AU2020204470B1

AU2020204470B1 - Dual-function protein for lipid and blood glucose regulation

Info

Publication number: AU2020204470B1
Application number: AU2020204470A
Authority: AU
Inventors: Zhao DONG; Xiong FENG; Shixiang JIA; Qiang Li; Jiyu Zhang; Chi Zhou
Original assignee: Ampsource Biopharma Shanghai Inc
Current assignee: Ampsource Biopharma Shanghai Inc
Priority date: 2020-07-03
Filing date: 2020-07-03
Publication date: 2021-05-27
Anticipated expiration: 2040-07-03

Abstract

The present disclosure relates to a dual-function protein for regulating blood glucose and lipid metabolism, wherein said dual-function protein comprises a human GLP-1 analog and human FGF21. In the present disclosure, provided is a method for preparing said dual function protein, and also provided is the use of said dual-function protein in the preparation of a biological substance for treating type 2 diabetes, obesity, dyslipidemia, fatty liver disease and/or metabolic syndrome. The dual-function protein provided in the present disclosure can synergistically regulate blood glucose and lipid levels in vivo, and satisfy multiple requirements for patients with type 2 diabetes such as lowering blood glucose, relieving hepatic steatosis, reducing body weight and improving metabolic disorders of circulating lipids.

Description

DUAL-FUNCTION PROTEIN FOR LIPID AND BLOOD GLUCOSE REGULATION

The present disclosure relates to a GLP-1-FGF21 dual-function protein and related

pharmaceutical combination, and also relates to the use of said dual-function protein for

preparing a medicament for treating type 2 diabetes, obesity, hyperlipidaemia, fatty liver disease

and/or metabolic syndrome, and treatment of these diseases using the medicament.

Glucagon-like peptide-1 (GLP-1) is an endocrine peptide consisting of 36 amino acids

secreted by mammalian intestinal L cells, and stimulates insulin secretion from pancreatic beta

cells in a glucose dependent manner by binding to and activating GLP-1 receptor (GLP-R),

inhibits glucagon release from pancreatic alpha cells to maintain normal glucose level. In

addition, it inhibits gastrointestinal movement and suppresses appetite (Knudsen LB, J Med

Chem, 2004, 4128-4134). Native human GLP-1 is easily inactivated in vivo by dipeptidyl

peptidase IV (DDP-IV), and has a short circulating half-life. Exendin-4 is isolated from the

saliva of toxic lizards from South Africa and has 39 amino acids with 53% homology to human

GLP-1 and exerts a similar biological activity. Compared with the human GLP-1, Gly replaces

Ala at the second position of N-terminus of Exendin-4 which enhances resistance of the peptide

to the proteolytic degradation induced by DDP-IV and extends the circulating half-life in vivo.

Exendin-4 has a special Trap-cage structure at the C-terminus, such that its binding affinity with

GLP-1 receptor is significantly higher than that of human GLP-1 (Neidigh JW et al.,

Biochemistry, 2001, 40:13188-13200). At the equimolar concentration, Exendin-4 exhibits a

stronger effect on promoting insulin release from pancreatic beta cells. Although there is slight

difference in structure and amino sequence, both the blood glucose metabolic regulators have

been marketed, wherein most predominant ones are Liraglutide of Novo Nordisk and Exenatide

of AstraZeneca. Treatment with Liraglutide or Exenatide 2-3 times daily can effectively control

the blood glucose level in type 2 diabetes patients. However, the high injection frequency results in high cost and poor clinical compliance for patients. In order to prolong the in vivo half-life and bioavailability of GLP-1 and Exendin-4 analogs, Fc or HSA fusion technologies have been used for developing long-acting drugs. At present, marketed products are Dulaglutide of Eli Lilly and

Albiglutide of GlaxoSmithKline. The most extensively used one in clinic is Dulaglutide, a

GLP-1 hIgG4 Fc fusion protein (Dulaglutide), wherein its average half-life is up to 90 hours

(Chinese patent CN 1802386 B), its clinical indication is type 2 diabetes, and the recommended

dose regimen is subcutaneous injection once a week. Clinical study showed that Dulaglutide

could effectively control postprandial blood glucose and glycosylated hemoglobin of diabetic

patients, and lower the body weight of obese patients by inhibiting appetite. However, varying

degrees of gastrointestinal adverse effects were observed. Epidemiological investigation showed

that the large percentage of patients with type 2 diabetes accompanied with nonalcoholic fatty

liver disease and lipid metabolism disorder (Radaelli MG et al., J Endocrinol Invest, 2017,

s40618). However, no clinical study demonstrated that human GLP-1 or Exendin-4 analogs have

the effect of treating fatty liver and hyperlipidaemia independent of weight loss (Petit JM,

Diabetes Metab, 2017, 43, 2S28-2S33). Therefore, GLP-1 products cannot completely satisfy all

clinical needs for patients with type 2 diabetes.

The family of fibroblast growth factors (FGFs) has 22 members and 7 subfamilies, wherein

the FGF19 subfamily exerts physiological activity in an endocrine manner, involves the

regulation and control of energy and cholic acid homeostasis, glucose and lipid metabolism, and

phosphate and vitamin D homeostasis (Moore DD et al., Science, 2007, 316:1436-1438 and

Beenken et al., Nature Reviews Drug Discover, 2009, 8:235). FGF21 is a member of FGF19

subfamily, and has 181 amino acids. The C-terminus of FGF21 binds first to a co-factor -Klotho

transmembrane protein, induces FGFR binding to the N-terminus of FGF21, then forms a stable

FGF21/j-Klotho/FGFR complex, which trigger subsequent signaling pathway in vivo (Yie J et

al., FEBS Lett, 2009, 583(1):19-24 and Micanovic R et al., J Cell Physiol, 2009, 219(2):

227-234). Under physiological conditions, FGF21 is to promote glucose utilization independent of insulin (Kharitonenkov A et al., J Clin Invest, 2005, 115(6): 1627-1635), to enhance insulin sensitization (Duthchak PA et al., Cell, 2012, 148, 387-393), to inhibit de novo lipogenesis and promote the fatty acid p-oxidation in liver, to decrease serum triglyceride level (Xu J et al.,

Diabetes, 2009, 58, 250-259). In addition, FGF21 could decrease total cholesterol and low

density lipoprotein-cholesterol contents in serum by inhibiting liver SREBP-2 synthesis to

relieve hypercholesteremia (Lin Z et al., Circulation, 2015, 131, 1861-1871).

