1
Description
HUMAN GRANULOCYTE-COLONY STIMULATING FACTOR
ISOFORMS
Technical Field
[ 1 ] The present invention relates to a novel human granulocyte colony stimulating
factor isoform, by which in vivo lifetime of human granulocyte colony stimulating
factor (rhG-CSF) is prolonged. More particularly, the present invention relates to a
human granulocyte colony stimulating factor isoform, obtained from the covalent
bonding of polyethylene glycol as a non-protein region to a human granulocyte colony
stimulating factor isoform having cysteine added to the N-terminal or the C-terminal
thereof.
Background Art
[2] Human granulocyte colony stimulating factor has the main biological function of
stimulating specific leucocytes, known as neutrophilic granulocytes or neutrophils, to
accomplish growth and development thereof in vivo (Welte et al., PNAS, 82,
1526-1530, 1985: Souza et al Science, 232, 61-65, 1986). Such neutrophilic
granulocytes function to protect a biological species against infection with mi-
croorganisms, when they are discharged to blood flow.
[3] The amino acid sequence of human granulocyte colony stimulating factor has been
reported by Nagata et al. (Nature, 319, 415-418,1986). Human granulocyte colony
stimulating factor is a protein capable of forming a complex with a receptor thereof in
a ratio of 2:2 via the dimerization of the receptor (Horan et al., Biochemistry, 35,
4886-96,1996).
[4] Aritomi et al. have shown the X-ray structure of the BN-BC domain complex of
human granulocyte colony stimulating factor with a receptor thereof (Nature, 401,
713-717, 1999). It is reported by Aritomi et al. that amino acids of human granulocyte
colony stimulating factor, which exist in contact region or adjacent region thereof
when the receptor is bound to human granulocyte colony stimulating factor, include
G4, P5, A6, S7, S8, L9, P10, Qll, S12, L15, K16, E19, Q20, L108, D109, D112,
Tl 15, Tl 16, Ql 19, E122, E123 and L124.
[5] Some isoforms of human granulocyte colony stimulating factor, obtained by
protein engineering technique have been disclosed (USP5581476, USP5214132,
USP5362853, USP4904584). Ridehall-Olsen et al., have reported that Lys40, Val48,
Leu49 and Phel44, present in the amino acid sequence of human granulocyte colony
stimulating factor, participate in the bonding with a receptor of human granulocyte
colony stimulating factor. Additionally, Rayton et al. have reported that Glu46, Leu49

2
and Phel44 are in contact with an immunoglobulin-like domain (Ig-like domain) of a
receptor of human granulocyte colony stimulating factor, while Lys40 and Val49 are
away from the domain. Thus, it is not certain that Lys40 and Val49 are in contact with
the receptor (J. biol. Chem., 2001, 276, 36779-36787). Further, Rayton et al., have
demonstrated that Glul9 of human granulocyte colony stimulating factor interacts with
Arg288 of a receptor thereof, thereby serving to transfer the signals of human
granulocyte colony stimulating factor (J. Biol. Chem., 1999, 274,17445-17451).
[6] It has been suggested to introduce at least one additional sugar chain artificially
into the human granulocyte colony stimulating factor protein by a genetic engineering
method (USP5218092). Such introduction of the sugar chain is performed by using a
method of exchanging, deleting and adding amino acids in the amino acid sequence of
a polypeptide.
[7J Additionally, many studies including bonding of polyethylene to human
granulocyte colony stimulating factor have been reported (Satake-Ishikawa et al. Cell
Structure and Function, 17, 157-160, 1992; USP5824778, USP5824784, W096/11953,
W095/21629, and WO94/20069).
[8] In vivo removal of a polypeptide or a polymer thereof is made by removal (or
clearance) and by receptor mediated degradation of protein in the kidney, spleen or
liver. Such removal is accomplished by the action of a protease specific to a substrate
or not. In general, in vivo protein removal depends on the size of protein (such a size
that glomerular filtration can be prevented), the charge of a protein molecule,
attachment of a sugar chain, protein receptor on the cell surface, or the like.
[9] Particularly, protein removal in the kidney depends on the physical properties of a
protein or a polymer thereof, such as size (molecular diameter), symmetry, shape/
rigidity or charge, or the like.
[10] Receptor mediated degradation of a protein is made when the protein loses its
function upon the bonding with a receptor. At the initial time, leucocytes are in-
sufficient, and bonding with a receptor on the surface of primordial hematopoietic stem
cells results in differentiation and growth into leucocytes. However, once the number
of leucocytes increases to reach a certain level, human granulocyte colony stimulating
factor is removed by the receptor present on the surface of leucocytes in order to
prevent excessive differentiation and growth caused by human granulocyte colony
stimulating factor. The ratio of the primordial hematopoietic stem cells to the receptors
of human granulocyte colony stimulating factor present on the surface of leucocytes is
approximately 1:5.
[11] Protein removal caused by the receptor present on the surface of leucocytes is
executed by the introduction of a receptor-bound polypeptide into cells, and by
lysosomal degradation of protein in the presence of proteases present in endosomes.

3
Degradation of protein by bonding of a receptor to human granulocyte colony
stimulating factor are described in detail by Saker and Roufenburger (Mol. Pharmacol.,
2003, 63, 147-158) and Saker (Nature biotech. 2002,20,908-913).
[12] To increase the half-life of a protein in blood, it is necessary to reduce protein
removal in the kidney and degradation of the protein by bonding of a receptor. It is
possible to reduce the protein removal in the kidney and to increase the half-life in
vivo by bonding a polymer capable of increasing the apparent molecular size to the
protein.
[13] Moreover, adhesion of a polymer to a protein can interrupt proteases effectively and
thus can prevent functions of non-specific proteases.
[14] Among such polymers, polyethylene glycol (PEG) is one of the polymers widely
used to prepare therapeutic protein products. Surface modification of a protein
molecule caused by bonding of a protein drug to a synthetic polymer can increase the
solubility of the drug to water or organic solvents. Accordingly, it is possible to
increase the biocompatibility of the drug, to reduce the immunoreactivity, to improve
the in vivo stability, and to retard clearance in the intestinal system, kidney, spleen or
liver. Clearance of small protein molecules is made by filtering in the intestinal system
or kidney. Hence, when a high- molecular weight PEG is bound to such small protein
molecules, it is possible to prevent such clearance (Knauf, M. J. et al, J. Hiol. Chem.
263:15064,1988). Introduction of a high-molecular weight polyethylene glycol (PEG)
into a protein drug can increase the stability of the protein molecule in a solution. Ad-
ditionally, it is possible to prevent adsorption of a non-specific protein by protecting
the intrinsic surface characteristics of the protein effectively. In this regard, there has
been an attempt to increase the in vivo half-life of a protein, to increase the solubility
of a protein and to reduce the in vivo immunoreactivity by boding PEG to a bio-
logically active peptide or protein. USP4,179,337 reported the result of such attempts
for the first time. Although bonding of PEG to a protein reduces many side effects due
to the aforementioned advantages, the PEG-bound protein undesirably shows a sig-
nificantly reduced in vivo activity because the active sites of the protein are blocked by
PEG.
[15] PEG, widely used up to date, is attached to the surface of protein by forming a
covalent bond with at least one free lysine residue. Herein, if a site, which is present on
the protein surface and is directly related to the activity of the protein, is bound to
PEG, the activity of the protein decreases. Additionally, since the bonding between
PEG and lysine residues occurs randomly, Various kinds of PEG-bound protein
conjugates are present in the form of a mixture, so that separation of a desired
conjugate in a pure state becomes complicated and difficult. For example, EP 0401384
discloses a material and a method for producing polyethylene glycol-added G-CSF.

