FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENTS RULES, 2003
COMPLETE SPECIFICATION
(See section 10, rule 13)
1. Title of the invention
"A Novel Process for Preparation of Biologically active Recombinant Mitogen"
2. Applicant(s)

Name
USV LIMITED

Nationality
Indian company incorporated under Companies Act, 1956

Address
B.S.D. Marg, Station Road, Govandi, Mumbai - 400 088 Maharashtra, India.

3. Preamble to the description


The following specification particularly describes the invention and the manner in which it is to be performed.

KFORM2

TECHNICAL FIELD
The present invention relates to an improved process for synthesis of biologically active rhPDGF-BB in prokaryotic cells. The present invention specifically relates to a process for obtaining biologically active recombinant PDGF in improved yields. Still further it relates to a simple and facile method of refolding-dimerizing rhPDGF-B to yield biologically active rhPDGF-BB which is efficient and inexpensive to perform and least time-intensive. Still further it relates to completely aqueous process for purification of rhPDGF-BB which is easily scalable and cost effective.
BACKGROUND AND PRIOR ART
Native Platelet Derived Growth Factor (PDGF), a potent mutagen is a dimeric molecule consisting of two polypeptide chains, one or more of which may be glycosylated. PDGF isolated from blood is predominantly a disulfide-linked homodimer or heterodimer of MW 28,000-31,000 composed of two chains, A and B (Heldin et al., 1979; Raines & Ross, 1982). The fully processed human B chain is encoded by the C-sis gene and consists of 112 amino acids. It has been found to have a high degree of homology with the p28sis protein product of the v-sis transforming gene of simian sarcoma virus (SSV) (Johnsson et al., (1984) Embo.3:921). Biologically active PDGF can exist as an AA or BB dimer, having a molecular weight of about 35,000 daltons or about 32,000 daltons, respectively. The human PDGF dimer is glycosylated and processed into a biologically active, three-dimensional conformation. This conformation is maintained by relatively weak noncovalent hydrogen bonds, hydrophobic and charge interactions, and strong covalent bonds between sulfur atoms. The PDGF dimer has 8 such disulfide linkages which may be both inter- and intrachain bonds. Reduction of either the AA or BB dimer into its
2

component monomenc chains destroys all biologic activity. The biological activity of PDGF is manifested only in the dimeric form, although there are some reports of the monomeric form exhibiting activity (U.S.Pat.No. 5,428,135 and references therein).
Several methods are known by which PDGF can be extracted from human platelets (Heldin et al., (1979) Proc. Natl. Acad. Sci. U.S.A. 76:3722-3726; Antoniades et al., (1979) Proc. Natl. Acad. Sci. U.S.A. 76:1809-1813). However in addition to being expensive to perform, these methods are generally inefficient, yielding only up to about 5% of the original starting material. Improved recoveries have been obtained by following the procedure of Antoniades (U.S.Pat.No. 4,479,896) and Lipton et al. (U.S.Pat.No.4,350,687), but yields are still limited by the availability of human platelets, the starting material. Furthermore, the therapeutic use of products derived from human blood carries the risk of transmission of a number of diseases such as Acquired Immune Deficiency Syndrome.
These problems have been circumvented by the expression of PDGF in genetically engineered eucaryotic cells such as yeast (U.S.Pat. No.4,769,328), Expression of PDGF homodimers and heterodimers is described in, for example, U.S.Pat. Nos. 4,766,073; 4,769328; 4,801,542; 4,845,075; 4,849,407; 5,045,633; 5,128,321; and 5,187,263. Eucaryotes have the ability to modify a protein posttranslationally in a variety of ways. In case of PDGF, such modification includes glycosylation, cleavage of leader sequences, folding, and formation of the intramolecular disulfide bonds requisite for biological activity. Although there has been some success in the production of recombinant PDGF, these results have been achieved primarily in eukaryotic cells, such as yeast and mammalian cells. Although reasonable quantities of product can apparently be produced in such systems, these cells, particularly mammalian cells, can be difficult and expensive to culture, and therefore , they are
3

not commercially preferred. It is generally preferable to be able to produce recombinant proteins in prokaryotic cells, as these are easily and cheaply produced in large quantities.
Prokaryotes have been used to express many protein products of DNA recombinant technology despite their inability to posttranslationally modify and process many eucaryoitc proteins. Methods are available in the art for expressing recombinant PDGF homodimers and heterodimers in bacteria (see, for example, Hoppe et al. (1990) Eur. J. Biochem 187:207-214; Fretto et al. (1993) J.Biol.Chem. 268:3625-3631, yeast (Kelly et al (1984) EMBO 7.4:3399-3405; Ostmanetal. (1989) Growth Factors 1:271-281), and mammalian cells (see, for example, Ostman et al. (1988) J.Biol.Chem. 263:16202-16208, EP177,957; U.S.Pat. No.5,219,759, U.S.Pat. No.6,017,731 and U.S.Pat. No.6,083,723. While production of PDGF with yeast and mammalian expression systems leads to very low levels of a biologically active protein, the utilization of bacterial expression systems results in substantially higher yields of the growth factor which is exclusively found as biologically non-active protein in inclusion bodies. Their growth rate, and hence their rate of protein expression are faster than those of eucaryotes, in addition to being easy to manipulate and inexpensive to culture.
Procedures have been developed for in vitro renaturation of biologically active eucaryotic proteins (WO86/05809). The success of the in vitro procedure has been limited severely by the complexity of the protein involved. However, to date, there has been little success in producing PDGF in prokaryotes, in particular E.coli. In fact, yield of PDGF in E.coli has been extremely difficult to detect and/or the resulting products are typically inactive. Previous attempts to express PDGF-B in E.coli have not led to biologically active products. Apparently folding of the polypeptide chain
4

and closing of disulfide bridges do not occur correctly in E.coli (Devare et al, Wang & Williams, 1984). Hoppe et al, 1989, expressed the sequence coding for the mature part of PDGF-B in E.coli as cro-p-gal-PDGF-B fusion protein which was exclusively found in the inclusion bodies. A monomeric PDGF-B fragment shortened by 12 amino acid residues from the amino terminus was excised from the fusion protein by CNBr cleavage. After protection of thiols by S-sulfonation, the fragment was purified by gel permeation chromatography and reversed phase HPLC. The monomeric protein was dimerized in the presence of a mixture of reduced and oxidized glutathione to yield biologically active rhPDGF-BB with an overall yield of- 0.7 mg of rhPDGF-BB/L of culture. Dalie et al, 1992, expressed hPDGF in E.coli from high-level cytoplasmic expression vector pTacBIq under transcriptional control of inducible E.coli Tac promoter and regulated by the lactose repressor (LacF). The production of hPDGF
was increased 50% in a Ion", htpR'strain (CAG629) and could account for approximately 1% of total cellular protein. From a 20L fermentation about 5 g of protein was extracted from the inclusion body preparation in the presence of 8M urea and 5 mM DTT to keep the protein soluble during purification. The inclusion body protein was isolated, solubilised and subjected to CM Sepharose column followed by refolding using reducing gradient of urea concentration and further purified by RP-HPLC to yield 16 mg of pure refolded protein accounting to 16% yield. Prior to refolding the sulfyhydryl groups were sulfonated. U.S.Pat. No. 5,334,532 claims a method for production of rhPDGF in prokaryotic cells, preferably E.coli encoding a fusion protein wherein the fusion partner was p-galactosidase with parts of cro repressor, with an expression amounting to 30% of the total proteins. The same disclosed that as native PDGF harbors only one methionine at position 12, CNBr cleavage resulted in excision of 12 amino acids from the N-terminal region which however did not affect the biological activity. Prior to dimerization the sulfyhydryl
5

groups were sulfonated. The yield after dimerization amounted to 0.5 to 0.7 mg of protein from a litre of culture. U.S. Pat.No.5,428,135 provides a method for refolding reduced rPDGF from a high expression host cell system, such as E.coli. The '135 patent discloses the process comprising solubilization of inclusion bodies in the presence of 8M urea at pH of 3.0, purifying the isolated denatured rPDGF using SE-Sepharose under reducing conditions, adjusting pH to 7.6 maintaining rPDGF in reduced state, subjecting the basic solution to gradient chromatography under reducing conditions. The free sulfhydryl groups of the reduced rPDGF were blocked using 0.1 M oxidized cysteine under reducing conditions preferably for about 18 hours at around 22°C followed by buffer exchanging and finally diluting to 0.5 mg/ml and 0.2 mg/ml protein concentration with non-denaturing aqueous buffer, preferably 20 mM Tris buffer of pH from 7.5 to 8.5. The disulphide exchange was then initiated using a reduced sulfhydryl compound, preferably 0.5 to 3.0 mM reduced cysteine at 30°C for about 16 hours.
Rinas et al, 1999, performed a procedure for renaturation of heterodimeric PDGF-AB from inclusion bodies of recombinant Escherichia coli using size-exclusion chromatography. Either prepurified or crude PDGF-AB inclusion bodies solubilized with guanidium hydrochloride were subjected to buffer exchange from denaturing to renaturing conditions during chromatography. Renaturation of PDGF-AB involves folding of the solubilized and unfolded molecules into dimerization competent monomers during size-exclusion chromatography and subsequent dimerization of folded monomers into the biologically active heterodimeric growth factor. Optimized conditions result in an overall yield of 75% active PDGF-AB with respect to size-exclusion chromatography and subsequent dimerization. The described approach allows renaturation at high protein concentration and circumvents aggregation which
6

