme was classified as a serine protease based on inhibition of its activity by diisopropylfluorophosphate (DFP), which covalently attaches to a reactive serine residue. Thermomycolase was optimally active at pH 8.5 and was stable over a broad pH range (6.0 to 9.5 for 20 h at 30°C). As purified thermomycolase autolyzed, it resulted in low-molecular-mass peptides. The physicochemical characterization of the protein was therefore carried out using a DFP-inhibited enzyme. The molecular mass was 11 to 17 kDa when determined by gel filtration and 32 to 33 kDa when estimated from sedimentation equilibrium of DFP-thermomycolase and by SDS-PAGE (256). The thermostability of the enzyme depended on the concentration of calcium ions, a property shared by other thermostable enzymes, e.g., thermolysin, the neutral protease of Bacillus thermoproteolyticus (257). The t_1/2 (time required for the activity of the enzyme to fall to 50% of its original value at a given temperature) was 110 min at 73°C in the presence of 10 mM calcium. Calcium (10⁴ M) was necessary to stabilize the enzyme against autolytic degradation at low temperatures: one calcium ion bound to the enzyme molecule with high affinity (258), markedly lowering the rate of autolytic degradation and of thermal or urea denaturation but causing no detectable change in the conformation of the enzyme. It was postulated that the bound calcium stabilized the enzyme by raising the local activation energy for unfolding.

An alkaline protease of Humicola lanuginosa (Thermomyces lanuginosus) was studied by two different groups, and their enzyme preparations differed in important properties. Shenolikar and Stevenson (226) purified the enzyme in one step based on its specific binding to an organomercury-Sepharose column, from which the enzyme was selectively eluted with a buffer containing mercuric chloride. The mercury-enzyme complex, reactivated by cysteine in the presence of EDTA, autolyzed rapidly. Based on the inhibition of the enzyme by some reagents, such as Hg²⁺ or p-chloromercuribenzoate, which react with free thiols, and its reactivation by cysteine in the presence of EDTA, the enzyme was identified as a unique thiol proteinase produced by a fungus (226). Both gel filtration and sedimentation analyses showed that the enzyme had a molecular mass of 237 kDa. It preferentially cleaved its substrate at the C-terminal end of the hydrophobic amino acid residues. However, Hasnain et al. (108) failed to find a protease that specifically bound to the affinity matrix used by Shenolikar and Stevenson (226). Rather, the protease activity was purified by hydrophobic affinity chromatography and shown to have a pH optimum of 8.0 and a molecular mass of 38 kDa and was inhibited by phenylmethylsulfonyl fluoride, an inhibitor that is specific to serine protease. Although the enzyme preparation obtained by Hasnain et al. was also inhibited by Hg²⁺ and p-chloromercuribenzoate, it was not inhibited by alkylating agents [iodoacetic acid, iodoacetamide or 5,5'-dithiobis(2-nitrobenzoic acid)], suggesting that it was a serine protease containing a partially buried cysteine. The authors (226) apparently favored the view that the fungus secreted only a single protease. Such differences in properties are generally explained on the basis of strain differences or culture-condition-induced modifications in enzyme structure. We shall encounter other examples of differences in enzyme properties from the same (?) fungus.

Arima et al. (19) purified an extracellular lipase from Humicola lanuginosa strain Y-38, isolated from compost in Japan. The enzyme was produced in a medium containing soybean oil, starch, corn steep liquor, and antifoaming agent. It was purified to homogeneity from 80-h-old culture medium by successive steps of ammonium sulfate precipitation, dialysis, ion-exchange chromatography, and gel filtration chromatography, with 30% recovery. The protein, a single polypeptide (molecular weight, 27,500), was optimally active at pH 8.0 and was stable in the pH range of 4 to 11. Its temperature optimum for activity was at 60°C. It showed appreciable activity at up to 65°C but was inactivated on heating at 80°C for 20 min (149). The enzyme could be stored frozen for more than 6 months. The enzyme molecule contained disulfide linkages but no free ---SH group.

