WO 2007/035283 PCT/US2006/035025
PROTEIN F10CDLATION USING SALTS
CROSS REFERENCE TO RELATED APPLICATIONS
This application claims the benefit of Unites States Provisional Application No.:
60/717,838, filed on September 15,2005, which is incorporated herein by reference in its
entirety.
TECHNICAL FIELD
The invention relates to separation methods, for example, methods of recovering a
purified product from a fluid including impurities such as one or more soluble impurities,
cells, cellular debris, or insoluble impurities.
BACKGROUND
Within the biotechno10gy industry, the purification of proteins on a commercial scale
is an important challenge to the deve10pment of recombinant proteins for therapeutic and
diagnostic purposes. Problems related to yield, purity, and throughput challenge the
manufacturing sector. "With the advent of recombinant protein techno10gy, a protein of
interest can be produced using cultured eukaryotic host cell lines engineered to express a
gene encoding the protein. What can result from a host cell culturing process, however, is a
mixture of the desired protein a10ng with impurities that are either derived from the protein
itself, such as protein variants, or from the host cell, such as host cell proteins, DNA, and
cellular debris. The use of the desired recombinant protein for pharmaceutical applications
may be contingent on being able to reliably recover adequate levels of the protein from these
impurities. Recombinant techno10gy can also produce proteins that are not found in nature,
for example, novel mutant proteins, fusion proteins, or proteins with hetero10gous signal
sequences that direct the secretion of the protein to the medium. Recombinant proteins can
be expressed in many eukaryotic cell types, including Chinese Hamster Ovarian cells (CHO),
baby hamster kidney (BHK), NSO mye10ma cells, and Pichia pastoris yeast cells.
Typically, to produce a recombinant protein, a recombinant DNA vector is created
that contains a gene that codes for the protein to be expressed with appropriate sequences to
direct the transcription and translation of the gene in the desired cell type. The vector can
also contain sequences such as selectable or counterselectable markers, for example, drug
-1-

WO 2007/035283 PCT/US2006/035025
resistance genes, and/or sequences designed to promote the stable retention of the protein
expression sequences. For mammalian cells, plasmid and viral vectors, for example,
retroviral vectors, can be used.
Following the creation of the vector, the vector is then introduced into the cells. The
vector can be transfected as naked DMA using standard methods, for example, lipofection,
calcium phosphate, DEAE-dextran, electroporation, or biolistics (gene gun). Viral vectors
can be introduced by infection with viral particles. The cells are then screened or selected for
those that contain the vector.
Cells that contain the vector and express the recombinant protein can be grown in a
liquid medium or on a solid support, and the protein isolated from the cell culture.
Mammalian cell density ranges between 106 cells/mL to 2x107 cells per mL or more. Most
proteins are secreted. Secreted protein concentrations can range between 4mg/L to 10g/L.
However, if the protein is produced intracellularly, the cells are broken to release the protein,
whereas if the protein is secreted, it can be isolated from the growth medium or the support
following removal of the cells and cell debris. The isolated protein can then be purified.
Conventional biopharmaceutical protein purification methods used to remove cells
and cellular debris include centrifugation, microfiltration, and depth filters. Filter aids, such
as diatomaceous earth, can be used to enhance performance of these steps, but they are not
always effective and sometimes significantly bind the product of interest. Their use may also
require the addition of a solid or a homogeneous suspension that can be challenging as part of
large-sale biopharmaceutical operations.
Polymeric fiocculants can be used to aid in the clarification of mammalian cell
culture process streams, but they can have limitations. For example, protamine sulfate
preparations typically used as processing aids are limited in application due to concerns
about inactivation of the protein of interest or product 10ss due to precipitation (Scopes,
1987). High quality reagent, such as that sold for medical use, can be expensive, hi certain
instances, removal to very low levels may require validation to ensure there are no
unexpected effects in patients. For example, chitosan is not a well-defined reagent and there
are concerns about its consistent performance in routine use in clarification applications.
Multiple charged polymers, such as DEAE dextran, acrylamide-based polymers often used in
waste-water treatment (NALCO Water Handbook, Chapter 8) and polyethylene amine (PEI)
— 2 —

WO 2007/035283 PCT/US2006/035025
have been considered for use in clarification applications. With respect to the latter two types
of polymers, the acrylamide reagents have the potential for contamination with toxic reagents
and polyethylene amine, while a highly effective clarification reagent, is often contaminated
with varying amounts of ethylenimine monomer, a suspected cancer agent (Scawen et al).
Moreover, many of these polymers, including PEI, tend to bind almost irreversibly to many
chromatography resins, thereby limiting downstream processing options. The regulatory and
raw material reuse concerns associated with these polymers have limited their application
primarily to academic studies.
Non-polymer based flocculants, such as alum and iron salts, have been utilized in the
wastewater treatment industry (NALCO Water Handbook). These substances may appear to
be non-useful in processing protein products, because they may bind to the protein product or
may catalyze chemical reactions resulting in modifications of the protein that could affect
safety or efficacy.
SUMMARY
The invention relates to separation methods. The separation methods can be used to
isolate a protein, such as a recombinant protein, from a fluid containing impurities such as
one or more soluble impurities, insoluble impurities, cells, or cellular debris.
In one aspect, the invention features separation methods that include the addition to a
fluid one or more (e.g., two or more) soluble solutions mat can form a precipitate that aids in
the removal of impurities. The precipitate may associate more strongly with impurities and
less strongly to a target product. The solution(s) can include soluble cations, e.g., metal ions
and/or soluble anions that are capable of interacting with, e.g., particulates, colloidal
material, cellular debris or cells, and form an insoluble precipitate, e.g., when mixed together.
The resulting precipitate can be clarified or removed using solid-liquid separation techniques,
such as microfiltration, depth filtration, or centrifugation. The treated fluid can have a
reduced impurity level in comparison to untreated fluid processed similarly.
Impurities may be related to those elements found in suspension within the fluid, ha
some embodiments, the impurities include colloidal material, particulate material, cells, cell
debris such as membrane fragments, and other large cellular complexes that are insoluble
under typical processing conditions. Impurities may also refer to cellular components that
remain soluble under typical processing conditions. DNA, host cell proteins and
-3-

WO 2007/035233 PCT/US2006/035025
phospholipids are examples of cell components that are present in solution during
clarification. Additionally, soluble product-related impurities, such as inactive isoforms or
aggregated species may be present.
Impurity levels can be assessed by a variety of methods. One method, which
provides a measure of the amount of debris in the fluid, is the nepha10metric measurement of
turbidity. Alternatively, the level of debris can be evaluated by measuring the area of
membrane filter required to process a known volume of the fluid. Specific impurities may
also be soluble in the fluid requiring specific biochemical tests to evaluate. DNA levels may
be measured using fiuorometric-based methods, such as by using the commercially available
dye Picogreen (Invitrogen, Product Number P-7581). Alternate approaches include
hybridization methods, such as slot-blot techniques, or polymerase chain reaction (PCR
methods). Host cell protein levels may be evaluated by sodium dodecyl sulfate
polyacrylamide gel electrophoresis (SDS-PAGE), reversed phase chromatography, t>r
enzyme linked immunosorbent assay (ELISA) methods. Phospholipids may be analyzed by
thin layer chromatography or high performance liquid chromatography.
The resulting precipitate can interact with both impurities in suspension and soluble
impurities and this interaction can decrease the levels of these impurities in the purified fluid.
As a result, the separation methods can provide cost, and/or time savings, as well as
increased quality of product, for example, for an industrial process that uses mammalian cell
culture for the production of recombinant proteins.
In another aspect, the invention features an increase in the performance of subsequent
chromatographic steps, for example, the performance of a Protein A chromatographic step.
Protein A chromatography is typically performed by direct application of cell-free
conditioned medium to resin on which Staphylococcus aureus protein A has been
immobilized. The resin is subsequently washed with an aqueous solution of neutral pH
(approximately pH 6-8) and bound protein is often eluted with an acidic buffer. Prior to
subsequent processing, the eluate pool is adjusted to neutral pH. The Protein A eluate pool
often precipitates upon neutralization, especially when high density cell cultures are used for
the load. In accordance with the invention, host cell protein removal by the Protein A step is
greater when the load has been treated with a metal and an anion in comparison to an
untreated fluid. In embodiments, the precipitation of the Protein A peak upon neutralization
— 4 —

WO 2007/035283 PCT/US2006/035025
is often less when the load fluid has been treated with a cation and an anion providing a
processing improvement.
In another aspect, the invention features selection of two soluble agents that, when
mixed, form a solid that may improve process fluid purity with high product recovery under
appropriate conditions. These agents may include, but are not limited to, calcium,
manganese, magnesium, aluminum, cobalt, nickel, carbonate, fluoride, sulfite, phosphate,
silicate and alginates. These compounds represent a combination of multivalent metal ions,
with monovalent anions, or alternatively, multivalent or polyvalent anions, that are the
preferred ligands to the metal. Salts of these cations and anions, when mixed under
appropriate conditions potentially to form complexes that are sparingly soluble, may have
utility in clarification applications.
In another aspect, the invention features a method, including forming a solid having a
first cation and a first anion in a medium including a protein; and separating the solid from
the protein.
In another aspect, the invention features a method, including introducing a first cation
and a first anion into a medium having a protein; precipitating a solid having the first cation
and the first anion; and separating the solid from the protein.
The methods described herein can be used to facilitate the filtration of one or more
impurities from a medium, e.g., a fluid medium (e.g., a turbid suspension). For example,
these methods can be used in a medium having one or more turbidity-causing agents that
render the impurities difficult or inconvenient to remove using conventional filtration
methods. Thus, in another aspect, the invention features a method that includes (i) forming a
solid that includes a first cation and a first anion in a medium (e.g., a fluid medium) that
includes a target moiety (e.g., a moiety to be purified) and one or more turbidity-causing
agents such as precipitated or aggregated host cell proteins, lipids, cellular debris, whole
cells, precipitated DNA, or the precipitate that forms upon the neutralization of the Protein A
peak.and (ii) separating the solid and the turbidity-causing agent(s) from the solution by, e.g.,
filtration. In embodiments, the turbidity causing agent can be of non-cellular origin, such as
colloidal material, particulate material derived from, environmental sources such as sand, dirt,
ground stainless steel fines, or precipitated excipents such as antifoam or urea. The medium
(e.g., a turbid suspension) can have a relatively high turbidity, such as greater than 5NTU as
-5-

WO 2007/035283 PCT/US2006/035025
measured by a tubidity meter, or greater than 100NTU, or greater than 500NTU.. In some
embodiments, the presence of the solid can increase the filter capacity of the medium. In
some embodiments, the turbidity of the treated medium (e.g., the medium after performing
steps (i) and (ii)) can be less than the untreated medium. In some embodiments, the target
moiety can be a protein (e.g., a soluble protein, e.g., an antibody). The method can further
include recovering the target moiety from the solution after filtration.
Embodiments can include one or more of the following features.
The first cation can be calcium, magnesium, strontium, aluminum, scandium,
lanthanum, silicon, titanium, zirconium, thorium, manganese, cobalt, copper, chromium,
iron, nickel, zinc, or vanadium. The first cation can be calcium.
The first anion can be phosphate, carbonate, chromate, tungstate, hydroxide, halide,
succinate, tartrate, citrate, sulfite, molybdate, nitrate, fluoride, silicate, and alginate. The first
anion can be phosphate.
The first cation can be calcium and the first anion can be phosphate.
The solid can have a solubility product constant of no more than about 104M2.
The method can further include introducing from about 4 mM to about 200 mM of
the first cation or the first anion into the medium.
The product of the concentrations of the first cation and the first anion can be greater
than about 10"5M2, 10^M2, or 2.7 x 10-2M2.
The concentrations of the first cation and the first anion in the medium can be
different.
The concentrations of the first cation and the first anion in the medium can be
substantially the same.
The method can further include changing the pH of the medium.
The pH of the medium can be maintained between from about 5 to about 9.
The method can provide separation of at least about 50% of the protein in the
medium. The method can provide separation of at least about 70% of the protein in the
medium.
The method can further include decreasing the clarified turbidity of the clarified
medium by at least about 30% relative to a second clarified medium identical to the medium
and free of the solid. The method can further include decreasing the turbidity of the clarified
-6-

WO 2007/035283 PCT/US2006/035025
medium by at least about 50% relative to a second clarified medium identical to the medium
and free of the solid.
The medium can include cells. The medium can further include mammalian cells.
The medium can further include eukaryotic cells.
The method can further include centrifuging the medium, filtering the medium
through a microfiltration membrane, or filtering the medium through a depth filter.
The solid can further include a second cation species or a second anion.
The medium that includes the protein, after the solid is formed and separated, can be
applied to a Protein A column and eluted to provide an eluted peak having a lower turbidity
than a similarly eluted peak of a second medium identical to the first medium and free of
formation of the solid.
The medium that includes the protein, after the solid is formed and separated, can be
applied to a Protein A column and eluted to provide an eluted peak having a lower soluble
impurity level than an eluted peak of a second medium identical to the medium and free of
formation of the solid.
The first cation and the first anion can be introduced sequentially.
The first cation and the first anion can be introduced simultaneously.
The method can include introducing different concentrations of the first cation and
the first anion into the medium or introducing the same concentration of the first cation and
the first anion into the medium.
The method can further include adjusting the temperature of the medium.
The protein can be a secreted protein. The protein can be an antibody, an antigen-
binding fragment of an antibody, a soluble receptor, a receptor fusion, a cytokine, a growth
factor, an enzyme, or a clotting factor.
In embodiments where the protein is an antibody or a fragment thereof, it can include
at least one, and typically two full-length heavy chains, and/or at least one, and typically two
light chains. Alternatively, the antibodies or fragments thereof can include only an antigea-
binding fragment (e.g., an Fab, F(ab')2, Fv or a single chain Fv fragment). The antibody or
fragment thereof can be a monoc10nal or single specificity antibody. The antibody or
fragment thereof can also be a human, humanized, chimeric, CDR-grafted, or in vitro
generated antibody. In yet other embodiments, the antibody has a heavy chain constant
— 7 —