is observed when refolding is carried out by dilution. U.S.Pat No. 5,759,835 claimed a method for producing a biosynthetic, biologically active, PDGF species by expressing rhPDGF and related muteins as a fusion protein expressed in E.coli by using IPTG as an inducer, cleaving the active protein from the fusion using cyanogen bromide, refolding the protein using oxido shuffling system comprising of oxidised and reduced glutathione. Karanmuri et al.,2007 reports a simple, novel method for expression and purification of rhPDGF-BB from Escherichia coli. The method reported to produce the dimeric protein in high yield (10-12 mg/g wet cell mass) and with a purity of >95%. rhPDGF-BB was exclusively found in inclusion bodies representing approximately 30% of the total cell proteins which were extracted and the monomer purified in single step by RP-HPLC using a wide pore preparative polystyrene hydrophobic matrix (Source-30 column: 30 µm polystyrene/divinylbenzene matrix; GE Healthcare). The chromatogram revealed a peak with three shoulders. The purified rhPDGF-B monomer was then refolded using Tris-EDTA buffer and subsequently dimerized to produce biologically active rhPDGF-BB. The complete removal of soluble proteins from inclusion bodies greatly increased the refolding efficiency by reducing aggregate formation. Similarly, Watson and Kenney (1992), showed that multiple peak formation could be abolished by running the column at 60°C and attributed the multiple peak formation to differences in hydrophobicity due to cis-trans isomerization of the Pro-Pro bond at positions 41 and 42.
In view of PDGF's clinical applicability in the treatment of injuries in which healing requires the proliferation of fibroblasts or smooth muscle cells and its value as an important component of a defined medium for the growth of mammalian cells in culture, the production of useful quantities of protein molecules similar to authentic PDGF which possess mitogenic activity is clearly invaluable. In addition, the ability
7

to produce relatively large amounts of PDGF would be a useful tool for elucidating the putative role of PDGF in a series of clinical applications as rhPDGF-BB currently has both research and therapeutic utility.
The present invention encompasses the prior art problems by obtaining a high yielding rhPDGF-B clone using a stable prokaryotic expression platform, achieving relative economics by avoiding use of expensive chromatography for in vitro refolding-dimenzation of the protein, minimizing polishing steps for attaining maximum purity, reducing and controlling of clipped moeities formation subject to endoprotease digestions and optimal use of endopeptidases for recovery of intact rhPDGF-BB from the fusion protein which are still unmet challenges.
In order for rhPDGF-BB to be produced on a commercially useful scale, improved methods of producing commercially useful quantities of biologically active, genetically engineered complex rhPDGF-BB are necessary. The present invention provides an improved method for overexpressing rhPDGF-B using prokaryotic platform by chimeric DNA constructs and also a simple and facile method for refolding-dimerization and purification of rhPDGF-B into biologically active rhPDGF-BB dimer.
OBJECT OF INVENTION:
Accordingly, it is an object of the present invention to provide an improved process for synthesis of biologically active rhPDGF-BB in prokaryotic cells. Another object is to provide a process for obtaining biologically active recombinant human PDGF-BB in improved yields. Yet another object is to provide a simple and facile method of
8

refolding-dimerizing rhPDGF-B to yield biologically active rhPDGF-BB which is efficient and cost effective to perform. Still yet another object is to provide a completely aqueous process for purification of rhPDGF-BB which is easily scalable and cost effective. A further object is to provide a recombinantly-produced polypeptide with substantially the same biological activity as native PDGF. Yet another object is to provide a formulation containing a recombinantly produced polypeptide having PDGF-like biological activity for therapeutic application to treat diabetic foot ulcers.
SUMMARY OF THE INVENTION
One embodiment of the present invention is an improved process for recombinant synthesis of human Platelet derived growth factor BB comprising of inducibly expressing rhPDGF-BB in a prokaryotic host cell wherein the expression is atleast 15% of total cell proteins and purifying the said protein avoiding organic solvents to obtain biologically active rhPDGF-BB of purity >97% wherein the yield is greater than 0.5g per litre of the fermentation broth.
Second embodiment of the present invention is a process for obtaining biologically
active rhPDGF BB expressed in a prokaryotic host cell with >97% purity comprising:
i. expressing said protein to atleast 15% of total cell proteins as inclusion
bodies; ii. solubilizing inclusion bodies in the presence of chaotrophs selected from the group consisting of urea and guanidium hydrochloride, preferably 6M-8M guanidium hydrochloride; iii. capturing the chimeric tagged protein using an expanded bed affinity chromatography;
9

iv. refolding and dimerizing the chimeric tagged protein of step (iii) in the
presence of arginine and optionally a sugar or a sugar alcohol wherein
the time of refolding is not more than 12 hours; v. cleaving the chimeric tagged protein by optimally using an
endoprotease wherein the clipping is not more than 8%; vi. purifying the crude protein of step (v) to >97% by ion exchange
followed by negative affinity chromatography further followed by salt
exchange process.
Third embodiment of the present invention is a chimera of rhPDGF-BB as shown in SEQ ID NO.; 1 and having an amino acid sequence as shown in SEQ ID NO.:2.
Fourth embodiment of the present invention is an affinity handle selected from the tags encoded by SEQ ID NO. 3 to SEQ ID NO. 9.
Fifth embodiment of the invention is a pharmaceutical composition of rhPDGF-BB as
claimed and atleast one pharmaceutically acceptable excipient.
BRIEF DESCRIPTION OF ACCOMPANYING DRAWINGS
The manner in which the objects and advantages of the invention may be obtained will appear more fully from the detailed description and accompanying drawings, which are as follows:
Figure 1: Flow chart for cloning of rhPDGF-B cDNA in pLMAB vector. Figure 2: rhPDGF Expression construct.
Figure 3: Feed rate profile showing increasing feed rate as a function of fermentation time.
10

Figure 4: Expression of rhPDGF-B; Lane 1 :Moleculaj weight Marker (Medium
range);Lane 2: 20 hr Sample; Lane 3: 19 hr Sample; Lane 4 :18 hr Sample; Lane 5 :17 hr Sample; Lane 6 -- 16 hr Sample
Figure 5 : Refolding gel picture of tagged rhPDGF-B to dimeric tagged rhPDGF-BB; Lane 1 - Marker; Lane 2: Monomeric Tag-PDGF B prior to refolding; Lane 3: Tagged PDGF-BB after refolding; Lane 4: Insoluble pellet fraction.
Figure 6: Non-reducing SDS-PAGE profile of the purification process of
rhPDGF- BB; Lane 1: Marker, Lane 2: Refolded sample; Lane 3: RPC purified tagged rhPDGF - BB; Lane 4: Post EK digestion sample, Lane 5: Post Ion exchange sample and Lane 6: Ni - IDA flow through.
Figure 7: SEC - HPLC chromatogram of clipped rhPDGF- BB.
Figure 8: SEC - HPLC chromatogram of dimer rhPDGF- BB.
Figure 9: PDGF reference standard showing the resolution of aggregate (multimer) rhPDGF-BB from the monomer.
Figure 10: SDS-PAGE analysis of non-specific clipping of rhPDGF-BB by
enterokinase. Lane 1: Marker, Lane 2: PDGF-BB(+p met) showing ~ 50% clipping, Lane 3: PDGF-BB(+p met) showing -30% clipping, Lane 4: PDGF-BB(+(3 met) showing ~ 10% clipping
Figure 11: rhPDGF-BB Peptide Mapping overlay.
Figure 12: rhPDGF-BB Cell Proliferation Bioassay.
DETAILED DESCRIPTION OF THE INVENTION
PDGF is believed to be biologically active only in climeric form. These biologically active PDGF dimers can take the form of a PDG£-AB heterodimer, a PDGF-BB homodimer, or a PDGF-AA homodimer (Hannink et al, Mol. Cell. Biol, 6,
11

