Castilho Leda R., Moraes Angela Maria (Ed.) Animal Cell Technology: From Biopharmaceuticals to Gene Therapy

Подождите немного. Документ загружается.

volumes and many consecutive extractions to achieve total recovery. A

value of K close to zero, on the other hand, means that no extraction

occurs (Harrison et al., 2003).

Liquid–liquid extraction offers several advantages, such as low cost,

high yield, ease of continuous operation, and scale-up. Also, in the case of

aqueous two-phase systems, the process would not cause degradation of

proteins, enzymes, virus, or organelles (Cunha et al., 2003; Harrison et al.,

2003). Efﬁcient extraction techniques can lead to a reduction in volume,

providing not only puriﬁcation, but also a concentration of the sample.

They are suitable for processing cell suspensions, enabling an integration

of the solid–liquid separation and the primary puriﬁcation stages and a

consequent reduction in the number of stages in the downstream proces-

sing. However, resolution and puriﬁcation factors are relatively low, com-

pared with chromatographic techniques. Because of this, liquid–liquid

extractions are often designed in the early stages of a puriﬁcation process.

Signiﬁcant improvements in resolution can be achieved with the use of

afﬁnity ligands in one of the phases, although this can increase cost

(Johansson, 1998).

12.4.2 Separation processes based on differences in molar mass

Proteins have high molar mass, the value of which differs from one protein

to another. This allows the use of simple methods to separate proteins

from low molar mass molecules, as well as to separate a speciﬁc protein

from others.

Density gradient centrifugation

Just like other techniques, centrifugation can be applied at different stages

in the downstream processing of a given protein. The main application of

centrifugation in biotechnology (as seen in Chapter 11) is in the clariﬁca-

tion of protein-containing suspensions, usually for the removal of cells

and cell debris, but also after ﬂotation and precipitation operations.

However, density gradient centrifugation or zone centrifugation widens

the range of possible applications of centrifugation by allowing the

separation of small bioparticles such as organelles and viruses, as well as

proteins. In the most widely used method, a continuous density gradient

of sucrose is prepared with a device that adds a concentrated solution of

sucrose to water, in decreasing proportions along a tube. In this way, the

medium density is higher at the bottom of the tube, and the macromole-

cules to be fractionated are applied on the top. After centrifugation, the

different proteins form bands or layers, according to their molar mass,

shape, and density. Two consecutive centrifugation steps (the ﬁrst using

CsCl and the second using sucrose) were employed by Deml et al.

(1999), for purifying HBsAg from cell culture supernatant of genetically

modiﬁed DS-2 cells.

304 Animal Cell Technology

Dialysis

Desalting or buffer exchanges are often required between puriﬁcation

steps. At the laboratory scale, the protein solution is placed in a tube of a

semipermeable polymer membrane immersed in the desired buffer. The

membrane pore size determines the minimum molar mass of the com-

pounds that are retained. Small molecules with a molar mass below the

membrane cut-off will ﬂow freely across the membrane until the osmotic

pressure equilibrium is reached. Complete buffer exchange requires several

changes of the dialysis liquid. The process should be carried out at a

temperature around 48C, to avoid loss of activity.

In industry, dialysis is not widely used because it is slow and labor-

intensive. Here, desalting and/or buffer exchange is usually performed by

diaﬁltration or molecular exclusion chromatography, which are discussed

later.

Microﬁltration

The principle of microﬁltration is the application of hydrostatic pressure

on a microporous ﬁlter membrane, so that the pressure difference forces

solutes, water molecules, and particles smaller than the membrane pore

size to ﬂow across the pores, retaining and concentrating the larger

particles in the suspension.