4
EP0335423 discloses a modified polypeptide having an activity of human granulocyte
colony stimulating factor. However, according to the prior art, the polyethylene glycol
molecule cannot be bound to a predictable specific residue but bound to any reactive
group present in the protein, such as a lysine residue, or the N-terminal in a non-
specific manner, resulting in the formation of a non-homogeneous product. In order to
bond a specific region of protein with PEG, USP5,766,897 and WO00/42175 disclose
a method of bonding human growth hormone with PEG, wherein PEG is allowed to
react selectively with cysteine by using PEG-maleimide so as to prevent PEG from
reacting with the active region of human growth hormone. It is required that a free
cysteine residue that does not participate in disulfide bonding is present in human
growth hormone to permit use of the above type of PEG. However, all of the four
cysteine residues present in human growth hormone participate in disulfide bonding.
Therefore, according to the prior art, pegylation is performed by using a human growth
hormone derivative, into which a cysteine residue is artificially inserted. Additionally,
USP6,555,660 and USP6,753,165 and US Laid-Open patent No. 2005/0058621 and
KR2002-0079778 disclose mutation of cysteine at various positions in the amino acid
sequence of human granulocyte colony stimulating factor.
[ 16] Meanwhile, Bowen et al. have shown that molecular weight of polyethylene in
polyethylene-modified human granulocyte colony stimulating factor is related to the
lifetime of a drug. In an in vitro test, efficacy of a protein drug has an inverse relation
with the molecular weight of polyethylene bound to the protein. However, in vivo
activity increases as the molecular weight increases. It is thought that a polymer of a
physiologically active protein shows a low affinity in receptor-mediated degradation,
and thus shows an increased half life. Therefore, although such polymers of human
granulocyte colony stimulating factor accomplish recovery of nutrophils in vivo, they
show lower activity than the non-polymerized human granulocyte colony stimulating
factor in terms of in vitro activity.
[17] Recently, a modified human granulocyte colony stimulating factor that includes
20kDa polyethylene glycol bound to the N-terminal has been developed and com-
mercialized (Neulasta). The polyethylene glycol-polymerized human granulocyte
colony stimulating factor has an increased half life, and thus can reduce administration
frequency.
[18] Particular examples of commercially available human granulocyte colony
stimulating factor include E. coli-derived filgrastim (trade name: Gran and Neupogen),
lenograstim (trade name: Neutroginand Granocyte) produced in animal cells, such as
Chinese Hamster Ovary (CHO) cells, and E.coli-derived nartograstim (trade name:
Neu-up), in which mutation occurs at the N-terminal regions in five amino acid
sequences in order to increase the efficacy of human granulocyte colony stimulating

5
factor protein.
Disclosure of Invention
Technical Problem
[19] However, human granulocyte colony stimulating factor isoforms according to the
prior art still have problems in that such amino acid sequence-modified human
granulocyte colony stimulating factor isoforms have a low activity or insufficient half
life in vivo, or non-homogeneous products are generated upon the bonding of PEG to
human granulocyte colony stimulating factor.
Technical Solution
[20] Therefore, the present invention has been made in view of the above-mentioned
problems. It is an object of the present invention to provide a human granulocyte
colony stimulating factor isoform having an excellent activity and a significantly
increased half life by bonding polyethylene glycol to amino acid sequence-modified
human granulocyte colony stimulating factor, wherein the amino acid sequence is
modified by substitution with or addition of cysteine in a specific site in order to
facilitate bonding polyethylene glycol to human granulocyte colony stimulating factor,
and then the modified site is partially bound to polyethylene glycol. According to the
prior art, human granulocyte colony stimulating factor isoforms have been generally
produced by modifying the sequence in the polypeptide and performing pegylation at
the position of a suitable amino acid in a specific site so as to increase the in vivo
activity or lifetime. However, the inventors of the present invention have found that a
novel isoform of human granulocyte colony stimulating factor, which shows an
increased in vivo activity or lifetime compared to the isoforms according to the prior
art, can be obtained by adding amino acids to both terminals of the polypeptide of
human granulocyte colony stimulating factor and performing pegylation at the same
addition sites. The present invention is based on this finding.
[21] According to an aspect of the present invention, there is provided a human
granulocyte colony stimulating factor isoform, which has at least one amino acid added
before Thr of the N-terminal of human granulocyte colony stimulating factor
represented by Sequence No. 1 or after Pro of the C-terminal thereof, wherein at least
one of the amino acids added to human granulocyte colony stimulating factor is
cysteine and at least one cysteine is bound to polyethylene glycol.
[22] It is generally known to those skilled in the art that the N-terminal in Sequence No.
1 may include Met. In this case, the cysteine may be added between the Met and the
threonine of the N-terminal. The above isoform is also included in the scope of the
present invention.
[23] More particularly, 1-12 amino acids may be added to each terminal of the

6
sequence. Preferably, one cysteine is added. More preferably, the cysteine is 175C,
added to the C-terminal.
[24] According to a preferred embodiment of the present invention, polyethylene glycol
is covalently bonded to the cysteine added to human granulocyte colony stimulating
factor. Preferably, polyethylene glycol has a molecular weight of 20 kDa~40 kDa.
More preferably, polyethylene glycol is branched.
[25] According to another preferred embodiment of the present invention, at least one
amino acid in human granulocyte colony stimulating factor is substituted with another
amino acid before polyethylene glycol is bound to human granulocyte colony
stimulating factor.
[26] According to another aspect of the present invention, there is provided a phar-
maceutical composition comprising the aforementioned human granulocyte colony
stimulating factor isoform and a pharmaceutically acceptable carrier.
[27] According to still another aspect of the present invention, there are provided genes
encoding a protein of the aforementioned human granulocyte colony stimulating factor
isoform having cysteine added to the N-terminal or C-terminal thereof.
[28] According to yet another aspect of the present invention, there is provided an
oligodeoxynucleotide used as a primer for modifying the amino acid sequence of
human granulocyte colony stimulating factor into the aforementioned human
granulocyte colony stimulating factor isoform. Preferably, the primer is the
oligodeoxynucleotide represented by any one of Sequence Nos. 2~7.
Advantageous Effects
[29] The human granulocyte colony stimulating factor isoform according to the present
invention shows a significantly increased in vivo activity or lifetime, and thus can
reduce administration frequency.
Brief Description of the Drawings
[30] The foregoing and other objects, features and advantages of the present invention
will become more apparent from the following detailed description when taken in
conjunction with the accompanying drawings in which:
[31] FIG. 1 illustrates the genetic sequence and protein sequence of human granulocyte
colony stimulating factor;
[32] FIG. 2 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the N-terminal site;
[33] FIG. 3 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the C-terminal site;
[34] FIG. 4 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the middle part of the polypeptide;