1304-1314 (1986). Structurally, PDGF is a disulfide-bonded dimer, with the B chain slightly smaller than and not completely homologous to the A chain {Johnsson et a/.1982). Biological effects of the AA, BB, and AB dimers are similar, although the AA dimer appears less potent {Meyer-Ingold and Eichner 1995). The therapeutic product is the rhPDGF-BB form {Deuel and Huang 1984). There are eight cysteine residues in the primary structure of mature PDGF-B monomer {Claesson-Welsh and Heldin 1989). In the BB dimer, six of these residues are involved in intra-chain disulfide bonds {Oefner et al. 1992). The other two cysteines are in bonds joining the two chains but are not necessary for dimer formation or activity {Kenney et al. 1994).
The difficulty inherent in obtaining biologically active recombinant proteins from high expression host cell systems is, in the case of PDGF, further exacerbated by: 1) the large number of free cysteine residues in each PDGF monomer; 2) the dimeric form of the naturally occurring biologically active PDGF; and, 3) the extreme hydrophobicity of PDGF. As is expected with high expression systems, recombinant PDGF (rPDGF) produced in E.coli is predominantly found in inclusion bodies as the denatured monomer. Because of the limited solubility of this denatured rPDGF in aqueous solutions, it tends to aggregate before it can refold into its biologically active conformation, resulting in poor refolding efficiencies and, hence poor yields of biologically active rPDGF material. Much of the isolated material may be partially denatured and partially misfolded, due either to original misfolding in the bacterial host cell and/or to inclusion body isolation conditions.
Recombinant PDGF-BB produced in a yeast host cell is secreted as a fully folded, biologically active homodimeric protein consisting of two highly twisted antiparallel pairs of B chains {Oefner et al. (1992) EMBO J. 11" 3921-3926). Each B chain comprises 109 amino acid residues. During fermentation to produce a bulk drug
12

substance, several post-translational modifications occur to the secreted rhPDGF-BB. This results in a bulk drug substance comprising rhPDGF protein having considerable structural heterogeneity. Such modifications include, but are not limited to endoproteolytic digestion between residues Arg-32 and Thr-33 (referred to as "clipping") and exoproteolytic removal of C-terminal amino acids (referred to as "truncation") at Arg-32 and Thr-109, glycosylation of Ser and/or Thr residues, and oxidation of methionine. These post translational modifications lead to a number of structural forms, so called isoforms, of rhPDGF-BB present in the secreted product. The structural heterogeneity of the rhPDGF-BB yeast product is further complicated by the presence of a rigid Pro-Pro bond at residues 41 and 42. The rotation of this rigid bond is further hindered by its proximity to a disulfide bond and the bulky side chain of Trp-40. This rotational hindrance leads to the formation of stable cis-trans isomers of rhPDGF-BB at room temperature.
U.S. Pat. No. 5,428,135 disclosed the post-translational endoproteolytic processing or clipping between Arg-32 and Thr-33 which lead to three different and distinct basic isoforms of rhPDGF-BB: intact, single-clipped, and double-clipped. By "intact" was intended both B chains remain intact and were not endoproteolytically cleaved between Arg-32 and Thr-33. By "double-clipped" was intended both B chains were endoproteolytically cleaved between Arg-32 and Thr-33. Each respective isoform might be additionally modified by C-terminal truncation of Arg-32 or Thr-109, producing a known number of derived structural species for any given basic isoform. Thus, for the intact isoform, C-terminal truncation might occur only at residue 109, leading to three different structural species of that isoform. For the single-clipped isoform, C terminal truncation of both residues 32 and 109 could produce eight different structural species of that isoform. For the double-clipped isoform, ten
13

structural species of that isoform were possible. All together, the post translational processes of clipping followed by C-terminal truncation of Arg-32 or Thr-109 residues lead to a complex mixture of twenty-one possible structural isoforms of rhPDGF-BB in the bulk drug substance.
Besides clipping leading to multimer formation, formation of aggregates was another major problem during refolding of rhPDGF-B. The more the protein became diluted during the refolding step, the less prone it was to aggregation. Unfortunately, a high dilution to initiate refolding was not practical for industrial processes, because it resulted in a large volume of solution, which needed to be concentrated afterwards. Additional purification steps might be needed which were much more expensive for larger volumes. Therefore, the challenge was to develop protein refolding processes which worked at even high protein concentrations.
The prior art lists following strategies which have been shown to be efficient for suppression of aggregation {De Bernardez Clark et al, 1999):1) High dilution of denatured protein solutions, which as mentioned above was impractical as it required a concentration step of the highly dilute protein solution afterwards 2) Step addition of denatured protein {Rudolph and Fischer, 1990) 3) Immobilization of unfolded protein on a matrix and subsequent refolding of protein while bound to the matrix {Stempfer et al., 1996a, 1996b). 4) Use of specific antibodies to prevent accessibility of aggregation sites {Katzav-Gozansky et al,, 1996) 5) Aggregation mediated through hydrophobic side chains, which were exposed to solvent and formed patches. The patches interacted, leading to formation of aggregates. Mutating the target protein in a way that breaked up the hydrophobic patches was another strategy to suppress aggregation {Knappik and Pluckthun, 1995 and Nieba et al., 1997). 6) Application of hydrostatic pressure in the order of 1-3 kbar could be used to dissociate proteins
14

without disrupting protein structure (Schade et al.9 1980) Higher hydrostatic pressure could be used to unfold proteins and to prevent aggregation.7) Aggregation could be suppressed by using fusion constructs with hydrophilic proteins or peptides 8) Addition of molecular chaperones which interacted unspecifically with unfolded proteins, especially at hydrophobic sections of the unfolded protein and promoted formation of native protein (Hard, 1996). 9) Addition of low molecular additives to the refolding buffer (Rudolph and Lillie, 1996; De Bernardez Clark et al.t 1999).
Folding additives plays a crucial role in the development of cost-efficient processes, since it allows folding to take place at relatively high concentrations and are universally applicable and inexpensive in comparison to previously mentioned strategies.
Use of oxido shuffling systems is a more practical and more commonly used strategy for protein refolding. An oxido shuffling system promotes initial oxidation of free sulfhydryl groups as well as disulfide rearrangement for the formation of correct disulfide bonds. Oxido shuffling systems consisted of an oxidizing and a reducing sulfhydryl component. Mixtures of reduced (GSH) and oxidized (GSSG) glutathione are used mostly frequently, but the pairs cysteine/cystine, cyteamine/cystamine, 2-mercaptoethanol/2-hydroxyethyl disulfide are also used (De Bernardez Clark et ai, 1999). The molar ratio of reduced to oxidized thiol component usually is in the range of 1:1 to 10:1. Protein disulfides can also be formed after formation of mixed disulfides with glutathione or sulfonation of protein thiols (Rudolph et at, 1997). The factor limiting the use of glutathione is the high raw material cost involved.
L-arginine is reported to be an efficient folding additive when added during refolding of tissue plasminogen activator (tPA) to suppress auto-proteolysis. Surprisingly not
15

only auto-proteolysis was suppressed, but also the yield of native protein strongly increased. Subsequently, ability of arginine to increase refolding yields was tested with other proteins and found to be satisfactory (Buchrier and Rudolph,\99\; Brinkmann et ai, 1992; Lin and Trough, 1993).
L-Arginine is the most basic natural amino acid with a pi of about 10.8, and a derivative of the denaturing agent guanidine. L-Arginine monohydrochloride has been widely used as an additive in protein refolding. It was, e.g., effective in improving the yield of human plasminogen activator (Rudolph and Fischer 1990), recombinant Fab fragments (Buchner and Rudolph 1991), immunotoxins (Brinkmann et al. 1992), functional single-chain antibody fragments (Tsumoto et al 1998), Interleukin-21 (Asano et al. 2002), human matrix metalloproteinase-7 (Oneda and Inouye 1999), and recombinant human neurotrophins (Suenaga et ai 1998; Rattenhall et a/,2001). It was found to be the most effective amino acid in suppressing the aggregation of lysozyme after heat-induced denaturation (Shiraki et a/.2002). Arakawa and Tsumoto (2003) found that it had no significant effect on the thermal stability of RNAse A and hen egg white lysozyme, but improved the reversibility of the respective thermal transitions. The renaturation yields of denatured lysozyme was significantly improved by the addition of increasing concentrations of L-ArgHCl. There was a linear increase in the recovery of activity up to 87% in the presence of 0.9 M L-ArgHCl; while in its absence, on average, the renaturation yield was only 23% (Ravi Charan Reddy K et al, 2005). At low concentrations of L-ArgHCl, a considerable variability in final yield was observed ranging from 10% to 35% in the absence of L-ArgHCl. Increasing concentrations of L-ArgHCl increase the surface tension of its aqueous solutions.
The efficiency of the refolding process is determined by the competition between productive refolding and unproductive side reactions, i.e., the formation of misfolded
16

species and the aggregation of denatured protein (Goldberg et ah 1991; Kieftiaber et al. 1991). Refolding additives may act on either one or all of the reactions involved, i.e., they may facilitate the refolding of a protein in question by stabilizing its native state or accelerating the kinetics of the "correct" folding reaction, as well as by suppressing unspecific aggregation of the unfolded polypeptide and/or intermediates on the folding pathway. It has been suggested that amino acids (arginine, proline, lysine) (Rudolph and Fischer 1990; Samuel et ah, 2000), polyamines (putrescine, spermidine, and spermine) (Kudou et al. 2003), and mild detergents (Tandon and Horowitz 1988; Wetlaufer andXie 1995; Krause et al. 2002), but also the denaturing agents GuHCl and urea themselves (Orsini and Goldberg 1978), act as suppressors of aggregation, while substances like sugars, polyalcohols, and ammonium sulfate improve refolding yields by stabilizing the native conformation of proteins (Sawano et ah 1992; Michaelis et al. 1995).
The present invention shows efficient refolding with respect to refolding yields by combining a sugar with an additive arginine. Unexpectedly optimum usage of sucrose and arginine in refolding buffer substantially reduced the percentage of aggregate formation with prominent increase in percentage purity though hydrophobicity of rhPDGF posed major hindrance. A simple and facile process for purifying the rhPDGF-B using ion exchange process followed by affinity chromatography yielded rhPDGF-BB with >97% purity.
The invention encompasses an improved process for recombinant production of human platelet derived growth factor-BB by isolating a cDNA fragment by following cDNA cloning methods as described by ("DNA Cloning 1, a practical approach,2nd Edition, by Glover and Hames). The process involved isolation of total RNA using guanidium thiocyanate method by standard procedure (Ausuble et a/.,1987) as well as
17