Omar et al. (191, 192) reported that the productivity and thermostability of lipase differed with different strains of H. lanuginosa. These workers developed an optimized medium containing sorbitol, corn steep liquor, silicone oil as an antifoaming agent, and whale or castor oil as enzyme inducer. With the pH maintained between 7 and 8 and the temperature set at 45°C, maximum enzyme production by their strain occurred after 30 h. Following acetone precipitation and successive chromatographic steps, they obtained a more thermostable enzyme (stable at 60°C for 20 h) than was obtained by Arima et al. (19). Their preparation was optimally active at pH 7.0. The enzyme showed increased activity in organic solvent-aqueous reaction systems, but hydrolysis in complete organic phase reactions did not occur.

A lipase gene from H. lanuginosa was cloned and expressed in Aspergillus oryzae (122). The lipase expressed in the heterologous host was purified by a two-step procedure involving hydrophobic interaction chromatography and ion-exchange chromatography (198). The structure of H. lanuginosa lipase (73) was similar to that of R. miehei lipase. Structural studies have been directed toward an understanding of the phenomenon of interfacial activation. Modification in the lid (described below) of H. lanuginosa lipase by site-directed mutagenesis provided direct evidence of the importance of a number of residues within the lid in terms of substrate binding and specificity (120). By spectroscopy and molecular dynamics simulation, Peters et al. (198) showed that a single mutation of a serine residue in the active site leads to substantial alterations in the motion of the lid and binding affinity of enzyme.

Rhizomucor miehei, formerly called Mucor miehei, also produces active extracellular lipase. The isolation and purification methods for M. miehei lipase have been described by Huge-Jensen et al. (123). The lipase produced predominantly in form A (see below) was purified by anion-exchange chromatography followed by affinity chromatography and further purified by hydrophobic interaction chromatography (39). If the purification steps were carried out at pH 4.5 instead of pH 7.0, form A was partially deglycosylated and converted to form B. The two forms showed a high degree of antigenic similarity and were optimally active at pH 7.0. However, form A, in contrast to form B, required a prior alkaline (pH 10.5) treatment for maximal activation. The apparent molecular mass of both lipases was ~32 kDa. The carbohydrate contents of purified R. miehei lipase forms A and B were 11% and 4% (wt/wt), respectively. The lipases rapidly hydrolyzed a broad spectrum of lipids found in animal fat and vegetable oil. The enzyme remained active even after exhaustive drying (254).

Boel et al. (39) constructed an R. miehei cDNA library in E. coli. From the DNA sequence data and the deduced amino acid sequence, they inferred that, unlike the characterized bacterial and mammalian enzymes, the R. miehei lipase was synthesized as a zymogen with a signal peptide of 24 amino acid residues and a further 70-amino-acid propeptide. Maturation involved a proteolytic cleavage to remove the propeptide. In further work (122), a recombinant plasmid containing the cDNA of R. miehei lipase was expressed in the filamentous fungus Aspergillus oryzae and the enzyme was obtained in large quantities. Heterologous expression did not affect the characteristics of the enzyme, showing that the precursor was correctly processed in A. oryzae.

R. miehei lipase has 269 amino acid residues. It was the first lipase whose three-dimensional structure was deduced by X-ray analysis (42). The lipase is an /-type protein, having a core of central, mostly parallel -sheets connected by a variety of hairpins, loops, and helical segments. Although the overall protein structure is quite unrelated to that of the serine proteases, the lipase catalytic center has the same three amino acids (serine, histidine, and aspartic acid) which characterize serine proteases as well. The finding of this "catalytic triad," which set up an H⁺ shuttle or the charge relay system at the active site of lipase, as in the enzymes of the trypsin family, suggests their convergent evolution. An interesting feature of the protein molecule is that the catalytic site is covered by a short -helical loop that acts as a "lid." When the enzyme is adsorbed at the oil-water interface, the lid moves, allowing access of the substrate to the active site and at the same time exposing a large hydrophobic surface which apparently facilitates the binding of the lipase to the lipid interface (72).