In conclusion, FGF21 exerts multiple regulatory functions on metabolic diseases, such as

obesity, type 2 diabetes, nonalcoholic fatty liver and hyperlipidaemia. Meanwhile, FGF21 is the

only discovered member without any mitogenic effect in this superfamily, which greatly reduces

potential carcinogenicity risk in clinical applications (Wu X et al., Proc Natl Acad Sci USA,

2010, 170: 14158-14163). However, due to its unstable physicochemical property, native FGF21

does not possess druggability so far due to the following reasons: (1) native FGF21 protein has

pool stability and is easily degraded by proteases in vivo; (2) FGF21 conformation is unstable

with ease of aggregation, which increases difficulty in scale-up production; (3) native FGF21 has

short circulating half-life, about 0.5-1 h in mice and 2-3 h in Cynomolgus monkeys

(Kharitonenkov A et al., J Clin Invest, 2005, 115: 1627-1635). Various long-acting

protein-engineering technologies are commonly used for prolonging the in vivo half-life of

recombinant FGF21. For example, conjugation of FGF21 and PEG molecule increases the

molecular weight, lowers the glomerular filtration rate, and prolongs the in vivo retention time

(see WO 2005/091944, WO 2006/050247, WO 2008/121563 and WO 2012/066075); FGF21

fuses to long chain fatty acid (which can binds to serum albumin) (see WO 2010/084169 and

WO 2012/010553); preparation of an agonist antibody which specifically binds to FGFR or

FGFR/-klotho complex to mimic the mechanism of FGF21, and to activate FGF/FGFR

signaling pathway (see WO 2011/071783, WO 2011/130417, WO 2012/158704 and WO

2012/170438); FGF21 fuses to Fc fragment to improve half-life (see WO 2004/110472, WO

2005/113606, WO 2009/149171, WO 2010/042747, WO 2010/129503, WO 2010/129600, WO

2013/049247, WO 2013/188181 and WO 2016/114633). At present, there is no marketed drug of

long-acting FGF21 protein, but there are three long-acting FGF21 candidates in clinic trials,

LY2405319 of Eli Lilly, PF-05231023 of Pfizer and BMS986036 of Bristol-Myers Squibb. In the

clinical trials, for patients with type 2 diabetes, LY2405319 and PF-05231023 had weight loss

effect and decreased serum TG level, but had no positive therapeutic effect on blood glucose

(Gaich G et al., Cell Metab, 2013, 18:333-340 and Dong JQ et al., Br J Clin Pharmacol, 2015,

-1051-1063). BMS986036 exhibited a good therapeutic effect on nonalcoholic fatty liver, but

there was no experimental study on blood glucose control for patients with type 2 diabetes.

Above-mentioned results showed that although the use of long-acting FGF21 protein alone can

exert many pharmacodynamic activities such as in body weight, nonalcoholic fatty liver and

hyperlipidaemia. However, it cannot satisfy the blood glucose control requirement which is

crucial in the treatment of patients with type 2 diabetes.

Recently, some studies reported that the combination of GLP-1 and FGF21 has a synergistic

effect on blood glucose control. For example, CN 102802657 A disclosed that the combination of

GLP-1 and FGF21 can synergistically lower the blood glucose level in db/db mice. However, the

combined usage of drugs not only increases the administration frequency for patients and

reduces the patient compliance, but also greatly increases treatment costs. In addition, a

dual-function protein prepared by fusing GLP-1 and FGF21 was also reported. In order to solve

the issue that FGF21 is easily degraded in vivo, scientists generally introduce point mutations in

native FGF21 molecule, but this inevitably increases the potential immunogenicity of the

dual-function protein (WO 2017/074123 and CN 104024273 B). Furthermore, the reported

synergistic effect of FGF21 and GLP-1 generally exhibited in terms of blood glucose control, but

their synergistic effects in terms of other metabolic diseases, such as obesity, nonalcoholic fatty

liver and lipid metabolism disorder are not investigated by comparing with marketed long-acting

GLP-1 analogs. The reasons for lack of above-mentioned investigations may comprise the

following: (1) neither native GLP-1 nor FGF21 is stable in vivo, and defection of structural integrity and stability in any molecule will eliminate the synergistic effect; (2) in the process of fusion of GLP-1 and FGF21 into a single protein, their respective three-dimensional conformation needs to be maintained to the maximum extent for preventing mutual interference, such that the functional synergy will be achieved, and this should be carefully considered when designing the molecule; (3) the functions of both GLP-1 and FGF21 depend on binding to their respective receptors, and it needs to be confirmed by a number of in vitro and in vivo experiments to clarify the conditions under which the dynamic equilibrium can be achieved among them. So far, there is no such report in published patents or other non-patent documents.

In conclusion, if a GLP-1-FGF21 dual-function protein drug that has enhanced stability,

prolonged half-life and low immunogenicity can be developed in the art, then the multiple

requirements of patients with type 2 diabetes for reducing blood glucose, relieving hepatic

steatosis, reducing body weight and improving metabolic disorders of circulating lipids can be

met.

Summary of the Disclosure

The present disclosure provides a dual-function protein having a synergistic effect in terms

of blood glucose and lipid regulations and comprising human GLP-1 analog and human FGF21,

the preparation method therefor and the use thereof. The present disclosure solves the issues such

as the defects relating to unstable structure and short in vivo half-life of native GLP-1 or FGF21,

retains the strong hypoglycemic effect of GLP-1 and the physiological effects of FGF21 on

insulin sensitization, weight loss, fatty liver and hypercholesteremia treatment, and relieves

gastrointestinal adverse effects caused by GLP-1 to some extent.

In one embodiment of the present disclosure, a dual-function protein can synergistically

regulate blood glucose and lipids, wherein said dual-function protein comprises human

glucagon-like peptide-1 analog (abbreviated as GLP-1 analog hereafter), linker peptide 1

(abbreviated as LI hereafter), human fibroblast growth factor 21 (abbreviated as FGF21 hereafter), linker peptide 2 (abbreviated as L2 hereafter) and human immunoglobulin Fc fragment (abbreviated as Fc fragment hereafter) sequentially from the N-terminus to the

C-terminus; wherein the linker peptide 1 comprises a flexible peptide; the linker peptide 2

comprises a flexible peptide and rigid peptide, the rigid peptide comprises at least 1 rigid unit,

and the rigid unit comprises carboxyl terminal peptide of human chorionic gonadotropin

p-subunit or a truncated sequence thereof. In one embodiment of the present disclosure, a dual-function protein can synergistically

regulate blood glucose and lipids, wherein said dual-function protein consists of sequentially

human glucagon-like peptide-1 analog (abbreviated as GLP-1 analog hereafter), linker peptide 1

(abbreviated as LI hereafter), human fibroblast growth factor 21 (abbreviated as FGF21

hereafter), linker peptide 2 (abbreviated as L2 hereafter) and human immunoglobulin Fc

fragment (abbreviated as Fc fragment hereafter) from the N-terminus to the C-terminus; wherein

the linker peptide 1 consists of a flexible peptide; the linker peptide 2 consists of a flexible

peptide and rigid peptide, the rigid peptide consists of at least 1 rigid unit, and the rigid unit

comprises carboxyl terminal peptide of human chorionic gonadotropin p-subunit or a truncated

sequence thereof.