described by Promega Total RNA Isolation Systems' protocol (Cat.# Z51110). This RNA was then used for generating lst-strand by using a standardized method with the help of AMV transcriptase. Non-specific interactions between strands was reduced using Rnase H and Rnase A, followed by purification of the lst-strand DNA (20ng/ ml) using phenol/chloroform/isoamyl alcohol and 70% ethanol wash. rhPDGF-B specific PCR amplification was done using a suitable polymerase to generate double-stranded cDNA prior to cloning the gene of interest into E.coli. This fragment was then directly cloned into an inducible expression vector. The flowchart of cloning method is given in Figure. 1. The invention is further described in the following examples, which are intended to illustrate the invention. Use of an indigenously designed expression vector for inducible expression of rhPDGF in prokaryotic host under the control of strong araB promoter is an inherent novelty of the present invention. Expression yield in grams quantity is another novel feature of the invention. rhPDGF-BB exclusively found in the inclusion bodies represented approximately 15% to 25% of the total cell protein when inducibly expressed in a series of E.coli K12 strains selected from a group consisting of TOP10, BL21, BL21 (DE3), BL21( RIPL), Origami, Rossettagami, BL21-DE3(pLysS), BL21 (trxB ).
The culture from glycerol stock was streaked on a 2XYT plate. The culture is an E. coli strain containing plasmid with DNA fragment encoding for rhPDGF. The plate was incubated at 37°C for 16-24 hrs. Single colony of the culture form 2XYT plate is inoculated in 10ml of 2XYT liquid medium and incubated at 37° C and 200 to 220 rpm, on rotary shaker for 16 hrs. The grown culture was transferred to 100 mL of basal fermentation medium (seed medium) in a 500 ml conical flask and incubated at 37° C and 200 to 220 rpm, on a rotary shaker for 8 hrs. Hundred ml of seed culture was transferred to 900 ml of fermentation medium in 21it. jar fermentor procured from
18

B Braun. The fermentation parameters are maintained as follows: aeration from 4 to 10 1pm, temperature at 37°C, stirrer speed from 300 to 1200 rpm. pH is maintained between 6.9 to 8.00. pO2 was cascaded with stirrer so as to maintain dissolved oxygen above 50% in the initial 16 hrs. The feed solution contained 15-25% glucose and 10-20% yeast extract. Fifteen to thirty ml of TES was added for 600 ml of feed. The feed was according to a predetermine feed rate which is a combination of exponential and linearly increasing feed rate (Figure 3). Nutrient solution was fed as per predefined strategy. Inducer solution was added between 2 hrs to 20 hours of growth. Antifoam solution was fed as and when excess foaming is observed. At various time-points during the fermentation run, aliquots were withdrawn from the fermentor and the OD of the sample was determined by spectrophotometry and % expression of PDGF. The final OD was -80 equivalent to -45 g dry cell weight. The fusion protein production was 15 -30 % of total cellular protein as determined by SDS-PAGE.
After the cell are fermented and the fusion protein induced to a level that it is present maximally as inclusion bodies, the culture broth is harvested from the fermentor. This broth is then centrifuged at 10-20,000 rpm for 15-30 min to pellet the cells. The cell pellets are then resuspended in lysis buffer and homogenised using a motor driven homogeniser. This solution is then directly subjected to multiple cycles of high pressure cell disruption at 850-900 bar using Panda2K from Niro Soavi. The crude cell lysate is then centrifuged at 10-20,000 rpm for 20 min to pellet the inclusion bodies. The inclusion body proteins are then resuspended in lysis buffer and centrifuged as above. The pellet obtained is the resuspended in a buffer solution. The pellet is resuspended using a homogeniser and then subjected to centrifugation at 10-20K rpm for 20 min at 8-15°C. The pellet thus obtained is subjected to a further 3
19

cycles of washing with the same buffer. The pellet obtained after final wash is resuspended in lysis buffer and homogenised thoroughly. This homogenate is centrifuged at 12K rpm for 20 min to obtain highly purified tagged rhPDGF-B. This monomeric protein is then suspended in 6-8 M GuCl or 4-8 M urea buffer containing 20-100 mM Tris, pH 8.0 and allowed to dissolve for 12-18 hrs at ambient temperature. Thereafter, insoluble particles are removed by centrifugation at 10-20,000 rpm for 30-60 min. The clear supernatant containing monomeric tagged PDGF-B is then diluted with the same buffer to adjust protein concentration and the disulfide linkages are reduced using appropriate reducing agents. The reduction is carried out at 25-37°C for 2-6 hrs with gentle stirring. The reduced monomeric rhPDGF-B solution is then added to the refolding buffer for formation of correct disulfide linkages during renaturation of monomeric tagged rhPDGF-B. The refolding method of the present invention makes use of pH conditions to carry out the refolding and dimerisation of tagged rhPDGF-B to form tagged-rhPDGF-BB (Figure.5) More than 20 refolding buffers were tried prior to obtaining dimeric rhPDGF-BB. Maximal recovery of the tagged rhPDGF-BB is seen after several days, preferably 2-10 days of refolding under gentle stirring at ambient temperatures. The refolding mixture is then directly pumped into a column containing reverse phase media, preferably polymeric matrix of styrene-divinyl benzene type having 15-30 micron bead size. The column is preequilibrated with 10-40 mM sodium phosphate at an appropriate pH. After sample loading is complete, the column was washed with several column volumes of equilibration buffer and the bound proteins eluted with a gradient of equilibration buffer and a solution containing 50-90% organic solvent in water. Non reducing SDS-PAGE analysis of the RPC fractions are analyzed. Pure tagged rhPDGF-BB fraction is collected and the concentration of organic solvent brought down by dilution with water. Protein is estimated by dye-binding method of Bradford and to this solution, a
20

non-ionic chaotropic agent is added to give a final concentration of 2-4 M and the pH adjusted to between 4,5-8.0. Restriction protease enzyme was added at a concentration of 1:10-1:100 (units/µg protein) to remove the tag and produce a dimeric rhPDGF-BB with intact N-terminus. The enzymatic reaction was carried out at ambient temperature under gentle stirring for 12-24 hrs and subsequently the reaction mixture loaded onto a cation-exchange matrix, preferably cross-linked agarose beads containing sulfopropyl or sulfonyl ligands. After washing out the unbound molecules, bound protein is eluted with a gradient of high salt concentration, preferably sodium chloride or acetate. The peak fraction containing mixture of di-tagged rhPDGF-BB, mono-tagged-rhPDGF-BB and rhPDBGF-BB was then passed through a small column of metal ion bound affinity matrix. This resulted in the di-tagged rhPDGF-BB and mono-tagged-rhPDGF-B binding to the matrix with rhPDGF-BB flowing through the column unbound. This unbound fraction containing highly pure rhPDGF-BB was collected and further concentrated and buffer-exchanged to a buffer of choice using a 5-10 kDa cut off membrane. Alternatively the purified rhPDGF-BB is bound to reverse phase matrix and washed extensively with buffer of choice to remove all salt. The bound protein is then eluted with a mixture of buffer and solution having high concentration of organic solvent. The organic solvent is removed under vacuum and the concentrated protein solution having highly pure rhPDGF-BB is recovered. The Ni-IDA flow through fraction was buffer exchanged to produce the required API. Figure.6 describe optimization of the enzymatic digestion step and analysis thereof.
In another embodiment of present invention the cell pellets obtained from the fermentor was washed and subjected to lysis in lysis buffer followed by pressure disruption. A chaotrope was added for solubilization, followed by Ni-IDA
21

chromatography. The eluted material was subjected to reductive refolding at a fixed concentration (50 -500 mg) /Lit. for 1-7 days in a suitable refolding buffer and the refolded material purified over a Source 30 RPC column followed by enzyme digestion. A process has also been developed wherein the source 30 RPC step is deleted. The polishing steps involved a SP sepharose chromatography and TFF to get rhPDGF-BB of high purity. Figure. 11 show peptide map and Figure 7 & 8 show SEC —HPLC profile of clipped and dimeric rhPDGF-BB API with that of the reference standard preparation(Figure. 9). The present invention can be illustrated by the following, non-limiting examples.
EXAMPLES
Materials Used:
Restriction enzymes and T4 DNA ligase were purchased from MBI Fermentas International; Thermus aquaticus (Taq) DNA polymerase was obtained from MBI Fermentas International; PDGF standards were obtained from Sigma and R& D systems. All enzymes were used in accordance with the manufacturer's instructions.
Plasmid vectors pGEM-T-3.0 kB, pBADHIS-A-4.1 kB from Promega and Invitrogen respectively were used as cloning vectors. Vector pRA-4.1 kB was indigenously designed at USV and also used as a cloning vector. Plasmid vector pET19B from Novagen was used as a source for IPTG & lactose inducible T7 promoter, a translational initiation site, a strong translation termination region with Ampr gene or Kanrgene. E.coli K-12 strains TOP10, BL21, BL21 (DE3), BL21 (RIPL), Origami, Rossettagami, BL21-DE3 (pLysS), BL21 (trxB) were obtained from Invitrogen and Novagen.
22

