Perfusion cultures and modelling of oxygen uptake with three-dimensional chondrocyte pellets

Chondrocyte pellets were cultivated in a perfused flow chamber and supplied with medium by a constant flow rate from a conditioning vessel. In this conditioning vessel the medium was aerated and used medium was exchanged semi-continuously. The higher amount of DNA and glycosaminoglycane (GAG) in these pellets compared to control cultures under stationary conditions showed a positive effect of the reactor system, compared to standard culture conditions. A diffusion reaction model was applied to calculate the oxygen uptake of the cell pellet and to describe the oxygen profile within the pellet. The model included diffusion within the cell pellet and oxygen uptake of the cells. Calculated data were compared to experimental data obtained by tissue engineered chondrocyte cell pellets. Model calculations agreed rather well with experimental data.     

Adding talc microparticles to Aspergillus terreus ATCC 20542 preculture decreases fungal pellet size and improves lovastatin production

Changing fungal morphology with the use of morphological engineering techniques leads to improving the production of metabolites by filamentous fungi in the submerged culture. Adding mineral microparticles is one such simple method to change fungal pellet size. Here, it was studied for a lovastatin producer, Aspergillus terreus ATCC 20542. The experiments were conducted in shake flasks and 10 μm talc microparticles were added to the preculture. Intrapellet oxygen concentration profiles were determined by an oxygen microprobe. Talc microparticles caused a decrease of A. terreus pellets diameter from about 2000 to 900 μm, dependent on their concentration in the preculture. Smaller pellets produced more lovastatin, whose titre exceeded then 120 mg L^?1, utilising more lactose. The decrease in pellet size resulted in changes of oxygen concentration profiles in the pellets. The estimated critical pellet diameter, at which the non‐oxygenated zone was observed in the centre of the pellets, was 1700 μm. Smaller pellets were fully penetrated by oxygen. To conclude, facilitated diffusion of oxygen into the pellets of smaller diameter and their less dense structure made lactose utilisation by A. terreus more efficient, which ultimately increased lovastatin production in the runs with talc microparticles added, compared to the control runs.     

Verbesserung alter biotechnologischer Verfahren

By the example of the Penicillin G production, it is shown, how the production process can be improved by using pellet suspension instead of highly viscous filamentous mycel. In order to optimize the pellet geometry, dissolved oxygen concentration profiles were measured, and the overall mass transfer resistance was calculated in the pellets; histological investigations were carried out with pellets also. All measurements indicated that the optimal pellet diameter is 400 μm. When using this pellet, the specific power input was reduced from 4–5 kW/m³ to 0.8 kW/m³. The losses during the recovery of Penicillin G can be reduced below 1% by using reactive extraction. The thermodynamics, kinetics and reaction technique of reactive extraction of Penicillin G are also taken into account.     

Diffusion and reaction within porous packing media: a phenomenological model

A phenomenological model has been developed to describe biomass distribution and substrate depletion in porous diatomaceous earth (DE) pellets colonized by Pseudomonas aeruginosa. The essential features of the model are diffusion, attachment and detachment to/from pore walls of the biomass, diffusion of substrate within the pellet, and external mass transfer of both substrate and biomass in the bulk fluid of a packed bed containing the pellets. A bench-scale reactor filled with DE pellets was inoculated with P. aeruginosa and operated in plug flow without recycle using a feed containing glucose as the limiting nutrient. Steady-state effluent glucose concentrations were measured at various residence times, and biomass distribution within the pellet was measured at the lowest residence time. In the model, microorganism/substrate kinetics and mass transfer characteristics were predicted from the literature. Only the attachment and detachment parameters were treated as unknowns, and were determined by fitting biomass distribution data within the pellets to the mathematical model. The rate-limiting step in substrate conversion was determined to be internal mass transfer resistance; external mass transfer resistance and microbial kinetic limitations were found to be nearly negligible. Only the outer 5% of the pellets contributed to substrate conversion. (c) 1993 Wiley & Sons, Inc.     