In the present disclosure, said "human GLP-1 analog" refers to an analog, fusion peptide, or

derivative which are obtained by substituting, deleting or adding one or more amino acid

residues on the amino acid sequence of human GLP-1 (as shown in SEQ ID NO: 1) and

maintains human GLP-1 activity. For example, said GLP-1 analog comprises but is not limited to

the amino acid sequences as shown in SEQ ID NO: 2, 3, 4 or 5 in the sequence listing. All

sequences in the sequencing listing are incorporated herein in their entireties. In at least one

embodiment of the present disclosure, said GLP-1analog is shown in SEQ ID NO: 2, and in

another embodiment, said GLP-1 analog is shown in SEQ ID NO: 5.

In the present disclosure, said "linker peptide 1 (L)" is a short peptide between GLP-1

analog and FGF21 and has connecting function. In at least one embodiment, said linker peptide 1 is non-immunogenic, and generates enough distal distance between GLP-1 analog and FGF21, such that minimal steric hindrance effect is present, which does not affect or not affect severely correct folding and spatial conformation of GLP-1 analog and FGF21. A person skilled in the art can design linker peptides according to conventional methods in the art. In at least one embodiment, a flexible peptide comprising 2 or more amino acids is used, and the amino acids are selected from the following amino acids: Gly(G), Ser(S), Ala(A) and Thr(T); in at least one embodiment, said linker peptide 1 comprises G and S residues. The length of the linker peptide is very important for the activity of the dual-function protein, and in at least one embodiment, the linker peptide consists of 5-30 amino acids. In a at least one embodiment of the present disclosure, the amino acid sequence of said linker peptide 1 is GGGGGGGSGGGGSGGGGS.

In the present disclosure, said "FGF21" comprises the sequence as shown in SEQ ID NO: 6

in which the secreting leader signal of amino acid position 1-28 is deleted; or comprises the

isoform sequence of SEQ ID NO: 6 in which the secreting leader signal of amino acid position

1-28 is deleted and which has G141S or L174P substitution. In a at least one embodiment of the

present disclosure, said FGF21 comprises the amino acid sequence as shown in SEQ ID NO: 6 in

which the secreting leader signal of amino acid position 1-28 is deleted and has L174P

substitution.

In the present disclosure, said "linker peptide 2 (L2)" is a short peptide between FGF21 and

Fc fragment and having connect function. Said linker peptide consists of a flexible peptide and a

rigid peptide, wherein said flexible peptide comprises 2 or more amino acid residues which are

selected from Gly(G), Ser(S), Ala(A) and Thr(T). In at least one embodiment, said flexible

peptide comprises G and S residues. With regard to the present disclosure, preferably, the general

structural formula of the amino acid composition of said flexible peptide is

(GS)a(GGS)b(GGGS)c(GGGGS)d, wherein a, b, cand d are integers greater than or equal to 0,

and a + b + c + d > 1.

In some embodiments of the present disclosure, said flexible peptide comprised in said L2

is selected from the following sequences:

(i) GGGGS;

(ii) GSGGGSGGGGSGGGGS;

(iii)GSGGGGSGGGGSGGGGSGGGGSGGGGS;

(iv)GGGGSGGGGSGGGGSGGGGSGGGGSGGGGSGGGGS;

(v)GGGSGGGSGGGSGGGSGGGS;

(vi) GGSGGSGGSGGS.

In the present disclosure, said rigid peptide constituting said linker peptide 2 (L2) consists

of one or more rigid units, and said rigid units are selected from a full-length or truncated

sequence consisting of carboxyl terminal amino acids 113 to 145 of human chorionic

gonadotropin p-subunit (known as CTP rigid unit hereafter); specifically, said CTP rigid unit

comprises the amino acid sequence as shown in SEQ ID NO: 7 or the truncated sequences

thereof.

In at least one embodiment, said CTP rigid unit comprises at least 2 glycosylation sites; for

example, in one at least one embodiment of the present disclosure, said CTP rigid unit comprises

2 glycosylation sites, for example, said CTP rigid unit comprises 10 amino acids of N-terminus

of SEQ ID NO: 7, i.e. SSSS*KAPPPS*; or said CTP rigid unit comprises 14 amino acids of

C-terminus of SEQ ID NO: 7, i.e. S*RLPGPS*DTPILPQ; for another example, in another

embodiment, said CTP rigid unit comprises 3 glycosylation sites, for example, said CTP rigid

unit comprises 16 amino acids of N-terminus of SEQ ID NO: 7, i.e. SSSS*KAPPPS*LPSPS*R;

for another example, in another embodiment, said CTP rigid unit comprises 4 glycosylation sites,

for example, said CTP rigid unit comprises 28, 29, 30, 31, 32 or 33 amino acids and starts at

positions 113, 114, 115, 116, 117 or 118 and terminates at position 145 of human chorionic

gonadotropin p-subunit. Specifically, said CTP rigid unit comprises 28 amino acids of

N-terminus of SEQ ID NO: 7, i.e. SSSS*KAPPPS*LPSPS*RLPGPS*DTPILPQ. In the present disclosure, * represents glycosylation sites. All possibilities represent independent embodiments of the present disclosure.

In some embodiments, the CTP rigid unit comprised in L2 of the present disclosure can

preferably comprise one of the following sequences:

(i) CTP1 : SSSSKAPPPSLPSPSRLPGPSDTPILPQ;

(ii) CTP2 : PRFQDSSSSKAPPPSLPSPSRLPGPSDTPILPQ;

(iii) CTP: SSSSKAPPPS;

(iv) CTP 4 : SRLPGPSDTPILPQ;

(v) CTP 5 : SSSSKAPPPSLPSPSR.