Oligonucleotide Synthesis:
Oligonucleotides used for PCR were synthesized by standard methods using an Applied Biosystems DNA synthesizer and purified by standard methods.
Example 1 :
Total RNA Isolation from human placental mRNA:
Total RNA was isolated by guanidium thiocyanate method using Total RNA Isolation Systems" (Promega, Cat # Z5110). Poly (A+)RNA from human cells and tissues was purified by affinity chromatorgraphy on Oligo (dT) cellulose. Total RNA pellet was resuspended in 1ml nuclease-free water and stored at -70° C.
Example 2:
Synthesis and cloning of cDNA: Reverse transcription and l^-strand synthesis:
Total RNA was used to prime for reverse transcription to generate 1 "-strand cDNA following standard RT-PCR technique as follows: 1.0 mg RNA/ mRNA, 0.5 ug Primer(Oligo dT) were taken in nuclease free water and heated at 70°C for 5 minutes, then chilled on ice for 5 minutes. The sample was microfuged for 5 seconds. Following reagents were added serially: Nuclease free water, RT buffer (5X), dNTP mixture (12.5 mM), ribonuclease inhibitor (1:2 dilution of 40 U/µl) and sodium pyrophosphate (40 mM, 4mM final concentration).
The above contents were mixed gently and prewarmed at 42°C for 2 minutes, followed by addition of AMV Reverse Transcriptase (RT) (l0U/µl), mixed gently, and the mixture left at 42°C/ 1.0 hour (RT reaction) subsequent to which the reaction
23

was heated to 75°C for 10 minutes (inactivation of RT). 1.0 Unit of RNAse A / tube was added and incubated at 25°C for 30 minutes followed by addition of 0.5 Units of RNAse H / to each reaction tube and incubated for 30 minutes at 37°C. To clean up the Is' strand cDNA, 85 ml of 10 mM Tris-C1/1.0 mM EDTA( pH 7.5) was added. Further 100 ml of buffered phenol (pH 8.0)-Chloraform-Isoamyl alcohol(25:24:l) was added, vortexed and spun at -14,000 rpm for 5 minutes in the micro centrifuge. The aqueous phase was transferred to a fresh tube and 3M Sodium Acetate (pH 5.5) was added and mixed. 100% chilled (-20°C) ethyl alcohol was added and left at -70°C for 30 minutes. The sample was spun -14,000 rpm for 5 minutes at 4°C and the supernatant discarded. The pellet was washed with 95% ethyl alcohol (precooled at 4°C) by spinning at 14,000 rpm for 5 minutes at 4°C. The sample was resuspended in 50 (il nuclease free water. The yield of 1st strand cDNA obtained by said method was about 20 ng/ml.
Example 3:
PCR Amplification:
For raising the ORF encoding the human mature PDOF, gene-specific primers were used. The primers were designed using a DNA sequence (Collins et al, Nature, 316: 748-750, 1985) using a primer design software, Lasergene are given as Seq. I.D. No.:10toSeq. ID No.: 16.
Steps:
100 ng of ls'-strand cDNA was used for each 50µ1 of PCR. Standard PCR amplification was done using Tag polymerase/Pfu Polymerase/Superscript Elongase
24

to make double-stranded cDNA prior to cloning the gene of interest into an appropriate vector of choice. Pfu polymerase was used in presence of MgSO4.
The PCR was done using two sets of primers Set A :(Seq I.D. Nos.:10 and 11) which resulted in a fragment of length 368 bp encompassing the PDGF-B mature protein coding DNA and HIS-EK domain and Set B: (Seq I.D. Nos.:10& 12) which also resulted in a fragment of length 369 bp encompassing the PDGF-B mature protein coding DNA and HIS-EK domain,
PCR conditions were set such that the annealing temperature were minimum 3-4 degrees below the actual Tm of the primers. 25-100 pmoles of primers were used for each 50 ul of PCR reaction. PCR condition was divided into 4 stages for total of 40 cycles. In the stage I all the ingredients e.g. buffers, primers, 1st-strand cDNA, 2mM dNTP, enzyme were mixed and heated at 94°C for 2 minutes in the thermal cycler for denaturation. The annealing was done at 54-58°C for 2:05 minutes; extension at 72°C for 2 min (4 cycles); then at stage II denaturation at 94°C for 1:05 min, annealing at 58-60°C for 2:05 min, extension at 72°C for 2 minutes (35 cycles). Lastly final extension(stage 3) was done at 72°C for 10 minutes. At the end of the incubation,the PCR mixture was subjected to electrophoresis in a 0.8% -1.5 % agarose gel as required. 5-10 ul of PCR mixture was loaded in gel from each 50 ul total reaction mixture.
Example 4:
Verification of the putative clone by restriction enzyme analysis:
The PCR fragments were double digested with PDGF ORF specific restriction enzymes like Eco47III, BstXl, Nspl, Bglll,and BseMI-Maell, Nspl-Maell, and
25

BseMI-NspI. The 368 bp and 369 bp fragments obtained were of right size as adjudged by gel electrophoresis by running corresponding markers. The 368 bp and 369 bp fragments respectively were cloned in the cloning vector pGEM-T using T-A cloning procedure and further subsequently in a prokaryotic expression vector pBADHis A as Ncol-Xhol insert (Figure 1). The ligated plasmids were used for transforming E.coli K-12 strains JM109, ToplO, BL21, BL21 (DE3), BL21( RIPL), Origami, Rossettagami, BL21-DE3(pLysS) and BL21 (trxB). The transformed competent cells were further used for subcloning, restriction analysis, sequencing and expression checks. Positive clones were screened on the basis of restriction analysis and % expression which was further confirmed using Sequencer for full cDNA sequence analysis.
Example 5:
Construction of Expression vector:
i) Construction of Expression vector rhPDGF-B/pRA (4.383 Kb):
PDGF cDNA of 368 bp was cloned as T-A in the cloning vector pGEM-T (Promega). Also a 369 bp PDGF cDNA insert was cloned in directional orientation at Sall/Xhol sites in an expression vector pBAD His A( Invitrogen)
A 449 base pair fragment was amplified from the vector pRA(in house expression vector ) using the primers as given in Seq. I.D. No.:13 and Seq. I.D. No.:14. This 449 bp fragment had a Bam HI site introduced in the 3rd mature amino acid (GGT gly —► GGA gly). From this fragment, a 137 bp BamHI-BamHI piece was removed and purified that had the EK site fused with the first amino acid (Ser) of the mature human PDGF-B.
26

Then a 357 base pair fragment was amplified from the PDGF- B-MAT clone in pBAD His A using the primers as given in given in Seq. I.D. No.15 and Seq. I.D. No.: 16.The fragment too has a Bam HI site introduced at the 3rd mature AA (GGT gly GGA gly).This fragment had two stop codons TAA.TGA'. The cutting and cloning sites were Bam HI and Xho I. The 357 bp PCR fragment was digested with Bam HI and Xho /and a 336 bp fragment was purified.
The 336 bp fragment was Hgated with 137 bp fragment to form a 473 bp fragment encoding the cDNA of PDGF-B and EK cut site fused with the first amino acid (Ser) of the mature RhPDGF-B preceeding the cDNA of PDGF-B.
The vector pRA (in-house 4059 bp) was digested with BamHI -Xhol and the large vector fragnent of 39 lObp was purified. This digestion removed a 149 bp BamHI-Xhol fragment from the vector. The 473 bp purified fragment was ligated with the BamHI-XhoI 3910 bp vector piece to generate the PDGF-B fusion clone(Figure 1 ).
ii) Construction of Expression vector rhPDGF-B pRA.Nco.Z (4.512 Kb):
To get better expression the PDGF-B insert in the pRA was modified by making a lacZ-Chimera. The codons for first 41 amino acids were appended in front of the HIS -EK domain preceding the PDGF-B ORE This was done by ligating a Ncol-Sal PCR fragment as ATG-lacZ-HIS with a Sall-Xhol fragment as EK-PDGF-STOP to result in the PDGF expression vector containing the ORF: ATG-lacZ-HIS- EK-PDGF-STOP. This vector was used to transform E. Coli Top 10 cells to get the LacZ PDGF clone This construct gave ~ 52 % expression in the fermentor amounting to - 15-20 gms of lacZ-PDGF chimera per liter.
iii) Construction of Expression Vector rhPDGF-B/pEB (4496bp):
27