WO 2007/035283 PCT/US2006/035025
region chosen from, e.g., IgGl, IgG2, IgG3, or IgG4. In another embodiment, the antibody
has a light chain chosen from, e.g., kappa or lambda. In one embodiment, the constant region
is altered, e.g., mutated, to modify the properties of the antibody (e.g., to increase or decrease
one or more of: Fc receptor binding, antibody glycosylation, the number of cysteine
residues, effectoT cell function, or complement function). Typically, the antibody or
fragment thereof specifically binds to a predetermined antigen, e.g., an antigen associated
with a disorder, e.g., a neurodegenerative, metabolic, inflammatory, autoimmune and/or a
malignant disorder. Exemplary antibodies that can be separated by the methods of the
invention include, but are not limited to, antibodies against an Aj3 peptide, interleukin-13 (JL-
13), interleukin-22 (IL-22), 5T4, and growth and differentiation factor-8 (GDF-8).
Other aspects, features and advantages will be apparent from the description of the
preferred implementations thereof and from the claims.
DESCRIPTION OF DRAWING
FIG. 1 is a flowchart of an embodiment of a separation method.
FIG 2 is a graphical representation that shows the effect of scale and mixing method
on flocculation.
FIG. 3 is a graphical representation that shows the effect of mixing speed on
flocculation.
FIG. 4 is a graphical summary of five pilot scale flocculation experiments.
FIG. 5 is a graphical summary that shows changes in % antibody recovery over time
for each of the five pilot scale flocculation experiments summarized in FIG.. 4.
DETAILED DESCRIPTION
Referring to the FIG. 1, a method 30 for separating a targeted protein, such as a
recombinant protein, is shown. Method 30 includes adding soluble salts (such as a calcium-
containing salt and a phosphate-containing salt) to a fluid containing the protein (step 32),
and impurities (which can include, but are not limited to, cellular debris, cells, DNA, host
cell protein, and product related impurities such as inactive isoforms or aggregated species).
The salt solutions can contain buffering agents to mirurnize pH changes or to optimize pH in
the fluid upon mixing the salt solutions. Upon contact (e.g., mixing), the soluble salts often
-8-

WO 2007/035283 PCT/US2006/035025
begin reacting to form an insoluble precipitate (such as solid calcium phosphate) that may
settle in the medium. As shown in the FIG. 1, the medium may subsequently be titrated to a
predetermined pH and the temperature adjusted (step 34), as necessary, to enhance
precipitation. The settling precipitate also settles the cells and the debris, and potentially
other impurities, while the target protein remains essentially soluble in the fluid. The fluid
suspension is incubated under appropriate conditions for a target duration in order to promote
precipitation and optimize clarification while maintaining high levels of target protein
recovery (Step 36). Subsequently, the precipitate is separated from the fluid containing the
target protein (step 38). This operation can occur in a variety of ways including gravity
settling, centrifugation or filtration where filtration alternatives include tangential flow
filtration, depth filtration, filtration through charged media, pad filtration where
diatomaceous earth is a component of the media. The amount of cation and anion added to
the media may not be sufficient to allow gravity settling of the precipitation or of debris but
may still facilitate filtration by acting as a filter aid. The solid-liquid separation may include
processing by a series of the aforementioned options, typically culminating with passage of
the fluid through a filter with a low nominal pore-size rating (such as 0.45, 0.2 or 0.1 uM),
which may be considered sterilizing in grade. The fluid clarified by the flocculation methods
described herein may require less filtration area, either as part of the primary clarification or
after an initial solid-liquid separation step such as centrifugation. Moreover, the turbidity of
post pad and post-sterilizing grade filter can be significantly reduced compared to a non-
f10cculated control.
Typically, subsequent purification can proceed through a series of chromatographic
steps, though other purification methods, such as crystallization and precipitation, can be
conceived. After clarification using the described method, performance of the first
chromatographic step, winch for antibodies is often a Protein A column, may be enhanced.
The overall removal of host cell-derived impurities, including host cell protein, can be
greater than when flocculation treatment is not conducted. In many instances, neutralized
Protein A eluate pool peak also has less precipitation as compared to the untreated control.
This decreased level of precipitation may require less filter area or may reduce the needed
processing time. The decreased level of precipitation also indicates the removal of an
undesired impurity.
-9-

WO 2007/035283 PCT/US2006/035025
Without wishing to be bound by theory, it is believed that the insoluble precipitate
enhances separation of the protein by selectively associating with the cells, cellular debris,
and other process stream impurities while not significantly interacting with the protein. A
non-limiting example of the hypothesized impurity removal process is shown in equation (1)
below:

Referring to equation (1), M* is a soluble [i.e., "(s)"] cation, and A- is soluble anion that can
interact with M4' to form an insoluble salt or complex; the shaded and filled circles each
represent a soluble or insoluble impurity; 1 is an insoluble precipitate that includes the
cation-anion salt or complex and impurities associated therewith (1 is sometimes referred to
herein as the "final complex or salt" or "solid having a first cation and a first anion"); and the
downward pointing arrow indicates that 1 is in precipitated form. For example, calcium
phosphate can interact ionically and/or by chelation with DNA, host cell protein, and cellular
debris, while not significantly interacting with the target protein. This selectivity allows the
protein to remain in the supernatant, and subsequently, to be readily separated from the
medium and other components of the conditioned medium.
. Still referring to FIG 1, separation method 30 includes introducing a first soluble salt
and a second soluble salt into a medium (step 32). The medium can be, for example, a
conditioned aqueous solution in which a recombinant protein has been formed. The first
soluble salt includes a first cation, and the second soluble salt includes a first anion. Upon
contact, the first cation and the first anion are capable of interacting in the medium and may
begin to form an insoluble precipitate. The fluid environment may be adjusted at any time in -
pH or temperature to optimize precipitation conditions (e.g., step 34) and the solution is
incubated for a target duration to allow the system to fully equilibrate (step 36).
In general, any cation (e.g., a first cation)/anion (e.g., a first anion) combination can
be selected that is capable of forming a relatively insoluble salt or complex in the fluid
-10-

WO 2007/035283 PCT/US2006/035025
containing the product of interest. In embodiments, such salts or complexes can be
identified as being insoluble, sparingly soluble, practically insoluble, very slightly soluble, or
slightly soluble in a solution comparable to the fluid containing the product of interest.
Exemplary cation/anion combinations include those in which the selected cation (e.g., a first
cation, e.g., M*" in equation 2 below) and the selected anion (e.g., a first anion, e.g., A" in
equation 2 below) are capable of forming a salt or complex (e.g., MA in equation 2 below)
that is relatively insoluble in water:

in which "(s)" and the downward pointing arrow are as defined with respect to equation (1).
In many instances, characterization of solubility in water as described in The Merck Index or
the Handbook of Physics and Chemistry or other similar references serves as an appropriate
indicator of potential performance.
In some embodiments, the selected cation and the selected anion are capable of
forming a salt or complex having a solubility product constant (KsP) in water of from about 1
x 10"4 M2 to about 1 x 10~50 M2 (e.g., from about 1 x 10"5 M2 to about 1 x 10"50 M2, from
about 1 x 10"6 M2 to about 1 x 10"50 M2, from about 1 x 10"4 Mz to about 1 x 10"40 M2). In
some embodiments, exemplary cations and anions can be identified as a solubility product
constant (Ksp) between the first cation and the first anion ([cation] x [anion]) of less than
about 10^M2, foi example, and preferably below about 10"5M2 or 10^M2. Substances with
Ksp values of less than 104M2 can be utilized in the methods since these substances, upon
mixing the a cation and the anion can result in a final solution of at most 10mM of each;
addition of the cation or anion in excess of 10mM can result in precipitation and subsequent
flocculation. Substances with higher Ksp values than those listed above can be used,
however, an excessive amount of cation and anion can be needed to form the solid.
The cation (e.g., a first cation) can be an alkaline earth metal, a transition metal, or a
main group element. These elements can be classified into hard acids, borderline acids, or
soft acids. Examples of first cations include calcium (Ca2+), magnesium (Mg2*), strontium
(Sr2+), aluminum (Al34), copper (Cu(I) or Cu(II)), scandium (Sc+3), lanthanum (La+3), silicon
-11-

WO 2007/035283 PCT/US2006/035025
(Si4+), titanium (Ti(III) or Ti(IV)), thorium, zirconium, manganese (Mn(II) or Mn(III)),
cobalt (Co(II) or Co(III)), chiomium (Cr(II) or Cr(III)), iron (Fe(D) or Fe(III)), nickel (Ni2+),
zinc (Zn24), and vanadium (V(III), V(IV), or V(V)). These represent the hard and borderline
acids. The first anion can be an atomic species or a molecular species. In some
embodiments, the first cation can be Ca2+, Nig2"1", Mn(II), Co(II), or Ni2+. In certain
embodiments, the first cation can be Ca2+.
Examples of first anions include anions that are the preferred ligands to the metal ion
used. For the hard acid and borderline acids, the anions may include fluoride, phosphate,
carbonate, silicate, chromate, tungstate, hydroxide, sulfite, nitrate, molybdate, succinate,
tartrate, and citrate, and to some extent sulfates and perch10rates (see Aquatic Chemistry,
Editors W. Stumm and JS Morgan, J Wiley, (1981), p 343; and R.G. Pearson, J. Amer. Chem.
Soc, v 85, p 3533 (1963).) In some embodiments, the first anion can be phosphate, sulfite,
carbonate, fluoride, molybdate, or silicate. In certain embodiments, the first anion can be
phosphate.
In some embodiments, the first cation can be Ca2+, and the first anion can be
phosphate, sulfite, carbonate, fluoride, molybdate, or silicate. In certain embodiments, the
first cation can be Ca2+, and the first anion can be phosphate.
In some embodiments, the first cation can be Mg2*, Mn(II), Co(II), or Ni2+, and the
first anion can be phosphate, carbonate, or fluoride.
The initial concentration for each of the first cation and the first anion (i.e., the
concentration of the first cation and the first anion that is introduced into the medium (before
the start of precipitation)) can range from about 2 millimolar to about 200 millimolar (e.g.,
from about 3 millimolar to about 200 millimolar, from about 4 millimolar to about 200 '
millimolar, from about 5 millimolar to about 200 millimolar, from about 4 millimolar to
about 100 millimolar, from about 4 millimolar to about 50 millimolar, from about 4
millimolar to about 40 millimolar, from about 4 millimolar to about 30 millimolar, from
about 4 millimolar to about 10 millimolar, from about 10 millimolar to about 80 millimolar,
from about 10 millimolar to about 40 millimolar, from about 10 millimolar to about 30
millimolar, from about 20 millimolar to about 80 millimolar, from about 20 millimolar to
about 40 millimolar, e.g., about 4 millimolar, about 6 millimolar, about 10 millimolar, about
13.3 millimolar, about 16 millimolar, about 20 millimolar, about 24 millimolar, about 30
-12-

WO 2007/035283 PCT/US2006/035025
millimolar, about 33.3 millimolar, about 40 millimolar, about 50 millimolar, or about 80
millimolar) depending upon the solubility of the final complex. In certain embodiments, the
concentration of the first cation introduced into the medium (before the start of precipitation)
can be about 6 millimolar, about 10 millimolar, about 20 millimolar, about 24 millimolar,
about 30 millimolar, about 40 millimolar, about 50 millimolar, or about 80 millimolar. In
certain embodiments, the concentration of the first anion introduced into the medium (before
the start of precipitation) can be 4 millimolar, about 10 millimolar, about 13.3 millimolar,
about 16 millimolar, about 20 millimolar, about 24 millimolar, about 30 millimolar, about 40
millimolar, about 50 millimolar, or about 80 millimolar. For example, the concentration of
the first cation introduced into the medium (before the start of precipitation) can be about 30
millimolar, and the concentration of the first anion introduced into the medium (before the
start of precipitation) can be about 20 millimolar. As another example, the concentration of
the first cation introduced into the medium (before the start of precipitation) can be about 24
millimolar, and the concentration of the first anion introduced into the medium (before the
start of precipitation) can be about 16 millimolar.
In some embodiments, the product of the above-described initial concentrations of the
first cation and the first anion can be from about 4 x 10"6 M2 to about 4 x 10"2 M2 (e.g., from
about 1.6 x 10"5 M2 to about 4 x 10"2 M2, from about 2.5 x 10"5 M2 to about 4 x 10"2 M2, from
about 1.6x10"5 M2 to about 6.4 x 10"3 M2, or from about 2.5 x 10'5 M2 to about 6.4 x 10"3
M2). In some embodiments, the product of the above-described initial concentrations of the
first cation and the first anion can be greater than about lx10"5M2, greater than about 2x10"
5M2, greater than about 1x10, greater than about 2x102, greater than about 10x10"
M2, or greater than about 2.7 x10"2M2. In some embodiments, these concentrations can
result in significant precipitation of the insoluble salt together with the impurities.
Concentrations that are too high may result in solid volume of greater than about 10% of the
total fluid volume. Concentrations in excess of 500mM cation or anion with low solubility
constants may give large solids volume. Adequate product recovery from a large solid
volume may be difficult using standard solid-liquid separation techniques. Examples of
precipitates include calcium phosphate, calcium sulfite, calcium carbonate, calcium fluoride,
calcium silicate, calcium molybdate, magnesium carbonate, magnesium phosphate,
-13-

WO 2007/035283 PCT/US2006/035025
magnesium fluoride, manganese phosphate, manganese carbonate, cobalt phosphate, nickel
phosphate, and nickel carbonate.
In some embodiments, the KsP of the final salt or complex in the fluid media (e.g., 1
in equation (1), i.e., the insoluble precipitate that includes the cation-anion salt or complex
and impurities associated therewith) can be from about 1 x 104 M2 to about 1 x 10'50 M2
(e.g., from about 1 x 1(T5 M2 to about 1 x 10"50 M2, from about 1 x 1(T6 M2 to about 1 x 10"50
M2, from about 1 x 10"4 M2 to about 1 x 10"40 M2). In some embodiments, the Ksp of the final
salt or complex in the fluid media can be less than about 10^M2, for example, and preferably
less than about 10"5M2 or 10^M2. For example, in the anti-IL13 #2 sample with 40mM
calcium and 20mM phosphate in Table 1, the supernatant after centrifugation (i.e., after
precipitation) contained 8.19mM calcium and 1.04mM phosphate. This level of soluble
calcium and phosphate corresponds to a Ksp of 8.5x10"6 M2. In the anti-AB #1 sample with
80mM calcium and 20mM phosphate in Table 1, the supernatant after centrifugation
contained 22.2mM calcium and 0.4mM phosphate. This level of soluble calcium and
phosphate corresponds to a KsP of 8.4x10"6 M2.
In some embodiments, the K^ of the final salt or complex in the fluid media (e.g., 1
in equation (1)) can be different (e.g., greater than) the Ksp of the cation-anion salt or
complex itself (i.e., no associated impurities, e.g., 2 in equation (2)) in water. For example,
referring to equations (1) and (2), the KsP of, e.g., 1 (e.g., MA = calcium phosphate) in the
fluid can be different from (e.g., greater than) the KsP of calcium phosphate itself in water
(e.g., 2 in equation (2) in which MA = calcium phosphate).
Other implementations of forming the precipitate can be performed. For example, the
first cation and the first anion can be introduced into the medium substantially
simultaneously or sequentially. In implementations in which the first cation is calcium and
the first anion is sulfite, introducing the sulfite into the medium before introducing the
calcium into the medium can enhance the precipitation of impurities (Example 2). The
concentrations of the first cation and the first anion can be substantially the same or different
(cation-rich or anion-rich). For example, the concentration of an ion can be about 1.5 times,
about two times, about three times, about four times, or about five times greater than the
concentration of the other ion. In some implementations, more than one cation and/or more
than one anion are introduced into the medium. The total concentration of the cations can
-14-