Microﬁltration membranes usually have a nominal pore diameter in the

range of 0.1–10 m. However, the membrane speciﬁcation is not an

absolute parameter. The membranes usually present a pore size distribu-

tion around the nominal value and the shape of the bioparticles can

determine whether they are retained or pass through the membrane. The

membranes are manufactured from polymers, such as Teﬂon

, polyester,

PVC (polyvinyl chloride), Nylon

, polypropylene, polyethersulfone, and

cellulose, or from inorganic materials, such as ceramic and sinterized

stainless steel.

In microﬁltration, the permeate ﬂux increases inversely with the suspen-

sion viscosity and proportionally to the applied pressure, provided that

there is no membrane fouling (Belford, 1988; Ho and Zydney, 2000). To

accelerate the process, it is possible to decrease the solution viscosity by

increasing the temperature, although not so much as to denature the

protein.

Microﬁltration is widely used for the removal of cells and fragments

from suspension. It is also used as a method of sterilization of solutions,

and has the advantage of high efﬁciency, simplicity, compactness, and

reliability.

Ultraﬁltration

In ultraﬁltration, water and other low molar mass molecules are forced

through a semi-permeable membrane by the application of high pressures

(1–7 bar) or of a centrifugal ﬁeld. This technique involves membranes

with pore diameters in the range of 1.0–20 nm, which are most commonly

characterized and selected based on their nominal molar mass cut-off

Produc t purification processes 305

(MWCO), usually deﬁned as the molar mass at which the membrane

rejects 90% of solute molecules. However, as in microﬁltration, the

molecular shape can affect permeability through the membrane pores. For

example, a membrane with a nominal cut-off of 100 kDa, which does not

allow globular molecules with a molar mass of 100 kDa to ﬂow through,

may allow ﬁbrous molecules with higher molar masses to ﬂow across the

pores. As in microﬁltration, the membrane pore size is not uniform, with a

normal distribution around an average value.

A common problem of this technique is the gradual decrease in

permeate ﬂux associated with membrane clogging or fouling, caused by

adsorption or physical deposition of particles and/or macromolecules on

membrane pores. Fouling can be minimized by prior clariﬁcation (particu-

late removal) of the feed solution, by the selection of operational condi-

tions that minimize interactions between membranes and macromolecules,

by the use of tangential ﬂow, or by performing intermittent back-ﬂushing

operations.

Diaﬁltration

Diaﬁltration consists of the application of microﬁltration or, more com-

monly, ultraﬁltration membranes for solvent exchange. In diaﬁltration

systems, schematically shown in Figure 12.2 , the permeate is collected

continuously at the same rate at which fresh solvent is added, so that

undesirable solutes are removed in the permeate stream and the retentate

volume is kept constant. Although diaﬁltration is being increasingly used,

especially on a large scale, a disadvantage is the generation of large

volumes of liquid efﬂuents. For instance, in a system with a 100 m

membrane operating at a permeate ﬂux of 30 L m

–2

–1

, a volume of

9210 L of water is required to remove 99.99% of the salt present in a

protein solution (Harrison et al., 2003).

Permeate

Fresh

buffer

Feed

tank

Diafiltration

module

Figure 12.2

General scheme of a diaﬁltration system.

306 Animal Cell Technology

Molecular exclusion chromatography

Molecular exclusion chromatography, also known as gel ﬁltration, size

exclusion chromatography, gel permeation chromatography, or simply gel

chromatography, is another separation method based on differences in

molecular size.

In this method, molecules partition between a solvent (aqueous buffer)

and a stationary phase of deﬁned porosity. A protein mixture dissolved in

a suitable buffer ﬂows through a column packed with spherical porous

particles made of an inert material (usually a polymer or a gel). The

column is equilibrated with a pre-selected buffer appropriate for sample

elution. The ﬂow occurs by gravity or aided by a pump.

In a sample containing a mixture of molecules, the smallest permeate the

matrix pores and follow a slow trajectory along the column axis and are

collected later forming the later chromatogram peaks. The largest mole-

cules are excluded from the pores, migrating through the interstitial space

and are, therefore, eluted earlier. The intermediate sized molecules may

partially penetrate the pores, allowing intermediate elution times.

