Chapter-2 Enzyme Preparation and Use
Chapter-2 Enzyme Preparation and Use
Chapter-2 Enzyme Preparation and Use
Sources of enzymes
Biologically active enzymes may be extracted from any living organism. A very wide range of
sources are used for commercial enzyme production from Actinoplanes to Zymomonas, from
spinach to snake venom. Of the hundred or so enzymes being used industrially, over a half are from
fungi and yeast and over a third are from bacteria with the remainder divided between animal (8%)
and plant (4%) sources (Table 2.1). A very much larger number of enzymes find use in chemical
analysis and clinical diagnosis. Non-microbial sources provide a larger proportion of these, at the
present time. Microbes are preferred to plants and animals as sources of enzymes because:
Attempts are being made to overcome some of these difficulties by the use of animal and plant cell
culture.
Intra/extra Scale of
Enzymea EC numberb Source Industrial use
-cellularc productiond
Animal enzymes
Catalase 1.11.1.6 Liver I - Food
Chymotrypsin 3.4.21.1 Pancreas E - Leather
Lipasee 3.1.1.3 Pancreas E - Food
Rennetf 3.4.23.4 Abomasum E + Cheese
Trypsin 3.4.21.4 Pancreas E - Leather
Plant enzymes
Actinidin 3.4.22.14 Kiwi fruit E - Food
a-Amylase 3.2.1.1 Malted barley E +++ Brewing
b-Amylase 3.2.1.2 Malted barley E +++ Brewing
1
Bromelain 3.4.22.4 Pineapple latex E - Brewing
b-Glucanaseg 3.2.1.6 Malted barley E ++ Brewing
Ficin 3.4.22.3 Fig latex E - Food
Lipoxygenase 1.13.11.12 Soybeans I - Food
Papain 3.4.22.2 Pawpaw latex E ++ Meat
Bacterial enzymes
a-Amylase 3.2.1.1 Bacillus E +++ Starch
b-Amylase 3.2.1.2 Bacillus E + Starch
Asparaginase 3.5.1.1 Escherichia coli I - Health
Glucose isomeraseh 5.3.1.5 Bacillus I ++ Fructose syrup
Penicillin amidase 3.5.1.11 Bacillus I - Pharmaceutical
i
Protease 3.4.21.14 Bacillus E +++ Detergent
Pullulanasej 3.2.1.41 Klebsiella E - Starch
Fungal enzymes
a-Amylase 3.2.1.1 Aspergillus E ++ Baking
Aminoacylase 3.5.1.14 Aspergillus I - Pharmaceutical
k
Glucoamylase 3.2.1.3 Aspergillus E +++ Starch
Catalase 1.11.1.6 Aspergillus I - Food
Cellulase 3.2.1.4 Trichoderma E - Waste
Dextranase 3.2.1.11 Penicillium E - Food
Glucose oxidase 1.1.3.4 Aspergillus I - Food
Lactasel 3.2.1.23 Aspergillus E - Dairy
e
Lipase 3.1.1.3 Rhizopus E - Food
m
Rennet 3.4.23.6 Mucor miehei E ++ Cheese
Pectinasen 3.2.1.15 Aspergillus E ++ Drinks
Pectin lyase 4.2.2.10 Aspergillus E - Drinks
m
Protease 3.4.23.6 Aspergillus E + Baking
o
Raffinase 3.2.1.22 Mortierella I - Food
Yeast enzymes
Invertasep 3.2.1.26 Saccharomyces I/E - Confectionery
l
Lactase 3.2.1.23 Kluyveromyces I/E - Dairy
e
Lipase 3.1.1.3 Candida E - Food
Raffinaseo 3.2.1.22 Saccharomyces I - Food
a
The names in common usage are given. As most industrial enzymes consist of mixtures
of enzymes, these names may vary from the recommended names of their principal
component. Where appropriate, the recommended names of this principal component is
given below.
b
The EC number of the principal component.
2
c
I - intracellular enzyme; E - extracellular enzyme.
d
+++ > 100 ton year-1; ++ > 10 ton year-1; + > 1 ton year-1; - < 1 ton year-1.
e
triacylglycerol lipase;
f
chymosin;
g
Endo-1,3(4)-b-glucanase;
h
xylose isomerase;
i
subtilisin;
j
a-dextrin endo-1,6-a-glucosidase;
k
glucan 1,4-a-glucosidase;
l
b-galactosidase;
m
microbial aspartic proteinase;
n
polygalacturonase;
o
a-galactosidase;
p
b-fructofuranosidase.
In practice, the great majority of microbial enzymes come from a very limited number of genera, of
whichAspergillus species, Bacillus species and Kluyveromyces (also called Saccharomyces) species
predominate. Most of the strains used have either been employed by the food industry for many
years or have been derived from such strains by mutation and selection. There are very few
examples of the industrial use of enzymes having been developed for one task. Shining examples of
such developments are the production of high fructose syrup using glucose isomerase and the use
of pullulanase in starch hydrolysis.
Producers of industrial enzymes and their customers will share the common aims of economy,
effectiveness and safety. They will wish to have high-yielding strains of microbes which make the
enzyme constitutively and secrete it into their growth medium (extracellular enzymes). If the
enzyme is not produced constitutively, induction must be rapid and inexpensive. Producers will aim
to use strains of microbe that are known to be generally safe. Users will pay little regard to the way
in which the enzyme is produced but will insist on having preparations that have a known activity
and keep that activity for extended periods, stored at room temperature or with routine
refrigeration. They will pay little attention to the purity of the enzyme preparation provided that it
does not contain materials (enzymes or not) that interfere with their process. Both producers and
users will wish to have the enzymes in forms that present minimal hazard to those handling them or
consuming their product.
The first stage of the screening procedure for commercial enzymes is to screen ideas, i.e. to
determine the potential commercial need for a new enzyme, to estimate the size of the market and
to decide, approximately, how much potential users of the enzyme will be able to afford to pay for
it. In some cases, the determination of the potential value of an enzyme is not easy, for instance
when it might be used to produce an entirely novel substance. In others, for instance when the
novel enzyme would be used to improve an existing process, its potential value can be costed very
accurately. In either case, a cumulative cash flow must be estimated, balancing the initial screening
and investment capital costs including interest, tax liability and depreciation, against the expected
long term profits. Full account must be taken of inflation, projected variation in feedstock price and
source, publicity and other costs. In addition, the probability of potential market competition and
changes in political or legal factors must be considered. Usually the sensitivity of the project to
changes in all of these factors must be estimated, by informed guesswork, in order to assess the risk
factor involved. Financial re-appraisal must be frequently carried out during the development
process to check that it still constitutes an efficient use of resources.
If agreement is reached, probably after discussions with potential users, that experimental work
would be commercially justifiable, the next stage involves the location of a source of the required
enzyme. Laboratory work is expensive in manpower so clearly it is worthwhile using all available
databases to search for mention of the enzyme in the academic and patents literature. Cultures may
then be sought from any sources so revealed. Some preparations of commercial enzymes are quite
rich sources of enzymes other than the enzyme which is being offered for sale, revealing such
preparations as potential inexpensive sources which are worth investigating.
If these first searches are unsuccessful, it is probably necessary to screen for new microbial strains
capable of performing the transformation required. This should not be a 'blind' screen: there will
usually be some source of microbes that could have been exposed for countless generations to the
conditions that the new enzyme should withstand or to chemicals which it is required to modify.
Hence, thermophiles are sought in hot springs, osmophiles in sugar factories, organisms capable of
metabolising wood preservatives in timber yards and so on. A classic example of the detection of
an enzyme by intelligent screening was the discovery of a commercially useful cyanide-degrading
enzyme in the microbial pathogens of plants that contain cyanogenic glycosides.