The pEB( in house) vector included a synthetic araBAD promoter whose expression level was controlled by an inducible system. The expression level was otherwise suppressed by the repressor control system which was the internal/external glucose level of the cell itself rhPDGF-B in pEB with a synthetic araBAD promoter was used for recombinant expression.
For cloning into vector pEB, the vector was digested with Sall-Hindlll restriction enzymes. A 398 bp fragment encompassing the DNA sequence as EK-PDGF-B MATURE-STOP-was cut with Sal I-HindIII from the sequenced PDGF-B clone in pRA.lacZ and ligated with pEB vector digested with Sail-Hind III. The circularized vector thus, had the coding ORF as ATG-AFFINITY TAG-EK-PDGF-B-STOP(Fig 2). The circularized vector was then used for transforming E.coli TOP10, BL21, BL21 (DE3), BL21( RIPL), Origami, Rossettagami, BL21-DE3(pLysS), BL21 (trxB ) cells.
Example 6:
Expression check at 1L flask scale:
E.coli transformed cells with rhPDGF-B/pEB (4496 bp) maintained in glycerol stocks were grown in 2X YT medium plate containing 50 ug/ml of sodium salt of ampicillin or kanamycin and 16 hour culture was used to inoculate 500 ml of 2X YT liquid media in 1L flask at 37°C to a density of 0.5 O.D. at 600 nm, at which time 1% arabinose was added and incubation was continued at 37°C for 6 hours.
Example 7:
Expression at 2L Fermentor with different inducers:
Ecoli TOP 10 and BL21 cells transformed with rhPDGF-B- pEB (4496 bp) expressing recombinant human Platelet Derived Growth Factor (rhPDGF) were
28

purified and maintained in glycerol stocks. An aliquot of the culture was removed from the stock and streaked on 2XYT plate (containing 50µg/ml of sodium salt of ampicillin) to separate single colonies, after growing the cells for 24 hours at 37°C. A single colony from the plate was used to inoculate 10ml of 2XYT liquid medium containing ampicillin or kanamycin for 8-16 hours at 37°C on a rotary shaker (200-220 rpm), subsequently 10 ml of the propogated culture was seeded to 100 ml of the basal medium contained in a 500 ml flask for 8-16 hours at 37°C on a rotary shaker (200-220 rpm). The seed medium was used to inoculate the fermentor (2L, B Braun Biotech International) each containing 600 ml of the basal medium.
The partial pressure of oxygen was maintained at 50% air pressure throughout the fermentation by automatic adjustment of the stirrer speed. The feed medium (glucose, 20% w/v; yeast extract, 20% w/v; KH2P04, 0.75% w/v; K2HP04, 0.75% w/v; Na2HPO4, 1% w/v; and TES (Trace element solution), 2.5%v/v) was pumped in to the fermentor at the pre determined feed rate (Figure 3). The fermentation media was induced after 4-16 hours from the start of fermentation with an inducer, selected from the group consisting of 5% to 40% lactose, 0.5% to 10% arabinose, 0.1% to 2.0% IPTG, when the cell density reached an OD in the range of 15 to 55 at 600 nm. Excessive foaming was controlled by the addition of antifoam solution till the foam subsided (Dow Corning 1510, Antifoam). Fermentation was performed for 8-20 hours, during which samples were withdrawn for measurement of OD and accumulation of rhPDGF within the cells. rhPDGF accumulation was measured by scanning Coomassie stained SDS-PAGE gels of whole cell lysate using standard methods.
Preferably the fermentation medium seeded with transformed E. coli TOP 10 cells were grown to a density of 50 to 55 at 600 nm and was induced with 1% arabinose
29

(40% w/v arabinose solution) after 16 hours of fermentation. The fermentation was carried over a period of 20 hours. The % rhPDGF-B expression pattern over a period of 20 hours is presented in Table 1. A SDS-PAGE chromatogram depicting the rhPDGF-B expression is shown in Figure 4.
Table 1: % rhPDGF-B expression pattern with 1% Arabinose (w/v) as inducer

Time (hrs) OD 600nm % rhPDGF-B Expression
0 0.28 -
16 58 2.56
17 62 6.18
18 68.5 9.08
19 73.5 14.61
20 70.5 15.09
The fermentation medium seeded with transformed E.coH BL 21 cells were grown to a density of 10 to 15 at 600 nm and was induced with 40% lactose after 4 hours of fermentation. The fermentation was carried over a period of 8 hours. The % rhPDGF-B expression pattern over a period of 8 hours is presented in Table 2.
30

Table 2: % rhPDGF-B expression pattern with 40% Lactose(w/v) as inducer

Time (hrs) OD 600nm % rhPDGF-B Expression
0 0.264 0
4 11.6 0
5 13.6 5.1
6 17.2 7.45
7 20.2 11.9
8 22.7 18.2
The fermentation medium seeded with transformed E.coli BL 21 cells were grown to a density of 10 to 15 at 600 nm and was induced with 1% IPTG (20% w/v IPTG solution) after 4 hours of fermentation. The fermentation was carried over a period of 8 hours. The % rhPDGF-B expression pattern over a period of 8 hours is presented in Table 3.
Table 3:% rhPDGF-B expression pattern with 1% IPTG (w/v) as inducer

Time (hrs) OD 600nm % rhPDGF-B Expression
0 0.28 0
4 12.3 0
5 14 6.5
6 17.8 13.1
7 19.9 17.5
8 21 23
31

Example 8:
Purification of Recombinant Human rhPDGF-BB:
E.coli TOP 10 cells transformed with rhPDGF-B pEB were grown in a fermentor in fed batch mode to an OD of 80 at 600 nm at 37°C, the feed was induced using 1% arabinose (40% w/v of arabinose solution) after 4 hours. Approximately 1.4 liters of fermentation broth was harvested at the end of fermentation cycle and was cooled to 10°C using chiller water. After achieving temperature harvest broth was concentrated and buffer exchanged using 50 mM Tris and 150 mM NaCl (TFF-I buffer) with 0.2p, tangential flow filtration cassette. Buffer exchanged cells were then homogenized in pressure homogenizer at 800 psi in three pass. Finally homogenized cells were suspended in TFF-I buffer and volume adjusted to 1.4 litre.
1.57 Kg of guanidine HC1 was added in 1.37 liter of homogenized cells and was kept at 37°C with stirring in incubator for 8 hours to achieve complete solubilization. 800 ml of streamline chelating matrix packed in streamline 50 column was used to perform expanded bed chromatography. 3 column volumes of water was used to wash the matrix. Nickel was loaded on matrix using 3 column volumes of 0.1M nickel sulphate wash followed by washing with 3 column volumes of water. Column was equilibrated with 50 mM Tris, 6M GuHCI and 30 mM imidazole. To the solubilized broth 37 ml of 2M Imidazole was added to achieve 30 mM concentration. Solubilized broth was then loaded in expanded mode and 2 column volumes in packed mode. Bound protein was buffer exchanged by 2 column volumes of 50 mM Tris, 8M urea and 30 mM imidazole. The bound protein was eluted by 50 mM Tris, 8M urea and 250 mM imidazole through an 50 ml inline column of streamline chelating. Total
32

protein content was determined using Bradford method. Total protein content was 4.26 g.
47.5 ml of 0.2 M EDTA and 307 mg of dithiothreitol were added to 950 ml of expanded bed column eluted protein and kept under stirring for 30 minutes. The refolding buffer contained Tris(10-100 mM) at an appropriate pH; Arginine (0 -500 mM); Sucrose(0-250 mM); EDTA(0-5 mM); L -Cystine( 0-100 mg/Liter); specifically at a protein concentration of 0.1-3 mg /Liter. Reduced sample was then transferred into 3 liter of refolding buffer having composition as 50 mM Tris, 500 mM sucrose, 2 mM EDTA, 250 mM arginine base and 100 mg/L of cystine. Protein was allowed to refold for 12 hours at room temperature with slow stirring(Figure 5) and its progress was monitored by gel electrophoresis under non-reducing conditions(Figure 6).
pH of refolded sample was adjusted to 6.5 and enterokinase was added in the ratio of 13.2 units/mg of protein and kept under stirring for 16 hours at room temperature.
500 ml of SP Sepharose FF packed in XK 50/40 column was used for ion exchange chromatography. Matrix was equilibrated with 3 column volumes of lOmM sodium phosphate, pH 6.5 at 40 ml/minute. Enterokinase digested material was loaded at 40 ml/minute. After loading, column was washed with 3 column volumes of equilibration buffer. Bound protein was eluted by using linear gradient of 100% Buffer (lOmM sodium phosphate and IM NaCl) in 10 column volumes. Product having conductivity of 45 mS/cm was eluted and pooled and the protein estimated by Bradford method. The yield of rhPDGF-BB dimer was 805 mg/ liter of fermentation broth.
50 ml of chelating sepharose was packed in XK26/20 column. Matrix was loaded with 3 column volumes of 0.1 M nickel sulphate followed by washing with 3 column
33

volumes of water. Column was equilibrated with 3 column volumes of 10 mM sodium phosphate (pH 6.5) and 0.2 M sodium chloride. IEC eluted peak was loaded on matrix to collect product flow-through. Column was washed with 2 column volumes of equilibration buffer. The yield of rhPDGF-BB dimer was 673 mg/ liter of fermentation broth.
Affinity interaction chromatography flow through protein was concentrated to 2 mg/ml and diafiltered against 50 mM sodium acetate, pH 5.5 in 0.1m2 TFF cassette. The yield of rhPDGF-BB dimer was 617 mg/ liter of fermentation broth. SDS-PAGE of the purification process is as in Figure.6
TFF protein was passed through 0.2u sterilized filter and collected in 500 ml PETG (glycol modified polyethylene terephthalate) bottle and stored at 5°C. The yield of the pure rhPDGF-BB dimer was 590 mg/ liter of fermentation broth with a purity of 97.78%. Table 4 shows consistency in the % of monomer content kept under control < 3%.
Table 4: Percentage of PDGF BB Dimer and Mononer