The effect of oxygen tension on human articular chondrocyte matrix synthesis: Integration of experimental and computational approaches

Significant oxygen gradients occur within tissue engineered cartilaginous constructs. Although oxygen tension is an important limiting parameter in the development of new cartilage matrix, its precise role in matrix formation by chondrocytes remains controversial, primarily due to discrepancies in the experimental setup applied in different studies. In this study, the specific effects of oxygen tension on the synthesis of cartilaginous matrix by human articular chondrocytes were studied using a combined experimental‐computational approach in a “scaffold‐free” 3D pellet culture model. Key parameters including cellular oxygen uptake rate were determined experimentally and used in conjunction with a mathematical model to estimate oxygen tension profiles in 21‐day cartilaginous pellets. A threshold oxygen tension (pO₂ ≈ 8% atmospheric pressure) for human articular chondrocytes was estimated from these inferred oxygen profiles and histological analysis of pellet sections. Human articular chondrocytes that experienced oxygen tension below this threshold demonstrated enhanced proteoglycan deposition. Conversely, oxygen tension higher than the threshold favored collagen synthesis. This study has demonstrated a close relationship between oxygen tension and matrix synthesis by human articular chondrocytes in a “scaffold‐free” 3D pellet culture model, providing valuable insight into the understanding and optimization of cartilage bioengineering approaches. Biotechnol. Bioeng. 2014;111: 1876–1885. © 2014 The Authors. Biotechnology and Bioengineering Published by Wiley Periodicals, Inc.     

Simulation of growth of a filamentous fungus in 3 dimensions

The tridimensional growth of a filamentous fungus was simulated, based on a model for the evolution of the microscopic morphology of Trichoderma reesei. When supplemented with a spatial representation of growth, the model correctly simulates the evolution from a single spore to a pellet. Diffusion of oxygen is included in the model. The simulated tridimensional structures have a fractal nature; and the fractal dimension, determined by a box-counting method, increases during growth. The fractal dimension only depends on the mass of the pellet and is not affected by model parameters such as tip extension rate and branching frequency. Realistic pictures are obtained and the radius of the pellet increases at a constant rate. The influence of model parameters (tip extension rate, branching frequency, minimum porosity) on dissolved oxygen concentration profiles, biomass concentration profiles, rate at which the pellet diameter increases, and the evolution of the fractal dimension was determined. The dissolved oxygen profiles were found to be very different from the profiles, obtained by assuming a homogenous biomass distribution within the pellet. Finally, the formation of pellets from spore aggregates is calculated and the size of the spore aggregate is found to only influence the time needed before the appearance of a pellet and not its morphology. (c) 1997 John Wiley & Sons, Inc.     

Modeling and measurements of fungal growth and morphology in submerged fermentations

Generalizing results from fungal fermentations is difficult due to their high sensitivity toward slight variation in starting conditions, poor reproducibility, and difference in strains. In this study a mathematical model is presented in which oxygen transfer, agitation intensity, dissolved oxygen tension, pellet size, formation of mycelia, the fraction of mycelia in the total biomass, carbohydrate source consumption, and biomass growth are taken into account. Two parameters were estimated from simulation, whereas all others are based on measurements or were taken from literature. Experimental data are obtained from the fermentations in both 2 L and 100 L fermentors at various conditions. Comparison of the simulation with experiments shows that the model can fairly well describe the time course of fungal growth (such as biomass and carbohydrate source concentrations) and fungal morphology (such as pellet size and the fraction of pellets in the total biomass). The model predicts that a stronger agitation intensity leads to a smaller pellet size and a lower fraction of pellets in the total biomass. At the same agitation intensity, pellet size is hardly affected by the dissolved oxygen tension, whereas the fraction of mycelia decreases slightly with an increase of the dissolved oxygen tension in the bulk. All of these are in line with observations at the corresponding conditions.     

Improved enzyme production by bio-pellets of Aspergillus niger: targeted morphology engineering using titanate microparticles