The identification of a microbial source of an enzyme is by no means the end of the story. The
properties of the enzyme must be determined; i.e. temperature for optimum productivity,
4
temperature stability profile, pH optimum and stability, kinetic constants (K m, Vmax), whether there
is substrate or product inhibition, and the ability to withstand components of the expected feedstock
other than substrate. A team of scientists, engineers and accountants must then consider the next
steps. If any of these parameters is unsatisfactory, the screen must continue until improved enzymes
are located. Now that protein engineering (see Chapter 8) can be seriously contemplated, an
enzyme with sufficient potential value could be improved 'by design' to overcome one or two
shortcomings. However, this would take a long time, at the present level of knowledge and skill, so
further screening of microbes from selected sources would probably be considered more
worthwhile.
Once an enzyme with suitable properties has been located, various decisions must be made
concerning the acceptability of the organism to the regulatory authorities, the productivity of the
organism, and the way in which the enzyme is to be isolated, utilised (free or immobilised) and, if
necessary, purified. If the organism is unacceptable from a regulatory viewpoint two options exist;
to eliminate that organism altogether and continue the screening operation, or to clone the enzyme
into an acceptable organism. The latter approach is becoming increasingly attractive especially as
cloning could also be used to increase the productivity of the fermentation process. Cloning may
also be attractive when the organism originally producing the enzyme is acceptable from the health
and safety point of view but whose productivity is unacceptable (see Chapter 8). However, cloning
is not yet routine and invariably successful so there is still an excellent case to be made for applying
conventional mutation and isolation techniques for the selection of improved strains. It should be
noted that although the technology for cloning glucose isomerase into 'routine' organisms is known,
it has not yet been applied. Several of the glucose isomerase preparations used commercially
consist of whole cells, or cell fragments, of the selected strains of species originally detected by
screening.
The use of immobilised enzymes (see Chapter 3) is now familiar to industry and their advantages
are well recognised so the practicality of using the new enzymes in an immobilised form will be
determined early in the screening procedure. If the enzyme is produced intracellularly, the
feasibility of using it without isolation and purification will be considered very seriously and strains
selected for their amenability to use in this way.
It should be emphasised that there will be a constant dialogue between laboratory scientists and
biochemical process engineers from the earliest stages of the screening process. Once the
biochemical engineers are satisfied that their initial criteria of productivity, activity and stability
can be met, the selected strain(s) of microbe will be grown in pilot plant conditions. It is only by
applying the type of equipment used in full scale plants that accurate costing of processes can be
achieved. Pilot studies will probably reveal imperfections, or at least areas of ignorance, that must
be corrected at the laboratory scale. If this proves possible, the pilot plant will produce samples of
the enzyme preparation to be used by customers who may well also be at the pilot plant stage in the
development of the enzyme-utilizing process. The enzyme pilot plant also produces samples for
safety and toxicological studies provided that the pilot process is exactly similar to the full scale
operation.
5
Screening for new enzymes is expensive so that the intellectual property generated must be
protected against copying by competitors. This is usually done by patenting the enzyme or its
production method or, most usefully, the process in which it is to be used. Patenting will be
initiated as soon as there is evidence that an innovative discovery has been made.
Clearly defined media are usually out of the question for large scale use on cost grounds but may be
perfectly acceptable when enzymes are to be produced for high value uses, such as analysis or
medical therapy where very pure preparations are essential. Less-defined complex media are
composed of ingredients selected on the basis of cost and availability as well as composition. Waste
materials and by-products from the food and agricultural industries are often major ingredients.
Thus molasses, corn steep liquor, distillers solubles and wheat bran are important components of
fermentation media providing carbohydrate, minerals, nitrogen and some vitamins. Extra
carbohydrate is usually supplied as starch, sometimes refined but often simply as ground cereal
grains. Soybean meal and ammonium salts are frequently used sources of additional nitrogen. Most
of these materials will vary in quality and composition from batch to batch causing changes in
enzyme productivity.
Preparation of enzymes
Readers of papers dealing with the preparation of enzymes for research purposes will be familiar
with tables detailing the stages of purification. Often the enzyme may be purified several hundred-
fold but the yield of the enzyme may be very poor, frequently below 10% of the activity of the
original material (Table 2.2). In contrast, industrial enzymes will be purified as little as possible,
only other enzymes and material likely to interfere with the process which the enzyme is to
catalyse, will be removed. Unnecessary purification will be avoided as each additional stage is
costly in terms of equipment, manpower and loss of enzyme activity. As a result, some commercial
6
enzyme preparations consist essentially of concentrated fermentation broth, plus additives to
stabilise the enzyme's activity.
Table 2.2. The effect of number of steps on the yield and costs in a typical enzyme purification
process. The realistic assumptions are made that step yields are 75%, step purifications are three-
fold and step costs are 10% of the initial costs (later purification steps are usually intrinsically more
expensive but are necessarily of smaller scale).
Step Relative weight Yield (%) Specific activity Total cost Cost per weight Cost per activity
1.000 100 1 1.00 1 1.00
1 0.250 75 3 1.10 4 1.47
2 0.063 56 9 1.20 19 2.13
3 0.016 42 27 1.30 83 3.08
4 0.004 32 81 1.40 358 4.92
5 0.001 24 243 1.50 1536 6.32
The content of the required enzyme should be as high as possible (e.g. 10% w/w of the protein) in
order to ease the downstream processing task. This may be achieved by developing the
fermentation conditions or, often more dramatically, by genetic engineering. It may well be
economically viable to spend some time cloning extra copies of the required gene together with a
powerful promoter back into the producing organism in order to get 'over-producers' (see Chapter
8).
It is important that the maximum activity is retained during the preparation of enzymes. Enzyme
inactivation can be caused by heat, proteolysis, sub-optimal pH, oxidation, denaturants, irreversible
inhibitors and loss of cofactors or coenzymes. Of these heat inactivation, which together with
associated pH effects, is probably the most significant. It is likely to occur during enzyme
extraction and purification if insufficient cooling is available (see Chapter 1), but the problem is
less when preparing thermophilic enzymes. Proteolysis is most likely to occur in the early stages of
extraction and purification when the proteases responsible for protein turnover in living cells are
still present. It is also the major reason for enzyme inactivation by microbial contamination. In their
native conformations, enzymes have highly structured domains which are resistant to attack by
proteases because many of the peptide bonds are mechanically inaccessible and because many
proteases are highly specific. The chances of a susceptible peptide bond in a structured domain
being available for protease attack are low. Single 'nicks' by proteases in these circumstances may
have little immediate effect on protein conformation and, therefore, activity. The effect, however,
may severely reduce the conformational stability of the enzyme to heat or pH variation so greatly
reducing its operational stability. If the domain is unfolded under these changed conditions, the
whole polypeptide chain may be available for proteolysis and the same, specific, protease may
destroy it. Clearly the best way of preventing proteolysis is to rapidly remove, or inhibit, protease
7
activity. Before this can be achieved it is important to keep enzyme preparations cold to maintain
their native conformation and slow any protease action that may occur.
Some intracellular enzymes are used commercially without isolation and purification but the
majority of commercial enzymes are either produced extracellularly by the microbe or plant or
must be released from the cells into solution and further processed (Figure 2.1). Solid/liquid
separation is generally required for the initial separation of cell mass, the removal of cell debris
after cell breakage and the collection of precipitates. This can be achieved by filtration,
centrifugation or aqueous biphasic partition. In general, filtration or aqueous biphasic systems are
used to remove unwanted cells or cell debris whereas centrifugation is the preferred method for the
collection of required solid material.