WO 2007/035283 PCT/US2006/035025
range from about 5 millimolar to about 200 millimolar, as can the total concentration of the
anions. Where polymeric solutions are used, the concentration, in mM, may be substantially
lower based on the polymer molecular weight; the concentration may instead depend upon
the monomer molecular weight. In other implementations, only one or more cation or only
one or more anion is introduced into the medium. For example, when the medium containing
the protein already includes anion(s) or cation(s) capable of forming a precipitate, then the
appropriate cation(s) or anion(s), respectively, can be added to form the precipitate.
Alternatively, ions can be added to react with the ions already in the medium to form a first
precipitate, and additional cation/anion combination(s) can be added into the medium to form
other precipitate(s).
As shown in FIG. 1, the methods can optionally include titrating the medium to an
appropriate pH and/or adjusting the temperature (step 34).
In some embodiments, the pH of the medium may be adjusted to a predetermined pH
to increase precipitation (step 34). The pH of the medium can be increased or decreased, for
example, by titrating with a base (e.g., NaOH) or an acid, such as phosphoric acid or
hydroch10ric acid. The predetermined pH can be a function of, for example, the cation(s) and
anion(s) in the medium, other materials in the medium, and/or the composition of the
medium. The predetermined pH can range from about five to about nine, for example, from
about 6.5 to about 9.
In other embodiments, the pH of the medium is not adjusted.
In some embodiments, the medium may be heated or chilled to optimize performance
(step 34). As with adjusting the pH, the temperature and time for which the medium is
heated or incubated can be a function of, for example, the cation(s) and anion(s) in the
medium, other materials in the medium, and/or the composition of the medium.
The medium can be incubated at room temperature or heated up, for example, to
about 37°C. The incubation or heating period (step 36) can range from about one hour to
about twelve hours. While the medium is incubating or heating, the medium can be mixed
(e.g., at low speeds to reduce shearing of the materials), or the medium can be mixed for an
initial period of time and allowed to sit unmixed so that the precipitate can settle, which
allows the protein-containing supernatant to be easily separated.
-15-

WO 2007/035283 PCT/US2006/035025
In other embodiments, the medium is not heated, for example, if precipitation is
sufficient to provide good separation.
Next, the medium is centrifuged (step 38) to help separate the precipitate from the
supernatant, which reduces the turbidity of the supernatant. Other methods of solids removal
are possible, such as depth filtration or microfiltration. As illustrated below in the examples,
the turbidity of the medium can be reduced by at least about 30% relative to an untreated
control solution in which no flocculation occurred. In some implementations, the turbidity of
the medium is reduced by at least by about 50%, at least about 80%, at least about 90%, at
least about 95%, or at least about 98% or higher. As used herein, turbidity is measured using
a nepha10meter (such as those made by HACH, lowland, CO) according to standard
procedures.
After primary clarification such as through centrifugation, additional debris removal
can occur through the use of filtration (step 38). As illustrated below in the examples, the
separation methods described herein can provide high yields, with a protein recovery of at
least about 50% (such as at least about 60%, at least about 70%, at least about 80%, or at
least about 90%). As used herein, recovery is calculated as the mass of the target protein in
the post-treated pool to the mass in the pre-treatment pool. The mass of product is the
product of target protein concentration and volume where concentration can be determined
by a variety of methods, such as high performance liquid chromatography assays. For target
proteins that are antibodies, concentration can often be determined using protein A-based
analysis methods. Those skilled in the art of biopharmaceutical cell culture, purification or
protein characterization methodo10gies can identify suitable assay methods.
The protein can be subsequently purified, according to conventional methods.
Proteins or Polypeptides
The present invention relates to the separation of proteins, e.g., soluble or secreted
proteins, from a fluid. The term "protein" as used herein refers to one or more polypeptides
that can function as a unit. The term "polypeptide" as used herein refers a sequential chain of
arnino acids linked together via peptide bonds. The term "polypeptide" is used to refer to an
amino acid chain of any length, but one of ordinary skill in the art will understand that the
term is not limited to lengthy chains and can refer to a minimal chain comprising two amino
-16-

WO 2007/035283 PCT/US2006/035025
acids linked together via a peptide bond. If a single polypeptide can function as a unit, the
terms "polypeptide" and "protein" maybe used interchangeably.
In certain embodiments, the proteins are produced recombinantly. The terms
"recombinantly expressed protein" and "recombinant protein" as used herein refer to a
polypeptide expressed from a host cell that has been manipulated by the hand of man to
express that polypeptide. In certain embodiments, the host cell is a mammalian cell. In
certain embodiments, this manipulation may comprise one or more genetic modifications.
For example, the host cells may be genetically modified by the introduction of one or more
hetero10gous genes encoding the polypeptide to be expressed. The hetero10gous
recombinantly expressed polypeptide can be identical or similar to polypeptides that are
normally expressed in the host cell. The hetero10gous recombinantly expressed polypeptide
can also be foreign to the host cell, e.g., hetero10gous to polypeptides normally expressed in
the host cell. In certain.embodiments, the hetero10gous recombinantly expressed polypeptide
is chimeric. For example, portions of a polypeptide may contain amino acid sequences that
are identical or similar to polypeptides normally expressed in the host cell, while other
portions contain amino acid sequences that are foreign to the host cell. Additionally or
alternatively, a polypeptide may contain amino acid sequences from two or more different
polypeptides that are both normally expressed in the host cell. Furthermore, a polypeptide
may contain amino acid sequences from two or more polypeptides that are both foreign to the
host cell. In some embodiments, the host cell is genetically modified by the activation or
upregulation of one or more endogenous genes.
Any protein that may desirably be separated in accordance with the present invention
will often be selected on the basis of an interesting or useful bio10gical or chemical activity.
For example, the present invention may be emp10yed to separate any pharmaceutically or
commercially relevant antibody, receptor, cytokine, growth factor, enzyme, clotting factor,
hormone, regulatory factor, antigen, binding agent, among others. The following list of
proteins that can be separated according to the present invention is merely exemplary in
nature, and is not intended to be a limiting recitation. One of ordinary skill in the art will
understand that any protein may be expressed in accordance with the present invention and
will be able to select the particular protein to be produced based as needed.
-17-

WO 2007/035283 PCT/US2006/035025
Afitibodies and Binding Fragments
Antibodies, also known as immunog10bulins, are typically tetrameric glycosylated
proteins composed of two light (L) chains of approximately 25 kDa each and two heavy (H)
chains of approximately 50 kDa each. Two types of light chain, termed lambda and kappa,
may be found in antibodies. Depending on the amino acid sequence of the constant domain
of heavy chains, immunog10bulins can be assigned to five major classes: A, D, E, G, and M,
and several of these may be further divided into subclasses (isotypes), e.g., IgGi, IgG2, IgG3,
IgG*, IgAi, and IgA2- Each light chain includes an N-terminal variable (V) domain (VL) and
a constant (C) domain (CL). Each heavy chain includes an N-terminal V domain (VH), three
or four C domains (CHs), and a hinge region. The CH domain most proximal to VH is
designated as CHI. The VH and VL domains consist of four regions of relatively conserved
sequences called framework regions (FR1, FR2, FR3, and FR4), which form a scaffold for
three regions of hypervariable sequences (complementarity determining regions, CDRs).
The CDRs contain most of the residues responsible for specific interactions of the antibody
with the antigen. CDRs are referred to as CDR1, CDR2, and CDR3. Accordingly, CDR
constituents on the heavy chain are referred to as HI, H2, and H3, while CDR constituents on
the light chain are referred to as LI, L2, and L3. CDR3 is typically the greatest source of
molecular diversity within the antibody-binding site. H3, for example, can be as short as two
amino acid residues or greater than 26 amino acids. The subunit structures and three-
dimensional configurations of different classes of immunog10bulins are well known in the
art. For a review of the antibody structure, see Antibodies: A Laboratory Manual, Cold
Spring Harbor Laboratory, eds. Harlow et al., 1988. One of skill in the art will recognize
that each subunit structure, e.g., a CH, VH, CL, VL, CDR, FR structure, comprises active
fragments, e.g., the portion of the VH, VL, or CDR subunit the binds to the antigen, i.e., the
antigen-binding fragment, or, e.g., the portion of the CH subunit mat binds to and/or
activates, e.g., an Fc receptor and/or complement. The CDRs typically refer to the Kabat
CDRs, as described in Sequences of Proteins ofImmuno10gical Interest, US Department of
Health and Human Services (1991), eds. Kabat et al. Another standard for characterizing the
antigen binding site is to refer to the hypervariable 10ops as described by Chothia. See, e.g.,
Chothia, D. et al. (1992) J. Mol. Biol 227:799-817; and Tomlinson et al. (1995) EMBO J.
14:4628-4638. Still another standard is the AbM definition used by Oxford Molecular's
-18-

WO 2007/035283 PCT/US2006/035025
AbM antibody modelling software. See, generally, e.g., Protein Sequence and Structure
Analysis of Antibody Variable Domains. In: Antibody Engineering Lab Manual (EcL: Duebel,
S. and Kontermann, R., Springer-Verlag, Heidelberg). Embodiments described with respect
to Kabat CDRs can alternatively be implemented using similar described relationships with
respect to Chothia hypervariable 10ops or to the AbM-defined 10ops.
As used herein, the term "antibody" includes a protein comprising at least one, and
typically two, VH domains or portions thereof, and/or at least one, and typically two, VL
domains or portions thereof. In one embodiment, the antibody is a tetramer of two heavy
immunog10bulin chains and two light immunog10bulin chains, wherein the heavy and light
immunog10bulin chains are inter-connected by, e.g., disulfide bonds. The antibodies, or
portions thereof, can be obtained from any origin, including, but not limited to, rodent,
primate (e.g., human and non-human primate), camelid (e.g., camel or llama), as well as
recombinantly produced, e.g., chimeric, humanized, and/or in vitro generated, as described in
more detail herein.
Examples of binding fragments encompassed within the term "antigen-binding
fragment' of an antibody include (i) a Fab fragment, a monovalent fragment consisting of the
VL, VH, CL and CHI domains; (ii) a F(ab')2 fragment, a bivalent fragment comprising two
Fab fragments linked by a disulfide bridge at the hinge region; (iii) a Fd fragment consisting
of the VH and CHI domains; (iv) a Fv fragment consisting of the VL and VH domains of a
single arm of an antibody, (v) a dAb fragment, which consists of a VH domain; (vi) a
camelid or camelized variable domain; (vii) a single chain Fv (scFv); and (viii) a bispecific
antibody. Furthermore, although the two domains of the Fv fragment, VL and VH, are coded
for by separate genes, they can be joined, using recombinant methods, by a synthetic linker
that enables them to be made as a single protein chain in which the VL and VH regions pair
to form monovalent molecules (known as single chain Fv (scFv); see, e.g., Bird et al. (1988)
Science 242:423-26; Huston et al. (1988) Proc. Natl. Acad. Sci. U.S.A. 85:5879-83). Such
single chain antibodies are also intended to be encompassed within the term "antigen-binding
fragment" of an antibody. These fragments are obtained using conventional techniques
known to those skilled in the art, and the fragments are evaluated for function in the same
manner as are intact antibodies.
-19-

WO 2007/035283 PCT/US2006/035025
The antigen-binding fragment can, optionally, include a moiety that enhances one or
more of, e.g., stability, effector cell function or complement fixation. For example, the
antigen binding fragment can include a pegylated moiety, albumin, or a heavy and/or a light
chain constant region (or a portion thereof).
Other than "bispecific" or "bifunctional" antibodies, an antibody is understood to
have each of its binding sites identical. A "bispecific" or "bifunctional antibody," or an
antigen-binding fragment thereof, is an artificial hybrid antibody or fragment thereof having
two different antigen-binding sites. Bispecific antibodies, or antigen-binding fragments
thereof, can be produced by a variety of methods including fusion of hybridomas, linking of
Fab1 fragments, or recombinantly. See, e.g., Songsivilai & Lachmann, Clin. Exp. Immunol.
79:315-321 (1990); Kostelny et a/., /. Immunol. 148,1547-1553 (1992).
Numerous methods known to those skilled in the art are available for obtaining
antibodies or antigen-binding fragments thereof. For example, monoc10nal antibodies may
be produced by generation of hybridomas in accordance with known methods. Hybridomas
formed in this manner are then screened using standard methods, such as enzyme-linked
irnmunosorbent assay (ELISA) and surface plasmon resonance (Biacore™) analysis, to
identify one or more hybridomas that produce an antibody that specifically binds with a
specified antigen. Any form of the specified antigen may be used as the immunogen, e.g.,
recombinant antigen, naturally occurring forms, any variants or fragments thereof, as well as
antigenic peptide thereof.
One exemplary method of making antibodies includes screening protein expression
libraries, e.g., phage or ribosome display libraries. Phage display is described, for example,
in Ladner et al, U.S. Patent No. 5,223,409; Smith (1985) Science 228:1315-1317; WO
92/18619; WO 91/17271; WO 92/20791; WO 92/15679; WO 93/01288; WO 92/01047; WO
92/09690; and WO 90/02809.
In addition to the use of display libraries, the specified antigen can be used to
immunize a non-human animal, e.g., a rodent, e.g., a mouse, hamster, or rat. In one
embodiment, the non-human animal includes at least a part of a human immunog10bulin
gene. For example, it is possible to engineer mouse strains deficient in mouse antibody
production with large fragments of the human Ig 10ci. Using the hybridoma techno10gy,
antigen-specific monoc10nal antibodies derived from the genes with the desired specificity
-20-

WO 2007/035283 PCT/US2006/035025
maybe produced and selected. See, e.g., XENOMOUSE™, Green et al. (1994) Nature
Genetics 7:13-21, US 2003-0070185, WO 96/34096, published Oct. 31,1996, and PCT
Application No. PCT/US96/05928, filed Apr. 29,1996.
In another embodiment, a monoc10nal antibody is obtained from the non-human
animal, and then modified, e.g., humanized, deimmunized, chimeric, may be produced using
recombinant DNA techniques known in the art. A variety of approaches for making chimeric
antibodies have been described. See e.g., Morrison et al, Proc. Natl. Acad. Sci. U.S.A.
81:6851,1985; Takeda et al, Nature 314:452,1985, Cabffly et al, U.S. Patent No.
4,816,567; Boss et al, U.S. Patent No. 4,816,397; Tanaguchi et al, European Patent
Publication EP171496; European Patent Publication 0173494, United Kingdom Patent GB
2177096B. Humanized antibodies may also be produced, for example, using transgenic mice
that express human heavy and light chain genes, but are incapable of expressing the
endogenous mouse irnmunog10bulin heavy and light chain genes. Winter describes an
exemplary CDR-grafting method that may be used to prepare the humanized antibodies
described herein (U.S. Patent No. 5,225,539). All of the CDRs of a particular human
antibody may be replaced with at least a portion of a non-human CDR, or only some of the
CDRs may be replaced with non-human CDRs. It is only necessary to replace the number of
CDRs required for binding of the humanized antibody to a predetermined antigen.
Humanized antibodies or fragments thereof can be generated by replacing sequences
of the Fv variable domain that are not directly involved in antigen binding with equivalent
sequences from human Fv variable domains. Exemplary methods for generating humanized
antibodies or fragments thereof are provided by Morrison (1985) Science 229:1202-1207; by
Oi et al. (1986) BioTechniques 4:214; and by US 5,585,089; US 5,693,761; US 5,693,762;
US 5,859,205; and US 6,407,213. Those methods include isolating, manipulating, and
expressing the nucleic acid sequences that encode all or part of irnmunog10bulin Fv variable
domains from at least one of a heavy or light chain. Such nucleic acids may be obtained
from a hybridoma producing an antibody against a predetermined target, as described above,
as well as from other sources. The recombinant DNA encoding the humanized antibody
molecule can then be c10ned into an appropriate expression vector.
In certain embodiments, a humanized antibody is optimized by the introduction of
conservative substitutions, consensus sequence substitutions, germline substitutions and/or
-21-