The present study describes the design of bio-pellet morphologies of the industrial working horse Aspergillus niger strains in submerged culture. The novel approach recruits the intended addition of titanate microparticles (TiSiO(4), 8 μm) to the growth medium. As tested for two recombinant strains producing fructofuranosidase and glucoamylase, the enzyme titer by the titanate-enhanced cultures in shake flasks was increased 3.7-fold to 150 U/mL (for fructofuranosidase) and 9.5-fold to 190 U/mL (for glucoamylase) as compared to the control. This could be successfully utilized for improved enzyme production in stirred tank reactors. Stimulated by the particles, the achieved final glucoamylase activity of 1,080 U/mL (fed-batch) and 320 U/mL (batch) was sevenfold higher as compared to the conventional processes. The major reason for the enhanced production was the close association between the titanate particles and the fungal cells. Already below 2.5 g/L the micromaterial was found inside the pellets, including single particles embedded as 50-150 μm particle aggregates in the center resulting in core shell pellets. With increasing titanate levels the pellet size decreased from 1,700 μm (control) to 300 μm. Fluorescence based resolution of GFP expression revealed that the large pellets of the control were only active in a 200 μm surface layer. This matches with the critical penetration depth for nutrients and oxygen typically observed for fungal pellets. The biomass within the titanate derived fungal pellets, however, was completely active. This was due a reduced thickness of the biomass layer via smaller pellets as well as the core shell structure. Moreover, also the created loose inner pellet structure enabled a higher mass transfer and penetration depths for up to 500 μm. The creation of core-shell pellets has not been achieved previously by the addition of microparticles, for example, made of talc or alumina. Due to this, the present work opens further possibilities to use microparticles for tailor-made morphology design of filamentous fungi, especially for pellet based processes which have a long and strong industrial relevance for industrial production.     

Determination of the respiration kinetics for mycelial pellets of Phanerochaete chrysosporium.

In mycelial pellet cultures of the white rot basidiomycete Phanerochaete chrysosporium, low oxygen concentration negatively affects the production of the extracellular lignin peroxidases and manganese peroxidases which are key components of the lignin-degrading system of this organism. To test the hypothesis that oxygen limitation in the pellets is responsible for this effect, oxygen microelectrodes were used to determine oxygen concentration gradients within the mycelial pellets of P. chrysosporium. Pellets were removed from oxygenated cultures, allowed to equilibrate with air, and probed with oxygen microelectrodes. The oxygen profiles were modelled assuming that O2 uptake follows a Michaelis-Menten relationship. The Vmax and Km values for oxygen uptake were 0.76 +/- 0.10 g/m3 of pellet per s and 0.5 +/- 0.3 g/m3, respectively. These kinetic values were used to predict respiration rates in air-flushed cultures, oxygen-flushed cultures, and cultures with large pellets (diameter greater than 6 mm). The predicted respiration rates were independently validated by experimentally measuring the evolution of carbon dioxide from whole cultures.     

Determination of the respiration kinetics for mycelial pellets of Phanerochaete chrysosporium.

In mycelial pellet cultures of the white rot basidiomycete Phanerochaete chrysosporium, low oxygen concentration negatively affects the production of the extracellular lignin peroxidases and manganese peroxidases which are key components of the lignin-degrading system of this organism. To test the hypothesis that oxygen limitation in the pellets is responsible for this effect, oxygen microelectrodes were used to determine oxygen concentration gradients within the mycelial pellets of P. chrysosporium. Pellets were removed from oxygenated cultures, allowed to equilibrate with air, and probed with oxygen microelectrodes. The oxygen profiles were modelled assuming that O2 uptake follows a Michaelis-Menten relationship. The Vmax and Km values for oxygen uptake were 0.76 +/- 0.10 g/m3 of pellet per s and 0.5 +/- 0.3 g/m3, respectively. These kinetic values were used to predict respiration rates in air-flushed cultures, oxygen-flushed cultures, and cultures with large pellets (diameter greater than 6 mm). The predicted respiration rates were independently validated by experimentally measuring the evolution of carbon dioxide from whole cultures.     

Effects of dissolved oxygen tension and mechanical forces on fungal morphology in submerged fermentation

The effects of dissolved oxygen tension and mechanical forces on fungal morphology were both studied in the submerged fermentation of Aspergillus awamori. Pellet size, the hairy length of pellets, and the free filamentous mycelial fraction in the total biomass were found to be a function of the mechanical force intensity and to be independent of the dissolved oxygen tension provided that the dissolved oxygen tension was neither too low (5%) nor too high (330%). When the dissolved oxygen concentration was close to the saturation concentration corresponding to pure oxygen gas, A. awamori formed denser pellets and the free filamentous mycelial fraction was almost zero for a power input of about 1 W/kg. In the case of very low dissolved oxygen tension, the pellets were rather weak and fluffy so that they showed a very different appearance. The amount of biomass per pellet surface area appeared to be affected only by the dissolved oxygen tension and was proportional to the average dissolved oxygen tension to the power of 0.33. From this it was concluded that molecular diffusion was the dominant mechanism for oxygen transfer in the pellets and that convection and turbulent flow in the pellets were negligible in submerged fermentations. The biomass per wet pellet volume increased with the dissolved oxygen tension and decreased with the size of the pellets. This means that the smaller pellets formed under a higher dissolved oxygen tension had a higher intrinsic strength. Correspondingly, the porosity of the pellets was a function of the dissolved oxygen tension and the size of pellets. Within the studied range, the void fraction in the pellets was high and always much more than 50%.     