Therefore, the molecules are eluted in the inverse order of their

molecular size, as indicated in Figures 12.3 and 12.4, and the differences in

the elution times of different proteins are related to the fraction of pores

accessible to the solutes. Equations can be obtained that relate the fraction

of pores of different dimensions, the gel structure, and the molecular size

of solutes with so-called distribution coefﬁcients.

The distribution coefﬁcient K

(Equation 2) is deﬁned as the volume

fraction of pores, in a stationary phase, which is effectively permeated by a

solute of a given size. V

is the interstitial volume of the porous medium,

measured by the elution volume of a high molar mass solute that is totally

excluded from the matrix pores. V

is the elution volume of the product of

interest. V

represents the total solvent volume within the pores, available

for small solutes.

 V

)

(2)

Stationary phase

Molecules larger than the pores

Molecules with intermediate size

Molecules smaller than the pores

Figure 12.3

Schematic illustration of the separation principle involved in molecular exclusion chromatography.

Produc t purification processes 307

Since V

is difﬁcult to measure, it is common to use an alternative

distribution coefﬁcient, called K

(Equation 3). In this approach, V

replaced by the difference between the total column volume (V

) and the

interstitial volume (V

) of the column packed with the chromatographic

matrix (Ladisch, 2001; Moraes and Rosa, 2005).

 V

)

 V

)

(3)

For a given matrix, K

or K

can be plotted against the log of the molar

mass of different solutes. For a given molar mass range, a linear relation-

ship will be observed. Therefore, by applying to the column molecules of

known molar mass, and similar shape and density, it is possible to

determine the molar masses of solutes present in a sample (Kumpalume

and Ghose, 2003; Moraes and Rosa, 2005).

Ideally, the column matrix should be inert with regard to the molecules

to be separated. If electrostatic or hydrophobic interactions occur between

proteins and the stationary phase, it is recommended, respectively, to

increase or decrease the ionic strength of the liquid phase, otherwise partial

adsorption of the protein may cause a delay in elution from the column.

This would cause a longer elution time than that expected based on the

molar mass.

V-V

0.5

1.0

High

molar

mass

Solute

that interacts

with the

resin

Low

molar

mass

Intermediate

molar masses

Figure 12.4

Elution proﬁle of a mixture containing solutes of different molar masses and the variables used to

describe solute and matrix behavior (adapted from Ladisch, 2001).

308 Animal Cell Technology

The most commonly used matrices for molecular exclusion chromato-

graphy are cross-linked polymer gels, such as agarose, dextran, and

polyacrylamide. Since dextran and agarose are biodegradable, they should

be stored in the presence of antimicrobial agents.

The selection of a gel matrix should take into account pH and tempera-

ture stability, which should be compatible with the characteristics of the

target protein. The selection of the most suitable gel should also take into

account the main goal to be achieved. If the process is intended to separate

proteins from low molar mass solutes (, 5 kDa), a small pore size matrix

is recommended, so that proteins are completely excluded from the porous

medium. Such a strategy is used for desalting samples.

Matrices composed of small particles have high resolution, since the

molecular diffusion in the interstitial region is small and, therefore, peak

broadening is low. However, these systems are associated with low ﬂow

rates, since small particles result in larger pressure drops along the column.

Because the application of high pressure can lead to particle deformation

and, consequently, to bed compaction, particles that are large and rigid,

capable of withstanding high ﬂow rates, are commonly preferred for large-

scale processes, even if they are associated with lower resolution.

In molecular exclusion chromatography, resolution is directly related to

the sample loading volume, since the higher the sample volume, the higher

the volume in which the protein is eluted. Usually, sample volume should

be 0.5–5% of the total bed volume. For volumes below 0.5% the sample

becomes too diluted, while volumes above 5% work well only for

molecules with large differences in size. For desalting, due to the pro-

nounced difference in the molecular size of the components, sample

volumes can reach up to 30% of the column capacity without a signiﬁcant

decrease in resolution.