In some embodiments, the CTP rigid unit provided in the present disclosure has at least 70%

homology to the native CTP amino acid sequence; In some embodiments, the CTP rigid unit

provided in the present disclosure has at least 80% homology to the native CTP amino acid

sequence; In some embodiments, the CTP rigid unit provided in the present disclosure has at

least 90% homology to the native CTP amino acid sequence; In some embodiments, the CTP

rigid unit provided in the present disclosure has at least 95% homology to the native CTP amino

acid sequence.

In some embodiments of the present disclosure, L2 comprises 2, 3, 4 or 5 above-mentioned

CTP rigid units. In some embodiments of the present disclosure, L2 of said dual-function protein

comprises 2 CTPr igid unit: SSSSKAPPPSSSSSKAPPPS (CTP3-CTP3, or represented as

(CTP3 )2).

In the present disclosure, said "Fc fragment" is selected from the Fc fragments of human

immunoglobulins IgG, IgM, IgA and variants thereof; in at least one embodiment, is selected

from the Fc fragments of human IgGI, IgG2, IgG3 or IgG4 and variants thereof, wherein said

human IgG Fc fragment (represented as vFc) comprises at least one amino acid modification

located in wild type human IgG Fc, and the Fc variants have non-lytic characteristics, and show

an extremely minimal Fc-mediated effector functions (ADCC and CDC functions) and/or enhanced binding affinity with FcRn receptor; most preferably, human IgG Fc variant is selected from the group of:

(i) vFcyl: hinge, CH2 and CH3 regions of human IgG containing Leu234Val, Leu235Ala

and Pro331Ser mutations (the amino acid sequence as shown in SEQ ID NO: 8);

(ii) vFcy2-1: hinge, CH2 and CH3 regions of human IgG2 containing Pro331Ser mutation

(the amino acid sequence as shown in SEQ ID NO: 9);

(iii) vFcy2-2: hinge, CH2 and CH3 regions of human IgG2 containing Thr250Gln and

Met428Leu mutations (the amino acid sequence as shown in SEQ ID NO: 10);

(iv) vFcy2-3: hinge, CH2 and CH3 regions of human IgG2 containing Pro331Ser,

Thr250Gln and Met428Leu mutations (the amino acid sequence as shown in SEQ ID NO: 11).

(v) vFcy4: hinge, CH2 and CH3 regions of human IgG4 containing Ser228Pro and

Leu235Ala mutations (the amino acid sequence as shown in SEQ ID NO: 12).

The Fc variants provided by the present disclosure comprises, but is not limited to above 5

variants of (i) to (v), and also can be the combination or overlap of functional variants among

same IgG subtypes, for example, the variant of above-mentioned (iv) is a new combination

variant of IgG2 Fc obtained by overlapping the mutation sites in (ii) and (iii).

The Fc variant (vFc) in the dual-function protein of the present disclosure contains human

IgG, such as the hinge region and CH2 and CH3 regions of human IgGI, IgG2 and IgG4. Such

CH2 region contains amino acid mutations at positions 228, 234, 235 and 331 (determined by

EU numbering system). It is believed that these amino acid mutations can reduce Fc effector

function. Human IgG2 does not bind to FcyR, but shows a very weak complement activity. The

complement activity of Fcy2 variant having Pro331Ser mutation should be lower than that of

native Fcy2, and is still an FcyR non-binder. IgG4 Fc has some defects in activation of

complement cascade, and its binding affinity with FcyR is lower than that of IgGI by about one

order of magnitude. Compared with native Fcy4, the Fcy4 variant having Ser228Pro and

Leu235Ala mutations should show the minimal effector function. Compared with native Fcyl,

Fcyl having Leu234Val, Leu235Ala and Pro331Ser mutations also shows a reduced effector

function. These Fc variants are more suitable for preparing a dual-function protein of FGF21 and

analogs thereof than native human IgG Fc. However, positions 250 and 428 (positions

determined by EU numbering system) contain amino acid substitutions, such that the binding

affinity of Fc region with neonate receptor FcRn is increased, thus further prolonging the

half-life (Paul R et al., J Biol Chem, 2004, 279:6213-6216); the combination or overlap of

above-mentioned two types of functional variants obtains new variants which have reduced

effector function and prolonged half-life. The Fc variants of the present disclosure comprises, but

is not limited to above-mentioned mutations; the substitutions at other sites can also be

introduced, such that Fc has a reduced effector function and/or enhanced affinity with FcRn

receptor, at the same time, without causing reduced functions/activities of Fc variants or adverse

conformational changes, and see Shields RL et al., J Biol Chem, 2001, 276(9):6591-604 for

common mutation sites.

In one at least one embodiment of the present disclosure, the amino acid sequence of said

dual-function protein is shown in SEQ ID NO: 13. In another at least one embodiment of the

present disclosure, the amino acid sequence of said dual-function protein is shown in SEQ ID

NO: 15.

The dual-function protein of the present disclosure is glycosylated; preferably, said

dual-function protein is glycosylated by being expressed in mammalian cells; in at least one

embodiment, said dual-function protein is glycosylated by being expressed in Chinese hamster

ovary cells.

According to another embodiment of the present disclosure, provided is a DNA encoding

the above-mentioned dual function protein. In one at least one embodiment of the present

disclosure, the DNA sequence encoding said dual-function protein is shown in SEQ ID NO: 14.

According to still another embodiment of the present disclosure, provided is a vector. The

vector comprises the above-mentioned DNA.

According to still another embodiment of the present disclosure, provided is a host cell. The

host cell comprises the above-mentioned vector, or is transfected with the above-mentioned

vector.

In a particular embodiment of the present disclosure, the host cell is a CHO-derived cell

strain DXB-11.

According to still another embodiment of the present disclosure, provided is a

pharmaceutical composition. The pharmaceutical composition comprises a pharmaceutically

acceptable carrier, excipient or diluent, and an effective amount of the above-mentioned

synergistic dual function protein.

According to another embodiment of the present disclosure, provided is a method for

preparing or producing said dual-function protein from mammalian cell lines (such as a

CHO-derived cell line), comprising following steps:

(a) introducing a DNA encoding the above-mentioned dual-function protein into a

mammalian cell;

(b) screening a high-yield cell strain expressing more than 20 g/10 6 cells within a period of

every 24 hours in the growth medium thereof from step (a);

(c) culturing the screened cell strain in step (b);

(d) harvesting the fermentation broth obtained from step (c), and purifying the dual function

protein.

In at least one embodiment, said mammalian cell in step (a) is CHO cell; in at least one

embodiment, said mammalian cell in step (a) is CHO-derived cell line DXB-11.