Centrifugation
8
Centrifugation separates on the basis of the particle size and density difference between the liquid
and solid phases. Sedimentation of material in a centrifugal field may be described by
(2.1)
where v is the rate of sedimentation, d is the particle diameter, ρ s is the particle density, ρl is the
solution density, w is the angular velocity in radians s-1, r is the radius of rotation, h is the dynamic
viscosity, Fs is a correction factor for particle interaction during hindered settling and q is a shape
factor (=1 for spherical particles). Fs depends on the volume fraction of the solids present;
approximately equalling 1, 0.5, 0.1 and 0.05 for 1%, 3%, 12% and 20% solids volume fraction
respectively. Only material which reaches a surface during the flow through continuous centrifuges
will be removed from the centrifuge feedstock, the efficiency depending on the residence time
within the centrifuge and the distance necessary for sedimentation (D). This residence time will
equal the volumetric throughput (f) divided by the volume of the centrifuge (V). The maximum
throughput of a centrifuge for efficient use is given by
(2.2)
The efficiency of the process is seen to depend on the solids volume fraction, the effective
clarifying surface (V/D) and the acceleration factor (w 2r/g, where g is the gravitational constant,
981 cm s-2; a rotor of radius 25 cm spinning at 1 rev s-1 has an acceleration factor of approximately
1 G). Low acceleration factors of about 1 500 g may be used for harvesting cells whereas much
higher acceleration factors are needed to collect enzyme efficiently. The product of these factors
(w2rV/gD) is called the sigma factor (S) and is used to compare centrifuges and to assist scale-up.
Laboratory centrifuges using tubes in swing-out or angle head rotors have high angular velocity (w)
and radius of rotation (r) but small capacity (V) and substantial sedimentation distance (D). This
type of design cannot be scaled-up safely, primarily because the mechanical stress on the centrifuge
head increases with the square of the radius, which must increase with increasing capacity.
For large-scale use, continuous centrifuges of various types are employed (Figure 2.2). These allow
the continuous addition of feedstock, the continuous removal of supernatant and the discontinuous,
semicontinuous or continuous removal of solids. Where discontinuous or semicontinuous removal
of precipitate occurs, the precipitate is flushed out by automatic discharge systems which cause its
dilution with water or medium and may be a problem if the precipitate is required for further
treatment. Centrifugation is the generally preferred method for the collection of enzyme-containing
solids as it does not present a great hazard to most enzymes so long as foam production, with
consequent enzymic inactivation, is minimised.
9
Figure 2.2. Basic designs of industrial centrifuges, showing the flow of material within the bowls.
Motor drives, cooling jackets and sludge collection vessels are not shown. (a) Tubular bowl
centrifuge. This is generally operated vertically, the tubular rotor providing a long flow path
enabling clarification. The sludge collects and must be removed. (b) Continuous scroll centrifuge.
This is operated horizontally. The helical screw scrolls the solids along the bowl surface and out of
the liquid; the sludge being dewatered before discharge. The clarified liquor overflows over an
adjustable weir at the other end of the bowl. The screw conveyer rotates at a slightly different speed
to the bowl. (c) Continuous multichamber disc-stack centrifuge. The bowl contains a number of
parallel discs providing a large clarifying surface with a small sedimentation distance. The sludge is
removed through a valve.
Small particles of cell debris and precipitated protein may be sedimented using tubular bowl
centrifuges, of which Sharples centrifuges (produced by Pennwalt Ltd.) are the best known. These
semi-continuous centrifuges are long and thin enabling rapid acceleration and deceleration,
minimising the down-time required for the removal of the sedimented solids. Here the radius and
effective liquid thickness are both small allowing a high angular velocity and hence high
centrifugal force; small models can be used at acceleration factors up to 50,000 g, accumulating 0.1
Kg of wet deposit whereas large models, designed to accumulate up to 5 Kg of deposit, are
restricted to 16,000 g. The capacities of these centrifuges are only moderate.Multichamber disc-
10
stack centrifuges, originally designed (by Westfalia and Alpha-Laval) for cream separation, contain
multiple coned discs in a stack which are spun and on which the precipitate collects. They may be
operated either semi-continuously or, by using a centripetal pressurising pump within the centrifuge
bowl which forces the sludge out through a valve, continuously. The capacity and radius of such
devices are large and the thickness of liquid is very small, due to the large effective surface area.
The angular velocity, however, is restricted giving a maximum acceleration factor of about 8,000 g.
A different design which is rather similar in principle is the solid bowl scroll centrifuge in which an
Archimedes' screw collects the precipitate so that fluid and solids leave at opposite ends of the
apparatus. These can only be used at low acceleration (about 3,000 g) so they are suitable only for
the collection of comparatively large particles.
Although many types of centrifuge are available, the efficient precipitation of small particles of cell
debris can be difficult, sometimes near-impossible. Clearly from Equation 2.2, the efficiency of
centrifugation can be improved if the particle diameter (d) is increased. This can be done either by
coagulating or flocculating particles. Coagulation is caused by the removal of electrostatic charges
(e.g. by pH change) and allowing particles to adhere to each other. Flocculation is achieved by
adding small amounts of high-molecular-weight charged materials which bridge oppositely-
charged particles to produce a loose aggregate which may be readily removed by centrifugation or
filtration. Flocculation and coagulation are cheap and effective aids to precipitating or otherwise
harvesting whole cells, cell debris or soluble proteins but, of course, it is essential that the agents
used must not inhibit the target enzymes. It is important to note that the choice of flocculant is
determined by the pH and ionic strength of the solution and the nature of the particles. Most
flocculants have very definite optimum concentrations above which further addition may be
counter-effective. Some flocculants can be rapidly ruined by shear.
A comparatively recent introduction designed for the removal of cell debris is a moderately
hydrophobic product in which cellulose is lightly derivatised with diethylaminoethyl functional
groups. This material (Whatman CDR; cell debris remover) is inexpensive (essential as it is not
reusable), binds to unwanted negatively charged cell constituents, acts as a filter aid and may be
incinerated to dispose of hazardous wastes.
Filtration
Filtration separates simply on the basis of particle size. Its efficiency is limited by the shape and
compressibility of the particles, the viscosity of the liquid phase and the maximum allowable
pressures. Large-scale simple filtration employs filter cloths and filter aids in a plate and frame
press configuration, in rotary vacuum filters or centrifugal filters (Figure 2.3). The volumetric
throughput of a filter is proportional to the pressure (P) and filter area (A F) and inversely
proportional to the filter cake thickness (DF) and the dynamic viscosity
(2.3)
11
where k is a proportionality constant dependent on the size and nature of the particles. For very
small particles k depends on the fourth power of their diameter. Filtration of particles that are easily
compressed leads to filter blockage and the failure of Equation 2.3 to describe the system. Under
these circumstances a filter aid, such as celite, is mixed with the feedstock to improve the
mechanical stability of the filter cake. Filter aids are generally used only where the liquid phase is
required as they cause substantial problems in the recovery of solids. They also may cause loss of
enzyme activity from the solution due to physical hold-up in the filter cake. It is often difficult for a
process development manager to decide whether to attempt to recover enzyme trapped in this way.
Problems associated with the build-up of the filter cake may also be avoided by high tangential
flow of the feedstock across the surface of the filter, a process known as crossflow
microfiltration (Figure 2.4). This method dispenses with filter aids and uses special symmetric
microporous membrane assemblies capable of retaining particles down to 0.1 - 1 mm diameter
(cf. Bacillus diameter of about 2 mm).
Figure 2.3. The basic design of the rotary vacuum filter. The suspension is sucked through a filter
cloth on a rotating drum. This produces a filter cake which is removed with a blade. The filter cake
may be rinsed during its rotation. These filters are generally rather messy and difficult to contain
making them generally unsuitable for use in the production of toxic or recombinant DNA products.
There have been recent developments that improve their suitability, however, such as
the Disposable Rotary Drum Filter.