With regard to the total protein concentration in the injected sample, it

should be ensured that the ratio between the viscosity of the sample and

the eluent is not higher than 2. High viscosities can cause problems, such

as the occurrence of regions with distorted and irregular ﬂow patterns.

Hence, protein concentration in the loaded sample should ideally be in the

range of 10–20 mg mL

–1

, although in some cases it can reach 70 mg mL

–1

Therefore, to achieve high resolutions in molecular exclusion chromato-

graphy, it is common to use columns with large height-to-diameter ratios

(in the range of 20–40), although at an industrial scale this can become

difﬁcult.

12.4.3 Separation processes based on differences in electrical

charge

Separation of proteins based on differences in their electrical charge

depends on their acid-base properties, which are mostly determined by the

number and type of ionizable side groups in the peptide chain. Since

proteins are different from each other with respect to their composition

and amino acid sequence, they also have distinct acid-base properties.

Information on these properties allows a prediction of the behavior of a

given protein when exposed to an electrical ﬁeld.

Product purification processes 309

The acid-base properties of proteins are exploited in two methods,

electrophoresis and ion exchange chromatography. These are widely used

in the analysis and separation of protein mixtures.

Electrophoresis

Protein molecules are electrically charged at pH values different from their

isoelectric point (pI). As a consequence, they can migrate when exposed to

an electrical ﬁeld, at a rate dependent upon their electrical charge densities.

In the separation of proteins by electrophoresis, the sample is submitted

to an electrical ﬁeld, causing the electrically charged proteins to move in

the direction of the applied current. If the experiment is carried out in

solution (free electrophoresis) and the proteins have different charge

densities, they will move with different velocities, allowing their separa-

tion. In practice, the process is normally carried out in a gel matrix, instead

of in solution, and the gel can act not only as an inert support for the

electrophoresis buffer, but also, if desired, as an active material that

interacts with the proteins.

A very popular electrophoretic technique is SDS-PAGE (sodium dode-

cyl sulfate-polyacrylamide gel electrophoresis), used for separating protein

molecules based on their molar masses, providing high resolution and also

allowing the determination of the protein molar masses. In this method,

the sample is pre-treated with the detergent SDS, which causes the

dissociation of protein subunits and the complete unfolding of the poly-

peptide chain, which interacts with SDS molecules forming elongated rod-

shaped complexes. This provides identical charge/size ratios to all pro-

teins, so that the migration velocity of the denatured proteins becomes

dependent only on the molar mass.

Another version of electrophoresis, that is also very efﬁcient for protein

separation, is isoelectric focusing or electrofocusing. In this technique, a

protein mixture is submitted to an electrical ﬁeld in a gel support, such as

polyacrylamide or agarose, with a gradient of increasing pH established

from anode to cathode. Since the loaded proteins will be positively

charged at pH values below their pI, and negatively charged at pH values

above their pI, a protein will migrate until it ﬁnds a location within the gel

where the pH is the same as its pI (La

s, 1998).

Electrophoretic techniques have a number of practical advantages, such

as high resolution, equipment of relative simplicity and low cost, feasi-

bility for multiple samples, high sensitivity, easy detection of speciﬁc

molecules, and availability of standards giving reasonably well deﬁned

bands. However, as a consequence of its low capacity, electrophoresis is

mostly used for analytical purposes rather than as a preparative technique.

More details on electrophoretic techniques are provided in Chapter 13.

Ion exchange chromatography

Among the chromatographic methods, ion exchange is the most com-

monly used in protein puriﬁcation, due to its simplicity for scale-up, wide

applicability, and low cost compared with other chromatographic meth-

ods. Ion exchange of proteins involves their adsorption onto the charged

310 Animal Cell Technology

groups of a solid support, followed by their elution and concentration in

an aqueous buffer of high ionic strength.

