Skip to main content

Full text of "USPTO Patents Application 09804625"

See other formats


WORLD INTELLECTUAL PROPERTY ORGANIZATION 
Internationa] Bureau 




PCX 

INTERNATIONAL APPUCATION PUBUSHED UNDER THE PATENT COOPERATION TOEATY (PCT) 



(51) International Patent Classification ^ : 
C12N 15/62, C07K 14/54, 1/18, 1/30 



Al 



(11) International Publication Numlier: WO 96/38570 

(43) International Publication Date: 5 December 1996 (05.12.96) 



(21) International AppUcation Number: PCT/US96/048 1 1 

(22) International FUing Date: 1 1 April 1996 (1 1.04.96) 



(30) Priority Data: 

08/464.176 



2 June 1995 (02.06.95) 



US 



(71) Applicant: GENETICS INSTITUTE, INC. [USAJS]; 87 Cam- 

bridgcPark Drive, Cambridge, MA 02140 (US). 

(72) Inventors: VICIK, Steven. M.; 23 Union Street. Natick. MA 

01760 (US). SCHAUER, NeU. L.; 4 Woodridge Road. 
Milfoid. MA 01757 (US). MHICER, James, R.; 12 
East Deny Road, Deny, NH 03038 (US). LEVALLIE. 
Edward, R.; 90 Green Meadow Drive. Tewksbury, MA 
01876 (US). BRIASCO. Catherine . A.; 1 Arborwood Drive, 
Buriington. MA 01803 (US). DEEIZ, Jeffftey, S.; 235 
Beech Avenue, Melrose. MA 02176 (US). WINTERS, 
Dwight; 263 Appletrec Avenue, Camarillo, CA 93012 (US). 
THOMAS. Jenifer, L.; ApartmCTt 10, 3 Brook Road. Salem. 
NH 03079 (US). 

(74) Agent: MEINERT, M., C; <3cnetics Institute, Inc., - Legal 
Affairs. 87 CambridgcPark Drive, Cambridge, MA 02140 
(US). 



(81) Designated States: AU, CA. JP. European patent (AT, BE. 
CH, DE, DK, ES. H, FR. GB, GR, IE, IT, LU. MC. NL, 
PT, SE), 



Published 

With international search report. 



(54) Titie: NOVEL FUSION PROTEIN RECOVERY AND PURIFICATION METHODS 
(57) Abstract 



Provided by the present invention are novel methods of protein recovery and purification methods and more specifically novel methods 
for the recovery and purification thioredoxin fusion proteins, especially of IL-1 1 



FOR THE PURPOSES OF INFORMATION ONLY 



Codes used to identify States party to the PCT on the front pages of pamphlets publishing international 
applications under the PCT. 



AM 


Annenia 


GB 


United Kingdom 


MW 


Malawi 


AT 


Austria 


GE 


Georgia 


MX 


Mexico 


AU 


Australia 


GN 


Guinea 


NE 


Niger 


BB 


Barbados 


GR 


Greece 


NL 


Netherlands 


BE 


Belgium 


HU 


Hungary 


NO 


Norway 


BF 


Burkina Faso 


IE 


Ireland 


NZ 


New Zealand 


BG 


Bulgaria 


n 


Italy 


PL 


Poland 


BJ 


Benin 


jp 


Japan 


FT 


Porti^al 


BR 


Brazil 


KE 


Kenya 


RO 


Romania 


BV 


Belarus 


KG 


Kyrgystan 


RU 


Russian Federation 


CA 


Canmlfl 


KP 


Democratic People's Republic 


50 


Sudan 


CF 


Central African Rcpablic 




of Korea 


SE 


Sweden 


CG 


Congo 


KR 


Republic of Korea 


SG 


Singapore 


CH 


Switzerland 


KZ 


Kazakhstan 


SI 


Slovenia 


a 


Cflie d'lvoiie 


U 


Liechtenstein 


SK 


Slovakia 


CM 


Cameroon 


LK 


Sri Lanka 


SN 


Senegal 


CN 


China 


LR 


Liberia 


sz 


Swaziland 


cs 


Czechoslovakia 


LT 


Lithuania 


TD 


Chad 


cz 


Czech Republic 


LU 


Luxembourg 


TG 


Togo 


DE 


Germany 


LV 


Latvia 


TJ 


Tajikistan 


DK 


Denmark 


MC 


Monaco 


TT 


Trinidad and Tobago 


EE 


Estonia 


MD 


Republic of Moldova 


UA 


Ukrame 


ES 


Spain 


MG 


Madagascar 


UG 


Uganda 


Ft 


Hnland 


ML 


Mali 


US 


United States of America 


FR 


France 


MN 


Mongolia 


uz 


Uzbekistan 


GA 


Gabon 


MR 


Mauritania 


VN 


Viet Nam 



wo 96/38570 



PCTAJS96/04811 



Novel fusion protein recovery and purification methods 

FIELD OF INVENTION 

5 The present invention relates generally to novel protein recovery and purification 

methods and more specifically to novel methods for the recovery and purification of 
thioredoxin-like fusion proteins, especially of IL-11. 

BACKGROUND OF THE INVENTION 
The advent of recombinant technology now allows for the production of high levels of 
10 proteins within suitably transformed host cells. Where the host cell does not secrete the protein 
of interest, it is necessary to release the protein from the cells and to then further purify the 
protein. 

The initial step in the purification of an intracellular protein is release of the protein 
to the extracellular medium. This is typically accomplished using mechanical disruption 

IS techniques such as homogenization or bead milling. While the protein of interest is generally 

effectively liberated, such techniques have several disadvantages: Engler, Protein Purification 
Process Engineering, Harrison eds.:37-55 (1994). Temperature increases, which often occur 
during processing, may result in inactivation of the protein. Moreover, the resulting suspension 
contains a broad spectrum of contaminating proteins, nucleic acids, and polysaccharides. 

20 Nucleic acids and polysaccharides increase solution viscosity, potentially complicating 

subsequent processing by centrifugation, cross-fiow filtration, or chromatography. Complex 
associations of these contaminants with the protein of interest can complicate the purification 
process and result in unacceptably low yields. As such, more selective means of releasing 
intracellular proteins facilitates further downstream processing. Several techniques have been 

25 reported to permeabilize cells and/or to extract intracellular proteins. These methods include 

the use of solvents, detergents, chaotropic agents, antibiotics, enzymes, and chelating agents 
to enhance cell permeability and/or promote extraction. Additions of certain compotmds, such 
as glycine, to the fermentation medium during culture growth have also been reported to 
promote release of certain intracellular enzymes. Finally, techniques such as freeze-thaw 

30 treatment or osmotic shock have also been shown to release subsets of intracellular proteins. 

However, these techniques are not necessarily applicable to all intracellular £. co/z proteins, 
V and all have limited application for large scale processing, and/or other disadvantages. 

For example, while solvents such as toluene and x:hloroform promote release of 
intracellular proteins, these substances are known to be toxic and/or carcinogenic. Belter et 



wo 96/38570 



PCTAJS96/04811 



fl/., Bioseparations - Downstream Processing for Biotechnology ill -94 (1988). Ames, J. of 
Bact. 1^:1181-1183 (1984). Windholtz et aL, The Merck Index 10th Edition: 300 & 1364 
(1983). Naglak et al.. Separation Processes in Biotechnology, Asenjo eds.: 177-205 (1990). 
Ionic detergents, such as SDS, often irreversibly denature isolated proteins (Scopes, Protein 

5 Purification Principles and Practice 3rd edition. Cantor eds., 22-43 (1994)). Although non- 
ionic detergents like Triton X-100 or Triton X-1 14 are not normally denaturing, the recovered 
proteins are often associated with detergent micelles which can require additional processing 
to yield detergent-free protein. Scopes, supra. Chaotropic agents, such as urea and guanidine 
hydrochloride, can be denaturing at the concentrations required for complete release (Naglak 

10 et al., supra., Hettwer, Biotechnology and Bioengineering 33:886-895 (1989)). Their 
effectiveness may be dependent on the growth phase of the culture (Ingram, L., Bacteriol. 
146:331-336 (1981)). Antibiotics, such as polymyxin B, which affect the permeability of E. 
coli (Hancock, Ann. Rev. Microbiol. 38:237-264 (198i4)), are not typically used in the 
pharmaceutical industry due to general concerns over the use of antibiotics in manufacturing 

15 processes. The use of lysozyme, which provides for a relatively gentle means of protein 

release, is limited because of its relatively high cost of the enzyme (Belter et aL, supra,, 
Naglak et aL , supra.) and because of the subsequent need to purify the protein of interest from 
the enzyme reagent {Naglak et al, supra,). In addition, chelating agents, often used to 
enhance the effectiveness of other permeabilizing/release techniques such as lysozyme (Naglak 

20 etal, supra, Bucke, Principles of Biotechnology, Wiseman eds.: 151-171 (1983)), toluene (De 
Smet et al., Biochim. Biophys. Acta 506:64-80 (1978)), or Triton X-100 (Leive, Annals New 
York Academy of Sciences 235: 109-129 (1974)) extraction, suffer from the disadvantage of non- 
specific release of the protein of interest. For exan^)le. the use of chelator releases up to 18% 
of intracellular £. coli protein and can result in imdesired complex formation with 

25 lipopolysaccharides (LPS) and phosphatidyl-ethanolamine. Naglak, et aL supra at 185. 

The use of other methods for protein release also have disadvantages. For example, 
osmotic shock, in which cells are suspended in a high osmolarity medium, recovered, and 
subsequently placed in a low osmolarity buffer (Nossal et aL, J. Biol. Chem. 241:3055-3062 
(1966)) requires additional processing steps with respect to other extraction alternatives (Moir 

30 et aL, Separation Processes in Biotechnology, Asenjo eds: 67-94 (1990)) or necessitates the 
handling of large liquid volumes at low temperatures (Naglak et aL, supra,). This renders the 
method unattractive for large scale processing. Freeze-thaw treatment also releases intracellular 
proteins, although relatively low yields often resuh in multiple cycles or additional processing 
requirements (Naglak et aL, supra.; Bucke, supra.). In addition, cell paste freezing is an 



2 



wo 96/38570 



PCTAJS96/04811 



added non-trivial processing requirement compared with other extraction alternatives. Finally, 
reagents, such as glycine, have been added during fermentation to promote protein release to 
the extracellular medium (Aristidou et al.. Biotechnology Letters 15:331-336 (1993)), While 
partial release of several intracellular proteins has been reported, this approach requires direct 
5 coupling of fermentation and release strategies and subsequent separation of the protein of 
interest from a potentially complex extracellular broth. 

Once the protein is released from the host cell, purification of the protein of interest 
from other cell components is required. Unfortunately, most extraction approaches, such as 
cell lysis, not only expose the protein to potential degradation by host cell proteases, but also, 

10 make isolation of the protein from other elements of the resulting suspension more difficult. 
For example, the presence of negatively charged molecules, such as DNA, RNA, 
phospholipids, and lipopolysaccharides (LPS), often require the use of anion exchange 
chromatography (Sassenfeld, TIBTECH 8:88-93 (1990).; Spears, Biotechnology vol 3 - 
Bioprocessing, Rehm eds:40-51 (1993)) and/or precipitation with poly cat ions, such as 

15 protamine sulfate (Kelley et al. , Bioseparation 1 :333-349 (1991); Scopes, Protein Purification 

Principles and Practice, 2nd edition. Cantor eds., 21-71 (1987)), streptomycin sulfate (Wang 
et al, eds. Fermentation and Enzyme Technology:253-256 (1979)), polyethylenimine (PEI) 
(Kelley et aL, supra.; Sassenfeld, supra.), and/or aqueous two phase extraction with 
immiscible polymer systems such as polyediylene glycol (PEG)/phosphate or PEG/dextran 

20 (Kelley et aL, supra., Strandberg et aL, Process Biochemistry 26:225-234 (1991).). 

Alternatively, the protein of interest may be precipitated away from non-proteinaceous 
polyanionic contaminants through the addition of a neutral salt such as ammonium sulfate or 
potassium chloride (Wheelwright, Protein Purification: Design and Scale up of Downstream 
Processing: 87-98 (1991); Englard etaL, Methods in Enzymology Volume J82, Deutscher eds.: 

25 285-300 (1990)) and/or a polymer such as PEG or dextran sulfate (Wang et aL, supra.; 
Wheelwright, supra.). Where the protein of interest is positively charged, it will tend to bind 
to any negatively charged molecules present thereby making purification of the protein virtually 
impossible. 

Typically, researchers have utilized the initial fractionation steps, described above, to 
30 separate the offending polyanions from the protein of interest. Unfortunately, -each of these 
initial separation methods suffers from severe disadvantages, especially when used in the 
manufacture of pharmaceutical reagents. For example, the large quantities of non- 
proteinaceous polyanionic contaminants found in bacterial lysates tend to reduce the binding 
capacities of anion exchange chromatography resins. In addition, regeneration protocols are 



3 



wo 96/38570 PCTAJS96/0481 1 

often rendered ineffective due to tenacious binding of the poly anions to the resins (Spears, 
supra,). Finally, the low ionic strength conditions that favor protein binding are ineffective at 
disrupting polyanion-protein interactions and result in a lack of separation (Scopes, Protein 
Purification Principles and Practice 3rd edition. Cantor eds. 171 (1994)). Prolamine sulfate 
5 preparations are plagued by concerns over protease and viral contaminations. Moreover, 
unwanted protein precipitation can occur using this reagent (Scopes, Protein Purification 
Principles and Practice 2nd edition. Cantor eds., 21-71 (1987)), In the processing of 
pharmaceutical proteins, streptomycin sulfate is generally not used due to general apprehension 
over the use of antibiotics as process reagents (Scawen et ai. Handbook of Enzyme 

10 Biotechnology 2nd edition, Wiseman eds.:15-53 (1985)). PEI preparations are often 
contaminated with varying amounts of the ethylenimine monomer, a suspected cancer agent 
(Scawen et aL, supra.). PEI also tends to bind irreversibly to many chromatography resins, 
thereby limiting their effectiveness and the number of potential chromatography r^ins available 
for use post-PEl clarification. In general, aqueous two phase extractions systems are difficult 

15 to predict and often require an empirical approach for determining conditions that move the 
protein of interest into the appropriate aqueous phase (Kelley et aL, supra.). Techniques that 
specifically precipitate the protein of interest often result in the entrapment of the non- 
proteinaceous contaminants in the precipitate rendering the separation ineffective (Scopes, 
supra. ; Wheelwright. Protein Purification: Design and Scale up of Downstream Processing: 

20 87-98 (1991)). 

Accordingly, there continues to exist a need in the art for both effective protein release 
methods (that minimize or eliminate release of non-proteinaceous contaminants) as well as 
effective purification methods (that remove non-proteinaceous contaminants, especially 
polyanionic contaminants) and an overall release and purification process that is readily 

25 executed at large scale. 

BRIEF SUMMARY OF THE INVENTION 

Provided by the present invention is a method for the release and purification of a 
thioredoxin-like fusion protein from the cell into a solution by adding chelator to the solution 
which releases the fusion protein from the cell. Optionally, the temperature prior to the 
30 addition of chelator may be substantially cooler than after the addition of chelator, e.g., 
representing a 20-40 degree C temperature differential, with a 35**C differential preferred. The 
method is particularly amenable to large scale processing, as defined to be greater than 10 L 
of cell suspension volume, and upwards of hundreds of liters of total volume. Divalent 
cation/alcohol solution is then added thereby forming a soluble fraction, which contains the 



4 



wo 96/38570 



PCTA3S96/04811 



fusion protein, and an insoluble fraction, which contains unwanted contaminants. The divalent 
cation includes for example magnesium, manganese and calcium, alone or in combination. The 
alcohol can be methanol, ethanol, propanol, isopropanol, iso-butanol, and tertiary-butanol. 
Preferably, the final alcohol concentration is from 5 to 30% with 14% ethanol being preferred. 