H. lanuginosa lipase (Lipolase; Novo Industri A/S) is being used in detergent formulations in conjunction with other microbial enzymes (e.g., protease, amylase, and cellulase). In addition, lipases have applications in the food industry and are used for the biocatalysis of stereoselective transformations (128). Lipase (Lipozyme; Novo Industri A/S) from R. miehei is used to produce a cocoa butter substitute from a cheaper edible oil, in which oleate exclusively occupies the sn-2 position but palmitate rather than stearate predominates at the sn-1 and sn-3 positions (104, 128). Efforts in protein engineering of lipases are being made to obtain improved binding to negative charges on the lipid surface, to open the lid and activate the enzyme, and to increase stability to high pH and to anionic and nonionic surfactants.

-Amylase (EC 3.2.1.1) hydrolyzes -1,4-glycosidic linkages in starch to produce maltose and oligosaccharides of various lengths. All species of thermophilic fungi studied so far secrete amylase (3, 4, 23, 46, 85, 129, 220). However, only T. lanuginosus -amylase has been characterized. The addition of Tween 80 to agitated submerged cultures increased -amylase production 2.7-fold (22). Although a multiplicity of -amylases is common in fungi, only one electrophoretically similar form of the enzyme was detected in culture filtrates of seven strains of T. lanuginosus (177). -Amylases from two strains have been purified, but the estimations of their molecular masses gave ambiguous results. While the enzyme purified from stationary cultures of strain 1457 had a molecular mass of 54 to 57 kDa by SDS-PAGE (130), the enzyme purified from shaker-grown cultures of strain IISc 91 gave different values by different methods: ~24 kDa by SDS-PAGE, ~72 kDa by gel filtration, and ~42 kDa by Ferguson analysis on native PAGE (177). Since the partial amino acid sequence of IISc 91 -amylase showed a single N-terminal amino acid, it was suggested that the native enzyme is a homodimer with an apparent molecular mass of ~42 kDa. This was the first report of a dimeric form of -amylase in fungi. The -amylases were stabilized by calcium. For example, 10 mM calcium increased the thermostability of IISc 91 -amylase by eightfold. After addition of Ca²⁺, its half-life at 65°C was 4.5 h. A novel observation was that IISc 91 -amylase underwent structural changes upon heating to 94°C: the native, dimeric enzyme was progressively and irreversibly inactivated, and the subunits dissociated and reassociated to produce an inactive, 72-kDa trimeric species which fragmented on further heating. The enzyme produced exceptionally high levels of maltose from raw potato starch (177).

Glucoamylase (E.C. 3.2.1.3) is an exo-acting enzyme which hydrolyzes -1,4-glycosidic linkages and, less frequently, -1,6-glycosidic linkages from the nonreducing end of starch, producing -D-glucose as the sole product. During growth in a medium containing starch, T. lanuginosus also produced glucoamylase (213, 214, 246), which was separated from -amylase by conventional procedures of protein purification. Although maltose had been reported to be a better inducer of glucoamylase in this fungus (103), growth in a starch medium allowed both glucoamylase and -amylase to be produced simultaneously. Both enzymes were obtained in milligram quantities from 2 liters of culture filtrates (177). Although glucoamylases of T. lanuginosus had similar carbohydrate contents (10 to 12%), different authors have reported different molecular masses for the enzyme: ~57 kDa by gel filtration and SDS-PAGE (217), 70 to 77 kDa by SDS-PAGE (131), ~45 kDa by SDS-PAGE (178), and 72 kDa by SDS-PAGE (76). Except for this difference in molecular mass, the thermostabilities of the glucoamylases were similar. For example, IISc 91 glucoamylase was stable for ~7 h at 60°C (177). Unlike -amylase of T. lanuginosus, glucoamylase of the same organism was less stable in the presence of added calcium.