Comparative chondrogenesis of human cells in a 3D integrated experimental–computational mechanobiology model

We present an integrated experimental–computational mechanobiology model of chondrogenesis. The response of human articular chondrocytes to culture medium perfusion, versus perfusion associated with cyclic pressurisation, versus non-perfused culture, was compared in a pellet culture model, and multiphysic computation was used to quantify oxygen transport and flow dynamics in the various culture conditions. At 2 weeks of culture, the measured cell metabolic activity and the matrix content in collagen type II and aggrecan were greatest in the perfused+pressurised pellets. The main effects of perfusion alone, relative to static controls, were to suppress collagen type I and GAG contents, which were greatest in the non-perfused pellets. All pellets showed a peripheral layer of proliferating cells, which was thickest in the perfused pellets, and most pellets showed internal gradients in cell density and matrix composition. In perfused pellets, the computed lowest oxygen concentration was 0.075 mM (7.5% tension), the maximal oxygen flux was 477.5 nmol/m²/s and the maximal fluid shear stress, acting on the pellet surface, was 1.8 mPa (0.018 dyn/cm²). In the non-perfused pellets, the lowest oxygen concentration was 0.003 mM (0.3% tension) and the maximal oxygen flux was 102.4 nmol/m²/s. A local correlation was observed, between the gradients in pellet properties obtained from histology, and the oxygen fields calculated with multiphysic simulation. Our results show up-regulation of hyaline matrix protein production by human chondrocytes in response to perfusion associated with cyclic pressurisation. These results could be favourably exploited in tissue engineering applications.     

Pellet morphology, culture rheology and lovastatin production in cultures of Aspergillus terreus

Pellet growth of Aspergillus terreus ATCC 20542 in submerged batch fermentations in stirred bioreactors was used to examine the effects of agitation (impeller tip speed u(t) of 1.01-2.71 ms(-1)) and aeration regimens (air or an oxygen-enriched mixture containing 80% oxygen and 20% nitrogen by volume) on the fungal pellet morphology, broth rheology and lovastatin production. The agitation speed and aeration methods used did not affect the biomass production profiles, but significantly influenced pellet morphology, broth rheology and the lovastatin titers. Pellets of approximately 1200 microm initial diameter were reduced to a final stable size of approximately 900 microm when the agitation intensity was >/=600 rpm (u(t)>/=2.03 ms(-1)). A stable pellet diameter of approximately 2500 microm could be attained in less intensely agitated cultures. These large fluffy pellets produced high lovastatin titers when aerated with oxygen-enriched gas but not with air. Much smaller pellets obtained under highly agitated conditions did not attain high lovastatin productivity even in an oxygen-enriched atmosphere. This suggests that both an upper limit on agitation intensity and a high level of dissolved oxygen are essential for attaining high titers of lovastatin. Pellet size in the bioreactor correlated equally well with the specific energy dissipation rate and the energy dissipation circulation function. The latter took into account the frequency of passage of the pellets through the high shear regions of the impellers. Pellets that gave high lovastatin titers produced highly shear thinning cultivation broths.     