According to still another embodiment of the present disclosure, provided is the use of said

dual-function protein in the preparation of a drug for treating FGF21 related diseases and GLP-1

related diseases, and other metabolic, endocrinic and cardiovascular diseases, comprising obesity,

types 1 and 2 diabetes, pancreatitis, dyslipidemia, nonalcoholic fatty liver disease, nonalcoholic

steatohepatitis, insulin tolerance, hyperinsulinemia, glucose intolerance, hyperglycemia, metabolic syndrome, acute myocardial infarction, high blood pressure, cardiovascular disease, atherosclerosis, peripheral arterial disease, stroke, cardiac failure, coronary heart disease, nephropathy, diabetic complication, neuropathy, gastroparesis, and conditions associated with the severe inactivation mutations of insulin receptor; preferably, said diseases comprise obesity, types 1 and 2 diabetes, dyslipidemia, nonalcoholic fatty liver disease, nonalcoholic steatohepatitis and metabolic syndrome.

Compared with existing products, the dual-function protein of the present disclosure has

many advantages, which are demonstrated in detail by using, e.g., dual-function protein FP4I-2

of the present disclosure:

1. The half-life in vivo is prolonged, and blood glucose-lowering activity in vivo is

maintained for a longer period of time. The glucose tolerance test performed on C57BL/6 mice

shows that at 144 h after a single dose of FP4I-2, FP4I-2 still exhibits a good ability for

promoting glucose utilization, which is better than the commercially available GLP-1 analog

Liraglutide and Exenatide and native FGF21. Compared with GLP-1-Fc fusion protein

Dulaglutide, FP4I-2 exhibits a more excellent blood glucose control effect in the type 2 diabetes

mice.

2. Improved safety and tolerability. Dulaglutide induces severe gastrointestinal adverse

effects after the first administration. In db/db mice, the initial 24 hour food intake of the mice

after first administration of FP4I-2 is significantly elevated relative to that of mice administrated

with Dulaglutide, which shows that the dual-function protein FP4I-2 can effectively relieve

appetite inhibition induced by gastrointestinal adverse effects.

3. Improved therapeutic effects of dual-function protein on fatty liver. Compared with

Dulaglutide, FP4I-2 can significantly reduce the liver mass of db/db mice, improve liver function,

and the mechanism does not completely depend on the appetite inhibition related to

GLP-lanalog, demonstrating the physiological activity of FGF21. Relative to native FGF21, the in vivo half-life of FP4I-2 is significantly prolonged. In the animal model, the dose frequency of

FP4I-2 is twice per week, which will improve the clinical feasibility.

4. The C-terminal sequence of FGF21 is crucial to its activity, most dual-proteins

reported in the prior art uses FGF21 N-terminal fusion, and free C-terminus is beneficial to

maintain activity; however, the C-terminus of FGF21 also contains various protease hydrolysis

sites, and is very easily degraded; exposed intact C-terminus is more easily attacked by protease

and degraded. In order to overcome this problem, the prior art avoids using native FGF21, but

introducing corresponding mutations to improve its stability, however, this inevitably increases

the potential immunogenicity of dual-function proteins. In contrast, the dual-function protein

constructed in the present disclosure uses native FGF21, and has an Fc fragment connected at its

C-terminus. The dual-function protein provided by the present disclosure not only has a

significantly prolonged in vivo half-life in circulation, but also has a synergistic effect in terms of

blood glucose and lipid regulations, which suggests that the constructed dual-function protein

well maintains the properties of this two active molecules, and has good stability. This benefits

result from the new type of linker peptide among FGF21 and Fc variants; the linker peptide

consists of a flexible peptide and a rigid peptide, and the CTP rigid unit contains multiple

-carbohydrate side chains, can forms a relatively stable three-dimensional conformation, which

can effectively separate FGF21 and Fc, thereby lowers the steric hindrance effect caused by Fc

fragment to the utmost extent and keep relatively good FGF21 biological activity. In addition,

carbohydrate side chain of CTP rigid unit can mask the enzymolysis site of FGF21. The

protective effect lowers the sensibility of enzymolysis to proteases and achieves the purpose for

protein stability.

5. Mutations on Fc only retains the long half-life property in circulation of Fc, reducing

or eliminating ADCC and CDC effects (such as P331S), thus increases the safety of drug use. In

addition, Fc variants (such as T250Q/M428L) having an enhanced binding affinity with neonate

receptor (FcRn) can further prolong the half-life of dual function protein.

DETAILED DESCRIPTION OF THE DISCLOSURE

Human GLP-1 analog

The term "human GLP-1 analog" used herein refers to an analog, fusion peptide, and

residues on the amino acid sequence of human GLP-1 (as shown in SEQ ID NO: 1) and maintain

human GLP-1 activity. For example, said human GLP-1 analog comprises but are not limited to

the amino acid sequences as shown in SEQ ID NO: 2, 3, 4 or 5 in the sequence listing.

Human FGF21

The term "human FGF21" used herein refers to a wild type human FGF21 polypeptide.

The sequence of the wild type FGF21 protein can be obtained from UNIPROT database,

and the accession number is Q9NSA1. The precursor protein consists of 209 amino acids,

comprising a signal peptide (amino acids 1-28) and a mature protein (amino acids 29-209).

US 2001012628 Al teaches the isoform or allelic form of the wild type FGF21 having the

substitution from Leu to Pro in the mature protein (in the present disclosure, as shown in

positions 47-227 of SEQ ID NO: 13); another isoform of wild type FGF21 having the

substitution from Gly to Ser (Gly at the position 141 of SEQ ID NO: 6 is substituted or replaced

by Ser).

WO 2003/011213 teaches another isoform (see SEQ ID NO: 2 disclosed in WO

2003/011213, which has a signal peptide of 27 amino acid residues) having a relatively short

signal peptide (in the present disclosure, Leu at position 23 of SEQ ID NO: 6 deleted).

In the present disclosure, the wild type FGF21 comprises SEQ ID NO: 6 and the sequence

of the mature protein portion (amino acids 29-209) of the isoform having L174P or G141S

substitutions after removing the leader peptide; in addition, also comprised is the full-length

sequence of the precursor protein with the above-mentioned 27 or 28 amino acid signal peptide

added before those above sequences.

hCG-p carboxyl terminal peptide (CTP)

CTP is a short carboxyl terminal peptide of human chorionic gonadotropin (hCG) p-subunit.