5 The resulting divalent cation concentration can range from 1 to 1000 mM, with preferred being 

50 to 200 mM, and most preferred being 200 mM magnesium. Also preferred are a 
combination of 125 mM magnesium and 75 mM calcium; 125 mM magnesium and 75 mM 
manganese; as well as 125 mM magnesium. 38 mM manganese and 38 mM calcium. 

In another embodiment of the present invention, zinc can be added to the first soluble 

10 fraction obtamed above, thereby forming a second insoluble fraction from which the fusion 
protein can be isolated. Optionally, chelator can be added to solubilize the protein from the 
second insoluble fraction. Suitable sources of zinc include zinc chloride, zinc sulfate, and zinc 
acetate. The final zinc concentration can be from 1 to 500 mM, with 50 mM zinc chloride 
being preferred. 

15 Protein of the present invention can be either recombinantly produced in a transformed 

host cell, e.g., E. coli, and purified from the host cell or can be purified from soluble sources 
such as plasma, urine, and the like. When purifying from soluble sources, the initial release 
step utilizing chelator can be omitted and the purification can begin immediately with the 
addition of the divalent cation/alcohol solution as described above. 

20 Specifically provided by the present invention is a method for release of a thioredoxin- 

IL-11 fusion protein from a cell into a solution by adding EDTA, followed by adding stock 
solutions of MgClj, ethanol, and CaCl2 resulting in a solution comprised of 125 mM 
magnesium chloride, 75 mM calcium chloride, and 14% ethanol, and isolating the fusion 
protein from the soluble fi-action formed. The final concentration of EDTA can range from 

25 0.1 to 100 mM, with 15 mM being preferred. In a presently preferred embodiment, the 

temperature prior to the addition of the chelator is substantially cooler than after the addition 
of chelator, for example going fi-om about 3**C to about 37®C. The r-esulting concentration of 
magnesium chloride and calciimi chloride can range from 1 to 1000 mM and the final ethanol 
concentration can range from 5 to 30%. 

30 In another embodiment of the present invention, the protein of interest can be purified 

by taking advantage of a difference in pi of the protein of interest, as compared to the pi of 
the fusion protein and the fusion partner. Prior to cleavage of the fusion protem, two 
complementary ion exchange resins are utilized back to back and after cleavage of the fusion 
protein, two additional con^)lementary ion exchange resins are utilissed back to back. The 



5 



wo 96/38570 PCTAJS96/04811 

purification process takes advantage of the fact that the pi of the fusion protein, i,e. , the fusion 
partner linked to the protein of interest, is not the same as the pi of the protein of interest sans 
the fusion partner. 

According to the present invention, an adsorptive step is coupled with a non-adsorptive 
5 step to rigorously remove oppositely charged molecules under the conditions employed and to 
enhance the effectiveness and selectivity imparted by the cleavage reaction. This entails both 
selection of appropriate resins and the corresponding optimization of pH and ionic strength 
conditions to leverage the pi differential of fusion proteins from the protein of interest and the 
fusion partner. To purify the fusion protein it is desirable to have the pH as close to the pi so 

10 as to exclude oppositely charged contaminants; after cleavage it is desirable to have the pH as 
close to the pi of the protein of interest to exclude all oppositely charged contaminants. 
Typically, adjusting the pH only in the chromatography steps does not provide the requisite 
level of contaminant removal. Accordingly, the invention provides for the addition of non- 
adsorptive steps as "ionic fillers" to obtain the requisite log removals of contaminants. The 

15 particular combinations of complementary ion exchange resins chosen provides much greater 
levels of purification than is possible with traditional ion exchange methodology. 

According to the method of invention, the thioredoxin-like fusion protein is bound to 
a first resin, eluted, applied to a second resin (to which it does not bind) and is colleaed in the 
unbound fraction; next, the thioredoxin fusion protein is cleaved and the cleaved protein is 

20 bound to a third resin, eluted, applied to a fourth resin and collected in the unbound fraction 
from the fourth resin. Where the first and fourth resins are anion exchange resins, the second 
and third resins are cation exchange resins. Where the first and fourth resins are cation 
exchange resins, the second and third resins are anion exchange resins. The resins are selected 
based upon the pi of the fusion protein and the pi of the protein of interest. For exan:^)le, 

25 where the fusion protein is negatively charged prior to cleavage, and the protein of interest is 
positively charged after cleavage, the first and fourth resins are anion exchange resins, and the 
second and third resins are cation exchange resins. Alternatively, where the fusion protein is 
positively charged prior to cleavage, and the protein of interest is negatively charged after 
cleavage, the first and fourth resins are cation exchange resins, and the second and third resins 

30 are anion exchange resins. Suitable anion exchange resins have positively charged groups such 
as diethyleaminoethane (DEAE), polyethyleneimine (PEI), and quaternary aminoethane (QAE). 
Suitable cation exchange resins have negatively charged groiips such as sulfonyl, sulfylpropyl 
(SP), carboxyl, and carboxy methyl. 



6 



wo 96/38570 



PCTAJS96/048n 



Also provided by the present invention is a method for purifyii^ IL-1 1 which involves 
binding thioredoxin-IL-11 to a first anion exchange resin, such as Toyopearl QAE, eluting with 
a first eluant (where the preferred eluants include 20-100 mM Tris buffers at pH 7.5-8.5, 
containing 100-500 mM NaCl, and 50-200 mM Histidine buffers at pH 5.5-6.6, containing 0- 
5 150 mM NaCl, with the most preferred eluant being 75 mM Histidine, 75 mM NaCl. at pH 

6.2), applying this eluate to a first cation exchange resin, such as S Sepharose Fast Flow, 
collecting the thioredoxin-IL-11 in an unbound fraction, then cleaving the thioredoxin-IL-11 
fusion protein to form positively charged IL-1 1 , binding to a second cation exchange resin such 
as CM Sepharose Fast Flow (where the preferred eluants include 50-300 mM glycine buffers 

10 at pH 9.0-10.0, containing 100-500 mM NaCl, with the most preferred eluant being 150 mM 
glycine, 150 mM NaCl, pH 9.5), applying this second eluate to a second anion exchange resin 
such as Toyopearl QAE, and then collecting the IL-11 in the unbound fraction. 

DETAILED DESCRIPTION OF THE INVENTION 
For the large-scale production of a protein which has been recombinantly produced, 

15 protein is typically first released from the cell. Proteins can be produced as a fusion protein, 

for example as a thioredoxin fusion in £. coli, and are thought to be directed towards sites at 
which the inner and outer cell membranes become contiguous, sometimes referred to as Bayer's 
patches, through the influence of the thioredoxin fusion partner. According to the method of 
the present invention, host cells are selectively permeabilized, using a chelator such as Tris- 

20 EDTA, to release the fusion protein of interest. Chelators other than EDTA can also be used. 

for example, DPTA, EGTA, CDTA, citrate, and the like. The process is readily optimized 
for the concentration range of the chelator, the temperature of the reaction, as well as the pH 
and ionic strength of the solution, as is well within the skill of one skilled in the art. As one 
skilled in the art readily appreciates, the appropriate concentration of chelator will vary 

25 according to the nature of the solute, solution pH, solution temperature, solution ionic strength, 

and the molar proponion of chelator to metal chelate. Preferred are aqueous solutions of a pH 
between 2 and 11, a temperature between 10*C and 60**C, the ionic strength between 0 and 1, 
and the molar ratio of chelator to chelate ranging from 0. 1 : 1 .0 to 1 .0:0. 1 . Optimal conditions 
can be readily ascertained by one skilled in the art by monitoring the release of the protein of 

30 interest from the host cells. Typically, final chelator concentrations range from 0.1 mM to 
100 mM, with 15 mM preferred. In a presently preferred embodiment, the temperature prior 
to the addition of the chelator is substantially cooler than after the addition of chelator, for 
example going fi-om about S^'C to about 37 °C. 



7 



wo 96/38570 



PCTAJS96/04811 



Lipopolysaccharides in the outer membrane of gram negative bacteria form ionic 
interactions with magnesiimi and/or calcium ions (Vaara, Microbiological Reviews 56:395-411 
(1992); Hancock, Ann. Rev. Microbiol. 38:237-264 (1984); Felix, Anal. Biochem. 120:211- 
234 (1982). Presumably, the chelator acts on divalent ions, e.g., magnesium and calcium ions, 
5 as a coimter-ion to negatively charged groups. This leads to a dissociation, which in turn, 
leads to the generation of gaps in the membrane allowing for the selective release of the fusion 
protein. 

After release of the fusion protein into the medium, the cellular debris can be removed 
if desired, for example, by centrifiagation. However, one of the advantages of the present 

10 invention is that it does not require removal of cellular debris, as is usually required with other 
clarification methods. Present in solution are negatively charged molecules such as DNA, 
RNA, phospholipids and lipopolysaccharides, as well as the fusion protein of interest. Where 
the fusion protein of interest is positively charged, it will bind to any negatively charged 
molecules making purification of the protein virtually impossible. Well known to those skilled 

15 in the art is the use of polyethylenimine (PEI) for the removal of DNA (Scawen, et al, supra). 
Unfortunately, PEI is always contaminated with residual amounts of the monomer, ethylenimine 
(EI) (Scawen et aL, supra.), which has been classified by OSHA to be a cancer-suspect agent. 
According to the present invention, the removal of the negatively charged molecules from 
solution can be accomplished utilizing a mixture of one or more divalent cations, such as 

20 magnesium, manganese, and/or calcium plus alcohol, e.g., Mg'^'^/ROH, where R is a short 
carbon chain, 1 to 4 carbons in length, such as methyl, ethyl and the various isomers of propyl, 
as well as secondary butyl and tertiary butyl, with ethyl being preferred. Divalent cations are 
typically added as salts with counteranions such as chloride, sulphate, acetate, and the like. 
This divalent cation/alcohol solution is sometimes referred to herein simply as the "alcohol 

25 solution." 

Upon addition of the alcohol solution, most, if not all of the protein of interest remains 
in solution, whereas the contaminants precipitate out of solution forming an insoluble fraction. 
This step is also concentration, pH, temperature, and time-dependent. Prolonged incubation 
under optimum conditions may result in irreversible precipitation of the protein of interest. 
30 Elevated temperatures, and/or alcohol concentrations, and/or pH extremes may increase 
precipitation of protein. Temperatures lower than 50°C and final alcohol concentrations between 
5 and 30% are preferred, with 32**C and 14% (volume/volume %) ethanoi most preferred. 



wo 96/38570 



PCTAJS96/04811 



Insufficient divalent cation and/or alcohol concentrations lead to ineffective clarification 
removal of non-proteinaceous contaminants). Preferably, the final divalent cation concentration 
is 1 to 1000 mM; or 50 to 300 mM, with the most preferred range being 50 to 200 mM. 

Optionally, for increased clarification, manganese and/or calciimi salts can be added 
5 to the Mg^'^/ROH solution, at a final concentration of 1 to 1000 mM, so that the total molarity 

remains in the range of 1 to 1000 mM, or 50 to 300 mM, with the most preferred range 
currently conten^lated being 50 to 200 mM. When multiple divalent cations are utilized, 
preferred final concentrations total 50 to 200 mM salt and include, for example, 125 mM 
magnesium plus 75 mM calcium or manganese, or various combinations thereof, for example, 

10 about 125 mM magnesium, plus about 38 mM manganese and about 38 mM calcium. As used 
herein, the term "molarity" means the number of gram molecular weights of a compound 
dissolved in one liter of solution. 

Addition of the alcohol solution removes greater than 80-90% of the DNA and LPS 
molecules from solution. The resulting solution can then be further treated to remove the 

15 protein of interest from solution. For example, the protein of interest can be precipitated using 
zinc (Zn*"") salts such as zinc chloride, zinc sulphate, zinc acetate and the like. The resulting 
precipitate can be concentrated and recovered by solid/liquid separation techniques including 
centrifiigation and filtration. Precipitated protein can be subsequently solubilized by a chelating 
agent such as EDTA, DPTA, EGTA, CDTA, or citrate and further purified. Optionally, the 

20 precipitate may be stored prior to solubilization. The process can be optimized for zinc 
concentration, chelator to metal concentration, pH, temperature, and time, as is readily 
appreciated by one skilled in the art and can be readily adjusted for the protein of interest. As 
is known in the art, irreversible protein precipitation may occur in the presence of organic 
solvents, such as methanol, ethanol, and acetone (Pennell. The Plasma Proteins Volume 7, 

25 Putnam eds.:9-42 (1960); Wang et al eds. Fermentation and Enzyme Technology:253-256 

(1979).; Wheelwright, Protein Purification: Design and Scale up of Downstream Processing: 
87-98 (1991).; Scopes, Protein Purification Principles and Practice 2nd edition. Cantor eds., 
21-71 (1987)). This phenomenon is typically minimized by operating at lower temperatures for 
shorter durations. Preferred conditions for precipitation of the protein of interest are generally 

30 pH=6.0 to 8.5, 1 to 500 mM Zn""", a temperature lower than 20 and duration of fewer 
than 24 hours. Most preferred conditions are precipitation at pH=7.0 and (fC for 15 minutes 
at a final ZnClj concentration of 50 mM. 



9 



wo 96/38570 



PCTAJS96/()4811 



Subsequently, the precipitate is recovered by solid/liquid separation methods known in 
the art which may then be efficiently stored. At this suge of processing, the protein is no 
longer in solution and it is less susceptible to proteolysis. In addition, the protein is 
concentrated at least an order of magnitude, which readily allows for long term (circa 2 
months) frozen storage at -20°C to -80 °C of large scale process intermediates. Accordingly, 
this step is for de-coupling recovery from purification portions of the process. 

Prior to further processing, the protein precipitate is solubilized with a chelating agent. 
Preferred solubilization conditions are pH=6.5 to 11, temperature lower than 40 °C, and a 
ratio of moles of chelator per moles Zn"*"^ of about 0.5 to 1000. Most preferred is 
solubilization at pH = 8.0 and 20 °C using the chelator EDTA at a ratio of 5-100 moles EDTA 
per mole Zn"^*. Typically, greater than 70% of the protein of interest is recovered through 
precipitation, solid liquid separation, and solubilization processing. 

Alternatively, or in addition, the fusion protein of interest can be further purified. 
Purification can be accomplished using two ion exchange resins prior to cleavage of the fusion 
protein, followed by two ion exchange resins after cleavage of the fusion protein. The process 
leverages any difference in the pi of the fusion protein, i.e., the fiision partner linked to the 
protein of interest, from the pi of the protein of interest sans fusion partner. For exanq>le, 
where the protein of interest has been expressed as a fusion partner, and where the fiised form 
of the protein has a pi that is substantially different from the pi of the cleaved form (for 
exanqjie, where the fused form of the protein has a basic pi, and where the cleaved form has 
an acidic pi; or vice versa, the fused form being acidic and the cleaved form being basic), it 
is possible to advantageously utilize this difference in pi for purification of the cleaved form 
using complementary ion exchange resins. 