WO 2007/035283 PCT/US2006/035025
backmutations. Such altered immunog10bulin molecules can be made by any of several
techniques known in the art, (e.g., Teng et al., Proc. Natl. Acad. Sci. U.S.A., 80:7308-7312,
1983; Kozbor et al, Immuno10gy Today, 4: 7279,1983; Olsson et al, Meth. EnzymoL, 92: 3-
16,1982), and may be made according to the teachings of PCT Publication WO92/06193 or
EP 0239400).
An antibody or fragment thereof may also be modified by specific deletion of human
T cell epitopes or "deimmunization" by the methods disc10sed in WO 98/52976 and WO
00/34317. Briefly, the heavy and light chain variable domains of an antibody can be
analyzed for peptides that bind to MHC Class II; these peptides represent potential T-cell
epitopes (as defined in WO 98/52976 and WO 00/34317). For detection of potential T-cell
epitopes, a computer modeling approach termed "peptide threading" can be applied, and in
addition a database of human MHC class II binding peptides can be searched for motifs
present in the VH and VL sequences, as described in WO 98/52976 and WO 00/34317. These
motifs bind to any of the 18 major MHC class IIDR allotypes, and thus constitute potential T
cell epitopes. Potential T-cell epitopes detected can be eliminated by substituting small
numbers of amino acid residues in the variable domains, or preferably, by single amino acid
substitutions. Typically, conservative substitutions are made. Often, but not exclusively, an
amino acid common to a position in human germline antibody sequences may be used.
Human germline sequences, e.g., are disc10sed in Tomlinson, et al. (1992) J. Mol. Biol.
227:776-798; Cook, G P. et al. (1995) Immunol. Today Vol. 16 (5): 237-242; Chothia, D. et
al. (1992) /. Mol. Biol. 227:799-817; and Tomlinson et al. (1995) EMBOJ. 14:4628-4638.
The V BASE directory provides a comprehensive directory of human immunog10bulin
variable region sequences (compiled by Tomlinson, LA. et al. MRC Centre for Protein
Engineering, Cambridge, UK). These sequences can be used as a source of human sequence,
e.g., for framework regions and CDRs. Consensus human framework regions can also be
used, e.g., as described in US 6,300,064.
In certain embodiments, an antibody can contain an altered immunog10bulin constant
or Fc region. For example, an antibody produced in accordance with the teachings herein
may bind more strongly or with more specificity to effector molecules such as complement
and/or Fc receptors, which can control several immune functions of the antibody such as
effector cell activity, lysis, complement-mediated activity, antibody clearance, and antibody
-22-

WO 2007/035283 PCT/US2006/035025
half-life. Typical Fc receptors that bind to an Fc region of an antibody (e.g., an IgG
antibody) include, but are not limited to, receptors of the FcyW, FcyRII, and FcyRIII and
FcRn subclasses, including allelic variants and alternatively spliced forms of these receptors.
Fc receptors are reviewed in Ravetch and Kinet, Annu. Rev. Immunol 9:457-92,1991; Capel
et al, Immunomethods 4:25-34,1994; and de Haas et al, J. Lab. Clin. Med. 126:330-41,
1995).
Non-limiting examples of antibodies that can be separated by the methods of the
invention, include but are not limited to, antibodies against AP, IL-13, IL-22, GDF8 and 5T4.
Each of these antibodies is described in more detail hereinbelow and the appended Examples.
Anti-Afi Antibodies
As described in the appended Examples, anti-AB antibodies can be separated by the
methods of the invention. The terms "AB antibody," "Ap antibody," "anti-Ap antibody,"
and "anti-AP" are used interchangeably herein to refer to an antibody that binds to one or
more epitopes or antigenic determinants of APP, AP protein, or both. Exemplary epitopes or
antigenic determinants can be found within the human amy10id precursor protein (APP), but
are preferably found within the Ap peptide of APP. Multiple isoforms of APP exist, for
example APP695, APP751, and APP770. Arnino acids within APP are assigned numbers
according to the sequence of the APP770 isoform (see e.g., GenBank Accession No. P05067).
AP (also referred to herein as beta amy10id peptide and A beta) peptide is a ~4-kDa internal
fragment of 39-43 amino acids of APP (Ap39, Ap40, AP41, Ap42, and Ap43). Ap40, for
example, consists of residues 672-711 of APP and Ap42 consists of residues 672-713 of
APP. As a result of proteolytic processing of APP by different secretase enzymes jv vivo or
in situ, Ap is found in both a "short form," 40 amino acids in length, and a "10ng form,"
ranging from 42-43 amino acids in length. Epitopes or antigenic determinants can be 10cated
within the N-terminus of the Ap peptide and include residues within amino acids 1-10 of Ap,
preferably from residues 1-3,1-4,1-5,1-6, 1-7,2-7, 3-6, or 3-7 of Ap42 or within residues 2-
4, 5,6,7, or 8 of Ap, residues 3-5,6,7, 8, or 9 of Ap, or residues 4-7, 8,9, or 10 of Ap42.
"Central" epitopes or antigenic determinants are 10cated within the central or mid-portion of
the AJ3 peptide and include residues within amino acids 16-24,16-23,16-22,16-21,19-21,
-23-

WO 2007/035283 PCT/US2006/035025
19-22,19-23, or 19-24 of Ap. "C-terminal" epitopes or antigenic determinants are 10cated
within the C-terminus of the Ap peptide and include residues within amino acids 33-40,33-
41, or 33-42 of Ap.
In various embodiments, an Ap antibody is end-specific. As used herein, the term
"end-specific" refers to an antibody which specifically binds to theN-terminal or C-terroinal
residues of an Ap peptide but that does not recognize the same residues when present in a
10nger AP species comprising the residues or in APP.
In various embodiments, an AP antibody is "C-terminus-specific." As used herein,
the term "C terminus-specific" means that the antibody specifically recognizes a free C-
terminus of an AP peptide. Examples of C terminus-specific AP antibodies include those
that: recognize an Ap peptide ending at residue 40, but do not recognize an Ap peptide
ending at residue 41,42, and/or 43; recognize an Ap peptide ending at residue 42, but do not
recognize an Ap peptide ending at residue 40,41, and/or 43; etc.
In one embodiment, the antibody may be a 3D6 antibody or variant thereof, or a
10D5 antibody or variant thereof, both of which are described in U.S. Patent Publication No.
2003/0165496A1, U.S. Patent Publication No. 2004/0087777A1, International Patent
Publication No. WO02/46237A3. Description of 3D6 and 10D5 can also be found, for
example, in International Patent Publication No. WO02/088306A2 and International Patent
Publication No. WO02/088307A2. 3D6 is a monoc10nal antibody (mAb) that specifically
binds to an N-terminal epitope 10cated in the human p-amy10id peptide, specifically, residues
1 -5. By comparison, 10D5 is a mAb that specifically binds to an N-terminal epitope 10cated
in the human |3-amy10id peptide, specifically, residues 3-6. In another embodiment, the
antibody may be a 12B4 antibody or variant thereof, as described in U.S. Patent Publication
No. 20040082762A1 and International Patent Publication No. WO03/077858A2. 12B4 is a
mAb that specifically binds to an N-terminal epitope 10cated in the human P-amy10id peptide,
specifically, residues 3-7. In yet another embodiment, the antibody may be a 12A11
antibody or a variant thereof, as described in U.S. Patent Application No. 10/858,855 and
International Patent Application No. PCT/US04/17514. 12A11 is a mAb that specifically
binds to an N-terminal epitope 10cated in the human P-amy10id peptide, specifically, residues
3-7. In yet another embodiment, the antibody may be a 266 antibody as described in U.S.
Patent Application No. 10/789,273, and International Patent Application No.
-24-

WO 2007/035283 PCT/US2006/035025
WO01/62801A2. Antibodies designed to specifically bind to C-terminal epitopes 10cated in
human (3-amy10id peptide, for use in the present invention include, but are not limited to,
369.2B, as described in U.S. Patent No. 5,786,160.
In exemplary embodiments, the antibody is a humanized anti Afi peptide 3D6
antibody that selectively binds A|} peptide. More specifically, the humanized anti AP peptide
3D6 antibody is designed to specifically bind to an NH2-terrninal epitope 10cated in the
human P-amy10id 1-40 or 1-42 peptide found in plaque deposits in the brain (e.g., in patients
suffering from Alzheimer's disease).
Anti-AB antibodies can be used to treat amy10idogenic diseases, in particular,
Alzheimer's Disease. The term "amy10idogenic disease" includes any disease associated
with (or caused by) the formation or deposition of insoluble amy10id fibrils. Exemplary
amy10idogenic diseases include, but are not limited to, systemic amy10idosis, Alzheimer's
disease, mature onset diabetes, Parkinson's disease, Huntington's disease, fronto-temporal
dementia, and the prion-related transmissible spongiform encepha10pathies (kuru and
Creutzfeldt-Jacob disease in humans and scrapie and BSE in sheep and cattle, respectively).
Different amy10idogenic diseases are defined or characterized by the nature of the
polypeptide component of the fibrils deposited. For example, in subjects or patients having
Alzheimer's disease, P-amy10id protein (e.g., wild-type, variant, or truncated P-amy10id
protein) is the characterizing polypeptide component of the amy10id deposit. Accordingly,
Alzheimer's disease is an example of a "disease characterized by deposits of AP" or a
"disease associated with deposits of A(3," e.g., in the brain of a subject or patient. The terms
"P-amy10id protein," "P-amy10id peptide," "P-amy10id," "Ap," and "Ap peptide" are used
interchangeably herein.
Anti-5T4 Antibodies
The 5T4 antigen has been previously characterized (see e.g., WO 89/07947). The Ml
nucleic acid sequence of human 5T4 is known (Myers et al. (1994) JBiol Client 169:9319-
24 and GenBank at Accession No. Z29083). The sequence for 5T4 antigen from other
species is also known, for example, murine 5T4 (WO00/29428), canine 5T4 (WO01/36486)
or feline 5T4 (US 05/0100958).
-25-

WO 2007/035283 PCT/US2006/035025
Human 5T4 is a glycoprotein of about 72 kDa expressed widely in carcinomas, but
having a highly restricted expression pattern in normal adult tissues. It appears to be strongly
correlated to metastasis in co10rectal and gastric cancer. Expression of the 5T4 antigen is
' also found at high frequency in breast and ovarian cancers (Starzynska et al. (1998) Enr. J.
Gastroenterol Hepatol 10:479-84; Starzynska et al. (1994) Br. J. Cancer 69:899-902;
Starzynska et al. (1992) Br. J. Cancer 66:867-9). 5T4 has been proposed as a marker, with
possible mechanistic involvement, for tumor progression and metastasis potential (Carsberg
et al. (1996) IntJ Cancer 68:84-92). 5T4 has also been proposed for use as an
immunotherapeutic agent (see WO 00/29428). Antigenic peptides of 5T4 are disc10sed in,
e.g., US 05/0100958, the contents of which are incorporated by reference.
Several pending applications relate generally to nucleic acids encoding the anti-5T4
. monoc10nal antibody, vectors and host cells thereof, for example, U.S. Application
Publication Nos. 2003/0018004 and 2005/0032216. A provisional patent application
pertaining generally to the humanized anti-5T4 H8 monoc10nal antibodies and calicheamicin
conjugates thereof, as well as methods of treatment using these calicheamicin conjugates has
been filed (U.S. Application Publication No. 2006/0088522). The contents of all of these
applications are incorporated by reference herein in their entirety.
Anti-IL13 Antibodies
Another exemplary antibodies that can be separated by the methods of the invention
are anti-IL-13 antibodies. Interleukin-13 (IL-13) is a previously characterized cytokine
secreted by T lymphocytes and mast cells (McKenzie et al. (1993) Proc. Natl. Acad. Sci.
USA 90:3735-39; Bost et al. (1996) Immuno10gy 87:663-41). The term "IL-13" refers to
interleukin-13, including full-length unprocessed precursor form of IL-13, as well as the
mature forms resulting from post-translational cleavage. The term also refers to any
fragments and variants of IL-13 that maintain at least some bio10gical activities associated
with mature IL-13, including sequences that have been modified (e.g., recombinantly
modified). The term "IL-13" includes human IL-13, as well as other vertebrate species.
Several pending applications disc10se antibodies against human and monkey IL-13, IL-13
peptides, vectors and host cells producing the same, for example, U.S. Application
-26-

WO 2007/035283 PCT/US2006/035025
Publication Nos. 2006/0063228A and 2006/0073148. The contents of all of these
publications are incorporated by reference herein in their entirety.
IL-13 shares several bio10gical activities with IL-4. For example, either IL-4 or IL-13
can cause IgE isotype switching in B cells (Tomkinson et al. (2001) J. Immunol. 166:5792-
5800). Additionally, increased levels of cell surface CD23 and serum CD23 (sCD23) have
been reported in asthmatic patients (Sanchez-Guererro et al. (1994) Allergy 49:587-92;
Di10renzo et al. (1999) Allergy Asthma Proc. 20:119-25). In addition, either IL-4 or IL-13
can upregulate the expression of MHC class II and the low-affinity IgE receptor (CD23) onB
cells and monocytes, which results in enhanced antigen presentation and regulated
macrophage function (Tomkinson et al., supra). These observations suggest that IL-13 may
be an important player in the deve10pment of airway eosinophilia and airway
hyperresponsiveness (AHR) (Tomkinson et al., supra; Wills-Karp et al. (1998) Science
282:2258-61). Accordingly, inhibition of IL-13 can be useful in ameliorating the patho10gy
of a number of inflammatory and/or allergic conditions, including, but not limited to,
respiratory disorders, e.g., asthma; chronic obstructive pulmonary disease (COPD); other
conditions involving airway inflammation, eosinophilia, fibrosis and excess mucus
production, e.g., cystic fibrosis and pulmonary fibrosis; atopic disorders, e.g., atopic
dermatitis, urticaria, eczema, allergic rhinitis; inflammatory and/or autoimmune conditions
of, the slrin (e.g., atopic dermatitis), gastrointestinal organs (e.g., inflammatory bowel
diseases (IBD), such as ulcerative colitis and/or Crohn's disease), liver (e.g., cirrhosis,
hepatocellular carcinoma); scleroderma; tumors or cancers (e.g., soft tissue or solid tumors),
such as leukemia, glioblastoma, and lymphoma, e.g., Hodgkin's lymphoma; viral infections
(e.g., from HTLV-1); fibrosis of other organs, e.g., fibrosis of the liver, (e.g., fibrosis caused
by a hepatitis B and/or C virus).
Anti-IL22 Antibodies
Another exemplary antibodies that can be separated by the methods of the invention
are anti-IL-22 antibodies. Interleukin-22 (IL-22) is a previously characterized class II
cytokine that shows sequence homo10gy to IL-10. Its expression is up-regulated in T cells by
IL-9 or ConA (Dumoutier L. et al. (2000) Proc Natl Acad Sci USA 97(18): 10144- 9).
Studies have shown that expression of IL-22 mRNA is induced in vivo in response to LPS
-27-