Although T. lanuginosus produced both -amylase and glucoamylase activities simultaneously, the enzymes did not exhibit the synergism observed in the mesophilic fungus Aspergillus sp. (1). The thermal resistance of T. lanuginosus glucoamylase was increased severalfold by its entrapment in polyacrylamide gels (215). Glucoamylase effected up to 76% conversion of soluble or raw potato starch to glucose in 24 h, indicating that it was insensitive to end product inhibition (177). The properties of the enzyme suggested its usefulness in the commercial production of glucose syrups.

Another thermophilic fungus with a high potential in starch saccharification is Humicola grisea var. thermoidea. A strain of this fungus, isolated from soil from Brazil, produced 2.5- to 3.0-fold-higher glucoamylase activity when grown in a rich medium containing maltose as the principal carbon source than when grown on starch (250). The major starch-hydrolyzing enzyme had a molecular mass of 63 kDa, with pH and temperature optima of 5.0 and 55°C, respectively. The efficiency (V_max/K_m) of purified protein to hydrolyze starch was twice that of maltose. Kinetic experiments suggested that both starch and maltose were hydrolyzed at the same catalytic site. Another strain of H. grisea var. thermoidea produced a glucoamylase (74 kDa) that was remarkably insensitive to end product inhibition; it retained 65% activity in the presence of 950 mM glucose (47). Moreover, the enzyme was maximally active at pH 6.0 and 60°C. Increasing the copy number of the encoding gene through transformation (11) increased glucoamylase production by this fungus nearly threefold.

The cellulase system in fungi is considered to comprise three hydrolytic enzymes: (i) the endo-(1,4)--D-glucanase (synonyms: endoglucanase, endocellulase, carboxymethyl cellulase [EC 3.2.1.4]), which cleaves -linkages at random, commonly in the amorphous parts of cellulose; (ii) the exo-(1,4)--D-glucanase (synonyms: cellobiohydrolase, exocellulase, microcrystalline cellulase, Avicelase [EC 3.2.1.91]), which releases cellobiose from either the nonreducing or the reducing end, generally from the crystalline parts of cellulose; and (iii) the -glucosidase (synonym: cellobiase [EC 3.2.1.21]), which releases glucose from cellobiose and short-chain cellooligosaccharides (33). Although -glucosidase has no direct action on cellulose, it is regarded as a component of cellulase system because it stimulates cellulose hydrolysis (see below).

Based on the view prevalent in the 1970s that the levels of extracellularly produced cellulase enzymes determine the extent of solubilization of cellulose, several fungi were isolated and screened for high total cellulase activity in an attempt to develop a practical process for the enzymatic conversion of cellulose into glucose (159, 160). Mandels (159) observed that some species of thermophilic fungi degraded cellulose rapidly but that their culture filtrates had low cellulase activity. This was contradicted by reports that the thermophilic fungi Sporotrichum thermophile (67) and Talaromyces emersonii (88) produced cellulase activity nearly comparable to that of the mesophilic fungus Trichoderma reesei, regarded as the best source of fungal cellulase. Although cellulase productivity varies among strains (189), using uniform procedures for the measurement of cellulase activity, Bhat and Maheshwari (32) demonstrated that the endoglucanase and exoglucanase activities in the culture filtrate of their best strain of S. thermophile were about 10-fold lower and the -glucosidase activity was about 1.6-fold lower than in T. reesei. Despite these lower activities, S. thermophile degraded cellulose faster and grew at five times the rate of T. reesei. That S. thermophile has a powerful cellulolytic system was corroborated by the observation that its growth rates on insoluble cellulose and glucose were similar. Of greater significance, these observations raised strong doubts about the notion that secreted levels of cellulase determine the rate or extent of cellulolysis.