More specifically, the fusion protein is first purified with the appropriate type of ion 
exchange resin: for example, an anion-exchange resin to bind a negatively charged ftision 
protein, (or conversely, a cation-exchange resin to bind a positively charged fusion protein). 
Next, the complementary type of ion exchange resin is used to further remove contaminants 
from the process stream having a pi and net charge at the pH of operation that is different from 
the fused form of the protein. For example, in the case of a fusion protein with an acidic pi, 
a cation exchange resin can be used under conditions i.e., where the pH is greater than the pi 
of the fusion protein in which the fusion protein flows through the resin without binding, while 
the resin binds contaminants with pis greater than the pH of operation (or, conversely, in the 
case of a fusion protein with a basic pi, an anion exchange resin can be used under conditions 
i,e,, where the pH is less than the pi of the fusion in which the fusion protein flows through 



10 



wo 96/38570 



PCT/US96/04811 



the resin without binding, while the resin binds contaminants with pis less than the pH of 
operation). The pH requirements may be modified by the addition of conq)ounds that alter the 
ionic strength of the solution as is understood by one skilled in the art of ion exchange 
chromatography. The fusion protein is simply collected in the unbound fraction, also 
sometimes referred to as the flow through fraction. It is these complementary combinations 
of ion-exchange purifications that result in a preparation of fusion protein that has been purified 
away from contaminants with markedly different pis. 

Next, the fusion protein can be cleaved into the constituent proteins, namely the protein 
of interest having a significantly different pi from the fused form, and the fusion partner. 
Purification of the protein of interest can then proceed with additional applications of ion- 
exchange chromatography. Where once the fused form did bind to the resin, the cleaved 
protein of interest now flows through the resin without binding, while contaminants, (including 
uncleaved fusion protein aiKl/or the fusion partner which is that part of the protein of interest 
released during cleavage) do bind to the resin (or, conversely, where once the fused form 
flowed through the resin without binding, the cleaved protein of interest now binds to the resin 
while contaminants do not). More specifically, the products of the cleavage reaction are first 
applied, for example, to a cation exchange resin (where the cleaved protein is positively 
charged, or conversely, to an anion exchange resin if the product protein is negatively 
charged). Next, the complementary ion exchange resin, for example, an anion exchange resin 
(or conversely, a cation exchange resin) is used under conditions where the cleaved protein 
flows through the resin without binding, while the negatively charged contaminants (conversely, 
positively charged contaminants) do bind to the resin. The protein of interest is collected in 
the flow through. Again, these additional ion exchange combinations further purify away 
contaminants having pis markedly different from the protein of interest. 



11 



wo 96/38570 



PCT/OS96/04811 



10 



15 



20 



25 



30 



35 



Many combinations of complementary ion exchange resins are possible. Set forth in 
Table 1, below are eight variations, four for the case of a negatively charged fusion protein 
(with a positively charged cleaved protein) and four for the case of a positively charged fusion 
protein (with a negatively charged cleaved protein). 

TABLE 1 

COMPLEMENTARY ION EXCHANGE COMBINATIONS 

Negativdy charged fusion protein 
Positively charged cleaved protein 



I 


u 


III 


IV 


Anion exchange (bind) 


Cation exchange (no bind) 


Cation exchange (no bind) 


Anion exchange (bind) 


i 


4 


4 


4 


Cation exchange 
(no bind) 


Anion exchange (bind) 


Anion exchange (bind) 


Cation exchange 
(no bind) 




4 


4 


4 


Cleave 


Cleave 


Cleave 


Cleave 


1 


4 


4 


4 


Cation exchange (bind) 


Cation exchange (bind) 


Anion exchange (no bind) 


Anion exchange (no bind) 


i 


4 


4 


4 


Anion exchange (no bind) 


Anion exchange (no bind) 


Cation exchange (bind) 


(Nation exchange (bind) 



Positively charged fusion protein 
Negativdy charged cleaved protein 



V 


VI 


VII 


vm 


Cation exchange (bind) 


Anion exchange (no bind) 


Anion exchange (no bind) 


Cation exdiange (bind) 


4 


4 


4 


4 


Anion exchange (no bind) 


Cation exchange (bind) 


Cation exchange (bind) 


Anion exchange (no bind) 


4 


4 


4 


4 


Cleave 


Cleave 


Cleave 


Cleave 


4 


4 


4 


4 


Anion exchange (bind) 


Anion exchange (bind) 


Cation exchange (no bind) 


Cadon exchange(no bind) 


4 


4 


4 


4 


Cation exchange(no bind) 


Cation exchange (no bind) 


Anion exchange (bind) 


Anion exchange (bind) 



The particular variation chosen depends upon which configuration of complementary 
exchange resins provides the greatest efficiency of purification. For example, for the 
purification of thioredoxin-IL-1 1 , surprisingly it has been found that the first ccHifiguration (I) 



12 



wo 96/38570 



PCTAJS96/04811 



gives the most efficient purification. Certain positively charged contaminants unexpectedly bind 
to the first anion exchange resin and are eluted along with the thioredoxin-IL-1 1 fusion protein. 
These positively charged contaminants then bind to the following cation exchange resin while 
the thioredoxin-IL-ll fusion protein does not and simply flows through. Thus, for certain 

5 applications, it may be preferable to select a pre-cleavage configuration in which the fusion 

protein is bound in the first step, thereby allowing most of the contaminants to remain in the 
flow-through fraction. Similarly, it may be preferable to select a post-cleavage configuration 
in which the cleaved protein of interest is bound in the first post-cleavage step, thereby 
allowing the majority of contaminants to remain in the flow-through fraction. 

10 Suitable eluants to elute the protein of interest from the anion exchange resin are well 

known to one skilled in the art and include, for exanq)le, any buffered solution capable of 
maintaining pH at the desired value and also containing from 0 to 1.0 M of an ionic sah 
capable of causing desorption of the protein from the resin. 

Suitable eluants to elute the protein of interest from the cation exchange resin are well 

15 known to one skilled in the art and include, for example, any buffered solution capable of 
" maintaining pH at the desired value and also containing from 0 to 1.0 M of an ionic salt 
capable of causing desorption of the protein from the resin. 

According to the present invention, the DNA sequence encoding a heterologous peptide 
or protein selected for expression in a recombinant system is desirably fused to a thioredoxin- 

20 like DNA sequence for expression in the host cell. A thioredoxin-like DNA sequence is 
defined herein as a DNA sequence encoding a protein or fragment of a protein characterized 
by an amino acid sequence having at least 30% homology with the amino acid sequence of £. 
coli thioredoxin. Incorporated by reference is McCoy, et al. , USPN 5.292,646; issued March 
8, 1994 which discloses an E. coli thioredoxin sequence (SEQ ID NO. 22, therein). 

25 Alternatively, a thioredoxin-like DNA sequence is defined herein as a DNA sequence encoding 

a protein or fragment of a protein characterized by having a three dimensional structure 
substantially similar to that of human or £. coli thioredoxin and optionally by containing an 
active-site loop. The DNA sequence of glutaredoxin is an example of a thior^oxin-like DNA 
sequence which encodes a protein that exhibits such substantial similarity in three-dimensional 

30 conformation and contains a Cys....Cys active site loop. The amino acid sequence of £. coli 
thioredoxin is described in H. Eklund et aL, EMBO J. 5:1443-1449 (1984). The three- 
dimensional structure of £. coli thioredoxin is depicted in Fig. 2 of A. Holn^ren, J. Biol. 
Chem. 26^:13963-13966 (1989). In Fig. 1 of McCoy, et al,, supra, nucleotides 2242-2568 
encompasses a DNA sequence encoding the £. coli thioredoxin protein (Lim et al., J. 

13 



wo 96/38570 PCT/US96/0481 1 

Bacterid., 163:311-316 (1985)) (McCoy, et al, supra). A comparison of the three 
dimensional structures of E. coli thioredoxin and glutaredoxin is published in Xia, Protein 
Science 7:310-321 (1992). These four publications are incorporated herein by reference for the 
purpose of providing information on thioredoxin-like proteins that is known to one of skill in 
the art. 

As the primary example of a thioredoxin-like protein useful in this invention, E. coli 
thioredoxin has the following characteristics. £. coli thioredoxin is a small protein, only 11.7 
kD, and can be produced to high levels (> 10%, corresponding to a concentration of 15 /iM 
if cells are lysed at 10 Ajjo/nil). The small size and capacity for a high level synthesis of the 
protein contributes to a high intracellular concentration. E. coli thioredoxin is further 
characterized by a very stable, tight structure which can minimize the effects on overall 
structural stability caused by fusion to the desired peptide or proteins. 

The three dimensional structure of £. coli thioredoxin is known and contains several 
surface loops, including a distinctive Cys....Cys active-site loop between residues Cysgj and 
Cysjfi which protrudes from the body of the protein. This Cys....Cys active-site loop is an 
identifiable, accessible surface loop region and is not involved in any interactions with the rest 
of the protein that contribute to overall structural stability. It is therefore a good candidate as 
a site for peptide insertions. Both the amino- and carboxyl-termini of £. coli thioredoxin are 
on the, surface of the protein, and are readily accessible for fusions. Human thioredoxin, 
glutaredoxin and other thioredoxin-like molecules also contain this Cys....Cys active-site loop. 

£. coli thioredoxin is also stable to proteases. Thus, £. coli thioredoxin may be 
desirable for use in £. coli expression systems, because as an £. coli protein it is characterized 
by stability to £. coli proteases. £. coli thioredoxin is also stable to heat up to SO'^C and to 
low pH. 

Other thioredoxin-like proteins encoded by thioredoxin-like DNA sequences useful in 
this invention share homologous amino acid sequences, and similar physical and structural 
characteristics. Thus, DNA sequences encoding other thioredoxin-like proteins may be used 
in place of £. coli thioredoxin according to this invention. For example, the DNA sequence 
encoding other species' thioredoxin, e.g., human thioredoxin, have been enq)loyed by these 
inventors in the compositions and methods of this invention. Human thioredoxin has a three- 
dimensional structure that is virtually superimposable on £. coli's thr^-dimensional structure, 
as determined by comparing the NMR structures of the two molecules. Human thioredoxin 
also contains an active-site loop structurally and functionally equivalent to the Cys....Cys 
active-site loop found in the £. coli protein. Human IL-11 fused in frame to the carboxyl 



14 



wo 96/38570 



PCT/US96/04811 



terminus of human thioredoxin (i.e., a human thioredoxin/IL-11 fusion) exhibited the same 
expression characteristics as the £. coli thioredoxin/IL-11 fusion exemplified in Exan^les 1-2. 
Consequently, human thioredoxin is a thioredoxin-Iike molecule and can be used in place of 
or in addition to E. coli thioredoxin in the production of protein and small peptides in 
5 accordance with the method of this invention. Insertions into the human thioredoxin active-site 
loop and onto the amino terminus may be as well tolerated as those in £. coli thioredoxin. 

Other thioredoxin-like sequences which may be employed in this invention include all 
or portions of the protein glutaredoxin and various species' homologs thereof. (A. Holmgren, 
cited above.) Although £. coli glutaredoxin and E. coli thioredoxin share less than 20% amino 

10 acid homology, the two proteins do have conformational and functional similarities (Eklund et 
aL, EMBO J. 5:1443-1449 (1984)) and glutaredoxin contains an active-site loop structurally 
and functionally equivalent to the Cys....Cys active-site loop of E. coli thioredoxin. 
Glutaredoxin is therefore a thioredoxin-like molecule as herein defined. 

The DNA sequence encoding protein disulfide isomerase (PDI), or that portion thereof 

15 containing the thioredoxin-like domain, and its various species' homologs (Edman et aL, 
Nature 577:267-270 (1985)) may also be employed as a thioredoxin-like DNA sequence, since 
a repeated domain of PDI shares > 30% homology with E. co/z thioredoxin and that repeated 
domain contains an active-site loop structurally and functionally equivalent to the Cys....Cys 
active-site loop of E. coli thioredoxin. These three publications are incorporated herein by 

20 reference for the purpose of providing information on glutaredoxin and PDI which is known 
and available to one of skill in the art. • 

Similarly the DNA sequence encoding phosphoinositide-specific phospholipase C (PI- 
PLC), fragments thereof and various species* homologs thereof (Bennett et aL, Nature 
534:268-270 (1988)) may also be employed in the present invention as a thioredoxin-like 

25 sequence based on their amino acid sequence homology with £. coli thioredoxin, or 
alternatively based on similarity in three-dimensional conformation and the presence of an 
active-site loop structurally and functionally equivalent to the Cys....Cys active-site loop of £. 
coli thioredoxin. All or a portion of the DNA sequence encoding an endoplasmic reticulum 
protein, such as ERp72, or various species homologs thereof are also included as thioredoxin- 

30 like DNA sequences for the purposes of this invention (Mazzarella et aL, J. Biol. Chem. 
265:1094-1101 (1990)) based on amino acid sequence homology, or alternatively based on 
similarity in three-dimensional conformation and the presence of an active-site loop structurally 
and functionally equivalent to the Cys....Cys active-site loop of £. coli thioredoxin. Another 
thioredoxin-like sequence is a DNA sequence which encodes all or a portion of an adult T-cell 



15 



wo 96/38570 PCTAJS96/04811 

leukemia-derived factor (ADF) or other species homologs thereof. N. Wakasugi et al:, Proc. 
Natl. Acad. Sci. USA 57:8282-8286 (1990). ADF is now believed to be human thioredoxin. 
Similarly, the protein responsible for promoting disulfide bond formation in the periplasm of 
E xoli, the product of the dsbh gene (Bardwell et al. Cell 67: 581-589 (1991), also can be 
considered a thioredoxin-like sequence. These four publications are incorporated herein by 
reference for the purpose of providing information on PI-PLC. ERp72, ADF, and dsbA which 
are known and available to one of skill in the art. 

It is expected from the definition of thioredoxin-like DNA sequence used above that 
other sequences not specifically identified above, or perhaps not yet identified or published, 
may be thioredoxin-like sequences either based on the 30% amino acid sequence homology to 
E. coli thioredoxin or based on having three-dimensional structures substantially similar to E. 
coli or human thioredoxin and having an active-site loop functionally and structurally equivalent 
to the Cys....Cys active-site loop of £. coli thioredoxin. One skilled in the art can determine 
whether a molecule has these latter two characteristics by comparing its three-dimensional 
structure, as analyzed for example by x-ray crystallography or 2- dimensional NMR 
spectroscopy, with the published three-dimensional structure for E, coli thioredoxin and by 
analyzing the amino acid sequence of the molecule to determine whether it contains an active- 
site loop that is structurally and functionally equivalent to the Cys....Cys active-site loop of £. 
coli thioredoxin. By "substantially similar" in three-dimensional structure or conformation is 
meant as similar to £. coli thioredoxin as is glutaredoxin. In addition a predictive algorithm 
has been described which enables the identification of thioredoxin-like proteins via computer- 
assisted analysis of primary sequence (Ellis et al. Biochemistry 31: 4882-91 (1992)). Based 
on the above description, one of skill in the art will be able to select and identify, or, if 
desired, modify, a thioredoxin-like DNA sequence for use in this invention without resort to 
undue experimentation. For example, simple point mutations made to portions of native 
thioredoxin or native thioredoxin-like sequences which do not effect the structure of the 
resulting molecule are alternative thioredoxin-like sequences, as are allelic variants of native 
thioredoxin or native thioredoxin-like sequences. 