7
[35] FIG. 5 illustrates the structure of a plasmid vector expressing a human granulocyte
colony stimulating factor isoform; and
[36] FIG. 6 is a graph showing the in vivo lifetime of human granulocyte colony
stimulating factor isoforms in rats.
Best Mode for Carrying Out the Invention
[37] Reference will now be made in detail to the preferred embodiments of the present
invention.
[38] FIG. 1 illustrates the genetic sequence and protein sequence of human granulocyte
colony stimulating factor. Amino acid sequence No. 1 is the same as the amino acid
sequence represented in FIG. 1 with the exception of Met in -1 position.
[39] FIG. 2 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the N-terminal site. AA represents
amino acids, and n is the number of amino acids. N is an integer ranging from 1 to 12.
The amino acids include at least one cysteine.
[40] FIG. 3 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the C-terminal site. AA represents
amino acids, and n is the number of amino acids. N is an integer ranging from 1 to 12.
The amino acids include at least one cysteine.
[41] FIG. 4 illustrates the amino acid sequence of a human granulocyte colony
stimulating factor isoform, amino acid-modified at the part comprising a sugar chain.
Herein, 133-threonine is substituted with cysteine, and polyethylene glycol is
covalently bonded with the same cysteine. This is a known isoform according to the
prior art (see Korean Laid-Open Patent No. 2002-0079778).
[42] FIG. 5 illustrates the structure of a plasmid vector expressing a human granulocyte
colony stimulating factor isoform. As a promoter, PR of a bacteriophage lamda is used.
To accomplish the expression in E. coli, widely used in laboratories, the plasmid vector
includes cl 857 genes capable of inducing such expression in the plasmid vector.
[43] FIG. 6 is a graph showing the in vivo lifetime of human granulocyte colony
stimulating factor isoforms in rats. PBS carrier is used as a control, and test samples
include: T133C20kDa(Br) (G-CSF isoform obtained by substituting 133-Thr with
cysteine and covalently bonding branched PEG with 20 kDa to the same cysteine
according to Korean Laid-Open Patent No. 2002-0079778); commercially available
Neulasta (Amgene Co.) having 20kDa PEG bound to the N-terminal; N+20kDa(Br)
(G-CSF isoform obtained by adding cysteine between the threonine and methionine of
the N-terminal (-1C) of G-CSF and covalently bonding branched PEG with 20 kDa to
the same cysteine); 175C+20kDa(Br) (G-CSF isoform obtained by adding cysteine
after 174-Pro of the C-terminal of G-CSF and covalently bonding branched PEG with

8
20 kDa to the same cysteine); and 175C+40kDa(Br) (G-CSF isoform obtained by
adding cysteine after 174-Pro of the C-terminal of G-CSF and covalently bonding
branched PEG with 40 kDa to the same cysteine). Each isoform is diluted with PBS,
and is administered to rats through the tail veins in a dose of lOOD/kg based on the body
weight measured upon the administration day. Administration is carried out only once
upon the test day.
[44] The inventors of the present invention have studied the protein structure of human
granulocyte colony stimulating factor and the effect of the amino acids forming human
granulocyte colony stimulating factor upon the activity thereof. Reference has been
made to the structure of human granulocyte colony stimulating factor, information
about the effect of the amino acids forming human granulocyte colony stimulating
factor upon the activity thereof, and to the results of GOR4 protein secondary structure
prediction used to determine the effect of the modification of amino acids upon the
protein structure (GOR program in Secondary Structure Prediction,
http:us.expasy.org).
[45] As used herein, the term "human granulocyte colony stimulating factor isoforms"
include analogs, variants, mutants, conjugates, or the like, which maintain the
indigenous activity of human granulocyte colony stimulating factor, even though they
experience modification of one amino acid residue into another amino acid residue in
the original amino acid sequence of human granulocyte colony stimulating factor.
[46] Three letter or single letter mnemonic codes for amino acids used herein are
defined as follows according to biological standards:
[47] Ala (A): alanine; Asx (B): asparagine or aspartic acid; Cys (C): cysteine; Asp (D):
aspartic acid; Glu (E): glutamic acid; Phe (F): phenylalanine; Gly (G): glycine; His
(H): histidine; He (I): isoleucine; Ls (K): lysine; Leu (L): leucine; Met (M):
methionine; Asn (N): aspargine; Pro (P): proline; Gin (Q): glutamine; Arg (R):
arginine; Ser (S): serine; Thr (T): threonine; Val (V): valine; Trp (w): tryptopane; Tyr
(Y): tyrosine; Glx (Z): glutamine or glutamic acid.
[48] As described herein, "(single letter for an amino acid) (amino acid position) (single
letter for another amino acid)" means that the former amino acid in the corresponding
amino acid position is substituted with the latter amino acid. For example, T133C
means that threonine, the 133r amino acid in the sequence of natural human
granulocyte colony stimulating factor, is substituted with cysteine.
[49] Herein, the primer used to induce the modification of an amino acid at a specific
site is represented by the expression (single letter for an amino acid) (amino acid
position) (single letter for another amino acid) 1 or 2, wherein suffix 1 represents the
primer for a single strand template progressing in the direction of 5'->3' in a double
strand template, while suffix 2 represents the primer for a single strand template

9
progressing in the direction of 3'->5'.
[50] The human granulocyte colony stimulating factor genes obtained by the above
process may be modified in the position at least one codon. As used herein, "mod-
ification" refers to a change in the amino acid sequence of human granulocyte colony
stimulating factor caused by substitution or addition of at least one codon in the genes
encoding human granulocyte colony stimulating factor. More particularly, such mod-
ification includes: addition of cysteine before threonine, the first amino acid in the
amino acid sequence of human granulocyte colony stimulating factor; addition of
cysteine after proline of the C-terminal of the same amino acid sequence; sequence
modification in the middle of the polypeptide sequence in addition to the modification
at the terminal sites; or the like.
[51] For example, T133C means substitution of threonine, the 133r amino acid of
natural human granulocyte colony stimulating factor, with cysteine. According to an
embodiment of the present invention, a synthetic oligonucleotide including a codon
that encodes modification of a target amino acid in human granulocyte colony
stimulating factor is constructed. In general, an oligonucleotide with a length cor-
responding to about 27-36 nucleotides is used. Although a shorter oligonucleotide may
be used, the oligonucleotide preferably has 12-15 nucleotides complementary to the
template at both sides of the nucleotide encoding modification. Such oligonucleotides
may be sufficiently hybridized with the template DNA. A synthetic oligonucleotide
used for the amino acid modification according to the present invention is shown in the
following Table 1. The oligonucleotide may be prepared by a method known to those
skilled in the art.
[52] According to a preferred embodiment of the present invention, DNA of human
granulocyte colony stimulating factor isoform, one of whose amino acids is modified,
is constructed. In this case, PCR is performed by using human granulocyte colony
stimulating factor DNA (Table 1) as a template and a synthetic oligonucleotide
encoding modification as a primer. If the double- strand template is separated in the
heating step of PCR, a complementary primer is hybridized with each single strand
template. DNA polymerase allows the nucleotides complementary to the template to be
linked continuously thereto from -OH groups of the primer encoding modification
along the direction of 5'-»3'. As a result, the second strand includes the primer
encoding modification, so that the resultant genes encode the target modification. The
second strand functions as a template DNA during repeated replication steps, so that
the genes encoding modification can be amplified continuously.
[53] Hereinafter, modification in the middle of the polypeptide of human granulocyte
colony stimulating factor will be explained in more detail with reference to T133C
taken as an example. T133C refers to substitution of threonine, the 133r amino acid of