Name B.N0.529 B.NO.530 B.N0.531 B.N0.532 B.N0.535
PDGF BB Dimer 99.31% 98.24% 98.42% 98.05% 98.13%
Monomer 0.69% 1.76% 1.58% 1.95% 1.87%
In this purification process, enterokinase digestion to remove the fusion tag resulted in clipped forms of PDGF-BB which amounted to nearly 50% of EK digested PDGF-BB. Efforts made towards reducing the clipped fragment formation resulted surprisingly in controlling the clipping to minimum 30%, more preferably to 10%
34

L

with simple optimization of enterokinase to protein ratio and reaction temperature against use of various chromatographic procedures of the prior art.
Reaction 1: EK Concentration at IIU per 50 ug protein. EK Digestion at RT for 0-tol6Hrs.
Reaction 2: EK Concentration at IIU per 100 ug protein. EK Digestion at RT for 0-to 16 Hrs.
Reaction 3: EK Concentration at IIU per 50 ug protein. EK Digestion at 4°C for 0-tol6 Hrs. Reaction 4: EK Concentration at IIU per 100 ug protein. EK Digestion at 4°C for 0-tol6 Hrs.
After EK digestion, further purification was carried out and the final pure protein was analysed by SDS-PAGE. 5µg pure PDGF BB from each reaction was analysed by reducing Gel (Figure 10).
However the formation of clipped fragments was not affected by the protein concentration which was evident from the following studies. Protein concentration for EK digestion was varied and EK concentration and incubation temperature were kept constant.
Reaction 1: Protein concentration 0.1 mg/ml, EK at 1IU: 150 ug protein, incubation at 4°C for 0-to 24 Hrs.
Reaction 2: Protein concentration 0.5 mg/ml, EK at 1IU:150 ug protein, incubation at 4°C for 0-24 24 Hrs.


Reaction 3: Protein concentration 0.25 mg/ml, EK at 1IU:150 ug protein, incubation at 4°C for 0-to 24 Hrs.
Reaction 4: Protein concentration 0.15mg/ml, EK at 1IU:150 ug protein, incubation at 4°C for 0-to 24 Hrs.
After EK digestion further purification was done. 1 ug pure PDGF BB from each reaction was reduced with (3-Mercapto Ethanol and checked on SDS-PAGE(Figure 10).
Example 9:
PDGF Characterization:
(a) SEC analysis- Size Exclusion Chromatography:
PDGF-BB sample analysis was carried out for Dimer and Monomer content with respect reference standard (Figure 9) using TSK GEL G3000SWXL as a stationary phase. l%(w/w) sodium dodecyl Sulfate in 0.063M Phosphate Buffer pH-7.0 was used as mobile phase at flow rate of 0.3ml/min and at a wavelength 210nm and 20-30ug of sample was injected (Figure 7).
(b) Tryptic mapping
0.5mg rhPDGF-BB was reduced, alkylated and digested with enzyme Trypsin, Tris HCL buffer pH8.5,Urea.
Peptide mapping of the rhPDGF BB Homodimer was carried out using Stationary phase TSK GEL ODS-80TS 250mm x 4.6mm, 5um. 0.1%TFA(w/v) in water and 0.085% w/v TFA in Acetonitrile were used as mobile phase A and mobile phase B
36

respectively at a flow rate 0.8ml/min and at wavelength 210nm. A step gradient method( 0 %B for l0min, 18%B in 36 minutes,27%B in 60 minutes, 40% B in 80 minutes, 70%B in 90 minutes) was used and l00µl of sample was injected(Figure 11).
(c)Determination of N-terminal amino acid sequence of Pure rhPDGF BB protein:
Pure rhPDGF BB was run on Tris tricine gel under reducing conditions. Protein bands were transferred to PVDF membrane and stained with CBB stain. rhPDGF-B monomer band at 12.5 KDa was removed from the membrane and sent to Anshul Biotech, Hyderabad for N terminal amino acid residue determination.
First five amino acids at the N terminal were;
Serine - Leucine - Glycine - Serine - Leucine
(d)Determination of endoproteolytic clipped rhPDGF by Size Exclusion Chromatography:
Endoproteolytic clipping analysis was carried out by reducing rhPDGF BB with dithiothreitol at 37°C for 2 hrs, using Superdex 200 10/300GL as a stationary phase.l%(w/w) sodium dodecyl Sulfate in 0.063M Phosphate Buffer pH-7.0 was used as mobile phase at flow rate of 0.3ml/min, wavelength of 210nm and 20-30ug of sample was injected (Figure 7).
Example 10:
rhPDGF- BB Cell Proliferation Bioassay:
The bioactivity of rhPDGF-BB was checked using mitogenic assay (Figure 12). The cell proliferation assay (mitogenecity assay) descried by- Raines et al 1985 and Rizza
37

et al 1985 was modified as per the assay requirements. Swiss Albino 3T3 mouse embryonic fibroblasts (received from NCCS, Pune,India) were seeded in 96 well microtiter plate in complete medium DMEM containing 5-10% fetal bovine serum and antibiotics at a concentration of 4-8 x 104 cells / ml and incubated for 6-8 hours in 37°C, 5% CO2 humidified incubator. The complete medium was then replaced with quiescent medium DMEM containing 0.1-1% calf serum for 2-18 hours. The quiescent medium was removed and various concentrations of rhPDGF-BB prepared in quiescent media were added to the wells. The control well did not contain rhPDGF-BB and served as base line control. The cells were then incubated 37°C, 5% C02 humidified incubator for 48-56 hours. After incubation, 20ul solution of MTT (5mg/ml) was added per well and incubated for 4-6 hours in 37°C, 5% CO2 humidified incubator. The formazan crystals developed were dissolved in acidified SDS solution (10-30% SDS, pH 2-4) and incubated at room temperature in dark for 12-18 hours. The 96 well plate was then put on shaker for 2-5 minutes and read at 570 nm in a Spectramax spectrophotometer.
The mean optimal densities (O.D.) were calculated and ED50 dose was determined by using the formula Cone. ED50 (x) = [ Calculated net OD for ED50 (y) - Intercept (b) ] / Slope(a)(x) = (y-b)/a
The ED50 values were compared with reference standard.
Example 11:
Wound Healing Assay:
The wound healing assay was performed according to the method described (Bartold and Ramen 1996 and Watanabe et al 1996) with some modifications. The human
38

muscle fibroblasts cell line used in the assay was developed in the lab. The human muscle fibroblasts were seeded in 6 well polystyrene culture plate at 1-3 x 10s cells/ ml DMEM containing 10% Fetal bovine serum. After cells reached confluence stage, cells were washed twice with plain DMEM and incubated with serum deprived medium for 24 hours in order to avoid any possible interference due to growth factors present in the serum. After 24 hours, artificial wound was created mechanically on the cell layer. The size of each wound was observed under microscope and only those wounds with the same circular shape and size which have cell free area were used for the assay. The wells were washed with plain medium and were treated with different concentrations of PDGF BB made in quiescent medium containing DMEM with 0.1-1 % calf serum and incubated in 37°C 5% CO2 humidified incubator for 24-96 hours. The proliferative activity of the cells towards the wound site after every 24 hours was observed under the microscope and was recorded by photomicrography (Data not shown ). PDGF-BB reference standard was used as positive control. The assay was found to be consistent. Wounds were completely filled with cells after 48 hours. The proliferative activity of rhPDGF-BB was compared with control wells as well as with reference standard.
While the present invention is described above in connection with preferred or illustrative embodiments, these embodiments are not intended to be exhaustive or limiting of the invention. Rather, the invention is intended to cover all alternatives, modifications and equivalents included within its spirit and scope, as defined by the appended claims.
39