DNA sequences which hybridize to the sequence for E, coli thioredoxin or its structural 
homologs under either stringent or relaxed hybridization conditions also encode thioredoxin-like 
proteins for use in this invention. An example of one such stringent hybridization condition 
is hybridization at 4 X SSC at 65°C, followed by a washing in O.IXSSC at 65*'C for an hour. 
Alternatively an exemplary stringent hybridization condition is in 50% formamide, 4XSSC at 
42 ""C. Examples of non-stringent hybridization conditions are 4XSSC at 50**C or hybridization 



16 



wo 96/38570 



PCTAJS96/048n 



with 30-40% formamide at 42**C. The use of all such thioredoxin-like sequences are believed 
to be encompassed in this invention. 

Construction of a fusion sequence of the present invention, which comprises the DNA 
sequence of a selected peptide or protein and the DNA sequence of a thioredoxin-like sequence, 
employs conventional genetic engineering techniques. See, Sambrook et aL, Molecular 
Cloning. A Laboratory Manual, 2nd edition, Cold Spring Harbor Laboratory, Cold Spring 
Harbor, New York (1989). Fusion sequences may be prepared in a number of different ways. 
For example, the selected heterologous protein may be fused to the amino terminus of the 
thioredoxin-like molecule. Alternatively, the selected protein sequence may be fused to the 
carboxyl terminus of the thioredoxin-like molecule. Small peptide sequences could also be 
fused to either of the above-mentioned positions of the thioredoxin-like sequence to produce 
them in a structurally unconstrained manner. 

This fusion of a desired heterologous peptide or protein to the thioredoxin-like protein 
increases the stability of the peptide or protein. At either the amino or carboxyl terminus, the 
desired heterologous peptide or protein is fused in such a manner that the fusion does not 
destabilize the native structtire of either protein. Additionally, fusion to the soluble 
thioredoxin-like protein improves the solubility of the selected heterologous peptide or protein. 

It may be preferred for a variety of reasons that peptides be fused within the active-site 
loop of the thioredoxin-like molecule. The region on the surface of thioredoxin surrounding 
the active-site loop has evolved, in keeping with the protein's major function as a non-specific 
protein disulfide oxido-reductase, to be able to interact with a wide variety of other protein 
surfaces, and so may be especially tolerant to the presence of inserted sequences. In addition 
the active-site loop region is bounded by segments of strong secondary structure, which 
provides many advantages for peptide fusions. Any small peptide inserted into the active-site 
loop of a thioredoxin-like protein is present in a region of the protein which is not involved in 
maintaining tertiary structure. Therefore the structure of such a fusion protein is stable. 
Indeed previous work has shown that E, coli thioredoxin can be cleaved into two fragments at 
a position close to the active-site loop, and yet the tertiary interactions stabilizing the protein 
reniain intact. 

The active-site loop of E, coli thioredoxin has the sequence NHj-.-Cysaj-Gly-Pro- 
CySj^. . .COOH. Fusing a selected peptide with a thioredoxin-like protein in the active-site loop 
portion of the protein constrains the peptide at both ends, reducing the degrees of 
conformational freedom of the peptide, and consequently reducing the number of possible 
alternative structures taken by the peptide. The inserted peptide is bound at each end by 



17 



wo 96/38570 



PCTAJS96/04811 



cysteine residues, which may form a disulfide linkage to each other as they do in native 
thioredoxin and further limit the conformational freedom of the inserted peptide. Moreover, 
this invention places the peptide on the surface of the thioredoxin-like protein. Thus the 
invention provides a distinct advantage for use of the peptides in screening for bioactive peptide 
5 conformations and other assays by presenting peptides inserted in the active-site loop in this 
structural context. 

Additionally the fusion of a peptide into the loop protects it from the actions of E. coli 
amino- and carboxyl-peptidases. Further, a restriction endonuclease cleavage site RsrII already 
exists in the portion of the E. coli thioredoxin DNA sequence encoding the loop region at 

10 precisely the correct position for a peptide gene fusion. (See, McCoy, et aL supra; Figure 4.) 
RsrII recognizes the DNA sequence CGG(A/T)CCG leaving a three nucleotide long 5'- 
protruding sticky end. DNA bearing the complementary sticky ends will therefore insert at this 
site in only one orientation. 

As used herein the term "fusion protein" includes, but is not limited to, any "protein 

15 of interest" that is covalently bonded to another protein, e.g. , the iiision partner. The cleavage 

of products of the fusion protein comprise the fusion partner and the protein of interest. As 
used herein the term "thioredoxin fusion protein" includes, but is not limited to, the expression 
product of the thioredoxin-like DNA (described supra), and another DNA encoding the protein 
of interest. 

20 As used herein the term "anion exchange resin" includes, but is not limited to, any t>pe 

of support to which are bound positively charged pendant groups, such as 
diethyleaminoethane (DEAE), polyethyleneimine (PEI), or quaternary aminoethane (QAE) 
groups. The groups may either be positively charged regardless of pH, or positively charged 
within a specific pH range, being neutral (with no charge) outside of that pH range. 

25 As used herein the term "cation exchange resin" includes, but is not limited to, any 

type of support to which are bound negatively charged pendant groups, such as sulfonyl, 
sulfylpropyl (SP), carboxyl, or carboxymethyl (CM) groups. The groups may either be 
negatively charged regardless of pH, or negatively charged within a specific pH rmige, being 
neutral (with no charge) outside of that pH range. 

30 As used herein the term "charged" includes, but is not limited to a chemical species 

having a non-zero net electrostatic charge, either positive or negative, without regard to the 
magnitude of the net charge. 

As used herein the term "negatively charged" includes, but is not limited to, any 
chemical species, whether charged or not, that adsorbs to an anion exchange chromatography 

18 



wo 96/38570 PCTAJS96/04811 

resin at the pH and ionic strength of the operating buffer; or any chemical species that does not 
adsorb to a cation exchange chromatography resin at the pH and ionic strength of the operating 
buffer. 

As used herein the term "positively charged" includes, but is not limited to, any 
5 chemical species, whether charged or not, that adsorbs to a cation exchange chromatography 
resin at the pH and ionic strength of the operating buffer; or any chemical species that does not 
adsorb to an anion exchange chromatography resin at the pH and ionic strength of the operating 
buffer. 

As described in greater detail below, the term "host cell" generally includes any 
10 transformed or non-transformed gram negative microorganism. 

As used herein, the term "chelator" includes but is not limited to, any compound which 
will form two or more intermolecular ordinary or coordinate bonds with metal ions in solution, 
so that one or more heterocyclic rings are formed with each bound metal ion, and includes, but 
is not limited to, such compounds as ethylenediaminetetraacetic acid (EDTA), 
15 diethylenetriaminopentaacetic acid (DPTA), ethylene glycol-bis(2-aminoethyl ether) tetraacetic 

acid (EGTA), 1 ,2-cyclohexanediaminetetraacetic acid (CDTA), and citric acid. 

As used herein, the term "divalent cation" includes, but is not limited to such species 
as Mg""^ (magnesium), Mn'^'^ (manganese), Ca**"* (calcium) and the like. 

As used herein, the term "collecting" includes, but is not limited to for example, the 
20 . process by which a process stream is pumped through a column of chromatography resin and 
the unbound fraction (sometimes also referred to as the flow through fraction) is collected . 

While the present method of the invention is exemplified by purification of 
recombinantly-produced proteins from transformed host cells, the method is also amenable to 
purification of proteins which are naturally occurring within a cell and can be used generally 
25 to purify proteins from any solution, regardless of source. 

The following examples illustrate practice of the invention. These exanq)les are for 
illustrative purposes only and are not intended in any way to limit the scope of the invention 
claimed. Example 1 describes construction of a fiision protein; Example 2 describes the 
expression of a fusion protein; Example 3 describes selective release of protein from a cell 
30 using chelator; Example 4 relates to removal of the negatively charged molecules from 
solution; Example 5 describes the selective precipitation with zinc; and Example 6 relates to 
purification of a fusion protein based upon a difference in pi upon cleavage of the protein of 
interest from its fusion partner. 



19 



wo 96/38570 



PCTA)S96/04811 



EXAMPLE 1 - PREPARATION OF FUSION PROTEIN MOLECULE 

A thioredoxin-like fusion protein can be made by constructing a fusion DNA 
comprising a thioredoxin-like sequence linked to a DNA encoding the polypeptide of interest 
and expressing the DNA construct in an appropriate host cell. For example, a thioredoxin-like 
fusion molecule of the present invention can be prepared using E. coli thioredoxin as the 
thioredoxin-like sequence and recombinant IL-11 (Paul et al,, Proc. Natl. Acad. Sci. U.S.A. 
57:7512-7516 (1990); see also, copending United States Patent Applications SN 07/526,474, 
and SN 07/441,100 and PCT Patent publication WO91/07495, published May 30, 1991 
incorporated herein by reference) as the selected heterologous protein. The E. coli thioredoxin 
(trxA) gene (McCoy, et al, supra,) was cloned based on its published sequence and employed 
to construct various related £. coli expression plasmids using standard DNA manipulation 
techniques, described extensively by Sambrook, et aL, Molecular Cloning. A Laboratory 
Manual, 2nd edition, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1989). 

A first expression plasmid pALtrxA-781 was constructed containing the E. colivmk 
gene without fusion to another sequence. This plasmid further contained sequences which are 
described in detail below for the related IL-1 1 fusion plasmid. This first plasmid. which directs 
the accumulation of > 10% of the total cell protein as thioredoxin in an £. coli host strain 
GI724, was further manipulated as described below for the construction of a trxA/IL-l 1 fusion 
sequence. 

Alternatively, a thioredoxin-like molecule modified to include metal-binding/chelating 
amino acid residues, such as, e.g., histidine residues at positions 2. 31 and 63, or, 
alternatively, at positions 31 and 63. was prepared as described in co-pending USSN 
08/165.301, incorporated herein by reference, using standard DNA manipulation techniques 
(reference above). 

The entire sequence of the related plasmid expression vector, pALtrxA/EK/ILl 1 APro- 
581 (McCoy, et aL, supra., as illustrated in Fig. 1) and contains the following principal 
features: 

Nucleotides 1-2060 contain DNA sequences originating from the plasmid pUC-18 
(Norrander et at.. Gene 26: 101-106 (1983)) including sequences containing the gene for j3- 
lactamase which confers resistance to the antibiotic ampicillin in host E. coli strains, and a 
colEl-derived origin of replication. Nucleotides 2061-2221 contain DNA sequences for the 
major leftward promoter (pL) of bacteriophage X (Sanger et al., J. Mol. Biol. 762:729-773 
(1982)), including three operator sequences, OlI, Ol2 and 0^3. The operators are the binding 
sites for Xcl repressor protein, intracellular levels of which control the amount of transcription 



20 



wo 96/38570 



PCTAJS96/04811 



initiation from pL. Nucleotides 2222-2241 contain a strong ribosome binding sequence derived 
from that of gene 10 of bacteriophage T7 (Dunn and Studier, J. Mol. Biol. 766:477-535 
(1983)). 

Nucleotides 2242-2568 contain a DNA sequence encoding the £. coli thioredoxin 
5 protein (Lim et al., J. Bacteriol. 765:311-316 (1985)). There is no translation termination 
codon at the end of the thioredoxin coding sequence in this plasmid. 

Nucleotides 2569-2583 contain DNA sequence encoding the amino acid sequence for 
a short, hydrophilic, flexible spacer peptide "-GSGSG-". Nucleotides 2584-2598 provide 
DNA sequence encoding the amino acid sequence for the cleavage recognition site of 
10 enterokinase (EC 3.4.4.8), "-DDDDK--" (Maroux et al., J. Biol. Chem. 246:5031-5039 
(1971)). 

As an alternative embodiment, a single additional codon can be inserted into the linker 
sequence of the plasmid to introduce a specific site for chemical cleavage of the thioredoxin-IL- 
1 1 fusion protein by hydroxylamine. The nucleotide triplet introduced between residues 2598 

15 and 2599 of pALtrxA/EK/ILllAPro-581, "-AAT-", encodes an asparagine residue. This 
asparagine, in combination with the glycine residue immediately following, comprises a new 
hydroxylamine cleavage site. Under appropriate conditions, detailed in Example 6, 
hydroxylamine cleavage will occur between the asparagine and glycine residues. As an 
additional feature of this alternative embodiment two naturally occurring asparagine residues 

20 present in wild-type thioredoxin, amino-acids 84 and 107, may be altered to glutamine by 
standard techniques to remove two other unwanted hydroxylamine cleavage sites, thus reducing 
secondary hydroxylamine cleavage products which could hanq>er subsequent purification 
procedures. 

Nucleotides 2599-3132 contain DNA sequence encoding the amino acid sequence of 
25 a modified form of mature human IL-11 (Paul et aL, Proc. Natl. Acad. Sci. USA 87:7512- 
7516 (1990)), deleted for the N-terminal prolyl-residue normally found in the natural protein. 
The sequence includes a translation termination codon at the 3'-end of the IL-11 sequence. 

Nucleotides 3133-3159 provide a "Linker" DNA sequence containing restriction 
endonuclease sites. Nucleotides 3160-3232 provide a transcription termination sequence based 
30 on that of the £. coli aspA gene (Takagi et al., Nucl. Acids Res. 75:2063-2074 (1985)). 
Nucleotides 3233-3632 are DNA sequences derived from pUC-18. 

As described in Example 2 below, when cultured under the appropriate conditions in 
a suitable £. coli host strain, this plasmid vector can direct the production of high levels 
(approximately 10% of the total cellular protein) of a thioredoxin/IL-11 fusion protein. By 

21 



wo 96/38570 PCT/US96/04811 

contrast, when not fused to thioredoxin, IL-11 accumulated to only 0.2% of the total cellular 
protein when expressed in an analogous host/vector system. 
EXAMPLE 2 - EXPRESSION OF A FUSION PROTEIN 

A thioredoxin/IL-11 fusion protein is produced according to the following protocol 

5 using the piasmid constructed as described in Example 1. pALtrxA/EK/ILllAPro-581 is 
transformed into the £. coli host strain GI724 (F, lacP, lacP", ampC::\cV) by the procedure 
of Dagert and Ehrlich, Gene 6:23 (1979). The untransformed host strain £. coli GI724 was 
deposited with the American Type Culture Collection, 12301 Parklawn Drive, Rockville, 
Maryland on January 31, 1991 under ATCC No. 55151 for patent purposes pursuant to 

10 applicable laws and regulations. Transformants are selected on 1.5% w/v agar plates 
containing IMC medium, which is composed of M9 medium (Miller, "Experiments in 
Molecular Genetics", Cold Spring Harbor Laboratory, New York (1972)) containing ImM 
MgS04 and supplemented with 0.5% w/v glucose, 0.2% w/v casamino acids and 100 
/zg/ml ampiciliin. 

15 GI724 contains a copy of the wild-type Xcl repressor § ene stably integrated into the 

chromosome at the ampC locus, where it has been placed under the transcriptional control of 
Salmonella typhimurium trp promoter/operator sequences. In GI724, Xcl protein is made only 
during growth in tryptophan-free media, such as minimal media or a minimal medium 
supplemented with casamino acids such as IMC, described above. Addition of tryptophan to 

20 a culture of GI724 will repress the trp promoter and turn off synthesis of Xcl, gradually causing 
the induction of transcription from pL promoters if they are present in the cell. 

GI724 transformed with pALtrxA/EK/ILll Pro-581 is grown at SO** C to an Agoo of 
20 in IMC medium (with 3x MgSOJ. The glucose concentration of the culture is maintained 
at approximately 0.2% (wt./vol.). The pH is maintained at 7.0 with 7.5 M ammonium 

25 hydroxide. Tryptophan is added to a final concentration of 1(X) n%lvA and the culture incubated 
for a further 4 hours at 37° C. During this time, thioredoxin/IL-11 fusion protein accumulated 
to approximately 10% of the total cell protein. 