12
A simple and familiar filtration apparatus is the perforate bowl centrifuge or basket centrifuge, in
effect a spin drier. Cell debris is collected on a cloth with, or without, filter aid and can be skimmed
off when necessary using a suitable blade. Such centrifugal filters have a large radius and effective
liquid depth, allowing high volumes. However, safety decrees that the angular velocity must be low
and so only large particles (e.g. plant material) can be removed satisfactorily.
Figure 2.4. Principles of (a) dead-end filtration and (b) cross-flow filtration. In dead-end filtration
the flow causes the build-up of the filter cake, which may prevent efficient operation. This is
avoided in cross-flow filtration where the flow sweeps the membrane surface clean.
Phases form when limiting concentrations of the polymers are exceeded. Both phases contain
mainly water and are enriched in one of the polymers. The limiting concentrations depend on the
type and molecular weight of the polymers and on the pH, ionic strength and temperature of the
solution. Some polymers form the upper hydrophobic phase in the presence of fairly concentrated
solutions of phosphates or sulphates (e.g. 10% polyethylene glycol 4000/12.5% potassium
phosphate buffer). A drawback to the useful dextran/polyethylene glycol system is the high cost of
the purified dextran used. This has been alleviated by the use of crude unfractionated dextran
preparations, much cheaper hydroxypropyl starch derivatives and salt-containing biphasic systems.
13
Aqueous biphasic systems are of considerable value to biotechnology. They provide the
opportunity for the rapid separation of biological materials with little probability of denaturation.
The interfacial tension between the phases is very low (i.e. about 400-fold less than that between
water and an immiscible organic solvent), allowing small droplet size, large interfacial areas,
efficient mixing under very gentle stirring and rapid partition. The polymers have a stabilising
influence on most proteins. A great variety of separations have been achieved, by far the most
important being the separation of enzymes from broken crude cell material. Separation may be
achieved in a few minutes, minimising the harmful action of endogenous proteases. The systems
have also been used successfully for the separation of different types of cell membranes and
organelles, the purification of enzymes and for extractive bioconversions (see Chapter 7).
Continuous liquid two-phase separation is easier than continuous solid/liquid separation using
equipment familiar from immiscible solvent systems, for example disc-stack centrifuges and
counter-current separators. Such systems are readily amenable to scale-up and may be employed in
continuous enzyme extraction processes involving some recycling of the phases.
Cells, cell debris proteins and other material distribute themselves between the two phases in a
manner described by the partition coefficient (P) defined as:
(2.4)
where Ct and Cb represent the concentrations in the top and bottom phases respectively. The yield
and efficiency of the separation is determined by the relative amounts of material in the two phases
and therefore depends on the volume ratio (Vt/Vb). The partition coefficient is exponentially related
to the surface area (and hence molecular weight) and surface charge of the particles in addition to
the difference in the electrical potential and hydrophobicity of the phases. It is not generally very
sensitive to temperature changes. This means that proteins and larger particles are normally
partitioned into one phase whereas smaller molecules are distributed more evenly between phases.
A partition coefficient of greater than 3 is required if usable yields are to be achieved by a single
extraction process. Typical partition coefficients for proteins are 0.01-100 whereas the partition
coefficients for cells and cell debris are effectively zero.
The influence of pH and salts on protein partition is complex, particularly when phosphate buffers
are present. A given protein distributes differently between the phases at different pH's and ionic
strength but the presence of phosphate ions affect the partition coefficient in an anomalous fashion
because these ions distribute themselves unequally resulting in electrostatic potential (and pH)
differences. This means that systems may be 'tuned' to enrich an enzyme in one phase, ideally the
upper phase with cell debris and unwanted enzymes in the lower phase.
An enzyme may be extracted from the upper (polyethylene glycol) phase by the addition of salts or
further polymer, generating a new biphasic system. This stage may be used to further purify the
enzyme. A powerful modification of this technique is to combine phase partitioning and affinity
partitioning. Affinity ligands (e.g. triazine dyes) may be coupled to either polymer in an aqueous
biphasic system and thus greatly increase the specificity of the extraction.
14
Cell breakage
Various intracellular enzymes are used in significant quantities and must be released from cells and
purified (Table 2.1). The amount of energy that must be put into the breakage of cells depends very
much on the type of organism and to some extent on the physiology of the organism. Some types of
cell are broken readily by gentle treatment such as osmotic shock (e.g. animal cells and some gram-
negative bacteria such asAzotobacter species), whilst others are highly resistant to breakage. These
include yeasts, green algae, fungal mycelia and some gram-positive bacteria which have cell wall
and membrane structures capable of resisting internal osmotic pressures of around 20 atmospheres
(2 MPa) and therefore have the strength, weight for weight, of reinforced concrete. Consequently a
variety of cell disruption techniques have been developed involving solid or liquid shear or cell
lysis.
The rate of protein released by mechanical cell disruption is usually found to be proportional to the
amount of releasable protein.
(2.5)
where P represents the protein content remaining associated with the cells, t is the time and k is a
release constant dependent on the system. Integrating from P = P m (maximum possible protein
releasable) at time zero to P = Pt at time t gives
(2.6)
(2.7)
(2.8)
(2.9)
It is most important in choosing cell disruption strategies to avoid damaging the enzymes. The
particular hazards to enzyme activity relevant to cell breakage are summarised in Table 2.3. The
most significant of these, in general, are heating and shear.
15
Table 2.3. Hazards likely to damage enzymes during cell disruption.
All mechanical methods require a large input of energy, generating heat. Cooling is
Heat essential for most enzymes. The presence of substrates, substrate analogues or polyols
may also help stabilise the enzyme.
Shear forces are needed to disrupt cells and may damage enzymes, particularly in the
Shear
presence of heavy metal ions and/or an air interface.]
Disruption of cells will inevitably release degradative enzymes which may cause
serious loss of enzyme activity. Such action may be minimised by increased speed of
Proteases processing with as much cooling as possible. This may be improved by the presence
of an excess of alternative substrates (e.g. inexpensive protein) or inhibitors in the
extraction medium.
Buffered solutions may be necessary. The presence of substrates, substrate analogues
pH
or polyols may also help stabilise the enzyme.
Some enzymes may suffer conformational changes in the presence of detergent and/or
solvents. Polyphenolics derived from plants are potent inhibitors of enzymes. This
Chemical
problem may be overcome by the use of adsorbents, such as polyvinylpyrrolidone,
and by the use of ascorbic acid to reduce polyphenol oxidase action.
Reducing agents (e.g. ascorbic acid, mercaptoethanol and dithiothreitol) may be
Oxidation
necessary.
Foaming The gas-liquid phase interfaces present in foams may disrupt enzyme conformation.
Heavy- Heavy metal ions (e.g. iron, copper and nickel) may be introduced by leaching from
metal the homogenisation apparatus. Enzymes may be protected from irreversible
toxicity inactivation by the use of chelating reagents, such as EDTA.
Media for enzyme extraction will be selected on the basis of cost-effectiveness so will include as
few components as possible. Media will usually be buffered at a pH value which has been
determined to give the maximum stability of the enzyme to be extracted. Other components will
combat other hazards to the enzyme, primarily factors causing denaturation (Table 2.3).
Much of the energy absorbed by cell suspensions is converted to heat so effective cooling is
essential. The amount of protein released by sonication has been shown to follow Equation 2.9. The
constant (k) is independent of cell concentrations up to high levels and approximately proportional
to the input acoustic power above the threshold power necessary for cavitation. Disintegration is
independent of the sonication frequency except insofar as the cavitation threshold frequency
depends on the frequency.
Equipment for the large-scale continuous use of ultrasonics has been available for many years and
is widely used by the chemical industry but has not yet found extensive use in enzyme production.