WO 2007/035283 PCT/US2006/035025
administration, and that IL-22 modulates parameters indicative of an acute phase response
(Dumoutier L. et al. (2000) supra; Pittman D. et al. (2001) Genes and Immunity 2:172), and
that a reduction of IL-22 activity by using a neutralizing anti-lL-22 antibody ameliorates
inflammatory symptoms in a mouse collagen-induced arthritis (CIA) model. Thus, IL-22
antagonists, e.g., neutralizing anti-IL-22 antibodies and fragments thereof, can be used to
induce immune suppression in vivo, for examples, for treating autoimmune disorders (e.g.,
arthritic disorders such as rheumatoid arthritis); respiratory disorders (e.g., asthma, chronic
obstructive pulmonary disease (COPD)); inflammatory conditions of, e.g., the skin (e.g.,
psoriasis), cardiovascular system (e.g., atherosclerosis), nervous system (e.g., Alzheimer's
disease), kidneys (e. g., nephritis), liver (e.g., hepatitis) and pancreas (e.g., pancreatitis).
The term "IL-22" refers to interIeukin-22, including full-length unprocessed precursor
form of IL-22, as well as the mature forms resulting from post-translational cleavage. The
term also refers to any fragments and variants of IL-22 that maintain at least some bio10gical
activities associated with mature IL-22, including sequences that have been modified. The
term "IL-22" includes human IL-22, as well as other vertebrate species. The arnino acid and
nucleotide sequences of human and rodent IL-22, as well as antibodies against IL-22 are
disc10sed in, for example, U.S. Application Publication Nos. 2005-0042220 and 2005-
0158760, and U.S. Patent No. 6,939,545. The contents of all of these publications are
incorporated by reference herein in their entirety.
Anti-GDF8 Antibodies
Yet another exemplary antibodies that can be separated by the methods of the
invention are anti-GDF8 antibodies. Growth and differentiation factor-8 (GDF-8), also
known as myostatin, is a secreted protein and is a member of the transforming growth factor-
beta (TGF-P) superfamily of structurally related growth factors, all of which possess
physio10gically important growth-regulatory and morphogenetic properties (Kingsley et al.
(1994) Genes Dev., 8:133-146; Hoodless et al. (1998) Curr. Topics Microbiol. Immunol,
228:235-272). Similarly to TGF-p\ human GDF-8 is synthesized as a 375 arnino acid 10ng
precursor protein. The precursor GDF-8 protein forms a homodimer. During processing the
amino-terminal propeptide is cleaved off at Arg-266. The cleaved propeptide, known as the
"latency-associated peptide" (LAP), may remain noncovalently bound to the homodimer,
-28-

WO 2007/035283 PCT/US2006/035025
thereby inactivating the complex (Miyazono et al. (1988) J. Biol. Chem. 263: 6407-6415;
Wakefield et al. (1988) J. Biol. Chem. 263:7646-7654; Brown et al. (1990) Growth Factors,
3: 35-43; and Thies et al. (2001) Growth Factors, 18:251-259). The complex of mature
GDF-8 with propeptide is commonly referred to as the "small latent complex" (Gentry et al.
(1990) Biochemistry, 29: 6851- 6857; Derynck et al. (1995) Nature, 316: 701-705; and
Massague (1990) Ann. Rev. Cell Biol, 12: 597-641). Other proteins are also known to bind
to mature GDF-8 and inhibit its bio10gical activity. Such inhibitory proteins include
follistatin and Mlistatin-related proteins (Gamer et al. (1999) Dev. Biol, 208:222-232).
The term "GDF-8" refers to growth and differentiation factor-8 and, where
appropriate, factors that are structurally or functionally related to GDF-8, for example, BMP-
11 and other factors be10nging to the TGF-fJ superfamily. The term refers to the full-length
unprocessed precursor form of GDF-8, as well as the mature and propeptide forms resulting
from post-translational cleavage. The term also refers to any fragments and variants of GDF-
8 that maintain at least some bio10gical activities associated with mature GDF-8, including
sequences that have been modified. The amino acid sequence human GDF- 8, as well as
many other vertebrate species (including murine, baboon, bovine, chicken) is disc10sed, e.g.,
US 2004-0142382, US 2002-0157125, and McPherron et al. (1997) Proc. Nat. Acad. Sci.
U.S.A., 94:12457-12461; the contents of all of which are hereby incorporated by reference k
their entirety). Examples of neutralizing antibodies against GDF-8 are disc10sed in, e.g., US
2004-0142382, and may be used to treat or prevent conditions in which an increase in muscle
tissue or bone density is desirable. Exemplary disease and disorders include muscle and
neuromuscular disorders such as muscular dystrophy (including Duchenne's muscular
dystrophy); amyotrophic lateral sclerosis; muscle atrophy; organ atrophy; frailty; tunnel
syndrome; congestive obstructive pulmonary disease; sarcopenia, cachexia, and other muscle
wasting syndromes; adipose tissue disorders (e.g., obesity); type 2 diabetes; impaired glucose
tolerance; metabolic syndromes (e.g., syndrome X); insulin resistance induced by trauma
such as burns or nitrogen imbalance; and bone degenerative diseases (e.g., osteoarthritis and
osteoporosis)
-29-

WO 2007/035283 PCT/US2006/035025
Soluble Receptors arid Receptor Fusions
In some embodiments, proteins separated by the methods of the invention can be
soluble receptors or fragments thereof. Examples of soluble receptors include the
extracellular domain of a receptor, such as soluble tumor necrosis factor alpha and beta
receptors (TNFR-1; EP 417,563 published Mar. 20,1991; TNFR-2, EP 417,014 published
Mar. 20, 1991; and reviewed inNaismith and Sprang, JInflamm. 47(l-2):l-7,1995-96, each
of which is incorporated herein by reference in its entirety). In other embodiments, the
soluble receptor includes the extracellular domain of interleukin-21 receptor (IL-21R) as
described in, for example, US 2003-0108549 (the contents of which are also incorporated by
reference).
In other embodiments, the methods of the invention are used to separate soluble
receptor fusions. The fusion protein can include a targeting moiety, e.g., a soluble receptor
fragment or a ligand, and an immunog10bulin chain, an Fc fragment, a heavy chain constant
regions of the various isotypes, including: IgGl, IgG2, IgG3, IgG4, IgM, IgAl, IgA2, IgD,
and IgE). For example, the fusion protein can include the extracellular domain of a receptor,
and, e.g., fused to, a human immunog10bulin Fc chain (e.g., human IgG, e.g., human IgGl or
human IgG4, or a mutated form thereof), hi one embodiment, the human Fc sequence has
been mutated at one or more arnino acids, e.g., mutated at residues 254 and 257 from the
wild type sequence to reduce Fc receptor binding. The fusion proteins may additionally
include a linker sequence joining the first moiety to the second moiety, e.g., the
immunog10bulin fragment. For example, the fusion protein can include a peptide linker, e.g.,
a peptide linker of about 4 to 20, more preferably, 5 to 10, amino acids in length; the peptide
linker is 8 amino acids in length. For example, the fusion protein can include a peptide linker
having the formula (Ser-Gly-Gly-Gly-Gly)y wherein y is 1,2,3,4,5, 6,7, or8. In other
embodiments, additional amino acid sequences can be added to the N- or C-terminus of the
fusion protein to facilitate expression, steric flexibility, detection and/or isolation or
purification.
la certain embodiments, the soluble receptor fusion comprises a soluble TNFR-Ig
(e.g., a soluble fragment of a TNF receptor, e.g., p55 or p75 human TNF receptor or
derivatives thereof, e.g., 75 kd TNFR-IgG (e.g., 75 kD TNF receptor fused to an Fc portion
of human IgGl).
-30-

WO 2007/035283 PCT/US2006/035025
A chirneric or fusion protein of the invention can be produced by standard
reconibinant DNA techniques. For example, DNA fragments coding for the different
polypeptide sequences are ligated together in-frame in accordance with conventional
techniques, e.g., by emp10ying blunt-ended or stagger-ended termini for ligation, restriction
enzyme digestion to provide for appropriate termini, filling-in of cohesive ends as
appropriate, alkaline phosphatase treatment to avoid undesirable joining, and enzymatic
ligation. In another embodiment, the fusion gene can be synthesized by conventional
techniques including automated DNA synthesizers. Alternatively, PCR amplification of gene
fragments can be carried out using anchor primers that give rise to complementary overhangs
between two consecutive gene fragments that can subsequently be annealed and reamplified
to generate a chimeric gene sequence (see, for example, Ausubel et al. (eds.) Current
Protocols in Molecular Bio10gy, John Wiley & Sons, 1992). Moreover, many expression
vectors are commercially available that encode a fusion moiety (e.g., an Fc region of an
imrnunog10bulin heavy chain). Immunog10bulin fusion polypeptide are known in the art and
are described in e.g., U.S. Pat. Nos. 5,516,964; 5,225,538; 5,428,130; 5,514,582; 5,714,147;
and 5,455,165.
Growth Factors and Cytoldnes
Another class of polypeptides that have been shown to be effective as pharmaceutical
and/or commercial agents and that can desirably be produced according to the teachings of
the present invention includes growth factors and other signaling molecules, such as
cytokines.
Growth factors are typically glycoproteins that are secreted by cells and bind to and
activate receptors on other cells, initiating a metabolic or deve10pmental change in the
receptor cell. Non-limiting examples of mammalian growth factors and other signaling
molecules include cytokines; epidermal growth factor (EGF); platelet-derived growth factor
(PDGF); fibroblast growth factors (FGFs) such as aFGF and bFGF; transforming growth
factors (TGFs) such as TGF-alpha and TGF-beta, including TGF-beta 1, TGF-beta 2, TGF-
beta 3, TGF-beta 4, or TGF-beta 5; insulin-like growth factor-I and -II (IGF-I and IGF-II);
des(l-3) -IGF-I (brain IGF-I), insulin-like growth factor binding proteins; CD proteins such
as CD-3, CD-4, CD-8, and CD-I 9; erythropoietin; osteoinductive factors; immunotoxins; a '
-31-

WO 2007/035283 PCT7US2006/035025
bone morphogenetic protein (BMP); an interferon such as interferon-alpha, -beta, and -
gamma; co10ny stimulating factors (CSFs), e.g., M-CSF, GM-CSF, and G-CSF; interleukins
(TLs), e.g., IL-1 to IL-13 (e.g., EL-11); tumor necrosis factor (TNF) alpha and beta; insulin
A-chain; insulin B-chain; proinsulin; follicle stimulating hormone; calcitonin; luteinizing
hormone; glucagon; clotting factors such as factor VIIIC, factor IX, tissue factor, and von
Willebrands factor; anti-clotting factors such as Protein C; atrial natriuretic factor, lung
surfactant; a plasminogen activator, such as urokinase or human urine or tissue-type
plasminogen activator (t-PA); bombesin; thrombin, hemopoietic growth factor;
enkephalinase; RANTES (regulated on activation normally T-cell expressed and secreted);
human macrophage inflammatory protein (MEM -alpha); mullerian-inhibiting substance;
relaxin A-chain; relaxin B-chain; prorelaxin; mouse gonadotropin-associated peptide;
neurotrophic factors such as bone-derived neurotrophic factor (BDNF), neurotrophin-3, -4, -
5, or -6 (NT-3, NT-4, NT-5, or NT-6), or a nerve growth factor such as "NGF-beta. One of '
ordinary skill in the art will be aware of other growth factors or signaling molecules that can
be expressed in accordance with methods and compositions of the present invention.
Specific alterations in the glycosylation pattern of growth factors or other signaling
molecules have been shown to have dramatic effects on their therapeutic properties. As one
example, a common method of treatment for patients who suffer from chronic anemia is to
provide them with frequent injections of recombinant human erythropopietin (rHuEPO) in
order to boost their production of red b10od cells. An ana10g of rHuEPO, darbepoetin alfa
(Aranesp®), has been deve10ped to have a 10nger duration than normal rHuEPO. The
primary difference between darbepoetin alfa and rHuEPO is the presence of two extra sialic-
acid-containingN-linked oligosaccharide chains. Production of darbepoetin alfa has been
accomplished using in vitro glycoengineering (see Elliott et al., Nature Biotechno10gy
21(4):414-21,2003, incorporated herein by reference in its entirety). Elliott et al. used in
vitro mutagenesis to incorporate extra glycosylation sites into the rHuEPO polypeptide
backbone, resulting in expression of the darbepoetin alfa ana10g. The extra oligosaccharide
chains are 10cated distal to the EPO receptor binding site and apparently do not interfere with
receptor binding. However, darbepoetin alfa's half-life is up to three-fold higher than
rHuEPO, resulting in a much more effective therapeutic agent.
-32-

WO 2007/035283 PCT/US2006/035025
Clotting Factors
Clotting factors have been shown to.be effective as pharmaceutical and/or
commercial agents. Hemophilia B is a disorder in which the b10od of the sufferer is unable
to clot. Thus, any small wound that results in bleeding is potentially a life-threatening event.
For example, Coagulation Factor IX (Factor IX or "FIX") is a single-chain glycoprotein
whose deficiency results in Hemophilia B. FIX is synthesized as a single chain zymogen that
can be activated to a two-chain serine protease (Factor IXa) by release of an activation
peptide. The catalytic domain of Factor IXa is 10cated in the heavy chain (see Chang et al., J.
Clin. Invest., 100;4,1997, incorporated herein by reference in its entirety). FIX has multiple
glycosylation sites including both N-linked and O-linked carbohydrates. One particular O-
linked structure at Serine 61 (Sia-a2,3-Gal-pi,4-GlcNAc-|}l,3-Fuc-al-O-Ser) was once
thought unique to FIX but has since found on a few other molecules including the Notch
protein in mammals and Drosophila (Ma10ney et al, Journal ofBiol. Chem., 275(13), 2000).
FIX produced by Chinese Hamster Ovary ("CHO") cells in cell culture exhibits some
variability in the Serine 61 oligosacchari.de chain. These different glycoforms, and other
potential glycoforms, may have different abilities to induce clotting when administered to
humans or animals and/or may have different stabilities in the b10od, resulting in less
effective clotting.
Hemophilia A, which is clinically indistinguishable from Hemophilia B, is caused by
a defect in human clotting factor VIII, another glycoprotein that is synthesized as a single
chain and then processed into a two-chain active form. The present invention may also be
emp10yed to control or alter the glycosylation pattern of clotting factor VIII in order to
modulate its clotting activity. Other clotting factors that can be produced in accordance with
the present invention include tissue factor and von Willebrands factor.
Enzymes
Another class of polypeptides that have been shown to be effective as pharmaceutical
and/or commercial agents and that can desirably be produced according to the teachings of
the present invention includes enzymes. Enzymes may be glycoproteins whose glycosylation
pattern affects enzymatic activity. Thus, the present invention may also be used to produce
-33-