All of the fusion protein is found to be in the soluble cellular fraction, and is purified, 
as follows, for the following analyses. Cells are lysed in a french pressure cell at 20,000 psi 

30 in 50 mM HEPES pH 8.0, 1 mM phenylmethylsulfonyl fluoride. The lysate is clarified by 
centrifugation at 15,000 x g for 30 minutes and the supernatant loaded onto a QAE-Toyopearl 
column. The flow-through fractions are discarded and the fusion protein eluted with 50 mM 



22 



wo 96/38570 



PCTAJS96/04811 



HEPES pH 8.0, 100 mM NaCI. The eluate is adjusted to 2M NaCl and loaded onto a column 
of phenyl-Toyopearl. The flow-through fractions are again discarded and the fusion protein 
eluted with 50 mM HEPES pH 8.0. 0.5 M NaCl. 

The fusion protein is then dialyzed against 25 mM HEPES pH 8.0 and is > 80% pure 
5 at this stage. By T1165 bioassay (Paul et aL, supra) the purified thioredoxin/lL-11 protein 

exhibits an activity of 8xlO^U/mg. This value agrees closely on a molar basis with the activity 
of 2xlO^U/mg found for COS cell-derived IHl in the same assay. One milligram of the 
fusion protein is cleaved at 37°C for 20 hours with 1000 units of bovine enterokinase 
(Leipnieks et aL, J. Biol. Chem. 254:1677-1683 (1979)) in 1 ml 10 mM Tris-Cl (pH 8.0)/10 
10 mM CaClj. IL-11 is recovered from the reaction products by passing them over a QAE- 
Toyopearl column in 25 mM HEPES pH 8.0, where IL-11 is found in the flow-through 
fractions. Uncleaved fusion protein, thioredoxin and enterokinase remain bound on the 
colunm. The IL-11 prepared in this manner has a bioactivity in the T1165 assay of 2.5x10^ 
U/mg. Its physical and chemical properties are determined as follows: 
15 (1) Molecular Weight 

The molecular weight of the IL-11 is found to be about 21 kD as measured by 10% 
SDS-PAGE under reducing conditions (tricine system) in accordance with the methods of 
Schagger, et aL, Anal Biochem, J66:36S'319 (1987). The protein ran as a single band. 

(2) Endotoxin Content 

20 The endotoxin content of the IL-1 1 is found to be less than 0. 1 nanogram per milligram 

IL-1 1 in the LAL (Limulus amebocyte lysate, Pyrotel, available from Associates of Cape Cod, 
Inc., Woods Hole, Massachusetts, U.S.A.) assay, conducted in accordance with the 
manufacturer's instructions. 

(3) Isoelectric Point 

25 The theoretical isoelectric point of IL-1 1 is pH 1 1 .70. As measure by polyacrylamide 

gel isoelectric focusing using an LKB Ampholine PAGplate with a pH range from 3.5 to 9.5, 
the IL-11 runs at greater than 9.5, An exact measurement can not be taken because IL-11 is 
too basic a protein for accurate pi determinations. 

(4) Fluorescence Absorption Spectrum 

30 The fluorescence absorption spectrum of the IL-11, as measured on a 0.1 % aqueous 

solution in a 1 cm quartz cell shows an emission maximum at 335-337 nm. 

(5) UV Absorption 

UV absorption of the IL-11 on a 0.1 % aqueous solution in a 1 cm quartz cell showed 
an absorbance maximum at 278-280 nm. 



23 



wo 96/38570 



PCTAJS96/04811 



(6) Amino Add Composition 

The theoretical amino acid composition for IL-11. based on its amino acid sequence 
is as follows: 



lino Add 


Number 


Mole % 


Ala 


20 


11.3 


Asp Acid 


11 


6.22 


Cysteine 


0 


0 


Glu 


3 


1.70 


Phe 


1 


0.57 


Gly 


14 


7.91 


His 


■ '4 


2.26 


He 


2 


1.13 


Lys 


3 


1.70 


Leu 


41 


23.16 


Met 


2 


1.13 


Asn 


1 


0.57 


Pro 


21 


11.86 


Gin 


7 


3.96 


Arg 


18 


10.17 


Ser 


11 


6.22 


Thr 


9 


5.09 


Val 


5 


2.83 


Trp 


3 


1.70 


Tyr 


1 


0.57 



A sample of homogenous IL-11 is subjected to vspor phase hydrolysis as follows: 6 
N HCl and 2 N phenol reagent were added to an hydrolysis vessel in which tubes containing 
45 /il of 1: 10 diluted (w/HjO) IL-11 , concentrated to dryness are inserted. Samples are sealed 
under vacuum and hydrolyzed for 36 hours at 1 lO'C. After the hydrolysis, samples are dried 
and resuspended in 500 /il Na-S sample dilution buffer. Amino acid analysis was performed 
on a Beckman 7300 automated amino acid analyzer. A cation exchange column is used for 
separation of amino acids following post-column derivatization with ninhydrin. Primary amino 
acids are detected at 570 ran and secondary amino acids are detected at 440 nm. Eight point 
calibration curves are constructed for each of the amino acids. 

Because certain amino acids are typically not recovered, results for only 5 amino acids 
are given below. Since the hydrolysis is done without desalting the protein, 100% recovery 
was achieved for most of the amino acids. 

The relative recovery of each individual amino acid residue per molecule of 
recombinant IL-11 is determined by normalizing GLX = 10 <the predicted number of 
glutamine and glutamic acid residue in IL-11 based on cDNA sequence). The value obtained 
for the recovery of GLX in picomoles is divided by 10 to obtain the GLX quotient. Dividing 
the value obtained for the recovery in picomoles of each amino acid by the GLX quotient for 



24 



wo 96/38570 PCTAJS96/04811 

that saiT^jIe gives a number that represents the relative recovery of each amino acid in the 

sample, normalized to the quantitative recovery of GLX residues. The correlation coefficient 

comparing the expected versus the average number of residues of each amino acid observed 

is greater than 0.985, indicating that the number of residues observed for each amino acid is 

5 in good agreement with that predicted sequence. 

Amino No. of Residues No. of R^idues Corrdation 

Adds Calculated Expected Coefficient 

I Asp 12.78 12 
10 2 Glu 10.00 10 

3 Gly 12.80 14 0.9852 

4 Arg 16.10 18 

5 Pro 18.40 21 

15 (7) Amino Terminus Sequencing 

im (buffered in 95% acetonitrile TFA) is sequenced using an ABI 471 A protein 
sequencer (ABI, Inc.) in accordance with the manufacturer's instructions. Amino terminus 
sequencing confirms that the thioredoxin fusion protein produced IL-1 1 contains the correct IL- 

II amino-acid sequence, and only one amino terminus is observed. 
20 (8) Peptide Mapping 

The IL-11 is cleaved with Endoproteinase Asp-N (Boehringer Mannheim) (1:500 ratio 
of Asp-N to IL-11) in 10 mM Tris, pH 8, 1 M urea and 2 mM 4-aminobenzamidine 
dihydrochloride (PABA), at 37"C for 4 hours. The sample is then run on HPLC on a C4 
Vydac column using an A buffer of 50 mM NaHP04, pH 4.3, in dHjO, a B buffer of 100% 

25 isopropanol with a gradient at 1 ml/min from 100%A to 25%A and 75%B (changing at 

1% /minute). The eluted peptide fragments are then sequenced using an ABI 471 A protein 
sequencer (ABI, Inc.) in accordance with the manufacturer's instructions. The peptide map 
confirmed that the IL-11 produced from the thioredoxin fusion protein contains the expected 
IL-11 N-terminal and C-ierminal sequences. 

30 (9) Solubility 

lL-1 1 protein is tested for solubility in the substances below with the following results: 

Water very soluble 

Ethyl Alcohol very soluble 

Acetone very soluble 

35 IM sodium chloride very soluble 

10% sucrose very soluble 



25 



wo 96/38570 PCT/US96/a481 1 

(10) Sugar Composition and Protein/Polysaccharide Content in % 

The absence of sugar moieties attached to the polypeptide backbone of the IL-1 1 protein 
is indicated by its amino acid sequence, which contains none of the typical sugar attachment 
sites. 

5 When the fusion construct is made having a hydroxylamine cleavage site, cleavage is 

carried out as follows. A thioredoxin/IL-11 fusion protein, modified as described above to 
contain a hydroxylamine cleavage site between the thioredoxin and IL-11 sequences, is 
chemically cleaved in a reaction with hydroxylamine. The modified fusion protein at a 
concentration of 2.5mg/ml is cleaved in a reaction with IM hydroxylamine in 0.1 M CHES 

10 buffer at pH 9.7. The reaction is allowed to proceed for llh at 35°C, and is terminated by 
cooling to 4°C and lowering the pH to pH 8.0 by the addition of Tris-HCl <pH 7.3). 
EXAMPLE 3 . SELECTIVE PROTEIN RELEASE 

The selective release of a fusion protein from the cytoplasm of intact, harvested, £. coli 
cells occurs as a consequence of destabilization of the outer cell membrane by a chelator such 

15 as TRIS/EDTA. The chelating agent, EDTA, is thought to penneabilize gram-negative bacteria 

by binding divalent cations that stabilize lipopolysaccharides (LPS) in the outer membrane. 
While in a presently preferred embodinient, the cells are first harvested prior to the release 
step, it is also possible to simply add chelator directly to the cell culture medium. An excess 
of chelator is added to ensure the supply of chelator is not exhausted by any entities present 

20 in the culture medium which are bound by chelator. 

Selective release is accomplished, for example, by adding a solution of 680 mM THIS, 
240 mM EDTA, pH 8.0 to the harvest/release vessel containing chilled (3° C) harvested cells. 
The final buffer composition is 50 mM Tris and 15 mM EDTA and the £. coli concentration 
in the resulting suspension is 250 grams (wet cell weight) per liter. The suspension is heated 

25 to 37 °C and incubated for 30 minutes. Base, such as 5 N NaOH, is used to adjust the pH of 
the release suspension to about pH 8.5. This treatment releases fusion protein and a spectrum 
of constitutive E. coli proteins. Approximately 50% of the total cell protein including greater 
than 80% of the protein of interest and 15 % of cellular DNA is released. The quantity of 
protein liberated from the cells after addition of the TRIS/EDTA solution to the harvest cell 

30 suspension is monitored by Bradford assay (Bradford, M., Analytical Biochemistry 72:248-254 
(1976)). 



26 



wo 96/38570 



PCTAJS96/048n 



Table 2 lists the controlled operating parameters for this release step including preferred 
urgets as well as preferred ranges. One skilled in the art will appreciate that broader ranges 
will be similarly effective. Table 3 lists the monitored operatmg parameters. 

TABLE 2 Release Step: Controiled Operating Parameters 



Parameter 


Preferred Target 


Preferred Range 


Release vessel temperature 


aye 


32-42 C 


Release vessel agitation 


170 rpm 


160-160 


Release vessel pH 


8.5 


6.0-8.8pH 


Release time 


30 rnin 


20-60 min 


TABLE 3 Release Step: Monitored 
Operating Parameters 




Parameter 


Preferred 
Target 




Total protein Released (Bradford at 30 min) 


12 mg/mL 




Total Protein Relative to Lysis 


50% 




Total DNA Released 


300 ugTml 




Total DNA Relative to Lysis 


15% 





The release process also liberates polyanionic species such as DNA. RNA, and LPS 
from the cells. The presence of these non-proteinaceous components in the release suspension 
renders it turbid and relatively unfilterable. They are removed via the clarification step, as 
described in Example 4. 

EXAMPLE 4 . CLARinCATION AND REMOVAL OF NEGATIVELY CHARGED 
MOLECULES 

Process material is clarified by selective precipitation of non-protemaceous components 
and results in a filterable process stream. Precipitation is initiated by sequential addition of 2 
M MgClj, 95% ethanol, and 2 M CaClj to the harvest/release vessel containing the release 
suspension. The concentration of reactants in the resulting suspension is 125 mM MgCl2, 75 
mM CaClj. and 14 % (volume/volume %) ethanol. The precipitation, which is executed for 
a minimum of five minutes, is controlled at 32 ''C and a pH of 7.5. The concentration of 
reactants and the pH, temperature, and reaction time are controiled within the limits presented 
in Table 4. Non-proteinaceous components are removed fi-om suspension by continuous 
centrifiigation at a maximum centrifugal force of 15,000 x g, a residence time of 1.4 minutes, 
and a sedimentation distance of 0.5 mm. If desired, residual particulates can be removed from 



27 



WO96/38570 PCT/US96/04811 

the supernatant stream by in-line 0.5-ftm filtration. Greater than 80% of the DNA and 90% 
of the LPS previously in solution is removed. Recovery of the protein of interest is circa 80% . 
The protein concentration (Bradford assay) and turbidity (ODgoo) are monitored as indicators 
of process performance. 

5 Table 4 lists the controlled operating parameters for this clarification step including 

preferred targets as well as preferred ranges. One skilled in the art can readily appreciate that 
broader ranges will be similarly effective. Table 5 lists performance characteristics of 
monitored parameters. 



25 



30 



TABLE 4 Clarification Step: Control Operating Parameters 


Parameter 


Preferred Target 


Preferred Range 


2 M Mga2 


0.083 (v/v)» 


0.066-0.10 (v/v)' 


2 M CaClj 


0.050 (v/v)' 


0.033-0.068 (v/v)» 


95% Ethanol 


0.20 (v/v)» 


0.17-0.23 (v/v)' 


Clarification vessel/Temperature 


32»C 


20-37«C 


Clarification vessel/ Agitation 


220 rpm 


21O-230rpm 


Clarification vessel/pH 


7.5 


6.8-7.8 


Clarification time 


5 min 


5-30 min 


- volume reagent/volume release suspension 






TABLE 5 Clarification Step: Monitored Operating Parameters 


Parameter 


Preiferred 
Target 




Turbidity of bulk filtered supernatant (OOeoo) 


<0.32 0D« 


» 


Total protein in bulk filtered supernatant 
(Bradford assay) 


6 mg/mL 




Fusbn protein concentration in bulk filtered supernatant 2 mg/mL 
(Reversed Phase Analysis) 




Total DNA after clarification 


30 ug/mt 




Fold LPS removal 


1000 fbkJ 





EXAMPLE 5 - SELECTIVE PRECIPITATION 

After the clarification step, optionally, it is possible to use a selective precipitation of 
the protein of interest. This is accomplished by adding 1 volume of IM ZnClj to 19 volumes 
of 4 **C clarified supernatant. The pH of the suspension is maintained at 7.0 with 2.5 M Tris 
35 base. After incubation of the suspension for 15 minutes, the resulting precipitate is recovered 



28 



wo 96/38570 



PCT/US96/04811 



by continuous centrifugation at a maximum centrifugal force of 15,000 x g, a sedimentation 
distance of 2.5 cm, with a residence time of 3 minutes. The recovered precipitate is 
subsequently frozen in liquid nitrogen and stored frozen at - 80 **C. The precipitate is 
subsequently solubilized in 20 mM Tris/100 mM EDTA (pH = 8.0) at 20°C at a ratio of 100 
5 grams of precipitate per liter buffer. . Greater than 75% of the protein of interest is recovered 

over the process as measured by reversed phase HPLC. 
EXAMPLE 6 . PURIFICATION 

After the clarification step (Example 4), a processing method which can be used 
(alternative to selective precipitation (Example 5)) is ultrafiltration/diafiltration using tangential- 

10 flow membrane filtration. This step concentrates the clarified fusion protein solution and 

exchanges the protein into a low-ionic strength buffer that is suitable for ion exchange 
chromatography. The membrane used in the tangential-flow device serves as a porous filter 
that separates substances on the basis of molecular weight. Solution conqjonents of high 
molecular weight (such as proteins) are retained by the membrane, and components of low 

15 molecular weight (such as inorganic salts and buffer compounds) freely pass through the porous 

membrane structure and are removed in the permeate. 