Reasons for this may be the conformational lability of some (perhaps most) enzymes to sonication
and the damage that they may realise though oxidation by the free radicals, singlet oxygen and
hydrogen peroxide that may be concomitantly produced. Use of radical scavengers (e.g. N 2O) have
been shown to reduce this inactivation. As with most cell breakage methods, very fine cell debris
particles may be produced which can hinder further processing. Sonication remains, however, a
popular, useful and simple small-scale method for cell disruption.
17
Figure 2.5. A cross-section through the Manton-Gaulin homogeniser valve, showing the flow of
material. The cell suspension is pumped at high pressure through the valve impinging on it and the
impact ring. The shape of the exit nozzle from the valve seat varies between models and appears to
be a critical determinant of the homogenisation efficiency. The model depicted is the 'CD Valve'
from APV Gaulin.
As narrow orifices which are vulnerable to blockage are key parts of this type of homogeniser, it is
unsuitable for the disruption of mycelial organisms but has been used extensively for the disruption
of unicellular organisms. The release of proteins can be described by Equation 2.9 but normally a
similar relationship is used where the time variable is replaced by the number of passes (N) through
the homogeniser.
(2.10)
In the commonly-used operating range with pressures below about 75 MPa, the release constant (k)
has been found to be proportional to the pressure raised to an exponent dependent on the organism
and its growth history (e.g. k=k'P2.9 in Saccharomyces cerevesiae and k=k'P2.2 in Escherichia coli,
where P represents the operating pressure and k' is a rate constant). Different growth media may be
selected to give rise to cells of different cell wall strength. Clearly, the higher the operating
18
pressure, the more efficient is the disruption process. The protein release rate constant (k) is
temperature dependent, disruption being more rapid at higher temperatures. In practice, this
advantage cannot be used since the temperature rise due to adiabatic compression is very
significant so samples must be pre-cooled and cooled again between multiple passes. At an
operating pressure of 50 MPa, the temperature rise each pass is about 12 deg. C.
In addition to the fragility of the cells, the location of an enzyme within the cells can influence the
conditions of use of an homogeniser. Unbound intracellular enzymes may be released by a single
pass whereas membrane bound enzymes require several passes for reasonable yields to be obtained.
Multiple passes are undesirable because, of course, they decrease the throughput productivity rate
and because the further passage of already broken cells results in fine debris which is excessively
difficult to remove further downstream. Consequently, homogenisers will be used at the highest
pressures compatible with the reliability and safety of the equipment and the temperature stability
of the enzyme(s) released. High pressure homogenisers are acceptably good for the disruption of
unicellular organisms provided the enzymes needed are not heat labile. The shear forces produced
are not capable of damaging enzymes free in solution. The valve unit is prone to erosion and must
be precision made and well maintained.
Bead mills are available in various sizes and configurations from the Mickle shaker which has a
maximum volume of about 40 ml to continuous process equipment capable of handling up to 200
Kg wet yeast or 20 Kg wet bacteria each hour. The bead mills that have been studied in most detail
are the Dyno-Mill and the Netsch-Molinex agitator, both of which consist of a cylindrical vessel
containing a motor-driven central shaft equipped with impellers of different types. Both can be
operated continuously, being equipped with devices which retain the beads within the milling
chamber. Glass Ballotini or stainless steel balls are used, the size range being selected for most
effective release of the enzyme required. Thus 1 mm diameter beads are satisfactory for the rapid
release of periplasmic enzymes from yeast but 0.25 mm diameter beads must be used, for a longer
period, to release membrane-bound enzymes from bacteria.
The kinetics of protein release from bead mills follows the relationship given by Equation 2.9 with
respect to the time (t) that a particle spends in the mill. Unfortunately, however well designed these
mills are, when continuously operated there will be a significant amount of backmixing which
reduces the efficiency of the protein released with respect to the average residence time (t, see the
19
discussion concerning backmixing in reactors in Chapter 5). This is more noticeable at low flow
rates (high average residence times) and when the proportion of protein released is high. It may be
counteracted by designing the bead mill to encourage plug flow characteristics. Under these
circumstances the relationship can be shown to be
(2.11)
where i represents the degree of backmixing (i.e. i = 0 under ideal plug flow conditions and i = 1
for ideal complete backmixing). Equation 2.11 reduces to give the simplified relationship
of Equation 2.9 at low (near zero) values of i.
In addition to bead size, the protein release rate constant (k) is a function of temperature, bead
loading, impeller rotational speed and cell loading. Impeller speeds can be increased with
advantage until bead breakage becomes significant but heat generation will also increase. At a
constant impeller speed, the efficiency of the equipment declines with throughput as the degree of
backmixing increases. There will be an optimum impeller tip speed at which the increases in
disruption are balanced by increases in backmixing.
In general, increased bead loading increases the rate of protein release but also increases the
production of heat and the power consumption. Heat production is the major problem in the use of
bead mills for enzyme release, particularly on a large (e.g. 20 litres) scale. Smaller vessels may be
cooled adequately through cooling jackets around the bead chamber but larger mills require cooling
through the agitator shaft and impellers. However, if cooling is effective there is little damage to
the enzymes released.
Use of freeze-presses
The Hughes press and the 'X' press enable frozen cell pastes to be forced under high pressure (150 -
230 MPa, 10 - 15 tons per square inch) through narrow orifices, the disruption being produced by
phase and volume changes and by solid shear due to the ice crystals. The Hughes press can only be
used discontinuously on a small scale. The 'X' press may be used semi-continuously and is
amenable to scale-up but, although the method allows the breakage of even the most robust
organisms and the efficient recovery of heat-sensitive enzymes, freeze-pressing is not used on a
large scale for releasing enzymes from cells.
Enzymic lysis using added enzymes has been used widely on the laboratory scale but is less
popular for industrial purposes. Lysozyme, from hen egg-white, is the only lytic enzyme available
on a commercial scale. It has often used to lyse Gram positive bacteria in an hour at about 50,000 U
Kg-1 (dry weight). The chief objection to its use on a large scale is its cost. Where costs are reduced
by the use of the relatively inexpensive, lysozyme-rich, dried egg white, a major separation
problem may be introduced. Yeast-lytic enzymes from Cytophaga species have been studied in
some detail and other lytic enzymes are under development. If significant markets for lytic
enzymes are identified, the scale of their production will increase and their cost is likely to
decrease.Lysis by acid, alkali, surfactants and solvents can be effective in releasing enzymes,
provided that the enzymes are sufficiently robust. Detergents, such as Triton X-100, used alone or
in combination with certain chaotropic agents, such as guanidine HCl, are effective in releasing
membrane-bound enzymes. However, such materials are costly and may be difficult to remove
from the final product.
There are a number types of apparatus available. Stirred cells represent the simplest configuration
of ultrafiltration cell. The membrane rests on a rigid support at the base of a cylindrical vessel
which is equipped with a magnetic stirrer to combat concentration polarisation. It is not suitable for
large scale use but is useful for preliminary studies and for the concentration of laboratory column
eluates. Various large-scale units are available in which membranes are formed into wide diameter
tubes (1 - 2 cm diameter) and the tubes grouped into cartridges. These are not as compact as
capillary systems (area/volume about 25 m -1) and are very expensive but are less liable to blockage
21
by stray large particles in the feedstream. Cheaper thin-channel systems are available (area/volume
about 500 m-1) which use flat membrane sandwiches in filter press arrangements of various designs
chosen to produce laminar flow across the membrane and minimise concentration polarisation.
Capillary membranes represent a relatively cheap and increasingly popular type of ultrafiltration
system which uses micro-tubular membranes 0.2 - 1.1 mm diameter and provides large membrane
areas within a small unit volume (area/volume about 1000 m -1). Membranes are usually mounted
into modules for convenient manipulation. This configuration of membranes can be scaled up with
ease. Commercial models are available that give ultrafiltration rates of up to 600 L hr -1.