For the effective use of ion exchange for protein puriﬁcation, the

stationary phase should be able to bind to positively or negatively charged

proteins. There are two types of ion exchangers: anion exchangers (posi-

tively charged matrix) and cation exchangers (negatively charged matrix).

Low molar mass counter-ions are associated with the proteins, as well as

with the stationary phase. For protein binding onto the stationary phase,

the counter-ions should be dissociated. The most widely used counter-

ions are Na

and H

in cation exchangers, and Cl

–

and OH

–

in anion

exchangers.

The counter-ions can be arranged according to their intensity of inter-

action with the other ionic groups at the same concentration. Conse-

quently, a chloride would displace hydroxide ions as a counter-ion in ion

exchange. The counter-ions are not permanently bound to an ionic group,

but are in an equilibrium state, where continuous replacements take place.

So, ionic groups can become free to bind a protein. The higher the

counter-ion concentration, the lower will be the probability of the ionic

groups being available for protein binding. Sometimes, before the ion

exchange process, counter-ions should be replaced by others more suitable

for the speciﬁc application.

The support can be cellulose, agarose, dextran, silica, polyacrylate,

polyvinyl, or polystyrene, among other resins. The most suitable matrix

for a given application is often selected based on data provided by the

manufacturers.

The selection of the ionic functional groups of the support is based on

the strength of these groups, as listed in Table 12.2. The most widely used

functional groups are S, C, DEAE, and Q. The functional groups S and Q

are a strong acid and a strong base, respectively. Their pK values are

around 1 and 14, respectively, and thus, they are completely ionized at

practically any pH. The functional groups C and DEAE are a weak acid

Table 12.2 Chemical groups used in ion exchange for protein puriﬁcation

Formula Group

Strong anion exchangers

–CH

(CH

)

Trimethylaminomethyl (TAM)

–C

)

Triethylaminoethyl (TEAE)

–CH

(CH

)

Quaternary amine (Q)

Weak anion exchangers

–C

Aminoethyl (AE)

–C

H(C

)

Diethylaminoethyl (DEAE)

Strong cation exchangers

–SO

–

Sulfonate (S)

–CH

–

Sulfomethyl (SM)

–C

–

Sulfopropyl (SP)

Weak cation exchangers

–COO

–

Carboxy (C)

–CH

COO

–

Carboxymethyl (CM)

From Ladisch (2001), Karlsson et al. (1998).

Produc t purification processes 311

and a weak base, respectively, with pK values around 4 and 10. For the

weak groups, the prevailing pH affects the ionization, and the pK is used

as an indication of the adequate operational pH range. Therefore, to ensure

a suitable ionization, the cation exchanger C is used at a pH above 6, and

the anion exchanger DEAE at a pH below 9.

Besides inﬂuencing the adsorbent ionizing groups, the pH also affects

protein charge and stability. In practice, if the protein is more stable at a

pH below its pI, a cation exchanger is used, and conversely, if it is more

stable at a pH above its pI, an anion exchanger is used.

Some adsorbents present a remarkably high density of ionic groups on

the surface, allowing multiple adsorption points, which require high salt

concentrations to promote elution, which may cause denaturation.

The aqueous buffers used in ion exchange should contribute to the ion/

counter-ion dissociation. The buffer minimizes pH ﬂuctuations, avoiding

protein denaturation. The selection of the most suitable buffer depends on

the type of the ion exchanger, on product stability, and also on the

optimum pH for maximum adsorption. An important requirement is that

the buffer should not interact with the adsorbent and this is the reason

why the selected buffer, when charged, usually has the same charge as the

ion exchanger. Depending on the chromatographic step, adsorption or

elution, the pH should be adjusted to promote protein adsorption or its

displacement from the matrix.

During adsorption, the pH should be adjusted to one unit above or

below the protein pI, since larger differences result in a greater net charge

of the protein, and consequently, multiple adsorption points, requiring

severe conditions for elution. An ideal pH is suitable for adsorption at a

level which allows the elution to be performed only by a small pH change.