When a replacement buffer is added to the tangential-flow retentate at a rate 
approximately equal to the rate at which the buffer is drawn through the membrane and 
discarded, the initial buffer is continuously diluted (protein diafiltration). Under these 

20 conditions, compounds of low molecular weight are readily exchanged and the protein 

concentration remains constant. The addition of five retentate volumes of buffer results in 
^99% replacement of the initial buffer. When buffer is drawn from the tangential-flow device 
at a rate faster than that at which replacement buffer is added to the retentate, the protein 
solution is concentrated. 

25 The first Ultrafiltration/Diafiltration Step (UFDF #1) of the recovery process 

concentrates and buffer exchanges the fusion protein present in the filtered clarified supernatant. 
This process step uses a series of plate-and-frame membrane cartridges for example, Millipore 
Pellicon regenerated cellulose cassette filters, with a membrane molecular weight cutoff of < 10 
kDa. In addition, in-line prefilters (Millipore Milligard, 1.2 pan pore size) are used for 

30 continuous filtration of the retentate to remove any particulates which could foul the 
membranes. The temperature target for the step is 8**C. 

Before use, the ultrafiltration skid and membranes are flushed with 20 mM Tris, 0.3 M 
NaCl, pH 8.0 to equilibrate the system. A solution of 4 M NaCl is added at a ratio of 9.32% 
(w/v) to the filtered clarified supernatant (PCS) in a holding tank, which is then mixed before 

35 processing begins. The PCS is pumped across the membrane at a positive transmembrane 



AVO 96/38570 PCTAJS96/04811 

pressure, and the retentate is recirculated to the vessel while the permeate is directed to waste. 
The FCS is concentrated approximately three fold (Concentration I). After Concentration I is 
completed, a diluent stream of 20 mM Tris, 0.3 M NaCl, pH 8.0 is added to the holding vessel 
at a flow rate equal to the permeate flow rate (Diafiltration I). In this manner, the buffer of the 
FCS is continuously diluted with the diluent buffer. When the total permeate volume is ^ 5 times 
the Concentration I final volume, Diafiltration I is complete. Over 99% of the low-molecular 
weight solutes present in the original buffer are removed. 

When Diafiltration I is completed, a diluent stream of 20 mM Tris, pH 8.0 is added to 
the retentate vessel at a flow rate equal to the permeate flow rate (Diafiltration II). When the total 
permeate volume is ^ S times the Concentration I final volume, Diafiltration II is complete. 
After Diafiltration II is completed, the retentate is then concentrated an additional 10% 
(Concentration II). The equipment is flushed with a sufficient volume of 20 mM Tris, pH 8.0. to 
wash out the concentrated protein solution, bringing the total pool volume to approximately 0.4 
times the starting volume of FCS. The UFDF #1 pool is pumped out of the vessel and filtered 
through an autoclaved 0.2-/im filter attached to a clean, autoclaved pressure vessel. The filtered 
pool is stored at 2 - 10**C before ftirther processing. 

The average recovery of fiision protein from this step is 89% . Table 6 lists the operatmg 
parameters that are controlled during this step. 

TABLE 6 Operating Parameters Controlled for the UFDF #1 Step 



Procedure 



Parameter 



Recommended Target 



All procedures 



Equilibration 

Concentration I 
Diafilu-ation I 



Diafiltration D 



Concentration II 



Inlet feed pressure 

Retentate pressure 

Pressure drop over pre-filiers 

Holding unk temperature 

Retentate and permeate pH 

Retentate and permeate conductivity 

Final volume 

Retentate and permeate pH 
Permeate volume 

Permeate conductivity 
Retentate and permeate pH 
Permeate volume 

Permeate conductivity 
Final volume after flush 



80 psig 
40 psig 
£25 psig 
8 ^'C 
8.0 

17 - 23 mS/cm 

0.3 X starting volume of FCS 

8.0 

-^S X Concentration 
I fmal volume 

17-23 mS/cm 

8.0 

^5 X Concentration 
I final volume 

<2 mS/cm 

0.4 X starting volume of FCS 



30 



wo 96/38570 



PCT/US96/04811 



The purification of protein of interest, for example rhlL-ll, involves, first, 
purification of the fusion protein away from positively charged contaminants; second, cleavage 
of the fusion protein into its components, namely rhlL-ll and thioredoxin; and third, 
purification of the protein of interest, e.^., rhlL-U. 
5 Purification of the fusion form of the protein is accomplished with two ion- 

exchange chromatography steps. After cleavage of the fusion protein, two additional ion- 
exchange chromatography steps are used to purify the cleaved form of the fusion protein. In 
the first chromatography step, an anion-exchange resin, for example Toyopearl QAE, is used 
to adsorb the fusion protein (and other negatively charged proteins) from the process stream, 

10 and in the second step, a cation-exchange resin, for example S Sepharose Fast Flow, is used 
to adsorb positively charged proteins from the process stream while fusion protein flows 
through the column. 

Toyopearl QAE 550C is a strong anion-exchange resin composed of a rigid 
polymeric support that is covalently derivatized with a quaternary amine. Proteins and 

1 5 poly ionic substances (for example, polynucleotides and lipopolysaccharides) with a net negative 
charge at the pH of operation bind to this resin as a function of solution ionic strength. The 
Toyopearl QAE resin is used to adsorb fusion protein from the product stream under low-salt 
conditions through ionic interactions. Elution of the thioredoxin/rhIL-11 fusion protein is 
accomplished by both lowering the pH and increasing the ionic strength in the appropriate 

20 buffer. Appropriate eluants include 20-100 mM Tris buffers at pH 7.5-8.5 containing 100-500 

mM NaCl, or 50-200 mM Histidine buffers at pH 5.5-6.6, containing 0-150 mM NaCl. 

S Sepharose Fast Flow is a strong cation-exchange chromatography gel composed 
of a cross-linked agarose matrix that is derivatized with sulphonate groups. Contaminants, both 
proteins and polyionic substances with isoelectric points higher than the pH of the running 

25 buffer, bind to S Sepharose Fast Flow via charge interactions. 

The fusion protein is bound to and then eluted from the QAE column, passed 
through the S Sepharose Fast Flow column, and recovered in the S Sepharose Fast Flow 
unbound fraction. The conditions for elution of the QAE column have been optimized to 
minimize binding of thioredoxin/rhlL-ll fusion protein to the S Sepharose Fast Flow resin, 

30 while contaminants with a higher pi than the fusion protein bind to the S Sepharose Fast Flow 
colunui and are removed from the process stream. 

Both the Toyopearl QAE 550C column and the S Sepharose Fast Flow wlumn 
are equilibrated and run at ambient temperature. The S Sepharose Fast Flow column is 
equilibrated with 150 mM Histidine, 150 mM NaCl, pH 6.2 followed by a -second ^uilibration 

31 



wo 96/38570 PCT/US96/04811 

with 75 mM Histidine, 75 ixiM NaCl, pH 6.2. Next the Toyopearl QAE column is equilibrated 
separately from the S Sepharose Fast Flow column with 20 mM Tris pH 8.0. The UFDF #1 
concentrate, as described supra, is then loaded onto the equilibrated Toyopearl QAE column 
at a linear velocity of < 1.5 cm/min . 

After the load is completed, the Toyopearl QAE column is washed with 20 mM 
Tris, pH 8.0 at < 1.5 cm/min. The QAE column is then eluted with 75 mM NaCl, 75 mM 
Histidine, pH 6.2 at < 1 .0 cm/min. As the protein begins to elute from the QAE column, the 
column outlet is connected to the S Sepharose Fast Flow column. The tandem-column eluate 
peak is collected as a single pool. 



10 



32 



wo 96/38570 PCr/US96/04811 

Thioredoxin/rhIL-11 fusion protein elutes from the columns as a peak of 
approximately 5 column volumes. (The column volume is based on the dimensions of the 
Toyopearl QAE column.) Table 7 lists the operating parameters that are routinely monitored 
during this step. The average yield of lhioredoxin/rhIL-11 fusion protein from this step is 
91%. 

TABLE 7 Operating Parameters Monitored for the Toyopearl QAE/ 



S Sepharose Fast Flow Step 



Procedure 


Parameter 


Preferred Target 


fifth tmn PArktno* TovODesrI 
QAE 550C 


Bed height 


15-19 cm 


Column Packing: 
S Sepharose FF 


RpH hptaht 

DCU ilCIJ^Kl 


10-15 cm 


Equiltbration: Toyopearl QAE 
column 


linear velocity" 


£1.6 cm/min 




pH 


8.0 ± 0.2 




volume** 


^ 4 column volumes 


EquHib ration: 

S Sepharose FF column 


linear velocity* 


4 cnn/min 




pH (equilibration #1) 


6.2 ± 0.2 




volume** 


& 3 column volumes 




pH (equBibration #2) 


6.2 t 0.2 




volume" 


^ 3 column volumes 


Load: UFDF #1 Concentrate 


volume' 


1- 2 column volumes 




linear velocity* 


SI.O cm/min 


Wash: Toyopearl 
QAE column 


linear velocity* 


£ 1 .5 cm/min 




volume 


3-4 column volumes 


Elution 


linear velocity* 


-£ 1 .0 cm/min 



• Flow rate for the QAE column equilibration and load volume for the QAE column are based on the 

dimensions of the Toyopearl QAE column. 
^ Column equDibration is continued with additional buffer if the pH is out of the target range. 
^ Flow rate for the S Sepharose column equilibration is based on the dimensions of the S Sepharose 



column 

Following purification of the fusion protein by QAE Toyopearl and S Sepharose 
Fast Flow chromatography, the protein is chemically cleaved, in this example at the 
asparaginyl-glycyl peptide bond in the fusion linkage sequence, to generate rhIL-11 and 
thioredoxin. 

Cleavage of the asparaginyl-glycyl peptide bond under mildly basic conditions 
using hydroxylamine as the nucleophilic reagent is well documented, md a general method has 
been described (Bomstein, P. , and Balian. G. Cleavage at Asn-Gly bonds with hydroxylamine. 



33 



PCTAJS96;04811 

WO 96/38570 

Meth, EnzymoL 47(E): 132-145 (1977)). The asparaginyl-glycyl peptide bond is particularly 
sensitive to hydroxylaminolysis. although asparaginyl-leucyl, asparaginyl-methionyl, and 
asparaginyl-alanyl cleavages have been reported and may occur relatively slowly. Bomstein, 
supra. 

The Toyopearl QAE/S Sepharose FF eluate pool is added to the cleavage reaction 
vessel. The hydroxylamine cleavage solution, 3.0 M hydroxy lamine-HCl, 0.3 M CHES, pH 
9.7, is added to the vessel at a volume equal to one-half of the volume of the Toyopearl QAE/S 
Sepharose FF eluate, to produce a final reaction mixture concentration of 1.0 M 
hydroxylamine-HCl and 0.1 M CHES. The pH of the mixnire is adjusted to 9.3 (measured 
at 35**C) by addition of 10 N NaOH and the ten^jerature is brought to approximately SS^'C. 

After 9 hours of gentle agitation, the cleavage reaction is ended by reducing the 
temperature and pH of the reaction mixture. The temperanire of the reaction mixture is 
lowered to ^8**C while the pH of the mixture is controlled at 9.3 by addition of a 
neutralization solution, 2.0 M TRIS, pH 7.3. After cooling, additional neutralization solution 
is added until a volume equal to one-fifth of the volume of the reaction mixture has been added. 
The neutralized cleavage mixture is stored at :£8 before fiirther processing. The cleavage 
reaction opertting parameters are detailed in Table 8. 

The extent of thioredoxin/rhIL-1 1 fiision protein cleavage averages approximately 
73%. The most prominent proteins present in the process stream following the cleavage 
reaction are residual, uncleaved thioredoxin/rhlL-ll fiision protein and the two cleavage 
products, rhIL-11 and thioredoxin. 

TABLE 8 Operating Parameters for the Hydroxylamine Cleavage Step 



Procedure . 


Parameter 


Preferred Target 


Cleavage start 


volume of cleavage solution 


0.5 X load volume 




aghation rate* 


15 rpm 




starting pH* 


9.3 




temperature 


35«C 




time 


9hr 


Cleavage cooMown 


pH 


9.3 




final temperature 


-S8»C 


Cleavage end 


total volume of neutralization 


0.2 X total reaction mixture volume 


solution 





Agitation rate after all solutions have been added and the proper temperature and pH have been 
attained. 

The pH is determined at 35 "C. 



34 



wo 96/38570 FCT/OS96,04«11 

Following chemical cleavage of the fusion protein into its component parts of rhIL-11 
and thioredoxin, the process stream is concentrated and buffer exchanged into a lower ionic 
strength buffer using a second ultrafiltration/diafiltration step with tangential-flow membrane 
filtration. The second ultrafiltration/diafiltration step (UFDF #2) uses a series of plate-and- 
frame membrane cartridges, for example, MUlipore Pellicon regenerated cellulose cassette 
filters, with a MW cutoff of ^ 10 kDa. In addition, in-line prefihers (Millipore Milligard, 
1 .2 /tm pore size) are used for continuous filtration of the retentate to remove any particulates 
which could foul the membranes. The process step is carried out at a target temperature of < 
8"C. Before use, the ultrafiltration skid and membranes are equilibrated with 20 mM Tris, 0.2 
M NaCl, pH 8.0. After membrane equilibration, the cleavage mixture is pumped from the 
holding vessel across the membrane at a positive transmembrane pressure. The retentate is 
recirculated to the holding vessel and the permeate stream is directed to waste. This 
concentrates the rhlL-ll present in the cleavage reaction mixture, and reduces, the initial 
volume by a factor of approximately 6.5. 

After concentration is completed, a diluent stream of 20 mM Tris. 0.2 M NaCl, pH 8.0 
is added to the retentate vessel at a flow rate equal to the permeate flow rate. In this manner, 
the cleavage reaction buffer is continuously diluted with the diluent buffer. This diafiltration 
process is continued until the permeate volume is at least five times the concentrated retentate 
volume. After diafiltration is completed, the ultrafiltration equipment is flushed with sufficient 
20 mM Tris, 0.2 M NaCl, pH 8.0 to wash out the concentrated protein solution, 
(approximately 30-40% of the concentrate volume), and the rhIL-1 1 pool (concentrated solution 
and flush) is stored in the holding vessel until the next processing step begins. The average 
recovery of rhIL-11 from this step is 89%. Table 9 lists the operating parameters that are 
controlled during this step. 