Four reproduction-related polypeptide hormones follicle-stimulating hormone (FSH), luteinizing

hormone (LH), thyroid stimulating hormone (TSH) and chorionic gonadotropin (hCG) contain

the same alpha subunit and the different specific beta subunit from each other. The in vivo

half-life of hCG, compared with other three hormones, is prolonged, which is mainly due to the

specific carboxyl terminal peptide (CTP) of its beta subunit (Fares FA et al., Proc Natl Acad Sci

USA, 1992, 89: 4304-4308). Native CTP contains 37 amino acid residues, has 4 0-glycosylation

sites, and has sialic acid residue at the terminus. Negative charged, highly sialylated CTP can

resist the clearance of kidneys on same, thus prolong the in vivo half-life of same. However, the

inventors herein creatively connects a rigid peptide including at least one CTP rigid unit with a

flexible peptide with a suitable length, collectively as linker peptide 2 for connecting FGF21 and

Fc fragment.

N-terminal and C-terminal sequences of FGF21 are crucial to the functions of FGF21.

The spatial conformation of FGF21 is complex and fragile, such that FGF21 has a poor stability,

is easily degraded and aggregated; if FGF21 is fused to a ligand, the steric hindrance effect will

interfere with the correct folding of FGF21, making the activity of FGF21 significantly lowered

or even lost, or more easily to generate a polymer. Adding CTP rigid units between FGF21 and

Fc variants is equivalent to adding a section of rigid linker peptide. In one embodiment, it

ensures that the FGF21 fused at N-terminus will not affect the binding site of Fc variants and

FcRn, thereby not affecting the half-life; in addition, the Protein A binding site of Fc is very

important for the purification step in the preparation process, and connecting CTP rigid units

ensures that N-terminus fused FGF21 also will not "mask" its binding site with protein A. In

another embodiment, the addition of CTP rigid units also makes Fc fragments with about 25 KD

not interfere with the correct folding of N-terminus fused FGF21, and not cause the biological

activity/functions of FGF21 lowered or lost. This may be interpreted that a CTP rigid polypeptide with multiple carbohydrate side chains, which, relative to the random coil of

(GGGGS)n of flexible linker peptides, can form a stable three-dimensional conformation, and

this "barrier" effect promotes FGF21 and Fc fragment to fold and form a correct

three-dimensional conformation independently, thus not affecting their respective biological

activity. In another embodiment, the protective effect of carbohydrate side chain of CTP can

lower the sensibility of linker peptides to proteases, such that the dual-function protein is not

easily degraded in the linker region. In addition, CTP is derived from native human hCG, has no

immunogenicity, therefore, relative to non-native encoded amino acid sequence, is more suitable

for being used as a linker peptide.

IgG Fe variant

Non-lytic Fe variants

Fc elements are derived from the constant region Fc of immunoglobulin IgG, and have an

important effect in immune defense for eliminating pathogens. The effector function of IgG

mediated by Fc is exerted via two mechanisms: (1) binding to the Fc receptors (FcyRs) on the

cell surface, digesting pathogens by phagocytosis or lysis, or by killer cells through

antibody-dependent cytotoxic (ADCC)pathway, or (2) binding to Clq of the first complement

component Cl, triggering the complement dependent cytotoxic (CDC) pathway, thereby lysing

pathogens. Among four human IgG subtypes, IgGI and IgG3 can effectively bind to FcyRs, the

binding affinity of IgG4 with FcyRs is relatively low, and the binding of IgG2 with FcyRs is too

low to be determined, and therefore, human IgG2 almost has no ADCC effect. In addition,

human IgGI and IgG3 also can effectively bind to Cq, thereby activating the complement

cascade. The binding of human IgG2 with Clq is relatively weak, and IgG4 does not bind with

Clq (Jefferis R et al., Immunol Rev, 1998, 163: 59-76), and therefore, the CDC effect of human

IgG2 is also weak. There is no native IgG subtype which is very suitable for constructing

GLP-1-FGF21 dual function protein. In order to obtain a non-lytic Fc without effector function,

the most effective method is performing mutation modification on the complement and receptor binding domain of Fc fragment, regulating the binding affinity of Fc with related receptors, reducing or eliminating ADCC and CDC effects, only retaining the long half-life property of Fc in circulation, and not generating cytotoxicity. For additional mutation sites comprised in non-lytic Fc variants, one can refer to RL et al., J Biol Chem, 2001, 276(9):6591-604 or Chinese invention patent CN 201280031137.2.

Fe variants having an enhanced binding affinity with neonate receptor (FcRn)

The plasma half-life of IgG depends on its binding with FcRn, and generally, they bind at

pH 6.0, and dissociate at pH 7.4 (plasma pH). By the study of the binding site of the two, the

binding site on IgG with FcRn is modified, such that the binding ability thereof is increased at

pH 6.0. It is proven that the mutations of some residues in human Fcy domain which is important

for binding with FcRn can increase the serum half-life. It has been reported that the mutations of

T250, M252, S254, T256, V308, E380, M428 and N434 can increase or reduce FcRn binding

affinity (Roopenian et al., Nat Rview Immunology, 2007, 7:715-725). South Korea patent no.

KR 10-1027427 discloses Trastuzumab (Herceptin, Genentech) variants with an increased FcRn

binding affinity, and these variants comprise one or more amino acid modifications selected from

257C, 257M, 257L, 257N, 257Y, 279Q, 279Y, 308F and 308Y. South Korea patent no. KR

2010-0099179 provides bevacizumab (avastin, Genentech) variants, and these variants comprise

amino acid modifications of N434S, M252Y/M428L, M252Y/N434S and M428L/N434S, and

show an increased half-life in vivo. In addition, Hinton et al. also find that 2 mutants of T250Q

and M428L can increase the binding with FcRn by 3 and 7 times respectively. Mutating the 2

sites at the same time, the binding will be increased by 28 times. In rhesus monkeys, the mutants

of M428L or T250QM/428L show the plasma half-life in vivo is increased by 2 times (Paul R.

Hinton et al., J Immunol, 2006, 176:346-356). For additional mutation sites comprised in Fc

variants having an enhanced binding affinity with neonate receptor (FcRn), one can refer to

Chinese invention patent CN 201280066663.2. In addition, in some studies performing

T250Q/M428L mutations on the Fc fragment of five humanized antibodies, the interaction between Fc and FcRn is improved, and in the subsequent in vivo pharmacokinetic test, it finds that using subcutaneous injection administration, the pharmacokinetic parameters of Fc mutation antibodies are improved compared with wild type antibodies, such as an increased in vivo exposure, lowered clearance rate and improved subcutaneous bioavailability (Datta-Mannan A et al., MAbs. Taylor & Francis, 2012, 4(2): 267-273).