The steady improvement in the performance, durability and reliability of membranes has been a
boon to enzyme technologists, encouraging wide use of the various ultrafiltration configurations.
Problems with membrane blockage and fouling can usually be overcome by treatment of
membranes with detergents, proteases or, with care, acids or alkalis. The initial cost of membranes
remains considerable but modern membranes are durable and cost-effective. Ultrafiltration, done
efficiently, results in little loss of enzyme activity. However, some configurations of apparatus,
particularly in which solutions are recycled, can produce sufficient shear to damage some enzymes.
Concentration by precipitation
Precipitation of enzymes is a useful method of concentration and is ideal as an initial step in their
purification. It can be used on a large scale and is less affected by the presence of interfering
materials than any of the chromatographic methods described later.
Salting out of proteins, particularly by use of ammonium sulphate, is one of the best known and
used methods of purifying and concentrating enzymes, particularly at the laboratory scale.
Increases in the ionic strength of the solution cause a reduction in the repulsive effect of like
22
charges between identical molecules of a protein. It also reduces the forces holding the solvation
shell around the protein molecules. When these forces are sufficiently reduced, the protein will
precipitate; hydrophobic proteins precipitating at lower salt concentrations than hydrophilic
proteins. Ammonium sulphate is convenient and effective because of its high solubility, cheapness,
lack of toxicity to most enzymes and its stabilizing effect on some enzymes (see Table 2.4). Its
large-scale use, however, is limited as it is corrosive except with stainless steel, it forms dense
solutions presenting problems to the collection of the precipitate by centrifugation, and it may
release gaseous ammonia, particularly at alkaline pH. The practice of using ammonium sulphate
precipitation is more straightforward than the theory. Reproducible results can only be obtained
provided the protein concentration, temperature and pH are kept constant. The concentration of the
salt needed to precipitate an enzyme will vary with the concentration of the enzyme. However,
fractionation of protein mixtures by the stepwise increase in the ionic strength can be a very
effective way of partly purifying enzymes.
(2.11)
where S is the enzyme solubility, Kintercept is the intercept constant, Ksalt is the salting out constant
and T is the ionic strength which is proportional to the concentration of a precipitating salt.
Kintercept is independent of the salt used but depends on the pH, temperature, enzyme and the other
components in the solution. Ksaltdepends on both the enzyme required and the salt used but is
largely independent of other factors. This equation (2.11) may also be used to give the minimum
salt concentration necessary before enzyme will start to precipitate; the concentration change
necessary to precipitate the enzyme varying according to the magnitude of the salting out constant.
Some enzymes do not survive ammonium sulphate precipitation. Other salts may be substituted but
the more favoured alternative is to use organic solvents such as methanol, ethanol, propan-2-ol and
acetone. These act by reducing the dielectric of the medium and consequently reducing the
solubility of proteins by favouring protein-protein rather than protein-solvent interactions. Organic
solvents are not widely used on a large scale because of their cost, their flammability, and the
tendency of proteins to undergo rapid denaturation by these solvents if the temperature is allowed
to rise much above 0�C. On safety grounds when organic solvents are used, special flameproof
laboratory areas are used and temperatures maintained below their flashpoints.
Except when enzymes are presented for sale as ammonium sulphate precipitates, the precipitating
salt or solvent must be removed. This may be done by dialysis, ultrafiltration or by using a
desalting column of, for instance, Sephadex G-25.
Chromatography
Enzyme preparations that have been clarified and concentrated are now in a suitable state for
further purification by chromatography. For enzyme purification there are three principal types of
chromatography utilising the ion-exchange, affinity and gel exclusion properties of the enzyme,
23
usually in that order. Ion-exchange and affinity chromatographic methods can both rapidly handle
large quantities of crude enzyme but ion-exchange materials are generally cheaper and, therefore,
preferred at an earlier stage in the purification where the scale of operation is somewhat greater.
Gel exclusion chromatography (also sometimes called 'gel filtration' or just 'gel chromatography'
although it does not separate by a filtering mechanism, larger molecules passing more rapidly
through the matrix than smaller molecules) is relatively slow and has the least capacity and
resolution. It is generally left until last as an important final purification step and also as a method
of changing the solution buffer before concentration, finishing and sale. Where sufficient
information has been gathered regarding the size and variation of charge with pH of the required
enzyme and its major contaminants, a rational purification scheme can be devised. A relatively
quick analytical method for obtaining such data utilises a two-dimensional process whereby
electrophoresis occurs in one direction and a range of pH is produced in the other; movement in the
electric field determined by the size and sign of a protein's charge, which both depend on the pH.
As the sample is applied across the range of pH, this method produces titration curves (i.e. charge
versus pH) for all proteins present.
A large effort has been applied to the development of chromatographic matrices suitable for the
separation of proteins. The main problem that has had to be overcome is that of ensuring the matrix
has sufficiently large surface area available to molecules as large as proteins (i.e. they are
macroporous) whilst remaining rigid and incompressible under rapid elution conditions. In
addition, matrices must generally be hydrophilic and inert. Although the standard bead diameters of
most of these matrices are non-uniform and fairly large (50 - 150mm), many are now supplied as
uniform-sized small beads (e.g. 4 - 6 mm diameter) which allows their use in very efficient
separation processes (high performance liquid chromatography, HPLC), but at exponentially
increasing cost with decreasing bead size. Relatively high pressures are needed to operate such
columns necessitating specialised equipment and considerable additional expense. They are used
only for the small-scale production of expensive enzymes, where a high degree of purity is required
(e.g. restriction endonucleases and therapeutic enzymes).
Column manufacturers now supply equipment for monitoring and controlling chromatography
systems so that it is possible to have automated apparatus which loads the sample, collects fractions
and regenerates the column. Such equipment must, of course, have fail-safe devices to protect both
column and product.
Ion-exchange chromatography
Enzymes possess a net charge in solution, dependent upon the pH and their structure and isoelectric
point. In solutions of pH below their isoelectric point they will be positively charged and bind to
cation exchangers whereas in solutions of pH above their isoelectric point they will be negatively
charged and bind to anion exchangers. The pH chosen must be sufficient to maintain a high, but
opposite, charge on both protein and ion-exchanger and the ionic strength must be sufficient to
maintain the solubility of the protein without the salt being able to successfully compete with the
protein for ion-exchange sites. The binding is predominantly reversible and its strength is
determined by the pH and ionic strength of the solution and the structures of the enzyme and ion-
24
exchanger. Normally the pH is kept constant and enzymes are eluted by increasing the solution
ionic strength. A very wide range of ion-exchange resins, cellulose derivatives and large-pore gels
are available for chromatographic use.
Ion-exchange materials are generally water insoluble polymers containing cationic or anionic
groups. Cation exchange matrices have anionic functional groups such as -SO 3-, -OPO3- and
-COO- and anion exchange matrices usually contain the cationic tertiary and quaternary ammonium
groups, with general formulae -NHR2+and -NR3+. Proteins become bound by exchange with the
associated counter-ions.
Ion-exchange polystyrene resins are eminently suitable for large-scale chromatographic use but
have low capacities for proteins due to their small pore size. Binding is often strong, due to the
resin hydrophobicity, and the conditions needed to elute proteins are generally severe and may be
denaturing. Nevertheless such resins are a potential means of concentrating or purifying enzymes.
Ion-exchange cellulose and large pore gels are much more generally suitable for enzyme
purification and, indeed, many were designed for that task. A variety of charged groups, anionic or
cationic, may be introduced. The practical level of substitution of cellulose is limited as
derivatisation above one mole per kilogram may lead to dissolution of the cellulose. Consequently,
proteins may be eluted from them under mild conditions. Ion-exchange cellulose can be used in
both batch and column processes but on a large scale they are used mainly batchwise. This is
because the increased speed of large-scale batchwise processing and the avoidance of the deep-bed
filtering characteristics of columns outweigh any advantage due to the increase in resolution on
columns. Careful preparation before use and regeneration after use is essential for their effective
use.