TABLE 9 Operating Parameters Controlled for the UFDF #2 Step 



Procedure Parameter Recommended Target 

All procedures Inlet feed pressure £30psjQ 

Retentate pressure 1 0 psig 

Holding tank temperature 2 
Pressure drop over pre-fitters £60 psig 
Concentration Rnal volume 0.1 5 x starting volume 

Diafiltration Permeate volume ^ 5 times the retentate volume 



35 



wo 96/38570 PCT/US96/04811 

Following the UFDF #2 stq), the rhlHl is purified with two ion-exchange 
chromatography steps. In the first step, a cation-exchange resin, in this example CM Sepharose 
Fast Flow, is used to adsorb the rhIL-11 from the process stream. All steps are performed at 
a temperature of 2-8°C. The colunm is first equilibrated with a buffer containing 20 mM Tris, 
0.5 M NaCi, pH 8.0, and then is equilibrated with a buffer containing 20 mM Tris, pH 8.0. 
The unpurified rhIL-11 pool is pumped through the CM Sepharose column at a velocity of 
<3.2 cm/min. The column is then washed with a solution of 0.15 M Glycine, pH 9.5. The 
CM Sepharose FF column is eluted with 0.15 M Glycine, 0.15 M NaCl, pH 9.5 into a 
collecting vessel that contains a specified volume of purified water at ambient temperature for 
dilution of the peak. The column eluate peak is collected as a single pool, to which additional 
purified water at ambient temperature is added to achieve a final dilution of one part elution 
peak with 3 parts purified water. The recovery of rhIL-11 from this step averages 
approximately 75 % . The column operating parameters are detailed in Table 10. 



TABLE 10 Operating Parameters for CM Sepharose Fast Flow 



Procedure 


Paranteter 


Preferred Value 


Column Packing: 




11.13 cm 


CM Sepharose FF 


Bed height 


Equilibration 1 


linear velocity 


s3.2 cnrt/min 




pH 


7.8 - 8.2 




volume" 


^3.5 column volumes 




conductivity 


&30 mS/cm 


Equilibration 2 


linear velocity 


^3.2 cm/min 




pH 


7.8 - 8.2 




volunne* 


&6 column volumes 




conductivity 


£4 mS/cm 


Load 


linear velocity 


£3.2 cm/min 




volunne 


4-5 column volumes 


Dilution of load 


volume 


4 X lodd volume 




conductivity 


£4 mS/cm 


Wash 


Qnear velocity 


£3.2 cm/min 




pH 


9.3 - 9.7 




volume 


& 1 0 column volumes 




conductivity 


£2 mS/cm 


Elution 


linear velocity 


£2 an/min 



' Column equilibration is continued with additional buffer if the pH or conductivity targets are not 



met 



36 



wo 96/38570 PCT/US96/04811 

The final chromatographic step in the purification process is an anion exchange 
chromatography step, in this example Toyopearl QAE 550C, that adsorbs anionic contaminants 
from the rhIL-11 product stream; the rhIL-11 is recovered in the Toyopearl QAE unbound 
fraction. 

Column operations are carried out at room temperature. The Toyopearl QAE 
550C column is equilibrated first with 1 M Glycine, pH 9.5 and then with 40 mM Glycine, pH 
9.5. The diluted CM Sepharose peak is pumped through the column and chased with 
additional 40 mM Glycine, pH 9.5 to wash the remaining rhIL-11 product in the load out of 
the column. The unbound load effluent and the wash are collected into a storage vessel. 
This pool is then neutralized by the addition of 5% (v/v) 870 mM NaH2P04, pH 5.0. Column 
operating parameters are detailed in Table 11. 

Recovery of rhIL-11 for this step averages approximately 957c. 
Thioredoxin/rhIL-1 1 fusion protein that is not removed before this step is consistently removed 
10 a level below detection by SDS-PAGE. The eluate pool contains highly purified rhlHl, 
with minor amounts {<5%) of shorter length rWL-ll molecules. 

TABLE 11 Operating Parameters for the Toyopearl QAE Step 



Procedure 


Parameter 


Recommended Target 


Column Packing: Toyopearl 




20.22 cm 


QAE 550C 


Bed height 


Equilibration 1 


volume 


1.3 column volumes 




linear velocity 


£2.6 cnn/min 


Equilibration 2 


pH 


9,3-9.7 




conductivity 


£1.1 mS/cm 




volume' 


&3 column volumes 




linear velocity 


£2.6 cm/min 


Load 


volume 


£7,5 column volumes 




temperature 


ilS'C 




linear velocity. 


£2.6 cm/min 


Wash 


linear velocity 


£2.6 cm/min 


Neutralization 


volume of neutralization 
solution 


5% (v/v) of fiowthrough pool 



• Column equilibration is continued with additional 40 mM Glycine, pH 9.6, if the pH or 
conductivity targets are not met. 

While the present invention has been described in terms of specific methods and 

compositions, it is understood that variations and modifications will occur to those skilled in 

the art upon consideration of the present invention. The effective reagent concentrations, pH, 

and temperatures are process and protein specific and can be readily adjusted according to the 



37 



wo 9608570 PCr/US96/04Wl 

nature of the anionic contaminants and chemical characteristics and stability of the protein of 
interest as is readily apparent to one skilled in the art. 

Numerous modifications and variations in the invention as described in the above 
illustrative examples are expected to occur to those iskilled in the art and, consequently, only 
such limitations as appear in the appended claims should be placed thereon. Accordingly, it 
is intended in the appended claims to cover all such equivalent variations which come within 
the scope of the invention as claimed. 



38 



wo 96/38570 PCT/US96/04811 

WHAT IS CLAIMED: 

1 . A method for release and purification of a thioredoxin-likc fusion protein 
from a cell into a solution comprising the steps of: 

adding chelator to said solution, to release said protein from said cell to said 

solution, and 

adding a divalent cation/alcohol solution to said solution to form a first soluble 
fraction and a first insoluble fraction. 

2. The method of claim 1 further comprising the step of: 
isolating said protein from said first soluble fraction. 

3. The method of claim 1 further comprising the step of: 
raising the temperature of said solution. 

4. The method of claim 3, wherein said temperature is raised about 20 to 40 
degrees Celsius. 

5. The method of claim 4, wherein said temperature is raised from about 3°C 

to about 37X. 

6. The method of claim 1 , wherein said cell is £. coli. 

7 . The method of claim 1 , wherein said chelator is a member selected from the 
group consisting of EDTA, EGTA, CDTA, DPTA, and citric acid. 

8. The method of claim 6. wherein said chelator is EDTA. 

9. The method of claim 6, wherein the final concentration of said EDTA is 0. 1 

to 100 mM. 

10. The method of claim 6, wherein the final concentration of said EDTA is 15 

mM. 

11 . The method of claim 1 , wherein said divalent cation is a member selected 
from the group consisting of magnesium, manganese, and calcium. 

12. The method of claim 1. wherein said divalent cation/alcohol solution is a 
magnesium/numganese/alcohol solution. 

13. The method of claim 1, wherein said divalent cation/alcohol solution is a 

magnesium/calcium/alcohol solution. 

14. The method of claim 1, wherein said divalent cation/alcohol solution is a 
magnesium/manganese/calcium/alcohol solution. 

15. The method of claim 1, wherein said divalent cation/alcohol solution is 

magnesium/alcohol. 



39 



PCTAJS96/04811 

WO 96/38570 

16. The method of claim 1 , wherein said alcohol is a member selected from the 
' group consisting of methanol, ethanol, propanoic isopropanol, iso-butanol. and tertiary-butanol. 

17. The method of claim 1, wherein said alcohol is ethanol. 

18. The method of claim 1, wherein the final concentration of said alcohol is 

from 5 to 30%. 

19. The method of claim 17, wherein the final concentration of said alcohol is 

14%. 

20. The method of claim 1, wherein said divalent cation final concentration 

ranges from 1 to 1000 mM. 

21. The method of claim 20, wherein said divalent cation final concentration 

ranges from 50 to 200 mM. 

22. The method of claim 20, wherein said magnesium final concentration is 200 

mM. 

23. The method of claim 20, wherein said magnesium final concentration is 125 
mM, and calcium final concentration is 75 mM. 

24 . The method of claim 20. wherein said magnesium final concentration is 125 
mM, and manganese final concentration is 75 mM. 

25 . The method of claim 20, wherein said magnesium final concentration is 125 
mM. manganese final concentration is 38 mM and said calcium final concentration is 38 niM. 

26. The method of claim 1, further comprising the step of: 

adding zinc to said first soluble fraction to form a second insoluble fraction and 

a second soluble fraction. 

27 . The method of claim 20, further comprising the step of isolating said protein 

from said second insoluble fraction. 

28. The method of claim 20, further comprising the step of adding chelator to 

solubilize said protein from said second insoluble fraction. 

29. The method of claim 20, wherein said zinc is a member selected from the 
group consisting of zinc chloride, zinc sulphate, and zinc acetate. 

30. The method of claim 20, wherein said zinc final concentration is 1 to 500 

mM. 

31. The method of claim 20, wherein said zinc final concentration is 50 

zinc chloride. 

32. The method of claim 1 , wherein said protein is recombinantly produced in 
a transformed host cell. 



40 



PCTAJS96/04811 

WO 96/38570 

33. A method for purification of a protein in solution comprising the steps cf 
adding a divalent cation/alcohol solution to said solution to form a first 
soluble fraction and a first insoluble fraction. 

34. The method of claim 33, further comprising the step of: 
isolating said protein from said first soluble fraction. 

35. The method of claim 33, wherein said divalent cation is a member selected 
from the group consisting of magnesium, manganese, and calcium. 

36. The method of claim 33, wherein said alcohol is a member selected from 
the group consisting of methanol, ethanoL propanol, isopropanol, iso-butanol, and tertiary- 
butanol, 

37. The method of claim 33, wherein said alcohol is ethanol. 

38. The method of claim 33, wherein the final concentration of said alcohol is 
from 5 to 30%. 

39. The method of claim 33, wherein the final concentration of said alcohol is 

14%. 

40. The method of claim 33, wherein the final concentration of said divalent 

cation is from 1 to 1000 mM. 

41 . The method of claim 33, wherein said magnesium final concentration is 200 

niM. 

42. The method of claun 33, wherein said magnesium final concentration is 125 
mM, and said calcium final concentration is 75 mM. 

43. A method for purification of a protein in solution comprising the steps of: 
adding zinc to said solution to form a soluble fi-action and an insoluble 
fraction, 

isolating said protein from said insoluble fraction. 

44. The method of claim 43, wherein said zinc final concentration is from 1 to 

500 mM. 

45. The method of claim 43, wherein said zinc final concentration is 50 mM 

zinc chloride. 



41 



PCTAJS96/04811 

WO 96/38570 

46. A method for purification of a protein in solution comprising the steps of: 
adding divalent cation/alcohol to said solution to form a first soluble fraction 

and a first insoluble fraction, 

adding zinc to said first soluble fraction to form a second insoluble fraction 

and a second soluble fraction, and 

isolating said protein from said second insoluble fraction. 

47. A method for release of a thioredoxin IL-1 1 fusion protein from a cell into 
a solution comprising the step of adding EDTA to said solution. 

48. The method of claim 47, wherein the final concentration of said EDTA 

ranges from 0.1 to 100 mM. 

49. The method of claim 48, wherein the final concentration of said EDTA is 

15 mM. 

50. The method of claim 47, frirther comprising the step of : 

adding a solution comprising MgClj, CaCl^ and ethanol to form a soluble fraction 
and ah insoluble fraction, and 

isolating said protein from said soluble fraction. 

51 . The method of claim 50 wherein the final concentration of said MgClj and 
CaCl2 ranges from 1 to 1000 mM and the final concentration of said ethanol ranges from 5 to 
30%. 

52 . The method of claim 5 1 wherein said MgClj final concentration is 125 mM , 
said CaCla final concentration is 75 mM and said ethanol final concentration is 14%. 



42 



wo 9608570 PCT/US96/048n 

53. A method for purifying a protein of interest, 

wherein a protein prior to cleavage is negatively charged and comprises a fusion 
partner and said protein of interest and after cleavage is a positively charged protein of interest, 
comprising the steps of: 

binding said negatively charged protein to a first anion exchange resin, 
eluting said negatively charged protein with a first eluant to form a first eluate, 
applying said first eluate to a first cation exchange resin, 
collecting said negatively charged protein, in an unbound fraction, from said first 
cation exchange resin, 

cleaving said negatively charged protein, to form a positively charged protein, 
binding said positively charged protein to a second cation exchange resin, 
eluting said positively charged protein with a second eluant to form a second 

applying said second eluate to a second anion exchange resin, and 



eluate. 



collecting said positively charged protein in an unbound fraction, from said 
second anion exchange resin. 

54. The method of claim 53, wherein said protein is a thioredoxin fusion 

protein. 

55, The method of claim 54, wherein said thioredoxin fusion protein comprises 

thioredoxin and IL-11. 

56 The method of claim 53, wherein said first and said second anion exchange 
resins are anion exchange resins having positively charged members selected from the group 
consisting of diethyleaminoethane (DEAE), polyethyleneimine (PEI). and quaternary 

aminoethane (QAE). 

57. The method of claim 53, wherein said first and second cation exchange 
resins are cation exchange resins having negatively charged members selected fi-om the group 
consisting of sulfonyl, sulfylpropyl (SP), carboxyl, and carboxy methyl. 

58 . The method of claim 55, wherein said first eluant is a member selected from 

the group consisting of: 

(a) 20tol00mMTris, at pH 7.5 to 8.5, 100-500 mMNaCl and 

(b) 50 to 200 mM histidine buffer, at pH 5.5 to 6.6, 0 to 150 mM NaCL 

59. The method of claim 55, wherein said second eluant is 50 to 300 mM 
glycine buffer, pH 9.0 to 10.0. and 100 to 500 mM NaCl. 



43 



PCT/US96/04811 

WO 96/38570 

60. A method for purifying a protein, comprising the steps of: 
binding a fusion protein to a first resin. 

eluting said fusion protein with a first eluant to form a first eluate, 
applying said first eluate to a second resin, 

collecting said fusion protein in an unbound fraction from said second resin, 

cleaving said fiision protein to form cleaved protein, 

binding said cleaved protein to a third resin, 

eluting said protein with a third eluant to form a third eluate, 

applying said third eluate to a fourth resin, and 

collecting said protein in an unbound fraction from said fourth resin. 

61 . The method of claim 60, wherein said fusion protein is negatively charged 
prior to cleavage. 

62. The method of claim 60, wherein said fiision protein is a thioredoxin fusion 

protein. 

63. The method of claim 62, wherein said thioredoxin fiision protein is 

thioredoxin/IL-11 fusion protein. 

64. The method of claim 61, wherein said first and fourth resin is an anion 
exchange resin and wherein said second and third resin is a cation exchange resin. 

65. The method of claim 60, wherein said fiision protein is positively charged 
prior to cleavage. 

66. The method of claim 65, wherein said first and fourth resin is a cation 
exchange resin and wherein said second and third resin is an anion exchange resin. 

67. A method for purifying a protein, comprising the steps of: 
applying a fusion protein to a first resin, 

collecting said fiision protein in an unbound fraction from said first r^sin, 

binding said fusion protein to a second resin. 

eluting said fiision protein with a first eluant to form a first eluate, 

cleaving said fiision protein to form cleaved protein, 

binding said cleaved protein to a third resin, 

eluting said protein with a second eluant to form a second eluate, 

applying said second eluate to a fourth resin, and 

collecting said protein in an unbound fi-action from said fourth resin. 

68. The method of claim 67, wherein said fiision protein is a thioredoxin fiision 

protein. 



44 



WO96/38S70 PCnyUS96/04811 

69. The method of claim 68, wherein said thioredoxin fusion protein is 
thioredoxin/IL-11 fusion protein. 

70. The method of claim 67, wherein said fusion protein is negatively charged 

prior to cleavage. 

71. The method of claim 70, wherein said first and third resin is a cation 
exchange resin and wherein said second and fourth resin is an anion exchange resin. 