10
natural human granulocyte colony stimulating factor, with cysteine. To obtain T133C,
PCR is performed by using natural granulocyte colony stimulating factor DNA (FIG.
1) as a template and by introducing each of the primer pairs formed of N-term with
T133C2 and T133C1 with C-term. By doing so, two DNA fragments, having a
modified codon corresponding to cysteine that substitutes for threonine, the 133r
amino acid, can be obtained. Then, PCR is further carried out by introducing the two
DNA fragments and using primer pairs of N-term and C-term. In this manner, it is
possible to obtain modified genes of human granulocyte colony stimulating factor
isoform T133C, in which threonine is substituted with cysteine at the position of the
133r amino acid.
[54] According to another example of the present invention, amino acid addition is
made at the N-terminal or the C-terminal. Construction of human granulocyte colony
stimulating factor having at least one modified amino acid can be performed at the N-
terminal and the C-terminal. Such modification includes the sequence of Met-
(AA)n-Thr- at the N-terminal (see FIG. 2) and the sequence of -Gin-Pro-(AA)n at the
C-terminal (see FIG. 3). Herein, AA means any amino acid and particular examples
thereof include an amino acid that has a low molecular weight and shows no specific
function, such as glycine, serine or alanine. At least one AA includes cysteine. Suffix n
means the number of amino acids and ranges from 1 to 12. At least about 12 amino
acids are required for the in vivo recognition of the modified human granulocyte
colony stimulating factor, to which amino acids are added, as a heterogeneous protein,
followed by the production of an antibody. This is why the maximum number of amino
acids is limited to 12.
[55] Therefore, in order to construct human granulocyte colony stimulating factor genes
having additional amino acids, the N-terminal has the amino acid sequence of Met-
(AA)n-Thr- and includes the DNA base sequence of ATG-(NNN)n-ACT-, while the C-
terminal has the amino acid sequence of -Gln-Pro-(AA)n- and includes the DNA base
sequence of -CAG-CCG-(NNN)n-TAA. Herein, NNN represents the DNA base
sequence that encodes the corresponding amino acid.
[56] Then, human granulocyte colony stimulating factor genes having at least one
modified amino acid are constructed by using the primer including the DNA base
sequence obtained as described above and human granulocyte colony stimulating
factor genes as shown in FIG. 1.
[57] For example, if one glycine and one cysteine are added to the N-terminal, the
amino acid sequence at the N-terminal may include the sequence of Met-
(Gly-Cys-)Thr-. The primer used herein may be designed to have the sequence of
GTAATAAATA-ATG-(GGT-TGT-)ACT-. Then, modified human granulocyte colony
stimulating factor genes having the sequence of Met-(Gly-Cys-)Thr- may be

11
constructed from the above primer, by using the template of human granulocyte colony
stimulating factor genes as shown in FIG. 1 and the C-term primer as shown in Table
1.
[58] The human granulocyte colony stimulating factor genes obtained as described
above are further subjected to PCR by using N-term and C-term primers.
[59] According to still another embodiment of the present invention, it is possible to add
an amino acid to the N-terminal or C-terminal and to substitute at least one amino acid
positioned in the middle of the polypeptide of human granulocyte colony stimulating
factor with another amino acid, before polyethylene is bonded thereto. For example, -
1C/T133C includes addition of cysteine to the N-terminal and substitution of the
threonine, the 133 amino acid of the polypeptide of human granulocyte colony
stimulating factor, with cysteine. To obtain -1C/T133C, PCR is performed by using
preformed DNA of -1C human granulocyte colony stimulating factor isoform as a
template and by introducing primer pairs of N-term with T122C2 and T133C1 with C-
term. By doing so, two DNA fragments, having a modified codon corresponding to
cysteine that substitutes for threonine, the 133r amino acid, can be obtained. Then,
PCR is further carried out by introducing the two DNA fragments and using primer
pairs of N-term and C-term. In this manner, it is possible to obtain modified genes of
human granulocyte colony stimulating factor isoform -1C/T133C, in which threonine
is substituted with cysteine at the position of the 133r amino acid and cysteine is added
before the threonine of the N-terminal. Construction of isoforms at a position other
than 133 may be performed by using the corresponding DNA primer in the same
manner as -1C/T133C isoform.
[60] Modification of amino acid in the middle of the polypeptide of the amino acid-
added isoform including at least one cysteine at the N-terminal or the C-terminal can
be performed at a site known to those skilled in the art. For example, methods of sub-
stituting at least one amino acid in the middle of the polypeptide of human granulocyte
colony stimulating factor with another amino acid are disclosed in Korean Patent
Publication No. 0248111 (N-terminal polyethylene glycol-bound G-CSF or analogues
thereof, and preparation thereof), Korean Laid-Open Patent Nos. 2002-0007297
(G-CSF conjugates), 2002-0079778 (G-CSF conjugates) and 2004-0084884(G-CSF
conjugates), US Patent Nos. 5214132 (Polypeptide derivatives of human granulocyte
colony stimulating factor) and 5416195 (Polypeptide derivatives of granulocyte colony
stimulating factor), US Laid-Open Patent Nos. 2003-0158375 (G-CSF polypeptides
and conjugates), 2004-0214287 and 2005-0058621. More particularly, according to
USP5214132, USP5416195 and US2004-0214287, 1, 3,4, 5 and 17-position amino
acids of human granulocyte colony stimulating factor are substituted with Ala, Thr,
Tyr, Arg and Ser, respectively, or the cysteine, 17-position amino acid, is substituted

12
with alanine or serine.
[61] The contents of the above patents and patent applications are herein incorporated
by reference. Also, methods of producing human granulocyte colony stimulating factor
isoforms, pegylation methods, or the like, as disclosed in the prior art are also herein
incorporated by reference.
[62] The DNA sequence that encodes the human granulocyte colony stimulating factor
isoforms according to the present invention may be synthesized by the conventional
method known to those in the art, for example, by using an automatic DNA synthesizer
(available from Biosearch, Applied Biosystem , etc.).
[63] According to still another aspect of the present invention, there is provided a re-
combinant expression vector including the DNA sequence encoding the human
granulocyte colony stimulating factor isoform and a host cell transformed or
transfected with the above expression vector.
[64] As used herein, the term "polypeptide" may be used interchangeably with "protein"
or "physiologically active protein".
[65] As used herein, the term "isoform" means a physiologically active protein, which
has the main function of the original physiologically active protein but is modified
with a genetic engineering means or others. Herein, the term "modification" refers to
substitution, addition and loss in an amino acid sequence, and includes substitution
with cysteine or addition of cysteine for carrying out bonding with polyethylene
glycol. Further, such modification also includes a covalent bonding formed between
cysteine and polyethylene glycol via a chemical reaction.
[66] As used herein, the term "in vivo lifetime" is defined as follows: a physiologically
active protein loses its function when it is administered to humans or animals, because
the total amount and activity of the physiologically active protein decrease due to the
protease-mediated removal and the removal in the kidney. Therefore, the term "in vivo
lifetime" means the retention time of a protein during which the protein can perform its
function in vivo, and is represented by the in vivo half life of the physiologically active
protein.
[67] As used herein, the term "vector" refers to a DNA molecule as a carrier capable of
delivering foreign genes stably into a host cell. To be a useful vector, it is necessary for
a vector to allow replication, to have a means, through which it can be introduced into
a host cell, and to include a means for detecting itself. Additionally, the term "re-
combinant expression vector" refers to a cyclic DNA molecule formed by a vector
linked operatively to foreign genes so that the foreign genes can be expressed in a host
cell.
[68] As a recombinant expression vector, a DNA vector having human granulocyte
colony stimulating factor inserted thereto, may be produced. When constructing a re-