Batchwise operations involve stirring the pretreated and equilibrated ion-exchanger with the
enzyme solution in a suitable cooled vessel. Adsorption to the exchangers is usually rapid (e.g. less
than 30 minutes) but some proteins can take far longer to adsorb completely. Stirring is essential
but care must be taken not to generate fine particles (fines). Unadsorbed material may be removed
in a variety of manners. Basket centrifuges are a particularly convenient means of hastening the
removal of the initial supernatant and the elution of the adsorbed material. This is usually done
using stepwise increases in ionic strength and/or changes in pH but it is possible to place the
exchangers, plus adsorbed material, in a column and elute using a suitable gradient. However,
whilst ion-exchange cellulose are widely used for column chromatography on the laboratory scale,
their compressibility causes difficulty when attempts are made to use large scale columns.
Some of the problems with derivatised cellulose may be overcome using more recently introduced
materials. Derivatives of cross-linked agarose (Sepharose CL-6B) and of the synthetic polymer
Trisacryl have high capacities (up to 150 mg protein ml -1) yet are not significantly compressible. In
addition, they do not change volume with pH and ionic strength which allows them to be
regenerated without removal from the chromatographic column.
Affinity chromatography
25
This is a term which now covers a variety of methods of enzyme purification, the common factor of
which is the more or less specific interaction between the enzyme and the immobilised ligand. In its
most specific form, the immobilised ligand is a substrate or competitive inhibitor of the enzyme.
Ideally it should be possible to purify an enzyme from a complex mixture in a single step and,
indeed, purification factors of up to several thousand-fold have been achieved. An alternative,
equally specific approach is to use an antibody to the enzyme as the ligand. Such specific matrices,
though, are very expensive and cannot be generally employed on a large scale. Additionally, they
often do not perform as well as might be expected due to non-specific binding effects. In general,
affinity chromatography achieves a higher purification factor (with a median value in reported
purifications of about ten fold) than ion-exchange chromatography (with a median performance of
about three fold), in spite of it generally being used at a later stage in the purification when there is
less purification possible.
A less specific approach, suitable for many enzymes, is to use analogues of coenzymes, such as
NAD+, as the ligand. This method has been used successfully but has now been superceded by the
employment of a series of water soluble dyes as ligands. These are much cheaper and, usually by
trial and error, have been found to have surprising degrees of specificity for a wide range of
enzymes. This dye-affinity chromatography was allegedly discovered by accident, certain enzymes
being found to bind to the blue-dyed dextran used, as a molecular weight standard, to calibrate gel
exclusion columns.
Another fortuitous discovery was hydrophobic interaction chromatography, found when it was
noted that certain proteins were unexpectedly retained on affinity columns containing hydrophobic
spacer arms. Hydrophobic adsorbents now available include octyl or phenyl groups. Hydrophobic
interactions are strong at high solution ionic strength so samples need not be desalted before
application to the adsorbent. Elution is achieved by changing the pH or ionic strength or by
modifying the dielectric constant of the eluent using, for instance, ethanediol. A recent introduction
is cellulose derivatised to introduce even more hydroxyl groups. This material (Whatman HB1) is
designed to interact with proteins by hydrogen bonding. Samples are applied to the matrix in a
concentrated (over 50% saturated, > 2M) solution of ammonium sulphate. Proteins are eluted by
diluting the ammonium sulphate. This introduces more water which competes with protein for the
hydrogen bonding sites. The selectivity of both of these methods is similar to that of fractional
precipitation using ammonium sulphate but their resolution may be somewhat improved by their
use in chromatographic columns rather than batchwise.
Careful choice of matrices for affinity chromatography is necessary. Particles should retain good
flow and porosity properties after attachment of the ligands and should not be capable of the non-
specific adsorption of proteins. Agarose beads fulfil these criteria and are readily available as
ligand supports (see also Chapter 3). Affinity chromatography is not used extensively in the large-
scale manufacture of enzymes, primarily because of cost. Doubtless as the relative costs of
materials are lowered, and experience in handling these materials is gained, enzyme manufacturers
will make increased use of these very powerful techniques.
Gel exclusion chromatography invariably causes dilution of the enzyme which must then be
concentrated using one of the methods described earlier.
To achieve stability, the manufacturer uses all the subtleties at their disposal. Formulation is an art
and often the precise details of the methods used to stabilise enzyme preparations are kept secret or
revealed to customers only under the cover of a confidentiality agreement. Sometimes it is only the
formulation of an enzyme that gives a manufacturer the competitive edge over rival companies. It
should be remembered that most industrial enzymes contain relatively little active enzyme (< 10%
w/w, including isoenzymes and associated enzyme activities), the rest being due to inactive protein,
stabilisers, preservatives, salts and the diluent which allows standardisation between production
batches of different specific activities.
27
The key to maintaining enzyme activity is maintenance of conformation, so preventing unfolding,
aggregation and changes in the covalent structure. Three approaches are possible:
1. use of additives,
2. the controlled use of covalent modification, and
3. enzyme immobilisation (discussed further in Chapter 3).
In general, proteins are stabilised by increasing their concentration and the ionic strength of their
environment. Neutral salts compete with proteins for water and bind to charged groups or dipoles.
This may result in the interactions between an enzyme's hydrophobic areas being strengthened
causing the enzyme molecules to compress and making them more resistant to thermal unfolding
reactions. Not all salts are equally effective in stabilising hydrophobic interactions, some are much
more effective at their destabilisation by binding to them and disrupting the localised structure of
water (the chaotropic effect, Table 2.4). From this it can be seen why ammonium sulphate and
potassium hydrogen phosphate are a powerful enzyme stabilisers whereas sodium thiosulphate and
calcium chloride destabilise enzymes. Many enzymes are specifically stabilised by low
concentrations of cations which may or may not form part of the active site, for example
Ca2+ stabilises a-amylases and Co2+ stabilises glucose isomerases. At high concentrations (e.g. 20%
NaCl), salt discourages microbial growth due to its osmotic effect. In addition ions can offer some
protection against oxidation to groups such as thiols by salting-out the dissolved oxygen from
solution.
Low molecular weight polyols (e.g. glycerol, sorbitol and mannitol) are also useful for stabilising
enzymes, by repressing microbial growth, due to the reduction in the water activity, and by the
formation of protective shells which prevent unfolding processes. Glycerol may be used to protect
enzymes against denaturation due to ice-crystal formation at sub-zero temperatures. Some
hydrophilic polymers (e.g. polyvinyl alcohol, polyvinylpyrrolidone and hydroxypropylcelluloses)
stabilise enzymes by a process of compartmentalisation whereby the enzyme-enzyme and enzyme-
water interactions are somewhat replaced by less potentially denaturing enzyme-polymer
interactions. They may also act by stabilising the hydrophobic effect within the enzymes. Many
specific chemical modifications of amino acid side chains are possible which may (or, more
commonly, may not) result in stabilisation. A useful example of this is the derivatisation of lysine
side chains in proteases with N-carboxyamino acid anhydrides. These form polyaminoacylated
28
enzymes with various degrees of substitution and length of amide-linked side chains. This
derivatisation is sufficient to disguise the proteinaceous nature of the protease and prevent
autolysis.