72. The method of claim 67, wherein said fusion protein is positively charged 
prior to cleavage. 

73. The method of claim 72, wherein said first and third resin is an anion 
exchange resin and where said second and fourth resin is a cation exchange resin. 

74. A method for purifying a protein, comprising tiie steps of: 
applying a fusion protein to a first resin, 

collecting said fusion protein in an unbound fi-action from said first resin, 
binding said fusion protein to a second resin, 

eluting said fusion protein from said second resin with a first eluant to form a 



first eluate, 



cleaving said fusion protein to form cleaved protein, 

applying said cleaved protein to a third resin, 

collecting said protein in an unbound fraction from said third resin, 

binding said protein to a fourth resin, and 

eluting said protein from said fourth resin. 

75 . The method of claim 74 , wherein said fusion protein is a thioredoxin fusion 



protein. 



76. The method of claim 74, wherein said fusion protein is negatively charged 
prior to cleavage. 

77. The method of claim 75, wherein said thioredoxin fusion protein is 
thioredoxin/IL-11 fusion protein. 

78. The method of claim 76, wherein said first and fourth resin is a cation 
exchange resin and wherein said second and third resin is an anion exchange resin. 

79. The method of claim 74, wherein said fusion protein is positively charged 
prior to cleavage. 

80. The method of claim 79, wherein said first and fourth resin is an anion 
exchange resin and wherein said second and third resin is a cation exchange resin. 



45 



wo 96/38570 



eluate, 



PCTA3S96/04811 

81 . A method for purifying a protein, comprising the steps of: 
binding a fusion protein to a first resin, 

eluting said fusion protein from said first resin with a first eluant to form a first 
applying said first eluate to a second resin, 

collecting said fusion protein in an unbound fraction fi-om said second resin, 

cleaving said fusion protein to form cleaved protein, 

applying said cleaved protein to a third resin, 

collecting said protein in an unbound fraction from said third resin, 

binding said protein to a fourth resin, and 

eluting said protein from said fourth resin. 

82. The method of claim 8 1 , wherein said fusion protein is a thioredoxinjusion 



protein. 



83. The method of claim 81 , wherein said fusion protein is negatively charged 
prior to cleavage. 

84. The method of claim 82, wherein said thioredoxin fusion protein is 
thioredoxin/IL-11 fusion protein. 

85. The method of claim 83, wherein said first and third resin is an anion 
exchange resin and wherein said second and fourth resin is a cation exchange resin. 

86. The method of claim 81, wherein said fiision protein is positively charged 
prior to cleavage. 

87. The method of claim 86, wherein said first and third resin is a cation 
exchange resin and wherein said second and fourth resin is an anion exchange r-esin. 

88. A method for purifying IL-11, comprising the steps of: 
binding thioredoxin-lL-11 to a first anion exchange resin, 
eluting with a first eluant to form a first eluate, 
applying said first eluate to a first cation exchange resin, 

collecting said thioredoxin-IH 1 , in an unbound fraction, from said first cation 
exchange resin, 

cleaving said thioredoxin-IL-l 1 fusion protein to form positively charged IL-1 1 , 

binding said positively charged IL-1 1 to a second cation exchange resin, 

eluting said IL-11 with a second eluant to form a second eluate, 

applying said second eluate to a second anion exchange r-esin, 

collecting said IL-11, in an unbound fraction, from said second anion exchange 

resin. 



46 



wo 96/38570 PCT/US96/04811 

89. The method of claim 88. wherein said first and second anion exchange 
resins are anion exchange resins having positively charged members selected from the group 
consisting of diethyleaminoethane (DEAE), polyethyleneimine (PHI), and quaternary 
aminoethane (QAE). 

90. The method of claim 89, wherein said anion exchange resin is Toyopearl 

QAE. 

91 . The method of claim 88, wherein said first and second cation exchange 
resins are cation exchange resins having negatively charged members selected from the group 
consisting of sulfonyl, sulfylpropyl (SP), carboxyl, and carboxy methyl. 

92. The method of claim 91, wherein said first cation exchange resin is S 

Sepharose Fast Flow. 

93. The method of claim 91 , wherein said second cation exchange resin is CM 

Sepharose Fast Flow. 

94. The method ofclaim 88, wherein said first eluant is a member of the group 

consisting of: 

(a) 20 to 100 mM Tris, at pH 7.5 to 8.5, 100-500 mM NaCl and 

(b) 50 to 200 mM histidine buffer, at pH 5.5 to 6.6, 0 to 150 mM NaCl. 

95. The method of claim 88, wherein said second eluant is 50 to 300 mM 
glycine buffer, pH 9.0 to 10,0, and 100 to 500 mM NaCl. 

96. A method for purifying IL-1 1, comprising the steps of: 
binding thioredoxin-IL-1 1 to Toyopearl QAE, 
eluting with 75 mM NaCl, 75 mM histidine, pH 6.2, 
applying said first eluate to S Sepharose Fast Flow, 

collecting said thioredoxin-IL-11 in an unbound fraction from said S 
Sepharose Fast Flow, 

cleaving said thioredoxin-IL-11 fusion protein to form positively charged 

binding said positively charged IL-11 to CM Sepharose Fast How, 
eluting said lL-11 with 0.15 M glycine, 0.15 M NaCl, pH 9.5, 
applying said second eluate to Toyopearl QAE, and 
collecting said IL-11 in an unbound fraction from said Toyopearl QAE. 



47 



INTERNATIONAL SEARCH REPORT 



Inter. ■*> AffUaAin No 

PCT/US 96/04811 



tp^"'" Cffis/ir"' CQ7ia4/54 C07K1/.18 Ce7Kl/3G 

i Accofxling to Interaational Patent gasgfication (IPC) orlo both national d agification and IPC 

I B. FIELDS SEARCHED — 

Minimum documentation searched (dassification system followed by dassification symbols) 

IPC 6 C12N C07K 

Documentation searched other than minimiim 



documentotion to tiie extent that such documents are inchided in the fields seardied 



Electronic data base 



consulUd (hoing ft. iiKonation.1 «a«h (name of dat. b» «.d, who. praetiad. «a«h lcn». used) 



1 C DOCUM 
1 Category' 


ENT5 CONSIDERED TO BE RELEVANT - ■ r 

Olation of document, wi* indicalion, where approphale, of the relevant passages 


Relevant to claim No. 1 


A 


W0.A.94 G5318 (GENETICS INSTITUTE) 17 
March 1994 

see the whole document 


1-42. 
88-96 


A 


W0,A,94 025G2 (GENETICS INSTITUTE) 3 

February 1994 

see the whole document 


1-42, 
88-98 


A 


SrrrS. F.brlry 1993. m YORK 
US, 

pages 187-193. XP002G1Q947 . 
E R LAVALLIE ET AL.: "A thioredoxm gene 
fusion expression system that circumvents 
inclusion body formation in the E. con 
cytoplasma' 

see the whole document 


1-42. 
88-96 








-/-- 





I Further documents are lined in the contimution of box C. 

I * Spedal categories of dted documents : 

"A* document defining the general State of the art which is not 

considered to be of particular rdcvance 
'E' cariierdocuroenibmpuWiAcd on or afler the international 

filing date 

'L' document which may thnwdoubte on pfionQ^daing^ 
which is dted to establish the publication date of another 
dtation or other special reason (as spedfted) 

•O' document rcfcrnng to an oral disdosure, use. ciduhiticm or 
other means 

'P' document pubUdied prior to the international filing date but 
later than the priority date claimed 

i Date of the actual completion of the international search 

14 August 1996 



0 



Patent family members are listed in annex. 



T- Uter document published after the mtemijonal pliigfa'* . 
apriority date and not in conflict with the ap^^ 
dtS to underetand the principle or theory underiymg the 
invention 

•X* documemofpaiticdarrdevance;thedaiined in>^ 
cSSoTbe coSdered novd or cannot be COM 
involve an inventive step when the document IS taken alone 

•Y* document of particular rdcvance; the daimed invention 
cannot be considered to involve an inventive 
document is combined with one or m<« 5°Sii 
ments, sudi combination bdng obvious to a person stalled 
intheart 

document member of the same patent family 
Date of mailing of the international search report 

03. 0a96 



i Name and mailing address of the ISA 

European Patent OfTice, P.B. 5818 Patentiaan 2 
NL - 3380 HV Rijswiik 
Td. ( + 31-70) 340-2040. Tx. 31 651 ^ nl, 
Fax (t^ 31-70) 34O-3016 

Form PCT/ISA/21C (neond sheet) (July IW2) 



Authorized officer 



Masturzo, P 

page 1 of 2 



INTERNATIONAL SEARCH REPORT 



{nte ml Application No 

PCT/US 96/04811 



CXContimation) DOCUMENTS CONSIDERED TO BE RELEVANT 



Category • I Qtaliao of document, wift iniierton. whew »ppropri»«e, <rf die relevant passages 



BIOCHEMISTRY. 

vol. 34, no. 5, 7 February 1995. EASTON, 
PA US, 

pages 1787-1797, XP002O10948 
G WILLIAMS ET AL.: "Dissection of the 
extracellular human interferon gairnia 
receptor alpha-chain into two 
iiminoglobul in-like domains..." 
see the whole document 

JOURNAL OF BIOLOGICAL CHEMISTRY, 

vol. 270. no. 18, 5 May 1995, MD US, 

pages 10764-10770, XP002010949 

R A EDWARDS ET AL.: "Cloning and 

expression of a murine fascin homolog from 

mouse brain" 

see the whole document 

EP.A.0 499 541 (TRAMSGENE) 19 August 1992 
see the whole document 

W0,A,82 02841 (UNIVERSITY PATENTS) 2 

September 1982 

see the whole document 

ANALYTICAL BIOCHEMISTRY, 

vol. 173. no. 2, Septenfcer 1988, NEW YORK 
US, 

pages 440-444. XP002010950 

P G ZAWORSKI & G S GILL: "Precipitation 

and recovery of proteins from culture 

supernatants using zinc" 

see the whole document 



1-42 



1-42 



46 
46 

46 



Fam KT/ISA/310 (coMimntiiia of ttaai •!««) Mu* ••") 



page 2 of 2 



INTERNATIONAL SEARCH REPORT 



Into oal Apiilication No 

PCT/US 96/04811 





Patent document 
cited in search report 


Fubltcation 1 
date 1 


Patent family 
niember(s) 


Publication 
date 


1 


WO-A-9405318 


17-03-94 


lie A 

US-A- 
AU-B- 
EP-A- 
JP-T- 
US-A- 


5099993 
0671934 
8500838 
5437863 


24-10-95 

29- 03-94 
20-09-95 

30- 01-96 
01-08-95 


/ 


WO-A-9402502 


03-02-94 


US-A- 
AU-B- 


5292646 
4781493 


08-03-94 
14-02-94 




EP-A-499541 


19-08-92 


FR-A- 
CA-A- 

JP-A- 
US-A- 


2672895 
2061167 
4312600 
5276141 


21-08-92 
16-08-92 
04-11-92 
04-01-94 




WO-A-8202841 


02-09-82 


CA-A- 
EP-A- 


1210759 
0072843 


02-09-86 
02-03-83 



Foim PCT/lSA/aiO (jwteot funily twwi) <Ju|y IW2) 



INTERNATIONAL SEARCH REPORT 



niBtianal appfica»"on No. 

PCT/US 96/04811 



^«rrOb»crvatio«s where eert«n claims were found uase-rchdJt (CantiDu*tio« of it«» I of first O^) 
Thi. i»ur„«io«. report h« not been ..ubUshed,i„ respect of oeruin d.in.s under ArUd. PPX*) for *e fcUowing r«»,n. 

^ S^'^L^yr.lauiowbi^tmawr no. required u. be ^areted by Oris AuAorh^^ 



1. {Tj Claims Nos, 

^^?Sa"„rm^::g'f«mi;^.^o;^«^ ou, spedncany. 



ii^use they relate to part, of the m.erna.ion.1 appticaUon that do no|_c«nply wUh .he prescribed requirements ,0 such 



- see 



continuation-sheet PCT/ISV210 *) 



^ are depe„de«d«m.«Kl«. not drafted in accordance vdth*^ 

Box II Observationa where unity of mention is lacking (Continuation »f hem 2 of first sheet) 
This International Searching Authority found muWple invention, ta this intematiomd appHottion. as foUowr. 

- 6 Inventions 



- see 



continuation sheet PCT/ISA/210 - 



^""^ searchable daims. 



2. m /^.l..eard»bledaims could beseard^withoutelTortjustifying an «idiUon.l fee. tte 
^""^ of any additional fee. 

3. n AS only some of the required additional search f«s were ^ly paid ^^^.e applicant, this international search report 
^ covers only those claims for which fees were paid, specifically claims Nos.. 



4 n No required additional search fees were timely paid by ^U^^t Conse^^^^ international search report is 
^ Te^SS to Z invention first mentioned in the claims; it is covered by clamis Nos.. 



Remark on Protest 



I 1 The additional search fees were accompanied by the applicant's protest 

j — I No protest accompanied the payment of additional search fees. 



Form PCT/lSA/210 (continuation of first sheet (I)) (July 1992) 



internaOonai Application No. PCT/US96/ 0481 1 



FURTHER INFORMATION CONTINUED FROM PCTilSA/210 



») Due to the paucity of examples and the vague wording of the claims, the 
subject was not completely searched for economical reasons. 
The search was limited to processes of purification of thloredoxin- 
containing fusion proteins, and included all of the real examples pro 
vided. 



The Search Division has noticed the presence of multiple inventions in the present 
application, which have been subdivided as follows: 

1^ Claims 1-42 47-52- Purification of a thioredoxin-containing fusion protein by addition of 
alcohol and divalent cation to the solution containing it, after treatment with a chelator in order 
to disrupt cell membranes; 

2) Claims 43-45; A method to purify a thioredoxin-containing fusion protein by addition of zinc 
to the solution containing it; 

3) Claim 46; A method to purify a thioredoxin-containing fusion protein comprising adding zinc 
after the addition of alcohol and divalent protein; 

A\ Claims 53-59 64 71. 73. 78, 80, 85, 87; A method to purify a thioredoxin-containing fusion 
ploSy use Of L comprising two anion exchange 

and two cation exchange chromatographies; 

5) Claim 60-63. 65-66. 74-77, 79, 81 -84, 86; A method to purify a thioredoxin-containing - 
fusion protein by four chromatographic passages of wrhatever nature; 

6) Claims 88-96: A method to purify interleukin-1 1 starting from a thioredoxin-containing 
fusion protein by four chromatographic ion-exchange steps. 



This Page is Inserted by IFW Indexing and Scanning 
Operations and is not part of the Official Record 

BEST AVAILABLE IMAGES 

Defective images within this document are accurate representations of the original 
documents submitted by the applicant. 

Defects in the images include but are not limited to the items checked: 

□ BLACK BORDERS 

□ IMAGE CUT OFF AT TOP, BOTTOM OR SIDES 
□f FADED TEXT OR DRAWING 

1^ BLURRED OR ILLEGIBLE TEXT OR DRAWING 

□ SKEWED/SLANTED IMAGES 

□ COLOR OR BLACK AND WHITE PHOTOGRAPHS 

□ GRAY SCALE DOCUMENTS 

□ LINES OR MARKS ON ORIGINAL DOCUMENT 

□ REFERENCE(S) OR EXHIBIT(S) SUBMITTED ARE POOR QUALITY 

□ OTHER: ^ ■ 

IMAGES ARE BEST AVAILABLE COPY. 
As rescanning these documents will not correct the image 
problems checked, please do not report these problems to 
the IFW Image Problem Mailbox.