WO 2007/035283 PCT/US2006/035025
enzymes in a cell culture wherein the produced enzymes have a more extensive or otherwise
more desirable glycosylation pattern.
As but one non-limiting example, a deficiency in glucocerebrosidase (GCR) results in
a condition known as Gaucher's disease, which is caused by an accumulation of
glucocerebrosidase in lysosomes of certain cells. Subjects with Gaucher's disease exhibit a
range of symptoms including splenomegaly, hepatomegaly, skeletal disorder,
thrombocytopenia and anemia. Friedman and Hayes showed that recombinant GCR (rGCR)
containing a single substitution in the primary amino acid sequence exhibited an altered
glycosylation pattern, specifically an increase in fucose andN-acetyl glucosamine residues
compared to naturally occurring GCR (see United States Patent number 5,549,892).
Friedman and Hayes also demonstrated that this rGCR exhibited improved
pharmacokinetic properties compared to naturally occurring rGCR. For example,
approximately twice as much rGCR targeted liver Kupffer cells than did naturally occurring
GCR. Although the primary amino acid sequences of the two proteins differed at a single
residue, Friedman and Hayes hypothesized that the altered glycosylation pattern of rGCR
may also influence the targeting to Kupffer cells. One of ordinary skill in the art will be
aware of other known examples of enzymes that exhibit altered enzymatic, pharmacokinetic
and/or pharmacodynamic properties resulting from an alteration in their glycosylation
patterns.
Protein Production
Proteins separated by the methods of the invention can be produced recombinantly
using techniques well known in the art. Nucleotide sequence encoding the proteins are
typically inserted in an expression vector for introduction into host cells that may be used to
produce the desired quantity of modified antibody that, in turn, provides the polypeptides.
The term "vector" includes a nucleic acid construct often including a nucleic acid, e.g., a
gene, and further including minimal elements necessary for nucleic acid replication,
transcription, stability and/or protein expression or secretion from a host cell. Such
constructs may exist as extrachromosomal elements or maybe integrated into the genome of
a host cell.
— 34 —

WO 2007/035283 PCT/US2006/035025
The term "expression vector" includes a specific type of vector wherein the nucleic
acid construct is optimized for the high-level expression of a desired protein product.
Expression vectors often have transcriptional regulatory agents, such as promoter and
enhancer elements, Qptirnized_for high-levels of transcription in specific cell types and/ or
optimized such that expression is constitutive hased upon the use of a specific inducing
agent. Expression vectors further have sequences that provide for proper and/or enhanced
translation of the protein As known to those skilled in the art, such vectors may easily be
selected from the group consisting of plasmids, phages, viruses, and retroviruses. The term
"expression cassette" includes a nucleic acid construct containing a gene and having
elements in addition to the gene that allow for proper and or enhanced expression of that
gene in a host cell. For producing antibodies, nucleic acids encoding light and heavy chains
can be inserted into expression vectors. Such sequences can be present in the same nucleic
acid molecule (e.g., the same expression vector) or alternatively, can be expressed from
separate nucleic acid molecules (e.g., separate expression vectors).
The term "operably linked" includes a juxtaposition wherein the components are in a
relationship permitting them to function in their intended manner (e.g., functionally linked).
As an example, a promoter/enhancer operably linked to a polynucleotide of interest is ligated
to said polynucleotide such that expression of the polynucleotide of interest is achieved under
conditions which activate expression directed by the promoter/enhancer.
Expression vectors are typically replicable in the host organisms either as episomes or
as an integral part of the host chromosomal DNA. Commonly, expression vectors contain
selection markers (e.g., ampicillin-resistance, hygromycin-resistance, tetracycline resistance,
kanamycin resistance or neomycin resistance) to permit detection of those cells transformed
with the desired DNA sequences (see, e.g., Itakura et al., U.S. Patent No. 4,704,362). In
addition to the immunog10bulin DNA cassette sequences, insert sequences, and regulatory
sequences, the recombinant expression vectors of the invention may carry additional
sequences, such as sequences that regulate replication of the vector in host cells (e.g., origins
of replication) and selectable marker genes. The selectable marker gene facilitates selection
of host cells into which the vector has been introduced (see e.g., U.S. Pat. Nos. 4,399,216,
4,634,665 and 5,179,017, all by Axel et al). For example, typically the selectable marker
gene confers resistance to drugs, such as G418, hygromycin, or methotrexate, on a host cell
-35-

WO 2007/035283 PCT/US2006/035025
into which the vector has been introduced. Preferred selectable marker genes include the
dihydrofolate reductase (DHFR) gene (for use in dhff host cells with methotrexate
selection/amplification) and the neo gene (for G418 selection).
Once the vector has been incorporated into the appropriate host cell, the host cell is
maintained under conditions suitable for high level expression of the nucleotide sequences,
and the collection and purification of the desired antibodies. Any host cell susceptible to cell
culture, and to expression of proteins or polypeptides, may be utilized in accordance with the
present invention. In certain embodiments, the host cell is mammalian. Non-limiting
examples of mammalian cells that may be used in accordance with the present invention
include BALB/c mouse mye10ma line (NSO/1, ECACC No: 85110503); human retinoblasts
(PER.C6, CruCell, Leiden, The Netherlands); monkey kidney CV1 line transformed by SV40
(COS-7, ATCC CRL1651); human embryonic kidney line (293 or 293 cells subc10ned for
growth in suspension culture, Graham et al., J. Gen Virol, 36:59,1977); baby hamster
kidney cells (BHK, ATCC CCL 10); Chinese hamster ovary cells +/-DHFR (CHO, Urlaub
and Chasin, Proc. Natl. Acad. Sci. USA, 77:4216,1980); mouse sertoli cells (TM4, Mather,
Biol. Reprod., 23:243-251,1980); monkey kidney cells (CV1 ATCC CCL 70); African green
monkey kidney cells (VERO-76, ATCC CRL-1 587); human cervical carcinoma cells (HeLa,
ATCC CCL 2); canine kidney cells (MDCK, ATCC CCL 34); buffa10 rat liver cells (BRL
3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL 75); human liver cells (Hep
G2, HB 8065); mouse mammary tumor (MMT 060562, ATCC CCL51); TRI cells (Mather et
al., Annals N.Y. Acad. Sci., 383:44-68,1982); MRC 5 cells; FS4 cells; and a human
hepatoma line (Hep G2).
Additionally, any number of commercially and non-commercially available
hybridoma cell lines that express polypeptides or proteins maybe utilized in accordance with
the present invention. One skilled in the art will appreciate that hybridoma cell lines might
have different nutrition requirements and/or might require different culture conditions for
optimal growth and polypeptide or protein expression, and will be able to modify conditions
as needed.
Expression vectors for these cells can include expression control sequences, such as
an origin of replication, a promoter, and an enhancer (Queen et ah, Immunol. Rev. 89:49
(1986)), and necessary processing information sites, such as ribosome binding sites, RNA
-36-

WO 2007/035283 PCT/US2006/035025
splice sites, polyadenylation sites, and transcriptional terminator sequences. Preferred
expression control sequences are promoters derived from irnmunog10bulin genes, SV40,
adenovirus, bovine papil10ma virus, cytomega10virus and the like. See, e.g., Co et ah, (1992)
J. Immunol. 148:1149. Preferred regulatory sequences for mammalian host cell expression
include viral elements that direct high levels of protein expression in mammalian cells, such
as promoters and/or enhancers derived from FF-la promoter and BGH poly A,
cytomega10virus (CMV) (such as the CMV promoter/ enhancer), Simian Virus 40 (SV40)
(such as the SV40 promoter/enhancer), adenovirus (e.g., the adenovirus major late promoter
(AdMLP)), and polyoma. For further description of viral regulatory elements, and sequences
thereof, see, e.g., U.S. Pat. No. 5,168,062 by Stinski, U.S. Pat. No. 4,510,245 by Bell et al.
and U.S. Patent No. 4,968,615 by Schaffher et al. In exemplary embodiments, the antibody
heavy and light chain genes are operatively linked to enhancer/promoter regulatory elements
(e.g, derived from SV40, CMV, adenovirus and the like, such as a CMV enhancer/AdMLP
promoter regulatory element or an SV40 enhancer/AdMLP promoter regulatory element) to
drive high levels of transcription of the genes. In exemplary embodiments of the invention,
the construct include an internal ribosome entry site (IRES) to provide relatively high levels
of polypeptides of the invention in eukaryotic host cells. Compatible IRES sequences are
disc10sed in U.S. Pat. No. 6,193,980 that is also incorporated herein.
Alternatively, coding sequences can be incorporated in a transgene for introduction
into the genome of a transgenic animal and subsequent expression in the milk of the
transgenic animal (see, e.g., Deboer et al., US 5,741,957, Rosen, US 5,304,489, and Meade
et al., US 5,849,992). Suitable transgenes include coding sequences for light and/or heavy
chains in operable linkage with a promoter and enhancer from a mammary gland specific
gene, such as casein or beta lactog10bulin.
Prokaryotic host cells may also be suitable for producing the antibodies of the
invention. E. coli is one prokaryotic host particularly useful for c10ning the polynucleotides
{e.g., DNA sequences) of the present invention. Other microbial hosts suitable for use
include bacilli, such as Bacillus subtilis, enterobacteriaceae, such as Escherichia, Salmonella,
and Serratia, and various Pseudomonas species. In these prokaryotic hosts, one can also
make expression vectors, which will typically contain expression control sequences
compatible with the host cell (e.g., an origin of replication). In addition, any number of a
-37-

WO 2007/035283 PCT/US2006/03025
variety of well-known promoters will be present, such as the lactose promoter system, a
tryptophan (trp) promoter system, a beta-lactamase promoter system, or a promoter system
from phage lambda. The promoters will typically control expression, optionally with an
operator sequence, and have ribosome binding site sequences and the like, for initiating and
completing transcription and translation.
Expression of proteins in prokaryotes is most often carried out in E. coli with vectors
containing constitutive or inducible promoters directing the expression of either fusion or
non-fusion proteins. Fusion vectors add a number of amino acids to an antibody encoded
therein, often to the constant region of the recombinant antibody, without affecting
specificity or antigen recognition of the antibody. Addition of the amino acids of the fusion
peptide can add additional function to the antibody, for example as a marker {e.g., epitope tag
such as myc or flag).
Other microbes, such as yeast, are also useful for expression. Saccharomyces is a
preferred yeast host, with suitable vectors having expression control sequences {e.g.,
promoters), an origin of replication, termination sequences, and the like as desired. Typical
promoters include 3-phosphoglycerate kinase and other glycolytic enzymes. Inducible yeast
promoters include, among others, promoters from alcohol dehydrogenase, isocytochrome C,
and enzymes responsible for maltose and galactose utilization.
The vectors containing the polynucleotide sequences of interest (e.g., the heavy and
light chain encoding sequences and expression control sequences) can be transferred into the
host cell by well-known methods, which vary depending on the type of cellular host. For
example, calcium ch10ride transfection is commonly utilized for prokaryotic cells, whereas
calcium phosphate treatment, electroporation, lipofection, biolistics or viral-based
transfection may be used for other cellular hosts. (See generally Sambrook et al., Molecular
C10ning: A Laboratory Manual (Cold Spring Harbor Press, 2nd ed., 1989), incorporated by
reference herein in its entirety for all purposes.). Other methods used to transform
mammalian cells include the use of polybrene, protoplast fusion, liposomes, electroporation,
and microinjection (see generally, Sambrook et al., supra). For production of transgenic
animals, transgenes can be microinjected into fertilized oocytes, or can be incorporated into
the genome of embryonic stem cells, and the nuclei of such cells transferred into enucleated
oocytes.
-38-

WO 2007/035283 PCT/US2006/05025
When heavy and light chains are c10ned on separate expression vectors, the vectors
are co-transfected to obtain expression and assembly of intact immunog10bulins. Once
expressed, the whole antibodies, their dimers, individual light and heavy chains, or other
immunog10bulin forms of the present inyention can be separated as described herein and/or
further purified according to procedures known in the art, including ammonium sulfate
precipitation, affinity columns, column chromatography, HPLC purification, gel
electrophoresis and the like (see generally Scopes, Protein Purification (Springer-Verlag,
N.Y., (1982)). Substantially pure immunog10bulins of at least about 90 to 95% homogeneity
are preferred, and 98 to 99% or more homogeneity most preferred, for pharmaceutical uses.
The following examples are illustrative and not intended to be limiting.
Example 1
Flocculation with various cations and anions: Various monoc10nal antibodies (mAb)
(shown in Table 1) were produced by recombinant Chinese Hamster Ovary (CHO) cells
cultured in serum free media. Approximately 50mL of the cell-containing conditioned media
was aliquoted in to 125mL Erlenmeyer flasks (except the 30/20 calcium phosphate example
with anti-GDF8 #2 which had 400mL in a 10OOmL flask and the 30/20 calcium phosphate
example with anti-IL13 #1 which had 10OOmL in a 2000rnL flask). HEPES was added to
40mM to control the pH. A concentrated solution of metal cations was added to the solution
to achieve a final target concentration (see Table 1) and mixed gently. A concentrated
solution of anions was added to the mixture (see Table 1) to achieve the final anion
concentration and mixed gently (in some of the examples, indicated by an asterisk (*) in
Table 1 below, the anion was added first and the cation second). The pH was increased by
the addition of NaOH or decreased by the addition of HC1 to the targeted pH range. In many
of the examples, the pH did not need to be adjusted. The mixture was allowed to incubate on
a shaker at 18-25°C for one to four hours a10ng with negative controls. After the incubation,
the mixture was poured into a 50mL centrifuge tube. Each mixture was spun at 340g for ten
minutes. The clarified supernatant was aliquoted and the turbidity measured using a
nepha10meter (HACH, 10veland CO). The resulting turbidity is reported in Table 1 as a
percent reduction from that of the untreated control. The antibody concentration was
-39-

WO 2007/035283 PCT/US2006/035025
measured by a Protein A HPLC method general to antibodies. The recovery as compared to
the untreated control is reported in Table 1.
Table 1: FlQCCulation results for the cations calcium, magnesium, manganese, cobalt
(IT), and. nickel. . All treatments were performed with the addition of cation prior to anion,
except where noted by an asterisk (*), where the anion was added prior to the cation.
-40-

WO 2007/035283 PCT/US2006/035025


WO 2007/035283 PCT/US2006/035025

The residual levels of calcium and phosphate in the supematants of some samples
were measured. Calcium residuals were measured using a BioAssay Systems
QuantiChrom™ Calcium Assay Kit (DICA-500). Phosphate residuals were measured using a
BioAssay Systems Malachite Green Phosphate Assay Kit (POMG-25H). In the anti-IL13 #2
sample with 40mM calcium and 20mM phosphate in Table 1, the supernatant after
centrifugation contained 8.19mM calcium and 1.04mM phosphate. This level of soluble
calcium and phosphate translates to a Ksp of 8.5xl0"5 M2. hi the anti-AB #1 sample with
80mM calcium and 20mM phosphate in Table 1, the supernatant after centrifugation
-42-