Important lessons about the molecular basis of thermostability have been learned by comparison of
enzymes from mesophilic and thermophilic organisms. A frequently found difference is the
increase in the proportion of arginine residues at the expense of lysine and histidine residues. This
may be possibly explained by noting that arginine is bidentate and has a higher pK a than lysine or
histidine (see Table 1.1). Consequently, it forms stronger salt links with bidentate aspartate and
glutamate side chains, resulting in more rigid structures. This observation, among others, has given
hope that site-specific mutagenesis may lead to enzymes with significantly improved stability
(see Chapter 8). In the meantime it remains possible to convert lysine residues to arginine-like
groups by reaction with activated ureas. It should be noted that enzymes stabilised by making them
more rigid usually show lower activity (i.e. Vmax) than the 'natural' enzyme.
Enzymes are very much more stable in the dry state than in solution. Solid enzyme preparations
sometimes consist of freeze-dried protein. More usually they are bulked out with inert materials
such as starch, lactose, carboxymethylcellulose and other poly-electrolytes which protect the
enzyme during a cheaper spray-drying stage. Other materials which are added to enzymes before
sale may consist of substrates, thiols to create a reducing environment, antibiotics, benzoic acid
esters as preservatives for liquid enzyme preparations, inhibitors of contaminating enzyme
activities and chelating agents. Additives of these types must, of course, be compatible with the
final use of the enzyme's product.
Enzymes released onto the market should conform to a number of quality procedures including
regulatory requirements, which are legal and mandatory. This is provided by the quality assurance
(QA) within the company. Enzyme products must be consistent as appropriate to their intended use.
This may be ensured by good manufacturing practice (GMP) and further checked by quality
control (QC).
Customer service
A customer who is likely to use large quantities of an enzyme will be expected to specify the form
and activity in which the enzyme is supplied. The development of a new enzyme-catalysed process
is often a matter of teamwork between the customer and the enzyme company's development
scientists. Once the process is running, the enzyme company will probably be in contact with the
customer through three types of individuals:
1. Salespeople who ensure that the supply of enzyme continues and that the cost of the enzyme
is mutually acceptable.
2. Technical salespeople who liase with technical managers to ensure that their product is
performing up to specification. This association often leads to suggestions for the
improvement of the process. The technical sales team will be expected by the customer to
29
deal with problems to do with the enzyme. They may be able to solve problems themselves
or may require the services of the third group, the laboratory or pilot plant scientists.
3. The laboratory and pilot plant scientists will spend some time trouble-shooting as and when
necessary. They may test materials for customers, for instance if a new source of raw
material is under consideration.
All the enzyme company's employees must, of course, work as a team and it is in the interests of
both customer and manufacturer if their technical experts also cooperate closely. Both
organisations will learn from each other. Technical advances should accrue as a result of
suggestions from both sides.
All enzymes, being proteins, are potential allergens and have especially potent effects if inhaled as
a dust. Once an individual has developed an immune response as a result of inhalation or skin
contact with the enzyme, re-exposure produces increasingly severe responses becoming dangerous
or even fatal. Because of this, dry enzyme preparations have been replaced to a large extent by
liquid preparations, sometimes deliberately made viscous to lower the likelihood of aerosol
formation during handling. Where dry preparations must be used, as in the formulation of many
enzyme detergents, allergenic responses by factory workers are a very significant problem
particularly when fine-dusting powders are employed. Workers in such environments are usually
screened for allergies and respiratory problems. The problem has been largely overcome by
encapsulating and granulating dry enzyme preparations, a procedure that has been applied most
successfully to the proteases and other enzymes used in detergents. Enzyme producers and users
recognise that allergenicity will always be a potential problem and provide safety information
concerning the handling of enzyme preparations. They stress that dust in the air should be avoided
so weighing and manipulation of dry powders should be carried out in closed systems. Any spilt
enzyme powder should be removed immediately, after first moistening it with water. Any waste
enzyme powder should be dissolved in water before disposal into the sewage system. Enzyme on
the skin or inhaled should be washed with plenty of water. Liquid preparations are inherently safer
but it is important that any spilt enzyme is not allowed to dry as dust formation can then occur. The
formation of aerosols (e.g. by poor operating procedures in centrifugation) must be avoided as these
are at least as harmful as powders.
Activity-related toxicity is much rarer but it must be remembered that proteases are potentially
dangerous, particularly in concentrated forms and especially if inhaled. No enzyme has been found
to be toxic, mutagenic or carcinogenic by itself as might be expected from its proteinaceous
structure. However, enzyme preparations cannot be regarded as completely safe as such dangerous
30
materials may be present as contaminants, derived from the enzyme source or produced during its
processing or storage.
The organisms used in the production of enzymes may themselves be sources of hazardous
materials and have been the chief focus of attention by the regulatory authorities. In the USA,
enzymes must be Generally Regarded As Safe (GRAS) by the FDA (Food and Drug
Administration) in order to be used as a food ingredient. Such enzymes include a-amylase, b-
amylase, bromelain, catalase, cellulase, ficin, a-galactosidase, glucoamylase, glucose isomerase,
glucose oxidase, invertase, lactase, lipase, papain, pectinase, pepsin, rennet and trypsin. In the UK,
the Food Additives and Contaminants Committee (FACC) of the Ministry of Agriculture, Fisheries
and Food classified enzymes into five classes on the basis of their safety for presence in the foods
and use in their manufacture.
Group A. Substances that the available evidence suggests are acceptable for use in food.
Group B. Substances that on the available evidence may be regarded as provisionally acceptable
for use in food but about which further information must be made available within a specified time
for review.
Group C. Substances for which the available evidence suggests toxicity and which ought not to be
permitted for use in food until adequate evidence of their safety has been provided to establish their
acceptability.
Group D. Substances for which the available information indicates definite or probable toxicity
and which ought not to be permitted for use in food.
Group E. Substances for which inadequate or no toxicological data are available and for which it
is not possible to express an opinion as to their acceptability for use in food.
This classification takes into account the potential chemical toxicity from microbial secondary
metabolites such as mycotoxins and aflotoxins. The growing body of knowledge on the long-term
effects of exposure to these toxins is one of the major reasons for the tightening of legislative
controls.
The enzymes that fall into group A are exclusively plant and animal enzymes such as papain,
catalase, lipase, rennet and various other proteases. Group B contains a very wide range of enzymes
from microbial sources, many of which have been used in food or food processing for many
hundreds of years. The Association of Microbial Food Enzyme Producers (AMFEP) has suggested
subdivisions of the FACC's group B into:
Class ain8 microorganisms that have traditionally been used in food or in food processing,
including Bacillus subtilis, Aspergillus
niger, Aspergillus oryzae, Rhizopus oryzae, Saccharomyces cerevisiae,Kluyveromyces fragilis, Klu
yveromyces lactis and Mucor javanicus.
31
Class bin8 microorganisms that are accepted as harmless contaminants present in food,
including Bacillusstearothermophilus, Bacillus licheniformis, Bacillus coagulans,
and Klebsiella in8aerogenes.
It was proposed that Class a should not be subjected to testing and that Classes b and c should be
subjected to the following tests:
In addition Class c should be tested for microorganism pathogenicity and, under exceptional
circumstances, in vivo mutagenicity, teratogenicity, and carcinogenicity.
The cost of the various tests needed to satisfy the legal requirements are very significant and must
be considered during the determination of process costs. Plainly the introduction of an enzyme
from a totally new source will be a very expensive matter. It may prove more satisfactory to clone
such an enzyme into one of AMFEP's Class a organisms but this will first require new legislation to
regulate the use of cloned microbes in foodstuffs. Some of the safety problems associated with the
use of free enzymes may be overcome by using immobilised enzymes (see Chapter 3). This is an
extremely safe technique, so long as the materials used are acceptable and neither they, nor the
immobilised enzymes, leak into the product stream.
The production of enzymes is subject, in the UK, to the Health and Safety at Work Act 1974, to
ensure the health and safety of employees. Good manufacturing practice is employed and controls
ensure that enzyme production is performed by a pure culture of the producing microbes.
33