SEQUENCE LISTINGS:
SEQ ID NO.l Length: 480
Type:
Organism: Human
Feature:
Other Information : DNA Sequence of the chimera containing TAG-PDGF
ATG GCA CTG CAC GCA CAT CTG GAC CCT CAT CTG GTG ACG GAG CAC GCC CAC CTC GAT CCG CAC GCT AGC CTG CAC GCA CAT CTG GAC CCT CAT CTG GTG ACG GAG CAC GCC CAC CTC GAT CCG CAC GTC GAC GAC GAC GAC GAC AAG
AGC CTG GGA TCC CTG ACC ATT GCT GAG CCG GCC ATG ATC GCC GAG TGC AAG ACG CGC ACC GAG GTG TTC GAG ATC TCC CGG CGC CTC ATA GAC CGC ACC AAC GCC AAC TTC CTG GTG TGG CCG CCC TGT GTG GAG
GTG CAG CGC TGC TCC GGC TGC TGC AAC AAC CGC AAC GTG CAG TGC CGC CCC ACC CAG GTG CAG CTG CGA CCT GTC CAG GTG AGA AAG ATC GAG ATT GTG CGG AAG AAG CCA ATC TTT AAG AAG GCC ACG GTG ACG CTG GAA GAC CAC CTG GCA TGC AAG TGT GAG ACA GTG GCA GCT GCA CGG CCT GTG ACC TAA TGA
40

SEQIDNO.2
Length: 158
Type:
Organism : Human
Feature :
Other Information : Amino Acid of the chimera containing TAG-PDGF
MALHAHLDPHLVTQHAHLDPHASLHAHLDPHLVTQHAHLDPHVDDDDDKS LGSLTIAEPAMIAECKTRTEVFEISRRLIDRTNANfLVWPPCVEVQRCSG CCNNRNVQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLACKCET VAAARPVT
SEQ ID NO. 3
Length: 122 AA
Type : 6X His- EK- PDGF
Organism: Artificial sequence
Other Information :
HHHHHHVDDDDDKSLGSLTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVW PPCVEVQRCSGCCNNRNVQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLED HLACKCETVAAARPVT..
SEQ ID NO.4
Length: 168 AA
Type : lacZ- 6 X His- EK-PDGF
Organism : Artificial sequence
Feature :
Other Information :
VNVDESWLQEGQTRIIFDGVNSAFHLWCNGRWVGYGQDSRLPMGGAHHHH HHVDDDDDKSLGSLTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVWPPCV EVQRCSGCCNNRNVQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLA CKCETVAAARPVT..
41

SEQID NO.5
Length :157AA
Type: DACi-EK-PDGF
Organism : Artificial sequence
Feature :
Other Information:
APHAHPDPHPVTEHAHPDPHASPHAHPDPHPVTEHAHPDPHVDDDDDKSLGS
LTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVWPPCVEVQRCSGCCNNRN
VQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLACKCETVAAARPVT..
SEQIDNO.6
Length: 157AA
Type: DAC2-EK-PDGF
Organism : Artificial sequence
Feature :
Other Information:
ATHAHTDPHTVTEHAHTDPHASTHAHTDPHTVTEHAHTDPHVDDDDDKSLG
SLTIAEPAMLAECKTRTEVFEISRRLTORTNANFLVWPPCVEVQRCSGCCNNRN
VQCRPTQVQLRPVQVRKIEIVRKXPffKKATVTLEDHLACKCETVAAARPVT..
SEQ ID NO.7
Length: 157 AA
Type: DAC3-EK-PDGF
Organism : Artificial sequence
Feature :
Other Information:
ASHAHSDPHSVTEHAHSDPHASSHAHSDPHSVTEHAHSDPHVDDDDDKSLGS
LTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVWPPCVEVQRCSGCCNNRN
VQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLACKCETVAAARPVT..
SEQID NO.8
Length :157 AA
Type : DAC4-EK-PDGF
Organism: Artificial sequence
Feature :
Other Information :
42

ALHAHLDPHLVTEHAHLDPHASLHAHLDPHLVTEHAHLDPHVDDDDDKSLG
SLTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVWPPCVEVQRCSGCCNNRN
VQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLACKCETVAAARPVT..
SEQ ID N0.9
Length: 157 AA
Type : DAC5-EK-PDGF
Organism : Artificial sequence
Feature :
Other Information:
AQHAHQDPHQVTEHAHQDPHASQHAHQDPHQVTEHAHQDPHVDDDDDKSL
GSLTIAEPAMIAECKTRTEVFEISRRLIDRTNANFLVWPPCVEVQRCSGCCNNR
NVQCRPTQVQLRPVQVRKIEIVRKKPIFKKATVTLEDHLACKCETVAAARPVT
SEQ ID NO. 10
Length : 31 bp
Type :
Organism : Artificial sequence
Feature:
Other Information:
GCTCGTGCCATGGGGAGCCTGGGTTCCCTGA
SEQ ID NO. 11
Length: 32
Type:
Organism : Artificial sequence
Feature :
Other Information:
TGCTCCTCGAGACCCCCCGGGCTTTAGGTCAC
SEQ ID NO. 12
Length: 33bp
Type:
Organism: Artificial sequence
Feature :
Other Information :
TGCTCCTCGAGACCCCCCGGGCTTTAGGTCACA
43

SEQIDNO.13
Length :32 bp
Type:_
Organism : Artificial sequence
Feature :
Other Information :
CCCGACAGCAGGGGATCATTTTGCGCTTCAGC
SEQIDN0.14
Length: 30 bp
Type:
Organism: Artificial sequence
Feature :
Other Information:
GCTCGGAGGATCCCAGGCTCTTGTCGTCGT
SEQIDNO.15
Length: 3\ bp
Type:
Organism: Artificial sequence
Feature :
Other Information:
GGGGGCCTGGGATCCCTGACCATTGCTGAGC
SEQIDNO.16
Length: 33 bp
Type:
Organism: Artificial sequence
Feature:
Other Information:
GCGGACCTCGAGACCCCCCGGTCATTAAGGTCA
44

We claim,
1. An improved process for recombinant synthesis of human Platelet derived growth factor BB comprising of inducibly expressing rhPDGF BB in a prokaryotic host cell wherein the expression is atleast 15% of total cell proteins and purifying the said protein avoiding organic solvents to obtain biologically active rhPDGF-BB of purity >97% wherein the purified protein yield is greater than 0.5g per litre of the fermentation broth.
2. The process of claim 1, wherein said rhPDGF BB is expressed as a chimeric protein in a prokaryotic host cell preferably E. coli K12 strain selected from the group consisting of TOPI 0, BL21, BL21 (DE3), BL2l(RIPL), Origami,Rossettagami, BL21-DE3(pLysS), BL21(trxB).
3. The process of claim 1 wherein said protein is expressed by an expression cassette comprising of a regulatory element operably linked to chimera consisting of an affinity handle which is placed upstream of the gene encoding rhPDGF B.
4. The process of claim 3 wherein the chimera has a nucleotide sequence as shown in SEQ ID NO.l and an amino acid sequence as shown in SEQ ID NO.:2.
5. The process of claim 3 wherein the regulatory element consists of a promoter selected from the group consisting of araB, Trp, Tac and T7, preferably araB.
6. The process of claim 3 wherein the affinity handle is selected from the tags encoded by SEQ ID NO. 3 to SEQ ID NO. 9.
7. The process of claim 1 wherein the inducer is selected from the group consisting of 0.5% to 10% arabinose, 0.1% to 2.0% IPTG and 5% to 40 % lactose, preferably 0.5% to 5% arabinose.
45

8. A process for obtaining biologically active rhPDGF BB expressed in a
prokaryotic host cell with >97% purity comprising:
i. expressing said protein to atleast 15% of total cell proteins as inclusion
bodies; ii. solubilizing inclusion bodies in the presence of chaotrophs selected
from the group consisting of urea and guanidium hydrochloride,
preferably 6M-8M guanidium hydrochloride; iii. capturing the chimeric tagged protein using an expanded bed affinity
chromatography; iv. refolding and dimerizing the chimeric tagged protein of step iii) in the
presence of arginine and optionally a sugar or a sugar alcohol wherein
the time of refolding is not more than 12 hours; v. cleaving the chimeric tagged protein by optimally using an
endoprotease wherein the clipping is not more than 8%; vi. purifying the crude protein of step v) to >97% by ion exchange
followed by negative affinity chromatography further followed by salt
exchange process.
9. The process of claim 8, wherein the refolding and dimerization of said protein is carried out in presence of 0-500 mM arginine wherein the % of aggregate formation is not more than 30, preferably not more than 10, more preferably not more than 3.
10. The process of claim 8, wherein the refolding and dimerizaion of said protein is carried out in presence of optionally a sugar selected from the group consisting of sucrose, glucose, preferably sucrose.
46

11. The process of claim 8, wherein the endoprotease is enterokinase comprising 10 to 15 units per mg of protein wherein the % of clipping is not more than 10%, preferably not more than 8%.
12. A pharmaceutical composition of rhPDGF-BB as claimed in claim 1 and atleast one pharmaceutically acceptable excipient.
13. The process as claimed in any of the preceding claims 1 to 12 as substantially described herein with respect to the foregoing examples 1 to 11 and drawings 1 to 11.
Dated this the 18th day of November, 2008
DrKG Rajendran Head-Knowledge Cell USV Limited Applicant
47