13
combinant expression vector, in order to increase the expression level of transfected
genes in a host cell, the corresponding genes should be linked operatively to a tran-
scription and translation expression regulatory sequence that functions in a selected
host cell, as generally known in the art. Preferably, the expression regulatory sequence
and the corresponding genes are included in a single expression vector that further
includes a bacterial selection marker and a replication origin.
[69] A suitable vector including not only the genes encoding human granulocyte colony
stimulating factor isoforms but also the above elements (e.g. regulatory sequence) can
be constructed by using basic recombinant DNA technique. In order to form a desired
vector, individually separated DNA fragments should be linked to each other. Thus,
each DNA fragment is cleaved first by using restriction enzymes, and then the DNA
fragments are linked to each other considering a specific order and orientation.
[70] DNA may be cleaved by using a specific restriction enzyme in a suitable buffer. In
general, about 0.2-ID of a plasmid or DNA fragment is used in about 20Q of a buffer
solution along with about 1-2 units of the corresponding restriction enzyme. Suitable
buffer, DNA concentration, reaction time and reaction temperature will be determined
by the producer of the restriction enzyme. Generally, the reaction is carried out at 37°C
for about 1-2 hours. However, some enzymes require a higher temperature. After the
reaction, enzymes and other impurities are removed by extracting the digestion
solution with a mixture of phenol and chloroform. DNA may be precipitated with
ethanol to recover it from aqueous layer. At this time, terminals of DNA fragments to
be linked to each other should be compatible with each other in order to link DNA
fragments so that a functional vector can be formed.
[71] The cleaved DNA fragments are classified and selected on the size basis via elec-
trophoresis. DNA electrophoresis may be performed through an agarose or poly-
acrylamide matrix. Selection of the matrix depends on the size of DNA to be separated.
After the electrophoresis, DNA is extracted from the matrix via electroelution, or
extracted directly from the matrix. When a low-melting agarose matrix is used, agarose
should be melted before the extraction of DNA.
[72] DNA fragments to be linked to each other should be added to a solution in an
equimolar amount. The solution contains ATP, ligase buffer, and about 10 units ligase
such as T4 ligase per 0.5D of DNA. To link DNA fragments to a vector, the vector
should be cleaved by a suitable restriction enzyme first to form a linear vector. Such
linear vector may be used after it is treated with alkali phosphatase or bovine intestinal
hydrolase. Such treatment with hydrolase prevents the vector from self-ligation. The
recombinant expression vector constructed as described above is used to perform trans-
formation or transfection of a host cell.
[73] Polynucleotides can be introduced into a host cell according to a known method

14
disclosed in laboratory manuals such as [Davis et al., Basic Methods in Molecular
biology (1986)] and [Sambrook, J., et al. (1989) "Molecular Cloning" A laboratory
Manual 2nd edition]. Preferred methods of introducing polynucleotides into a host cell
include calcium chloride transformation, electroporation, or the like.
[74] According to the present invention, host cells are incubated in a nutrition medium
suitable for the production of polypeptides by using a known technique. For example,
hose cells may be incubated in a laboratory or industrial fermenting instrument via
small-scale or large-scale fermentation/shaking flask incubation with a suitable
medium under the conditions permitting expression and/or secretion of polypeptides.
Incubation is performed in a suitable nutrition medium containing a carbon source, a
nitrogen source and inorganic salts by using a known technique. The medium is
generally known to those skilled in the art and is commercially available. Otherwise, a
medium prepared directly in a laboratory may be used. If polypeptides are secreted
directly to the nutrition medium, the polypeptides can be separated directly from the
medium. If polypeptides are not secreted to the nutrition medium, they can be
separated from the cell lysate.
[75] Separation of polypeptides can be performed in a conventional manner. For
example, polypeptides can be separated from the medium by a conventional method
including centrifugal separation, filtration, extraction, spray drying, evaporation or pre-
cipitation. Further, purification of polypeptides can be performed by various methods
known to those skilled in the art, including chromatography (e.g. ion exchange chro-
matography, affinity chromatography, hydrophobic chromatography or size exclusion
chromatography), electrophoresis, fractional solubility separation (e.g. ammonium
sulfate precipitation), SDS-PAGE or extraction.
[76] Then, the human granulocyte colony stimulating factor isoform purified as
described above is subjected to pegylation in the manner as described hereinafter.
[77] Suitable "non-protein segments" used for the pegylation according to the present
invention include polyethylene glycol, and branched polyethylene glycol is preferred.
Other examples of such non-protein segments include, but are not limited to: at least
one material selected from the group consisting of water soluble polymers such as
polypropylene glycol (PPG), polyoxyethylene (POE), polytrimethylene glycol,
polylacatic acid and derivatives thereof, polyacrylic acid and derivatives thereof,
poly (amino acid), polyvinyl alcohol), polyurethane, polyphosphazenes,
poly(L-lysine), polyalkylene oxide (PAO), polysaccharide, or the like, and non-
immune polymers such as dextran, polyvinyl pyrrolidone, polyvinyl alcohol (PVA),
polyacryl amide, or the like. The biocompatible polymer used in the preparation of the
polymer derivatives according to the present invention preferably has a molecular
weight of about 2,000-100,000, more preferably 20,000-40,000.

15
[78] In order to form a bond between a biocompatible polymer and G-CSF via the thiol
group of G-CSF, the biocompatible polymer should be activated. For this purpose, the
biocompatible polymer may be bonded with a reactive functional group. The term
"reactive polymer group" means a group or a moiety capable of activating a bio-
compatible polymer so that it can be bonded with a biologically active material. To
form a bond between a biocompatible polymer and a biologically active material, one
of the terminal groups of the biocompatible polymer is converted into a reactive
functional group. This process is referred to as "activation". For example, to form a
bond between poly(alkylene oxide) and a biologically active protein, one of the
hydroxyl terminal groups may be converted into a reactive group, such as carbonate.
As a result, activated poly(alkylene oxide) can be obtained.
[79] The reactive group that may be used to activate a biocompatible non-protein
segment polymer according to the present invention is selected from the group
consisting of maleimide, acetamide, pentenoic amide, butenoic amide, isocyanate,
isothiocyanate, cyanuric acid chloride, 1,4-benzoquinone, disulfide, or the like.
[80] Bonding PEG to a protein via a thiol group of the protein can be performed by a
method generally known to those skilled in the art.
[81] For example, PEG-maleimide is used to allow thiol groups to be bound to activated
double bonds via Michael reaction [Ishii et al., Biophys J. 1986, 50:75-80]. Ad-
ditionally, as generally known to those in the field of protein chemistry, PEG-
iodoacetamide reagent may be used. The latter method is advantageous in that a stable
PEG-bound cysteine derivative, i.e. carboxymethyl cysteine, can be obtained by a
strong acidic hydrolysis, wherein the derivative can be identified and determined by
standard amino acid analysis [Gard FRN. Carboxymethylation, Method Enzymol
1972; B25: 424-49]. In addition to the above methods, a method of producing a stable
symmetric disulfide by using PEG-ortho-pyrimidyl-disulfide {Woghiren et al. Bio-
conjugate Chem 1993, 50:75-80}; a method of reacting an activated PEG, i.e. sulfosuc-
cinimidyl 4-(N-maleimidomethyl)cyclohexane-l-carboxylate activated-PEG with thiol
group of cysteine to form a covalent bond [USP5,166,322]; a method of reacting an
activated PEG, i.e. maleimido-6-aminocaproylester-activated PEG 4000 with the IL-2
mutant having a cysteine substituent at a specific site so as to be conjugated with a
cysteine residue [USP5,206,344]; and a method of modifying thiol groups of a protein
by using gamma-maleimidobutyric acid and beta-maleimidopropionic acid [Rich et al.,
J. Med. Chem. 18, 1004, 1975] maybe used.
[82] The pegylation method in the following examples were performed by using
mPEG20,000-maleimide, branched mPEG20,000-maleimide and branched
mPEG40,000 available from NOF in a molar ratio of 20:1 to human granulocyte
colony stimulating factor. The pegylation reaction is performed in 0.1 M phosphate