The buffer ionic strength determines the degree to which the ionic

groups of both the protein and stationary phase are blocked. During

adsorption, the highest ionic strength that still enables adsorption of the

desired protein is used, whereas during elution, the lowest ionic strength

that promotes its desorption is recommended. If the ionic strength is

excessively low during adsorption, the protein will adsorb very strongly,

making elution difﬁcult. Keeping the ionic strength as high as possible

during adsorption minimizes the adsorption of contaminants, and keeping

it low during the elution minimizes the desorption of contaminants. Such

a strategy simpliﬁes the elution step.

Once the optimum adsorption/desorption conditions are established,

other issues should be considered, such as the need for matrix pretreatment

and the operational mode for adsorption and elution steps. Pretreatment

of the ion exchanger can involve, for example, the removal of ﬁne particles,

swelling, washing, or counter-ion replacement.

Adsorption can be carried out in tanks and in a batch system, as well as

in chromatographic columns. Batch adsorption can be carried out in the

early puriﬁcation step, allowing the processing of large product volumes,

despite having a low efﬁciency. Column adsorption, on the other hand,

presents limitations with regard to the ﬂow rates, but gives better resolu-

tion. There are two distinct methods of elution. The ﬁrst method is that

protein adsorbed on the static ion exchange matrix is completely eluted by

a small volume of a strong eluent. This method is useful for the concentra-

312 Animal Cell Technology

tion of a protein present in a large sample volume. The second method

involves a dynamic ion exchange in which protein separation relies on

relative migration velocities along the column. In this case, there are three

distinct modes of elution:

(i) isocratic, employed for weakly bound proteins, resulting in large

elution volumes;

(ii) in steps, based on discontinuous and sequential changes of pH and/or

salt concentration;

(iii) gradient, based on continuous changes in the eluent composition (pH

and/or ionic strength).

The protein concentration in the eluted fractions collected from a

chromatography column is frequently determined by measuring the absor-

bance in the UV region (usually at 280 nm).

An advantage in ion exchange chromatography is that the matrix can be

regenerated by the removal of bound contaminants and by the reconstitu-

tion of the counter-ions, providing the matrix in the condition required

for a new protein adsorption cycle. If the support matrix is stored wet, it is

susceptible to microbiological degradation, and the use of antimicrobial

agents during storage is recommended.

12.4.4 Separation processes based on differences in hydrophobicity

The surfaces of proteins are mostly hydrophilic. Although the majority of

the hydrophobic residues tend to be buried in the interior of the protein,

some hydrophobic regions are also found on the surface (Voet and Voet,

1995; Ladisch, 2001). The level of surface hydrophobicity differs from one

protein to another, mainly as a consequence of the amino acid composition

and sequence. Difference in surface hydrophobicity is the property

exploited in hydrophobic interaction chromatography (HIC) and reverse

phase chromatography (RPC).

In both chromatographic methods, the adsorbent contains hydrophobic

groups. However, in RPC adsorbent hydrophobicity is much higher than

that in HIC (Eriksson, 1998). Therefore, in HIC the adsorbent is a weak

hydrophobic phase, whereas in RPC it is a dense hydrophobic phase.

Hydrophobic interaction chromatography

The adsorbents used in hydrophobic interaction chromatography are

usually polymer resins, such as reticulated agarose, derivatized with

aliphatic groups such as butyl and octyl, or with aromatic groups, such as

phenyl.

The tendency for protein binding to hydrophobic groups is often

reduced in low salt concentrations. Therefore, to increase the hydrophobic

interaction intensity, high salt concentrations are used during equilibration

and sample application. Due to their effect on strengthening the hydro-

phobic interactions, the most effective salts are those used in salting-out

precipitation, such as ammonium sulfate.

Produc t purific ation processes 313