WO 2007/035283 PCI/S2006/035025
contained 22.2mM calcium and 0.4mM phosphate. This level of soluble calcium and
phosphate translates to a Ksp of 8.4X10"6 M2.
Example 2
Effect of order of anion/cation addition: A mAb, anti-GDF8 #1 was produced by
recombinant CHO cells cultured in serum free media. Approximately 50mL of the cell-
containing conditioned media was aliquoted into 3xl25mL Erlenmeyer flasks, A, B, and C.
Sample A was left untreated and served as a negative control. HEPES was added to 40mM
to control the pH in Samples B and C. 5M calcium ch10ride was added to Sample B to a
concentration of 50mM; after gentle mixing, 0.5M sodium sulfite was added to a
concentration of 33.3mM. For Sample C, the order of addition was reversed. 0.5M sodium
sulfite was added to a concentration of 33.3mM; after gentle mixing, 5M calcium ch10ride
was added to a concentration of 50mM. All three mixtures hadpHs between 7.4 and 7.6.
The mixtures were allowed to incubate on a shaker for one hour at 18-25°C. After the
incubation, the mixtures were poured into 50mL centrifuge tubes. Each mixture was spun at
340g for ten minutes. The clarified supernatant was removed and assayed for antibody
concentration by Protein A HPLC, and for turbidity using a nepha10meter.
Samples B and C both had antibody recoveries of 94% as compared to the untreated
sample. Sample B, with the cation added first, showed an increase in turbidity as compared
to the untreated sample, indicating a precipitate had formed, but that it was too small to be
easily removed by centrifugation. Sample C, with the anion added first, showed a 46%
reduction in turbidity as compared to the untreated sample, indicating that the precipitate that
formed was large enough to be easily spun out by centrifugation. The turbidity reduction
also indicates that some amount of cellular debris and/or colloidal material was bound up by
the precipitate and removed in the pellet.
Example 3
A pilot-scale experiment using calcium and phosphate as precipitants and the effects
on downstream clarification and chromatography steps: A mAb, anti-AB #1, was produced
by recombinant CHO cells cultured in serum free media in a 500L bioreactor. At the time of
harvest, the culture was brought to room temperature (18-25°C). 350L of the culture was left
-43-

WO 2007/035283 PCT/US2006/035025
untreated and served as a negative control. 150L of the culture was transferred to a 200L
carboy and slowly mixed with an overhead mixer. A buffer was added to 40mM to control
the pH. 2M potassium phosphate was then added to a concentration of 20mM. 5M calcium
ch10ride was then added to a concentration of 30mM. The pH of the mixture was 7.3. The
f10cculated culture was incubated for 160 minutes while mixing. At the end of the incubation
the pH of mixture was 7.0.
Both the f10cculated and untreated cultures were processed through an Alfa Laval
BTPX 205 disc stack centrifuge at a flow rate of 4.4 L/min and a bowl speed of 7630 RPM
(8000g). The steady state centrate turbidity of the f10cculated sample was 14 NTU, as
compared to 117 NTU for the untreated sample, an 88% reduction in turbidity. The recovery
of antibody titer in the f10cculated centrate was 96% as compared to the untreated centrate.
The level of host cell proteins (HCP) in the f10cculated centrate was reduced by 35% from
533,341ppm to 348,087ppm as compared to the untreated centrate.
Both the f10cculated and untreated centrates were processed through Millipore A1HC
Pad Filters to a capacity of 250 L of centrate per square meter of filter. The untreated sample
showed a steadily increasing breakthrough of turbidity, from 4 NTU at 24 L/m2 to 26 NTU at
121 L/m2 to 37 NTU at 254 L/m2. The final pad filtrate pool turbidity for the untreated
sample was 21 NTU. The f10cculated sample showed only a small rise in turbidity through
the pads, from 2 NTU at 21 L/m2 to 6 NTU at 150 L/m2 to 7 NTU at 254 L/m2. The final pad
filtrate pool turbidity for the f10cculated sample was 5 NTU, a 76% reduction in turbidity as
compared to the untreated sample.
After pad filtration and additional filtration through a 0.2um polishing filter, the
samples were chromatographed using a GE Healthcare MahSelect Protein A affinity column.
When antibodies are eluted from Protein A columns, the peak pools often are turbid, and that
turbidity typically increases when the peak is neutralized. The untreated peak pool had a
turbidity of 10 NTU, and increased to 22 NTU upon neutralization. The f10cculated peak
pool had a tuibidity of 3 NTU, and increased to 8 NTU upon neutralization, 63% lower than
the untreated peak.
-44-

WO 2007/035283 PCT/US2006/035025
Example 4
Flocculation with calcium and phosphate resulting in reduced turbidity in both the
clarified conditioned media and in the Protein Apeak. Significant reduction of both a cell-
related and a product-related impurity were also achieved.
A mAb, anti-5T4, was produced by recombinant CHO cells cultured in serum free
media. Approximately 3L of culture was left untreated and served as a negative control.
Another 3L of the culture was transferred to a 4L vessel. A buffer was added to 40mM to
control the pH and gently mixed at 18-24°C. 2M potassium phosphate was then added to a
concentration of 20mM and the solution gently mixed. 5M calcium ch10ride was then added
to a concentration of 30mM and the solution gently mixed. The pHofthe mixture was 12.
The f10cculated culture was transferred to three 2L Erlenmeyer flasks and incubated for 2
hours while mixing. At the end of the incubation the pH of mixture was 7.0.
A 50mL sample of both the f10cculated culture and the untreated culture were spun at
340g for ten minutes. The clarified supernatant was removed and assayed for antibody
concentration by Protein A HPLC, and for turbidity using a nepha10meter. The total
recovery of product-related material was 78% as compared to the untreated sample. The
turbidity of the untreated sample was 29 NTU. The turbidity of the f10cculated sample was 2
NTU, a 93% reduction in turbidity as compared to the untreated sample.
Both the f10cculated and untreated samples were processed through Millipore A1HC
Pad Filters. After pad filtration and additional filtration through a 0.2um polishing filter, the
samples were chromatographed using a GE Healthcare MabSelect Protein A affinity column.
As antibodies elute from Protein A columns at high concentrations, the material in the apex
of the product peak often precipitates, resulting in a turbid solution. The level of
precipitation in the peak apex of the untreated sample, as measured by absorbance at 600nm
in a spectrophotometer, was 1.85AU. The level of precipitation in the peak apex of the
f10cculated sample, as measured by absorbance at 600nm in a spectrophotometer, was
0.03AU, a 98% reduction in turbidity as compared to the untreated sample.
When antibodies are eluted from Protein A columns, the peak pools are occasionally
turbid. Upon neutralization, the turbidity typically increases significantly between pH 5.5
and pH 6.0, as material precipitates and falls out of solution. Some of this precipitate tends
to become soluble again as the pH is raised above 7. The untreated peak pool, with an
-45-

WO 2007/035283 PCT/US2006/035025
antibody concentration of 8.1 mg/mL had a turbidity of 8.5 NTU when eluted, increased to
839 NTU between pH 5.5 and 6.0, and decreased to 53 NTU at pH 7.0. Upon filtration
through a 0.2ujn sterilizing grade filter, the turbidity was only reduced to 40 NTU.
The f10cculated product pool eluted from the column in less volume and therefore
was more concentrated. The f10cculated peak pool, with an antibody concentration of 15.1
mg/mL had a turbidity of 5.6 NTU when eluted, increased to 31 NTU between pH 5.5 and
6.0, a 96% reduction in turbidity from the control. The turbidity increased slightly to 46
NTU at pH 7.0 at 15.1mg/mL. Upon filtration through a 0.2um sterilizing grade filter, the
turbidity was reduced to 8.1 NTU, a reduction of 80% from the control.
"When the neutralized Protein A peak was diluted to a concentration of 8.1 mg/mL to
match the untreated sample, the turbidity of the unfiltered f10cculated sample decreased from
46 NTU to 25 NTU, a 53% reduction as compared to the untreated sample. Upon filtration
through a 0.2um sterilizing grade filter, the turbidity was reduced to 4.1 NTU, a 90%
reduction in turbidity as compared to the untreated sample.
The level of high molecular weight (HMW) aggregate present in the Protein A peak
pool was measured by Size-Exclusion HPLC. The level of aggregate in the untreated sample
was 9.51%. The level of aggregate in the f10cculated sample was 1.05%, an 89% reduction
in aggregate as compared to the untreated sample. With the reduction in aggregate taken into
account, the 78% product recovery in the culture translates to an 85% recovery of the desired
monomer.
The levels of host cell proteins (HCP), unwanted impurities secreted by the CHO
cells, were measured at different steps of the process using an ELISA. The HCP levels are
reported as parts per million (ppm), equivalent to ng of HCP per mg of antibody. The HCP
level in the untreated culture was 2.53E6 ppm. The HCP level in the f10cculated conditioned
medium was 3.83E5 ppm, an 84% reduction from the untreated culture. Both the untreated
and f10cculated cultures had an approximately 60% reduction in HCP through the pad filters
to 1.02E6 and 1.62E5 respectively. Upon purification over the Protein A column, the levels
of HCP in the untreated sample were reduced by 90% to 1.03E5 ppm. The levels of HCP in
the f10cculated sample were reduced by 98% to 3.83E2 ppm. Overall, there was a 1.4 10g
removal of HCP for the untreated purification train. The f10cculated purification train
-46-

WO 2007/035283 PCT/US2006/035025
achieved a 3.8 10g removal of HCP resulting in a 250-fold reduction in HCP as compared to
the untreated purification train.
While a number of implementations have been described, the invention is not so
limited.
As an example, in some implementations, the flocculation methods described herein
can be performed without cells present, for example, after the cells have been removed. The
medium may contain non-cellular insoluble material (see Example 5 below).
Example 5
The use of calcium and phosphate to form a solid precipitate to aid in filtration of a
turbid protein-containing solution: A mAb, anti-GDF8 # 1, was produced by recombinant
CHO cells cultured in serum free media. The cells were removed by an Alfa Laval BTPX
205 disc stack centrifuge and the resulting centrate was processed through Millipore A1HC
Pad Filters. After pad filtration and additional filtration through a 0.2um polishing filter, the
samples were chromatographed using a GE Healthcare MabSelect Protein A affinity column.
When antibodies are eluted from Protein A columns, using a low pH buffer, the peak pools
are occasionally turbid. Upon neutralization, the turbidity typically increases significantly.
In this example, the Protein A peak was held unfirtered for 7 days at 4°C and then
warmed up to room temperature. The turbidity of the peak was 192 NTU. The peak was
split into two 800mL samples, with one sample left untreated. To the second sample was
added 4mM potassium phosphate and 6mM calcium ch10ride. The treated sample was then
shaken in a 2L Erlenmeyer flask for 1 hour at 18-25°C. After shaking the turbidity of the
treated sample was 460 NTU.
Both the untreated and treated samples were then filtered through 17.7 cm2 Millipore
Express SHC 0.5/0.2um polyethersulfone capsule filters. Based on the amount of solution
that was able to pass through each filter, a maximum filter capacity was calculated. The
maximum filter capacity is the number of liters of solution that can pass through 1 m2 of
filter before the filter will plug and no more solution can pass. The untreated sample was
able to achieve a maximum filter capacity of approximately 30 L/m2 while the treated sample
achieved a maximum filter capacity of approximately 1500 L/m2, a 50-fold increase in filter
capacity. The post-filter turbidity for both the untreated and treated samples was 9 NTU,
indicating that the treatment did not likely result in the removal of additional particulate
-47-

WO 2007/035283 PCT/US2006/035025
matter, but did serve as a filter-aid and allowed larger volumes of solution to pass through the
filter before it plugged.
Example 6
The effect of scale and mixing method on the use of calcium and phosphate to form a
solid precipitate to aid in the removal of particulate matter in a turbid protein-containing
solution: Anti-AB #2 mAb, noted as "mAb B" in FIGS. 2 and 3, was produced by
recombinant CHO cells cultured in serum free media in pilot scale bioreactors (160-500L cell
culture). Approximately 125L, 1.5L or 50mL of the cell-containing conditioned media was
aliquoted in to a: 200L carboy, a 2L beaker, or a 125mL Erlenmeyer flask, respectively.
HEPES was added to 40mM to control the pH. The 200L carboy and the 2L beaker were
mixed with impellers. The 125mL Erlenmeyer was mixed by a shaker. A concentrated
solution of potassium phosphate was added to each mixture to achieve a final concentration
of 20mM in the final solution. A concentrated solution of calcium ch10ride was added to each
solution to achieve a final target concentration of 30mM and mixed. The final pHs were
between 7.0 and 7.5. The solid and medium were incubated for greater than one hour under
mixing conditions at room temperature (20-23°C). 50mL aliquots of an untreated sample,
and the 125L, 1.5L, and 50mL treated samples were centrifuged for 10 minutes at 340xg.
The turbidities and antibody concentrations in the supernatants were measured.
i The effect on turbidity of the scale and method of mixing is shown in FIG. 2. The
turbidity is reduced by more than 90% from the control in all of the flocculation examples
independent of scale and independent of mixing method (impeller or shaker). Additionally,
the antibody recovery in all treated samples was 100% as compared to the untreated sample.
Thus, it does not appear that flocculation in the present invention is dependent upon scale or
i upon mixing method for these widely different conditions.
The levels of host cell proteins (HCP), unwanted impurities secreted by the CHO
cells, were measured in the untreated supernatant and the treated 125L sample using an
ELISA. The HCP levels are reported as parts per million (ppm), equivalent to ng of HCP per
mg of antibody. The treated 125L sample had a reduction in HCP of 50% as compared to the
i untreated sample.
-48-

WO 2007/035283 PCT/US2006/035025
Example 7
The effect of mixing speed on the use of calcium and phosphate to form a solid
precipitate to aid in the removal of particulate matter in a turbid protein-containing solution:
Anti-AB #2 mAb, noted as "mAb B" in FIG. 3, was produced by recombinant CHO cells
cultured in serum free media in pilot scale bioreactors (160-500L cell culture).
Approximately 125L of the cell-containing conditioned media was placed in to two 200L
carboys. HEPES was added to 40mM to control the pH. The mixing was performed with
impellers, one operated at a tip speed of 0.9m/s and the other at 2.5m/s. A concentrated
solution of potassium phosphate was added to the mixture to achieve a final concentration of
20mM in the final solution. A concentrated solution of calcium ch10ride was added to the
solution to achieve a final target concentration of 30mM and mixed. The final pH was
between 7.0 and 7.5. The solid and medium were incubated under mixing conditions at room
temperature (20-23°C), and samples were taken at various time points. The turbidity of each
sample supernatant was measured after centrifuging at 340xg for ten minutes. The effect on
turbidity of the impeller tip speed is shown in FIG. 3. The turbidity is reduced by more than
90% from the control in all of the fiocculation examples independent of the tip speeds
investigated after one hour. The faster tip speed appeared to have a faster reduction in
turbidity at the 15 minute time point. This difference is, however, not significant. Thus, it
does not appear that flocculation in the present invention is dependent upon tip speed for the
speeds investigated after one hour of incubation.
The levels of host cell proteins (HCP), unwanted impurities produced by the CHO
cells, were measured in the untreated supernatant and the treated 0.9m/s sample using an
ELISA. The HCP levels are reported as parts per million (ppm), equivalent to ng of HCP per
mg of antibody. The treated sample had a reduction in HCP of 47% as compared to the
i untreated sample.
Example 8
Large Scale Flocculation:
A pilot-scale experiment using calcium and phosphate as precipitants and the effects
i on downstream clarification and chromatography steps: A mAb, anti-IL-13 #1 (noted as
mAb E in FIGS. 4 and 5), was produced in five different batches by recombinant CHO cells
-49-