16
buffer (pH 7-8) at room temperature for 2 hours with stirring.
[83] Pharmaceutically acceptable carriers, excipients or stabilizers must be non-toxic to
the subjects and be compatible with other constituents of the pharmaceutical
preparation under the corresponding administration dose and concentration. For
example, the pharmaceutical preparation should not contain any materials known to be
harmful to polypeptides, such as oxidants or other materials.
[84] Suitable carriers include: buffers such as phosphoric acid, citric acid and other
organic acids; antioxidants including ascorbic acid; low-molecular weight
polypeptides; proteins such as blood plasma albumin, gelatin and immunoglobulin; hy-
drophilic polymers such as polyvinyl pyrrolidone; amino acids such as glycine,
glutamin, arginine, or lysine; monosaccharides, disaccharides or other carbohydrates
including glucose, mannose or dextrin; chelate factors such as EDTA; metal ions such
as zinc, cobalt or copper; sugar alcohols such as mannitol or sorbitol; salt-forming
counterions such as sodium; and/or non-ionic surfactants such as Tween, Pluronic or
polyethylene glycol (PEG).
[85] A sugar chain-bound human granulocyte colony stimulating factor isoform should
be sterilized before it is administered for the purpose of treatment. Such sterilization
may be accomplished easily by way of filtering through a sterilized filtration
membrane.
[86] In general, a therapeutic sugar chain-bound human granulocyte colony stimulating
factor isoform composition is stored in a container equipped with a sterilized access
port, for example, an intravenous injection bag or vial having a septum through which
a subcutaneous injection needle can pass.
[87] The human granulocyte colony stimulating factor isoform according to the present
invention may be administered directly to animals by suitable routes including
parenteral administration routes, locally or systematically. A particular administration
route will be determined by the history of a patient, including side effects recognized
or expected upon the administration of human granulocyte colony stimulating factor
isoform. Particular parenteral administration routes include subcutaneous, in-
tramuscular, intravenous, intraarterial and intraperitoneal routes. Most preferably, ad-
ministration is carried out by continuous infusion (e.g. mini-pump such as an osmotic
pump) or injection, for example, intravenous injection or subcutaneous injection. A
preferred administration route for human granulocyte colony stimulating factor
isoform is subcutaneous administration.
[88] The human granulocyte colony stimulating factor isoform according to the present
invention may be administered to a patient in a therapeutically effective amount. The
term "therapeutically effective amount" refers to the amount sufficient to obtain a
desired treating effect by using a given administration mode under a given condition.

The therapeutic composition of human granulocyte colony stimulating factor should be
formulated and administered conformed to the preferred medical practice, under the
consideration of the particular condition to be treated, clinical conditions of an
individual patient (particularly, side effects upon the single administration of human
granulocyte colony stimulating factor), delivery sites of the composition of the human
granulocyte colony stimulating factor isoform, administration modes, administration
schedules and other factors generally known to those skilled in the art. The thera-
peutically effective amount of the human granulocyte colony stimulating factor
isoform is determined considering the above factors. In general, daily dose of the
human granulocyte colony stimulating factor isoform according to the present
invention ranges from about ID/kg to 1000/kg.
[89] Reference will now be made in detail to the preferred embodiments of the present
invention. It is to be understood that the following examples are illustrative only, and
the scope of the present invention is not limited thereto.
[90] Construction of Human Granulocyte Colony Stimulating Factor
Isoforms
[91] To construct isoforms of natural human granulocyte colony stimulating factor,
human granulocyte colony stimulating factor genes were synthesized as a template.
The genes synthesized in this example have base sequences as shown in the following
Table 1, and DNA sequence and amino acid sequence of human granulocyte colony
stimulating factor, used as a template, are shown in FIG. 1. In Table 1, N-term~175C
correspond to Sequence Nos. 2-7, respectively.
[92] Table 1

[93]
[94] 1. Construction of -1C Human Granulocyte Colony Stimulating Factor Isoform
[95] PCR was performed to construct -1C isoform by using DNA of natural human
granulocyte colony stimulating factor as a template and primer pairs formed of -1C
primer and C-term primer. The amplified DNA fragment was further subjected to PCR
by using N-term primer and C-term primer so that restriction enzyme cleavage sites
(Xba I and BamH I) could be introduced, and then was introduced into an expression

18
vector. By doing so, it was possible to obtain genes of a human granulocyte colony
stimulating factor isoform having cysteine added before threonine of the N-terminal.
[96] The amino acid sequence of the N-terminal was methionine-
cysteine-threonine-proline-leucine-glycine.
[97] 2. Construction of T133C Human Granulocyte Colony Stimulating Factor Isoform
[98] T133C was constructed in order to provide a control as a known isoform or to
realize amino acid modification in the middle of the polypeptide of human granulocyte
colony stimulating factor. T133C, having cysteine substituting for threoninine, the 133
amino acid of natural human granulocyte colony stimulating factor, was obtained by
carrying out PCR using DNA of natural human granulocyte colony stimulating factor
as a template and primer pairs formed of N-term primer with T133C2 and T133C1
with C-term primer. As a result, it was possible to obtain two DNA fragments modified
with a codon corresponding to cysteine substituting for threonine, the 133r amino acid.
The two DNA fragments were further subjected to PCR by using primer pairs formed
of N-term primer and C-term primer to obtain modified genes of T133C, a human
granulocyte colony stimulating factor isoform having cysteine substituting for
threonine in the position of the 133r amino acid. Introduction into an expression vector
was performed in the same manner as described in the above 'Construction of -1C
Human Granulocyte Colony Stimulating Factor Isoform'.
[99] 3. Construction of-1C/T133C Human Granulocyte Colony Stimulating Factor
Isoform
[100] -1C/T133C was constructed in order to realize amino acid addition to the N-
terminal and amino acid modification in the middle of the polypeptide of human
granulocyte colony stimulating factor. To construct -1C/T133C, threoninine, the 133rd
amino acid of the preformed -1C human granulocyte colony stimulating factor isoform,
was substituted with cysteine. For this, PCR was performed by using DNA of the
preformed -1C human granulocyte colony stimulating factor isoform as a template and
primer pairs formed of N-term primer with T133C2 and T133C1 with C-term primer.
As a result, it was possible to obtain two DNA fragments modified with a codon cor-
responding to cysteine substituting for threonine, the 133 amino acid. The two DNA
fragments were further subjected to PCR by using primer pairs formed of N-term
primer and C-term primer to obtain modified genes of -1C/T133C, a human
granulocyte colony stimulating factor isoform having cysteine substituting for
threonine in the position of the 133r amino acid. Introduction into an expression vector
was performed in the same manner as described in the above 'Construction of -1C
Human Granulocyte Colony Stimulating Factor Isoform'.
[101] 4. Construction of 175C Human Granulocyte Colony Stimulating Factor Isoform
[102] To construct 175C, PCR was performed by using DNA of natural human