Impacts of pellets injected from the low-field side on plasma in ITER

Impacts of pellets injected from the low-field side (LFS) on plasma in ITER are investigated using the 1.5D BALDUR integrated predictive modeling code. In these simulations, the pellet ablation is described using the neutral gas shielding (NGS) model. The pellet ablation model is coupled with the plasma core transport model, which is a combination of the MMM95 anomalous transport model and NCLASS neoclassical transport model. The boundary conditions are assumed to be at the top of the pedestal, in which the pedestal parameters are predicted using a pedestal model based on the theoretical-based pedestal width scaling (either magnetic and flow shear stabilization width scaling, or flow shear stabilization width scaling, or normalized poloidal pressure width scaling) and the infinite-n ballooning mode pressure gradient limit. These pedestal models depend sensitively on the density at the top of the pedestal, which can be strongly influenced by the injection of pellets. The combination of the MMM95 and NCLASS models, together with the pedestal and NGS models, is used to simulate the time evolution of the plasma current, ion and electron temperatures, and density profiles for ITER standard type-I ELMy H-mode discharges during the injection of LFS pellets. It is found that the injection of pellets results in a complicated plasma scenario, especially in the outer region of the plasma and the plasma conditions at the boundary in which the pellet has an impact on increasing the plasma edge density, but reducing the plasma edge temperature. The LFS pellet has a stronger impact on the edge as compared to the center. For fusion performance, the pellet can result in either enhancement or degradation, depending sensitively on the pellet parameters; such as the pellet size, pellet velocity, and pellet frequency. For example, when a series of deuterium pellets with a size of 0.5 cm, velocity of 1 km/s, and frequency of 2 Hz are injected into the ITER plasma from the LFS, the plasma performance, evaluated in terms of Q _fusion, can increase to 72% of that before the use of pellets. It is also found that the injection of pellets results in an increase in the ion and electron densities, but does not enhance the central plasma density. On the other hand, it results in the formation of another peak of the plasma density in the outer region near the plasma edge. The formation of the density peak results in the reduction of plasma transports near the edge by decreasing the contributions of ion-temperature-gradient and trapped electron modes, as well as kinetic ballooning modes.     

A method to evaluate the isothermal effectiveness factor for dynamic oxygen into mycelial pellets in submerged cultures

Several models have been developed simulating O2 transfer in bioreactors, but three limitations are often found: (i) an inadequate kinetic representation of O2 consumption or wrong boundary conditions, (ii) unrealistic parameter values, and (iii) inadequate experimental systems. In our study we minimized those possible sources of error. Oxygen uptake rate, void fraction of the pellet, and external O2 mass transfer coefficient were experimentally obtained from bioreactor studies in which pellets of Gibberella fujikuroi were naturally formed. Michaelis-Menten kinetics and diffusion equations were used to describe the O2 consumption rate and to evaluate the effectiveness factor in dynamic mode. The nonlinear mathematical model proposed was solved by the orthogonal collocation technique. The O2 consumption rate in pellets of G. fujikuroi of 1.7-2.0 mm is only marginally inhibited by diffusion constraints under conditions tested. Simulation analysis showed that the effectiveness factor decreased as the Thiele modulus and pellet diameter increased. The proposed model was applied to experimental data reported for other fungal pellets and allowed to predict optimal conditions for O2 transfer into mycelial pellets.     

In situ analysis of oxygen consumption and diffusive transport in high‐temperature acidic iron‐oxide microbial mats

The role of dissolved oxygen as a principal electron acceptor for microbial metabolism was investigated within Fe(III)‐oxide microbial mats that form in acidic geothermal springs of Yellowstone National Park (USA). Specific goals of the study were to measure and model dissolved oxygen profiles within high‐temperature (65–75°C) acidic (pH = 2.7–3.8) Fe(III)‐oxide microbial mats, and correlate the abundance of aerobic, iron‐oxidizing Metallosphaera yellowstonensis organisms and mRNA gene expression levels to Fe(II)‐oxidizing habitats shown to consume oxygen. In situ oxygen microprofiles were obtained perpendicular to the direction of convective flow across the aqueous phase/Fe(III)‐oxide microbial mat interface using oxygen microsensors. Dissolved oxygen concentrations dropped from ～ 50–60 μM in the bulk‐fluid/mat surface to below detection (< 0.3 μM) at a depth of ～ 700 μm (～ 10% of the total mat depth). Net areal oxygen fluxes into the microbial mats were estimated to range from 1.4–1.6 × 10^?4 μmol cm^?2 s^?1. Dimensionless parameters were used to model dissolved oxygen profiles and establish that mass transfer rates limit the oxygen consumption. A zone of higher dissolved oxygen at the mat surface promotes Fe(III)‐oxide biomineralization, which was supported using molecular analysis of Metallosphaera yellowstonensis 16S rRNA gene copy numbers and mRNA expression of haem Cu oxidases (FoxA) associated with Fe(II)‐oxidation.     