Terms "FGF21-related conditions" and "GLP-1-related conditions" comprise obesity, types

1 and 2 diabetes, pancreatitis, dyslipidemia, nonalcoholic fatty liver disease, nonalcoholic

steatohepatitis, insulin tolerance, hyperinsulinemia, glucose intolerance, hyperglycemia,

metabolic syndrome, acute myocardial infarction, high blood pressure, cardiovascular disease,

atherosclerosis, peripheral arterial disease, stroke, cardiac failure, coronary heart disease,

nephropathy, diabetic complication, neuropathy, gastroparesis, and conditions associated with the

severe inactivation mutations of insulin receptor.

"Conditions associated with the severe inactivation mutations of insulin receptor" describe

the conditions of subjects with insulin receptor (or a direct downstream possible protein thereof)

mutation, wherein said mutation results in a severe insulin tolerance, but generally no obesity

which is common in type 2 diabetes. In many embodiments, subjects with these conditions

exhibit the symptom complex of types 1 and 2 diabetes. Therefore, the involved subjects are

divided into several types according to the severity, comprising: type A diabetes resistance, type

C insulin resistance (AKA HAIR-AN syndrome), Rabson-Mendenhall syndrome, Donohue's

syndrome or Leprechaunism. These conditions are associated with a very high endogenous

insulin level, and results in an elevated blood glucose level. Therefore, in the involved subjects,

there are many clinical features associated with "insulin toxicity", wherein the clinical features

comprise androgen excess, polycystic ovarian syndrome (PCOS), hirsutism and acanthosis

nigricans (overgrowth of wrinkly skin and pigmentation).

"Diabetic complications" is the dysfunction of other tissue/organs of the body induced by

chronic hyperglycemia, such as diabetic nephropathy, diabetic neuropathy, diabetic feet (foot ulcers and low blood circulation) and eye lesions (retinopathy). Diabetes also increases the risks of heart disease and osteoarticular diseases. The other long-term complications of diabetes comprise skin, digestive, sexual function, teeth and gums disease.

"Metabolic syndrome (MS)" is the morbidness caused by abnormal metabolic parameters,

comprising: (1) abdominal obesity or overweight; (2) atherosclerosis and dyslipidemia, such as

hypertriglyceride and reduction in high density lipoprotein cholesterol (HDL-C); (3)

hypertension; (4) insulin resistance and/or abnormal glucose tolerance. In some criteria,

microalbuminuria, hyperuricemia, pro-inflammatory state (C-reactive protein) and

pro-thrombogenesis state (increase in Fibrinogen and Plasminogen inhibitor-1) are also

comprised.

"Dyslipidemia" is a lipoprotein metabolic disorder, comprising the oversynthesis or defect

of lipoprotein. Dyslipidemia can exhibit as the elevated concentration of total cholesterol, low

density lipoprotein (LDL) cholesterol and triglycerides, and the reduced concentration of high

density lipoprotein (HDL) cholesterol.

"Nonalcoholic fatty liver disease (NAFLD)" is a liver disease which is not associated with

abused alcohol consumption and is characterized in hepatocellular steatosis.

"Nonalcoholic steatohepatitis (NASH)" is a liver disease which is not associated with

abused alcohol consumption and is characterized in hepatocellular steatosis accompanied by

lobular inflammation and fibrosis.

"Atherosclerosis" is an angiopathy and is characterized in lipid deposits irregularly

distributed on endangium of large and medium-sized arteries, which results in hemadostenosis,

and eventually develops into fibrosis and calcification.

BriefDescription of the Drawings

Fig. 1. shows the nucleotide sequence of dual-function protein FP4I-2 of Spe/EcoRI

fragment in PCDNA3.1 expression vector according to the embodiments of the present disclosure and the deduced amino acid sequence, consisting of alpha 1 microglobulin leader peptide (1-19), GLP-1 analog (20-47), Li (48-65), FGF21 mature protein (66-246), L2 (247-301) and IgG2 Fc (302-524).

Fig. 2a. Reduced SDS-PAGE electrophoretogram of GLP-1-FGF21 dual-function protein

FP4I-2.

Fig. 2b. SEC-HPLC spectrogram of GLP-1-FGF21 dual-function protein FP4I-2.

Fig. 3a. Glucose tolerance test curve of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 16 h after a single injection (means ±SEM, n = 8).

Fig. 3b. Glucose tolerance test iAUC of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 16 h after a single injection means± SEM, n = 8); statistical difference symbols:

compared with the control group, *P< 0.05, and **P < 0.01.

Fig. 4a. Glucose tolerance test curve of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 96 h after a single injection (means ±SEM, n = 8).

Fig. 4b. Glucose tolerance test iAUC of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 96 h after a single injection means± SEM, n = 8); statistical difference symbols:

compared with the control group, *P< 0.05, and **P < 0.01.

Fig. 5a. Glucose tolerance test curve of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 144 h after a single injection (means ±SEM, n = 8).

Fig. 5b. Glucose tolerance test iAUC of GLP-1-FGF21 dual-function proteins FP4I-1 and

FP4I-2 144 h after a single injection (means ±SEM, n = 8); statistical difference symbols:

compared with the control group, *P< 0.05, and **P < 0.01.

Fig. 6a. Glucose tolerance test curve of Exendin4-FGF21 dual-function proteins FP4I-3 16

h after a single injection (means ±SEM, n = 8).

Fig. 6b. Glucose tolerance test iAUC of Exendin4-FGF21 dual-function proteins FP4I-3 16

h after a single injection (means ±SEM, n = 8); statistical difference symbols: compared with the control group, *P< 0.05, and **P < 0.01; compared with the Dulaglutide group, #P < 0.05, and

11P < 0.01.

Fig. 7a. Glucose tolerance test curve of Exendin4-FGF21 dual-function proteins FP4I-3 96

h after a single injection (means ±SEM, n = 8).

Fig. 7b. Glucose tolerance test iAUC of Exendin4-FGF21 dual-function proteins FP4I-3 96

h after a single injection (means ±SEM, n = 8); statistical difference symbols: compared with the

control group, *P< 0.05, and **P < 0.01; compared with the Dulaglutide group, #P < 0.05, and

11P < 0.01.

Fig. 8a. Glucose tolerance test curve of Exendin4-FGF21 dual-function proteins FP4I-3 144

h after a single injection (means ±SEM, n = 8).