19
granulocyte colony stimulating factor as a template and primer pairs formed of N-term
primer and 175C primer. The amplified DNA fragment was further subjected to PCR
by using N-term primer and C-term primer so that restriction enzyme cleavage sites
(Xba I and BamH I) could be introduced, and then was introduced into an expression
vector. By doing so, it was possible to obtain genes of a human granulocyte colony
stimulating factor isoform having cysteine added to the N-terminal. Introduction into
the expression vector was performed in the same manner as described in the above
'Construction of -1C Human Granulocyte Colony Stimulating Factor Isoform'.
[103] The amino acid sequence of the C-terminal was -
alanine-glutamine-proline-cysteine.
[104] 5. Construction of 175C/T133C Human Granulocyte Colony Stimulating Factor
Isoform
[105] To construct 175C/T133C, threoninine, the 133r amino acid of the preformed -1C
human granulocyte colony stimulating factor isoform, was substituted with cysteine.
For this, PCR was performed by using DNA of the preformed 175C human
granulocyte colony stimulating factor isoform as a template and primer pairs formed of
N-term primer with T133C2 and T133C1 with C-term primer. As a result, it was
possible to obtain two DNA fragments modified with a codon corresponding to
cysteine substituting for threonine, the 133r amino acid. The two DNA fragments were
further subjected to PCR by using primer pairs formed of N-term primer and C-term
primer to obtain modified genes of 175C/T133C, a human granulocyte colony
stimulating factor isoform having cysteine substituting for threonine in the position of
the 133 amino acid. Introduction into an expression vector was performed in the same
manner as described in the above 'Construction of -1C Human Granulocyte Colony
Stimulating Factor Isoform'.
[106] The genes of the human granulocyte colony stimulating factor isoforms constructed
as described above were introduced into an expression vector so as to be expressed by
a lamda PR promoter. Constitutional elements of the expression vector are shown in
the following Table 2.
[107] Table 2

Origin
PR promoter NEB Ltd. lamda DNA#301-18
Self-replication machine pT7T319U
Antibiotic marker (Kanamycin) pACYC177
Par pSClOl
cl 857 represor NEB Ltd. lamda DNA#301-18

20

SD sequence Synthesized DNA
Transcription terminating site NEB Ltd. lamda DNA#301-18(toop)
[108]
[109] Expression and Purification of Human Granulocyte Colony
Stimulating Factor Isoforms
[110] Transformation of plasmid vectors containing genes of human granulocyte colony
stimulating factor isoforms into E. coli was performed to construct strains producing
human granulocyte colony stimulating factor proteins.
[Ill] To perform the transformation, the calcium chloride method generally known to
those skilled in the art was used. Herein, E. coil used as a host cell was currently
available E. coli (HB101, NM strain series, etc.).
[112] The transformed E. coli was inoculated into 5ml of a Luria Broth (LB) medium in a
15ml tube, and was incubated(cultured) overnight in an incubator at 30°C. The seed
cells culture was inoculated into a 500ml of Luria Broth (LB) medium in a 3,000ml
triangular flask and was incubated at 30°C. When the absorbance at 600nm reached
0.8-1.2, the incubator was heated to 42°C and the cells were further incubated for 4
hours to induce the expression of human granulocyte colony stimulating factor
isoforms.
[113] After the completion of the culture, the culture solution was subjected to
centrifugal separation to collect the cells, and the cells were washed with distilled
water. The washed cells were lysed in a microfluidizer (available from Microfluidics
Co.) under 12,000 psi four times, and inclusion bodies were recovered via centrifugal
separation, followed by washing three times with distilled water.
[114] The inclusion bodies were thoroughly suspended and dissolved in 10ml of 8M
urea, 50mM glycine (pH 11.0) at room temperature for at least 5 hours, and then the
suspension was subjected to centrifugal separation at 10,000rpm for 30 minutes. The
supernatant was collected to obtain a solution of human granulocyte colony stimulating
factor isoform protein. To the protein solution, glycine and distilled water were added
to control the glycine concentration and the urea concentration to 50mM and 2M, re-
spectively. The resultant protein solution was titrated to pH 9.0 and was subjected to
protein refolding overnight at 4°C.
[115] After the completion of the protein refolding, the solution was adjusted to pH 4.0,
diluted with distilled water by three times or more, and was loaded to a column packed
with 10ml of CM Sepharose FF (available from Amersham Biosciences), which have
been previously equilibrated by using 2mM HC1 in an amount of at least three times of
the column volume. After the completion of the loading, the column was equilibrated
by using 2mM HC1 in an amount of at least three times of the column volume, and

21
washed by using 50mM acetate sodium chloride buffer (pH 5.4) in an amount of at
least three times of the column volume. Then, human granulocyte colony stimulating
factor isoforms were eluted by using 50mM acetate sodium chloride buffer (pH 5.4)
containing 35mM sodium chloride in an amount of at least twelve times of the column
volume and were divided into 5ml fractions. The fractions were collected after the
HPLC analysis.
[116] The human granulocyte colony stimulating factor isoforms, purified by CM
Sepharose FF, were concentrated by using a Centricon Plus-80 filter unit with a
molecular weight of 10,000Da at 3000rpm for 25 minutes to obtain protein concentrate
having a final protein concentration of about 5mg/ml.
[117] The concentrated protein was loaded to a column packed with 300ml of Sephacryl
S-100 (Amersham Biosciences), which have been previously equilibrated by using
lOmM acetate sodium chloride buffer (pH 5.4) in an amount of at least three times of
the column volume. The buffer solution was eluted at a rate of lml/min. to remove
multimers of the human granulocyte colony stimulating factor isoforms contained in
the protein solution and to separate purified monomers, which, in turn, was used in the
subsequent polyethylene glycol addition reaction.
[118] Bonding of Polyethylene Glycol to Human Granulocyte Colony
Stimulating Factor Isoforms
[119] 1. Preparation of Branched mPEG (20.000Vmaleimide-G-CSF isoform
[120] 10 mg of each of the human granulocyte stimulating factor isoforms according to
Example 2 was introduced into 0.1M phosphate buffer (pH 7.0-8.0) and 200mg of
branched mPEG (20,000)-maleimide (available from NOF) was added thereto.
Reaction was performed at room temperature for 2 hours with stirring. The reaction
mixture was quenched by reducing pH to 3.0 with IN HC1 solution.
[121] 2. Preparation of Branched mPEG (40.000Vmaleimide-G-CSF Isoform
[122] 10 mg of each of the human granulocyte stimulating factor isoforms according to
Example 2 was introduced into 0.1M phosphate buffer (pH 7.0-8.0) and 400mg of
branched mPEG (40,000)-maleimide (available from NOF) was added thereto.
Reaction was performed at room temperature for 2 hours with stirring. The reaction
mixture was quenched by reducing pH to 3.0 with IN HC1 solution.
[123] Purification of Polymer of Human Granulocyte Colony Stimulating
Factor Isoform with Polyethylene Glycol
[124] The PEG-G-CSF polymer obtained from Example 4 was purified as follows.
[125] 1. CM Sepharose FF Chromatography
[126] After the completion of the reaction, each mPEG-G-CSF polymer was titrated to
pH 3.0, diluted with distilled water by at least 10 times, and loaded to a column packed
with 10ml of CM Sepharose FF (Amersham Bioscience), which have been previously