WO 2007/035283 PCT/US2006/035025
cultured in serum free media in a 190L bioreactor. The bioreactor was run with a 12-14 day
culture time, and the cells had a viability between 8x10 -11x10 viable cells/mL, and were
66-88% viable. At the time of harvest, the culture was brought to room temperature (18-
25°C). 150L of the culture was transferred to a 200L carboy (for the first four batches) or
left in the bioreactor (for the last batch) and slowly mixed with an overhead mixer. HEPES
buffer was added to 40mM to control the pH. 2M potassium phosphate was then added to a
concentration of 20mM. 5M calcium ch10ride was then added to a concentration of 30mM.
The pH of the mixture was between 7 and 7.5. The f10cculated culture was incubated for
between 2 and 3 hours while mixing.
The f10cculated cultures were processed through an Alfa Laval BTPX 205 disc stack
centrifuge at a flow rate between 4 and 5 L/min and a bowl speed of 7630 RPM (8000g).
The turbidity of the centrates obtained are compared to the supernatant of the unf10cculated
sample (obtained from the centrifugation of the untreated control at 340g for ten minutes) in
FIG. 4: In all cases, the flocculation reduces the turbidity by greater than 85%. The recovery
of antibody titer in the f10cculated supernatants is shown in FIG. 5 as a function of incubation
time, and is above 75% in all cases. 50mL samples were taken from batches 403 and 405 and
incubated in 125 mL Erlenmeyer flasks for additional time. After 7 hours, batch 403 had a
titer of 60% of the untreated sample, and batch 405 had a titer of 80% of the untreated
sample.
The levels of host cell proteins (HCP), unwanted impurities secreted by the CHO
cells, were measured in the untreated supernatants and treated centrates using an ELISA. The
HCP levels are reported as parts per million (ppm), equivalent to ng of HCP per mg of
antibody. The treated samples from the 5 batches showed reductions in HCP from 49%-69%
as compared to the untreated samples.
The first 3 batches were processed through Millipore A1HC Pad Filters to a capacity
of 270 L of centrate per square meter of filter at a flux of 120 liters per square meter per hour
with no rise in pressure or turbidity.
After pad filtration, the samples generally remained stable for many days as measured
by turbidity.
-50-

WO 2007/035283 PCT/US2006/035025
The last 2 batches went directly over 0.2um filters without going through pads. Filter
capacities of 730 and 160 L/m2 were achieved, respectively, without the use of pad filters
prior to the 0.2pxn filters. These filter capacities represent a significant improvement in
filterability over that of unf10cculated material. Upon holding the 0.2um-filtered centrates
(without pads) at room temperature (18-24°C), the turbidity begins to increase within a few
hours due to continued precipitation of the calcium and phosphate. After 24-48 hours the
precipitate settles, forming a crystalline layer of calcium phosphate at the bottom of the
container. The resulting clarified protein containing solution has excellent filterability
characteristics both before and after the precipitate settles. No antibody is 10st in the
precipitate.
Flocculation for the final batch was performed directly in the pilot bioreactor. The
reactor was effectively cleaned using standard clean-in-place (OP) procedures (water rinse
followed by a 0.1N NaOH wash at 60-80°C)
All 5 batches were processed at pilot scale through a Protein A column, anion
exchange step, virus retaining filter, and final ultrafutration/diafiltration (UF/DF) with no
operational issues. Neutralized Protein A peaks all had turbidities of <20 NTU and were
highly filterable. Product quality, such as levels of high molecular weight aggregate and low
molecular weight clips, as measured by size exclusion HPLC and SDS-PAGE gel
electrophoresis, and levels of acidic and basic species, as measured by cation exchange
HPLC, were comparable to a previous non-f10cculated pilot campaign with this antibody.
Example 9
The use of calcium and phosphate as precipitants and the effects on a downstream
Protein A chromatography step and subsequent filtration: Three mAbs, anti-AB #2, anti-
GDF8 #1, and anti-IL22 were produced by recombinant CHO cells cultured in serum free
media. For each, the cultures were split in half, with the first sample being left untreated. To
the second sample (the treated sample) was added HEPES to a level of 40mM, potassium
phosphate to a level of 20mM, and calcium ch10ride to a level of 30mM. The cells from all
samples were removed by centrifugation and the resulting supernatants were processed
through Millipore A1HC Pad Filters. After pad filtration and additional filtration through
0.2um polishing filters, the samples were chromatographed using GE Healthcare MabSelect
-51-

WO 2007/035283 PCT/US2006/035025
Protein A affinity columns. When antibodies are eluted from Protein A columns, using a low
pH buffer, the peak pools are occasionally turbid. Upon neutralization, the turbidity typically
increases significantly.
The neutralized peak turbidities of the untreated and treated samples are shown in
Table 2 below. All three treated samples showed a significant reduction in neutralized peak
turbidity as compared to the untreated samples.
Table 2: Decrease in Protein A Peak turbidity after calcium phosphate treatment of
cell culture prior to loading

Both the untreated and treated anti-IL22 neutralized peaks were then filtered through
2.8 cm2 Pall Acrodisc Supor 0.8/0.2umpolyethersulfone syringe filters. Based on the
amount of solution that was able to pass through each filter, a maximum filter capacity was
calculated. The maximum filter capacity is the number of liters of solution that can pass
through 1 m2 of filter before the filter will plug and no more solution can pass. The untreated
sample was able to achieve a maximum filter capacity of approximately 10-30 L/m2. 170
L/m2 of the treated sample passed through the filter without a reduction in flow, at which
point no treated sample remained. The treated sample passed through the filter too rapidly to
accurately calculate a maximum filter capacity. However, as there was no reduction in flow
out to a challenge of 170 L/m2, it can be assumed that the maximum capacity would have
been significantly greater than 170 L/m2.
All publications, patent applications, patents, and other references mentioned herein
are incorporated by reference herein in their entirety.
Other embodiments are within the scope of the following claims.
-52-

WO 2007/035283 PCT/US2006/035025
WHAT IS CLAIMED IS:
1. A method, comprising:
forming a solid comprising a first cation and a first anion in a medium comprising
a protein; and
separating the solid from the protein.
2. The method of claim 1, wherein the first cation is selected from the group
consisting of calcium, magnesium, strontium, aluminum, scandium, lanthanum, silicon,
titanium, zirconium, thorium, manganese, cobalt, copper, chromium, iron, nickel, zinc,
and vanadium.
3. The method of claim 1, wherein the first cation is calcium.
4. The method of claim 1, wherein the first anion is selected from the group
consisting of phosphate, carbonate, chromate, tungstate, hydroxide, halide, succinate,
tartrate, citrate, sulfite, molybdate, nitrate, fluoride, silicate, and alginate.
5. The method of claim 1, wherein the first anion is phosphate.
6. The method of claim 1, wherein the first cation is calcium, and the first
anion is phosphate.

7. The method of claim 1, wherein the solid has a solubility product constant
of no more than about 10^M2.
8. The method of claim 1, further comprising introducing from about 4 mM to
about 200 mM of the first cation or the first anion into the medium.
9. The method of claim 1, wherein the product of the concentrations of the first
cation and the first anion before forming the solid is greater than about 10" M2.
-53-

WO 2007/035283 PCT/US2006/035025
10. The method of claim 1, wherein the product of the concentrations of the first
cation and the first anion before forming the solid is greater than about 10"^2.
11. The method of claim 1, wherein the product of the concentrations of the first
cation and the first anion before forming the solid is greater than about 2.7 x 10'2 M2.
12. The method of claim 1, wherein the concentrations of the first cation and the
first anion in the medium are different.
13. The method of claim 1, wherein the concentrations of the first cation and the
first anion in the medium are substantially the same.
14. The method of claim 1, further comprising changing the pH of the medium.

15. The method of claim 1, wherein the pH of the medium is maintained
between about 5 to about 9.
16. The method of claim 1, wherein at least about 50% of the protein in the
medium is separated.
17. The method of claim 1, wherein at least about 70% of the protein in the
medium is separated.
18. The method of claim 1, further comprising decreasing the turbidity of the
clarified medium by at least about 30% relative to a second clarified medium identical to
the medium and free of the solid.
19. The method of claim 1, further comprising decreasing the turbidity of the
clarified medium by at least about 50% relative to a second clarified medium identical to
the medium and free of the solid.
-54-

WO 2007/035283 PCT/US2006/035025
20. The method of claim 1, wherein the medium further comprises mammalian
cells.
21. The method of claim 1, wherein the medium further comprises eukaryotic
cells.
22. The method of claim 1, further comprising centrifuging the medium,
filtering the medium through a microfiltration membrane, or filtering the medium through
a depth filter.
23. The method of claim 1, wherein the solid further comprises a second cation
species or a second anion.

24. The method of claim 1, wherein the medium comprising the protein, after
the solid is formed and separated, is applied to a Protein A column and eluted to provide
an eluted peak having a lower turbidity than a similarly eluted peak of a second medium
identical to the first medium and free of formation of the solid.
25. The method of claim 1, wherein the medium comprising the protein, after
the solid is formed and separated, is applied to a Protein A column and eluted to provide
an eluted peak having a lower soluble impurity level than an eluted peak of a second
medium identical to the medium and free of formation of the solid.
26. A method, comprising:
introducing a first cation and a first anion into a medium comprising a protein;
precipitating a solid comprising the first cation and the first anion; and
separating the solid from the protein.
27. The method of claim 26, wherein the first cation and the first anion are
introduced sequentially.
-55-

WO 2007/035283 PCT/US2006/035025
28. The method of claim 26, wherein the first cation and the first anion are
introduced simultaneously.
29. The method of claim 26, wherein the first cation is selected from the group
consisting of calcium, magnesium, strontium, aluminum, scandium, lanthanum, silicon,
titanium, zirconium, thorium, manganese, cobalt, copper, chromium, iron, nickel, zinc,
and vanadium.
30. The method of claim 26, wherein the first anion is selected from the group
consisting of phosphate, carbonate, chromate, tungstate, hydroxide, halide, succinate,
tartrate, citrate, sulfite, molybdate, nitrate, fluoride, silicate, and alginate.
31. The method of claim 26, wherein the solid has a solubility product constant
of no more than about 10^M2.
32. The method of claim 26, further comprising introducing from about 4 mM
to about 200 mM of the first cation or the first anion into the medium.

33. The method of claim 26, wherein the product of the concentrations of the'
first cation and the first anion is greater than about 10"5M2.
34. The method of claim 26, wherein the product of the concentrations of the
first cation and the first anion is greater than about 10^M2.
35. The method of claim 26, wherein the product of the concentrations of the
first cation and the first anion is greater than about 2.7 x 10"2 M2.
3 6. The method of claim 26, comprising introducing different concentrations of
the first cation and the first anion into the medium.
-56-

WO 2007/035283 PCT/US2006/035025
37. The method of claim 26, comprising introducing the same concentration of
the first cation and the first anion into the medium.
38. The method of claim 26, further comprising changing the pH of the
medium.
39. The method of claim 26, wherein the pH of the medium is maintained
between about 5 to about 9.
40. The method of claim 26, further comprising adjusting the temperature of the
medium.
41. The method of claim 26, wherein at least about 50% of the protein in the
medium is separated.

42. The method of claim 26, wherein at least about 70% of the protein in the
medium is separated.
43. The method of claim 26, further comprising decreasing the turbidity of the
clarified medium by at least about 30% relative to a second clarified medium identical to
the medium and free of the solid.
44. The method of claim 26, further comprising decreasing the turbidity of the
clarified medium by at least about 50% relative to a second clarified medium identical to
the medium and free of the solid.
45. The method of claim 26, wherein the medium further comprises mammalian
cells.
46. The method of claim 26, wherein the medium further comprises eukaryotic
cells.
-57-

WO 2007/035283 PCT/US2006/035025
47. The method of claim 26, further comprising centrifuging the medium.
48. The method of claim 26, wherein the solid further comprises a second cation
or a second anion.
49. The method of claim 26, wherein the medium comprising the protein, after
the solid is formed and separated, is applied to a Protein A column and eluted to provide
an eluted peak having a lower turbidity than a similarly eluted peak of a second medium
identical to the first medium and free of formation of the solid.
50. The method of claim 26, wherein the medium comprising the protein, after
the solid is formed and separated, is applied to a Protein A column and eluted to provide
an eluted peak having a lower soluble impurity level than an eluted peak of a second
medium identical to the medium and free of formation of the solid.

51. The method of claim 1, wherein the protein is a secreted protein.
52. The method of claim 51, wherein the protein is selected from the group
consisting of an antibody, an antigen-binding fragment of an antibody, a soluble receptor,
a receptor fusion, a cytokine, a growth factor, an enzyme, and a clotting factor.
53. The method of claim 52, wherein the protein is an antibody or an antigen-
binding fragment thereof.
54. The method of claim 53, wherein the antibody or antigen-binding fragment
thereof binds to an Ap peptide, interleukin-13, interleukin-22,5T4, or growth
differentiation factor-8.
55. The method of claim 26, wherein the protein is a secreted protein.
-58-

WO 2007/035283 PCT/US2006/035025
56. The method of claim 55, wherein the protein is selected from the group
consisting of an antibody, an antigen-binding fragment of an antibody, a soluble receptor,
a receptor fusion, a cytokine, a growth factor, an enzyme, and a clotting factor.
57. The method of claim 56, wherein the protein is an antibody or an antigen-
binding fragment thereof.
58. The method of claim 57, wherein the antibody or antigen-binding fragment
thereof binds to an Ap" peptide, interleukin-13, interleukin-22, 5T4, or growth and
differentiation factor-8.
59. A method, comprising: (i) forming a solid that includes a first cation and a
first anion in a medium comprising a target moiety and a turbidity-causing agent; and (ii)
separating the solid and turbidity-causing agent from the solution by filtration.

60. The method of claim 59, wherein the target moiety is a protein.
61. The method of claim 60, wherein the protein is a soluble protein.
62. The method of claim 26, wherein the first cation is calcium.
63. The method of claim 26, wherein the first anion is phosphate.
64. The method of claim 26, wherein the first cation is calcium, and the first
anion is phosphate.
65. The method of claim 1, further comprising adjusting the temperature of the
medium.
-59-

Separation methods,
for example, to isolate a recombinant
protein, are disclosed. The methods
include forming a solid containing a
first cation and a first anion in a medium
containing a protein, and separating the
solid from the protein.