A question that had remained unresolved was whether cellulase formation in fungi is directly correlated with mycelial growth; i.e., were the enzymes produced during the trophophase or the idiophase? To determine this, the growth of S. thermophile on cellulose was arrested at different times by cycloheximide addition (32). When fungal growth was curtailed, some cellulose remained insolubilized in the medium, although cellulase that had already been secreted prior to growth arrest was present in the culture medium. It was inferred that degradation of cellulose is intimately associated with fungal growth (32).

In general, crystalline cellulose was found to be a superior carbon source for induction of cellulase enzymes in thermophilic fungi than were its amorphous or "impure" forms (90, 93, 166, 167, 211, 217). The exceptions are Thermoascus aurantiacus (137, 139, 140), Humicola insolens (113), and H. grisea var. thermoidea (270), which produced high cellulase and xylanase activities even on hemicellulosic substrates without cellulose. In some strains of S. thermophile, the disaccharides cellobiose and lactose also induced cellulase, although less efficiently (50, 189). The time course of the appearance of various cellulase components varied in different species: thus, in H. insolens (271) and T. aurantiacus (138), all three cellulase components appeared simultaneously, while in Chaetomium thermophile var. coprophile, -glucosidase activity preceded endo- and exoglucanase activities (92) and in S. thermophile it lagged behind endoglucanase and exoglucanase, which were typically formed during active growth. Rather, the appearance of -glucosidase in medium coincided with the time of extensive autolysis of mycelium (32, 91). For purification of cellulases, aged cultures have been used in which cellulose was nearly completely solubilized and the -glucosidase activity was maximal. The cellulase components have been separated using combinations of ion-exchange chromatography and gel filtration chromatography and/or preparative gel electrophoresis, but the protein yields have generally been low. Among the exceptions was a soil isolate of T. aurantiacus from India, which was grown in shake flasks containing shredded paper and peptone. From about 2 liters of culture filtrate, 1,622 mg of desalted crude protein was obtained, which, on further processing, yielded 30 mg of -glucosidase, 335 mg of exoglucanase, 161 mg of endoglucanase, and 258 mg of crystalline xylanase (139). Yoshioka and Hayashida (271) purified cellulases from a strain of H. insolens grown on wheat bran. From 2.5 g of culture extract protein, 9, 11, and 18 mg, respectively, of pure endoglucanase, exoglucanase, and -glucosidase were obtained (110, 113, 271).

Like the mesophilic fungi, the thermophilic fungi produce multiple forms of the cellulase components. However, two different strains of T. aurantiacus produced one form each of endoglucanase, exoglucanase, and -glucosidase, but the forms from the two strains had somewhat different properties (139, 248). The multiplicity of individual cellulases might be a result of posttranslational and/or postsecretion modifications of a gene product or might be due to multiple genes. For example, T. emersonii produced multiple endoglucanases, exoglucanases, and -glucosidases (65, 66, 165-168). Its culture filtrate protein was resolved by ion-exchange chromatography into four endoglucanases (Table 4) which, unlike their variable carbohydrate contents (28 to 51%), had similar molecular masses (68 kDa by gel filtration and 35 kDa by SDS-PAGE), isoelectric points, pH and temperature optima, thermal stabilities, and specific activities (178). The different endoglucanases were thought to reflect differential posttranslational or postsecretion modifications of a single gene product. However, cloning and expression of seven endoglucanase genes of H. insolens demonstrated a genetic basis of multiple forms. The seven cloned endoglucanases produced in Aspergillus oryzae were purified from culture medium by affinity chromatography on cellulose (70, 223, 224) and were found to differ in cleavage site specificity and kinetic constants (k_cat/K_m).

View this table: [in this window] [in a new window]   TABLE 4.   Salient properties of -1,4 endoglucanases of thermophilic fungi

The endoglucanases (30 to 100 kDa) of thermophilic fungi are thermostable, with optimal activity between 55 and 80°C at pH 5.0 to 5.5 and with carbohydrate contents from 2 to 50% (Table 4