22
equilibrated by using 2mM HC1 in an amount of at least three times of the column
volume. After the loading, the column was equilibrated by using 2mM HC1 in an
amount of at least three times of the column volume, and washed by using 50 mM
acetate sodium chloride buffer (pH 5.0) in an amount of three times of the column
volume. In the cases of mPEG (20,000)-maleimide-G-CSF isoform and branched
mPEG (20,000)-maleimide-G-CSF isoform, mPEG-G-CSF polymers were eluted by
using 50 mM acetate sodium chloride buffer (pH 5.4) containing 40 mM sodium
chloride in an amount at least twelve times of the column volume. In the case of
Branched mPEG (40,000)-maleimide-G-CSF isoform, mPEG-G-CSF polymer was
eluted by using 50 mM acetate sodium chloride buffer (pH 5.4) containing 20 mM
sodium chloride in an amount at least twelve times of the column volume. The eluted
solution was fractionated into a 5 ml fraction and each fraction was collection after the
HPLC analysis.
[127] 2. Gel Permeation Chromatography
[128] The mPEG-G-CSF polymers, purified by CM Sepharose FF, were concentrated by
using a Centricon Plus-80 filter unit with a molecular weight of 10,000Da at 3000rpm
for 25 minutes to obtain protein concentrate having a final protein concentration of
about 5mg/ml.
[129] The concentrated protein was loaded to a column packed with 300ml of Sephacryl
S-100 (Amersham Biosciences), which have been previously equilibrated by using
lOmM acetate sodium chloride buffer (pH 5.4) in an amount of at least three times of
the column volume. The buffer solution was eluted at a rate of lml/min. to remove
mPEG-G-CSF polymers contained in the protein solution. Proteins were serially eluted
in the order of G-CSF multimer, PEG-G-CSF, G-CSF dimer and G-CSF monomer.
[ 130] Evaluation of Activity and in vivo lifetime of mPEG-G-CSF
Polymers in Rats
[131] The mPEG-G-CSF polymers prepared and purified according to Examples 4 and 5
were evaluated for the activity and in vivo lifetime by using 8-week aged rats (SD rats,
male, five rats per test sample). The test samples include: PBS carrier as a carrier, G-
CSF T133C20kDa(Br) prepared and purified according to Examples 1 to 4 (G-CSF
isoform obtained by substituting 133-Thr with cysteine and covalently bonding
branched PEG with 20 kDa to the same cysteine according to Korean Laid-Open
Patent No. 2002-0079778); commercially available Neulasta (Amgene Co.) having
20kDa PEG bound to the N-terminal; and the isoforms according to the present
invention, prepared and purified according to Examples 1 to 4, i.e. N+20kDa(Br),
(G-CSF isoform obtained by adding cysteine between the threonine and methionine of
the N-terminal (-1C) of G-CSF and covalently bonding branched PEG with 20 kDa to
the same cysteine); 175C+20kDa(Br) (G-CSF isoform obtained by adding cysteine

23
after 174-Pro of the C-terminal of G-CSF and covalently bonding branched PEG with
20 kDa to the same cysteine); and 175C+40kDa(Br) (G-CSF isoform obtained by
adding cysteine after 174-Pro of the C-terminal of G-CSF and covalently bonding
branched PEG with 40 kDa to the same cysteine). Each isoform was diluted with PBS
and administered to rats through the tail veins in a dose of lOOD/kg based on the body
weight measured upon the administration day. Administration was carried out only
once upon the test day.
[132] To provide analytic samples, 200D of blood was collected from the tail veins at a
predetermined time depending on the constitution of a test group. In the case of the
blank, blood sampling was performed before the administration. Sampling for PD/PK
analysis was performed 7, 24,48 and 72 hours after the administration. Neutrophils
were counted in each blood sample to determine the in vivo activity.
[133] Additionally, each blood sample was introduced into an EDTA anticoagulant tube
and subjected to centrifugal separation to separate blood plasma. G-CSF concentration
in the plasma was measured by using a human G-CSF ELISA kit (Quantokine, R&D
Systems). Each sample was diluted serially to the linear zone allowing detection of
concentration in the kit. Measurement was performed two times per sample. The
results are shown in the following Table 3.
[134] As can be seen from Table 3, the polyethylene glycol-bound human granulocyte
colony stimulating factor isoforms having cysteine added to the N-terminal or C-
terminal according to the present invention shows an increased in vivo activity and in
vivo lifetime. In particular, N+20kDA(Br) and 175C+20kDa(Br) show a significant
increase in activity from the point of 24 hours after the administration, when compared
to the control. Additionally, 175C+40kDa(Br) shows an increase in activity from the
point of 24 hours after the administration and a similar aspect of neutrophil production,
when compared to the control. However, 175C+40kDa (Br) can maintain its activity
until the time of 72 hours after the administration. Particularly, 175C+40kDa(Br)
provides a significantly improved lifetime, while 175C+20kDa(Br) and N+20kDa(Br)
provide a significantly improved activity.
[135] Table 3

Time Carrier Neulasta N+20kDa(
Br) T133C20k
Da(Br) 175C+20kD
a(Br) 175C+40kD
a(Br)
0 1.31 0.94 0.87 1.03 1.19 1.52
24 1.77 12.5 12.78 11.33 12.8 11.5
48 0.99 12.5 16.51 6.84 17.25 10.92
72 1.43 2.31 2.35 4.21 4.32 9.83

24
[136]
Industrial Applicability
[137] As can be seen from the foregoing, the human granulocyte colony stimulating
factor isoforms according to the present invention provide a significantly increased in
vivo lifetime and can be used for treating or preventing various human granulocyte
colony stimulating factor-related diseases.
[138] While this invention has been described in connection with what is presently
considered to be the most practical and preferred embodiment, it is to be understood
that the invention is not limited to the disclosed embodiment and the drawings. On the
contrary, it is intended to cover various modifications and variations within the spirit
and scope of the appended claims.
[139]

25
Claims
[1] A granulocyte colony stimulating factor isoform, which includes at least one
amino acid added before Thr of N-terminal or after Pro of C-terminal in human
granulocyte colony stimulating factor represented by amino acid Sequence No. 1,
wherein at least one of the added amino acids is cysteine, and polyethylene
glycol is bound to at least one of the added cysteine.
[2] The granulocyte colony stimulating factor isoform according to claim 1, wherein
1~12 amino acids are added to each terminal of human granulocyte colony
stimulating factor.
[3] The granulocyte colony stimulating factor isoform according to claim 1, wherein
one cysteine residue is added to human granulocyte colony stimulating factor.
[4] The granulocyte colony stimulating factor isoform according to claim 3, wherein
the polyethylene glycol is a branched polyethylene glycol and has a molecular
weight of 20kDa~40kDa.
[5] The granulocyte colony stimulating factor isoform according to claim 4, wherein
the cysteine is 175C added to the C-terminal.
[6] The granulocyte colony stimulating factor isoform according to claim 1, wherein
at least one amino acid of human granulocyte colony stimulating factor is
substituted with another amino acid before polyethylene glycol is bound thereto.
[7] A pharmaceutical composition comprising the human granulocyte colony
stimulating factor isoform as defined in any one of claims 1 to 6 and pharma-
ceutically acceptable carriers.
[8] Genes encoding proteins of the human granulocyte colony stimulating factors
isoform having cysteine added to the N-terminal or the C-terminal thereof as
defined in any one of claims 1 to 6.
[9] An oligodeoxynucleotide used as a primer for modifying the amino acid
sequence of human granulocyte colony stimulating factor into the human
granulocyte colony stimulating factor isoform as defined in any one of claims 1
to 6.
[10] The oligodeoxynucleotide according to claim 9, wherein the primer is an
oligodeoxynucleotide represented by any one of Sequence Nos. 2 to 7.

Disclosed is a novel physiologically active protein, which is a human granulocyte colony stimulating factor isoform, constructed in order to increase the in vivo lifetime of human granulocyte colony sedtimulating factor. The human granulocyte colony stimulating factor isoform comprises a polypeptide and polyethylene glycol (PEG) bound thereto as a non-protein polymer. A specific site of the polypeptide is selected so that polyethylene can be bound to the site while not adversely affecting the activity of the protein. The amino acid of the site is modified with cysteine and polyethylene glycol is bound to the modified site. A pharmaceutical composition comprising the isoforms, genes encoding the isoforms, and a primer for modifying the amino acid sequence are also disclosed.