Strategies for penicillin fermentation in tower-loop reactors

Since it has not been possible to produce penicillin in tower-loop reactors with highly viscous filamentous molds of Penicillium chrysogenum which are employed in stirred-tank reactors, a new strategy has been developed to avoid the formation of this morphology and to use the pellet form of the fungi. When employing definite impeller speeds in the subculture in connection with definite inoculum amounts and substrate concentrations in the main culture (bubble column), it is possible to generate a suspension of isolated small pellets, which shows a low broth viscosity up to a sediment content of 45% over the entire fermentation time. Volumetric mass-transfer coefficients k(L)as are by a factor of 4 to 5 higher in these pellet suspensions than in filamentous broths. It was easy to maintain the necessary oxygen supply for penicillin production in these pellet suspensions. Under these conditions the specific penicillin productivities were higher with regard to power input (up to 90%), biomass, and consumed substrate than in the stirred-tank reactors with highly viscous filamentous morphology of the fungi. Under nonoptimized operating conditions the absolute penicillin production in the tower loop was 35% lower than in the stirred-tank reactor due to lower possible biomass concentrations. The separation of the biomass, and therefore the penicillin recovery, is much simpler when employing pellets. It is shown how the particular mass transfer resistances at the gas/liquid and liquid/pellet interfaces and within the pellets change with the pellet diameter. There should be a particular pellet diameter at which penicillin productivity has its maximum. These investigations indicate that the use of tower-loop reactors can, in the future, be an alternative for more economical penicillin production methods.     

Hydrodynamic and kinetic study of cellulase production by Trichoderma reesei with pellet morphology

Numerical simulations and experimental validation were performed to understand the effects of hydrodynamics on pellet formation and cellulase production by filamentous T. reesei. The constructed model combined a steady-state multiple reference frame (MRF) approach describing mechanical mixing, oxygen mass transfer, and non-Newtonian flow field with a transient sliding mesh approach and kinetics of oxygen consumption, pellet formation, and enzyme production. The model was experimentally validated at various agitation speeds in a two-impeller Rushton turbine fermentor. Results from simulation and experimentation showed that higher agitation speeds led to increases in the pellet diameter and the proportion of pelletized (vs. filamentous) forms of the biomass. It also led to increase in dissolved oxygen mass transfer rate in shear-thinning fluid and cellulase productivity. The extent of these increases varied considerably among agitation speeds. Pellet formation and morphology were presumably affected within a viscosity-dependent shear-rate range. Cellulase activity and cell viability were shown to be sensitive to impeller shear. A maximum cellulase activity of 3.5 IU/mL was obtained at 400 rpm, representing a twofold increase over that at 100 rpm.     

Morphology and physiology of an alpha-amylase producing strain of Aspergillus oryzae during batch cultivations

The microscopic morphology, that is, total hyphal length and total number of tips, has been characterized during batch cultivations of Aspergillus oryzae. The specific growth rate estimated by measuring the total hyphal length (mu(h)) corresponds well with the specific growth rate estimated from dry weight measurements during cultures grown as free hyphal elements. The average tip extension rate can be described with a saturation type kinetics with respect to the average total hyphal length, and the branching frequency is closely related to the total hyphal length. For the applied strain of A. oryzae, pellet formation occurs by coagulation of spores. The agglomeration process is pH dependent and pellets are formed at pH values higher than 5, whereas low pH (<3.5) results in growth as freely dispersed hyphal elements. The maximum specific growth rate has a broad pH optimum between 3 and 7, whereas the alpha-amylase production has a sharper maximum at about pH 6. During batch cultivation with pellets the growth is described well by the cube-root law when pellet fragmentation can be neglected. The kinetic parameter k in the cube-root law is derived from the growth kinetics with no mass transfer limitation, k = mu(h)/3. Based on an oxygen balance, the active growth layer in the pellet is estimated to be 200 to 325 mum and, consequently, up to 50% of the biomass is limited by oxygen for large pellets. Ethanol production (up to 1 g L(-1)) was observed during batch cultivations with pellets, suggesting that ethanol is produced in the oxygen limited part of the biomass. A constitutive, low alpha-amylase production was observed at high glucose concentration. The specific alpha-amylase production was significantly higher for filamentous growth than for pellets and oxygen appears to be necessary for production of alpha-amylase. (c) 1996 John Wiley & Sons, Inc.