Fig. 8b. Glucose tolerance test iAUC of Exendin4-FGF21 dual-function proteins FP4I-3

144 h after a single injection (means ±SEM, n = 8); statistical difference symbols: compared

with the control group, *P < 0.05, and **P < 0.01; compared with the Dulaglutide group, #p <

0.05, and "P < 0.01.

Fig. 9. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on 24 h food

intake in db/db mice after first administration means± SEM, n = 6); statistical difference

symbols: compared with the control group, *P < 0.05, and **P < 0.01; compared with the

Dulaglutide group, 'P < 0.05, and "P < 0.01.

Fig. 10. Effect of multiple administrations of GLP-1-FGF21 dual-function proteins FP4I-1

and FP4I-2 on glycated hemoglobin in db/db mice (means ±SEM, n = 6); statistical difference

Dulaglutide group, 'P < 0.05, and "P < 0.01.

Fig. 11. Effect of multiple administrations of GLP-1-FGF21 dual-function proteins FP4I-1

and FP4I-2 on accumulative food intake in db/db mice means± SEM, n = 6); statistical

difference symbols: compared with the control group, *P< 0.05, and **P < 0.01; compared with

the Dulaglutide group, #P< 0.05, and '#P < 0.01.

Fig. 12. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on body weight

in obese mice induced by high-fat diet (means ±SEM, n = 7); statistical difference symbols:

compared with the obesity control group, *P < 0.05, and **P < 0.01; compared with the

Dulaglutide group, 'P < 0.05, and "P < 0.01; compared with the FP4I-1 group, P < 0.05, and

P < 0.01.

Fig. 13. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on liver mass in

obese mice induced by high-fat diet means± SEM, n = 7); statistical difference symbols:

Dulaglutide group, 'P < 0.05, and "P < 0.01.

Fig. 14. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on liver

triglyceride content in obese mice induced by high-fat diet (means ±SEM, n = 7); statistical

difference symbols: compared with the obesity control group, *P < 0.05, and **P < 0.01;

compared with the Dulaglutide group, 'P < 0.05, and "P < 0.01.

Fig. 15. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on serum

triglyceride in obese mice induced by high-fat diet (means ±SEM, n = 7); statistical difference

symbols: compared with the obesity control group, *P < 0.05, and **P < 0.01; compared with the

Dulaglutide group, 'P < 0.05, and "P < 0.01; compared with the FP4I-1 group, &P< 0.05, and

P < 0.01.

Fig. 16. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on serum total

cholesterol content in obese mice induced by high-fat diet (means ±SEM, n = 7); statistical

compared with the Dulaglutide group, 'P < 0.05, and "P < 0.01.

Fig. 17. Effect of GLP-1-FGF21 dual-function proteins FP4I-1 and FP4I-2 on serum low

density lipoprotein cholesterol content in obese mice induced by high-fat diet (means ±SEM, n

= 7); statistical difference symbols: compared with the obesity control group, *P< 0.05, and **P

< 0.01; compared with the Dulaglutide group, #P < 0.05, and .. P < 0.01.

The present disclosure is further described below in combination with specific embodiments.

It is to be understood that these embodiments serve only to illustrate the present disclosure and

are not limiting the scope of the present disclosure. In the following embodiments, experimental

methods without specifying specific conditions are generally performed under conventional

conditions, for example, those described in Sambrook et al., Molecular Cloning: A Laboratory

Manual (New York: Cold Spring Harbor Laboratory Press, 1989), or the conditions

recommended by the manufacturer.

Generally, the dual-function protein of the present disclosure is prepared synthetically. The

nucleotide sequence according to the present disclosure, a person skilled in the art can

conveniently use various known methods to prepare the encoding nucleic acid of the present

disclosure. These methods are for example but not limited to: PCR, and DNA artificial synthesis

etc., the specific methods can refer to J. Sambrook, "Molecular Cloning: A Laboratory Manual".

As an embodiment of the present disclosure, a method comprising fragment synthesis of

nucleotide sequences, followed by overlap extension of PCR can be used for constructing the

encoding nucleic acid sequence of the present disclosure.

Also provided in the present disclosure is an expression vector comprising a sequence

encoding the dual-function protein of the present disclosure and a regulatory element

transcriptionally linked thereto. Said "transcriptionally linked" or "transcriptionally linked to"

refer to such a condition that some parts of a linear DNA sequence can regulate or control the

activity of other parts in the same linear DNA sequence. For example, if a promoter controls the

transcription of a sequence, then the promoter is transcriptionally linked to the encoding

sequence.

The expression vector can use commercially available ones, for example but not limited to:

vectors pcDNA3, pIRES, pDR and pUC18 which can be used for expression in an eukaryotic system. A person skilled in the art can select a suitable expression vector according to the host cell.

According to the restriction map of the known expression vector, a person skilled in the art

can insert the sequence encoding the dual-function protein of the present disclosure into a

suitable restriction site to prepare the recombinant expression vector of the present disclosure

following conventional methods via restriction digestion and ligation.

Also provided in the present disclosure is a host cell expressing the dual-function protein of

the present disclosure, wherein said host cell comprises the sequence encoding the dual-function

protein of the present disclosure. In at least one embodiment, said host cell is a eukaryotic cell,

for example but not limited to CHO, a COS cell, a 293 cell and a RSF cell etc. As a at least one

embodiment of the present disclosure, said cell is a CHO cell, which can well express the

dual-function protein of the present disclosure, and the dual-function protein with a good binding

activity and stability can be obtained.

Also provided in the present disclosure is a method for preparing the dual-function protein

of the present disclosure using recombinant DNA, the steps thereof comprise:

1) Providing a nucleic acid sequence encoding the synergistic dual function protein;

2) Inserting the nucleic acid sequence of 1) into a suitable expression vector, and obtaining

a recombinant expression plasmid;

3) Introducing the recombinant expression plasmid of 2) into a suitable host cell;

4) Culturing the transformed host cell under a condition suitable for expression;

5) Collecting the supernatant, and purifying the dual-function protein product.

To introduce said encoding sequence into the host cell one can use multiple known

technologies in the art, for example but not limited to: calcium phosphate precipitation,

protoplast fusion, liposome transfection, electroporation, microinjection, reverse transcription

method, phage transduction method, and alkali metal ion method.

With respect to the culture of and expression in the host cell can refer to Olander RM Dev

Biol Stand, 1996, 86:338. Cells and debris in the suspension can be removed by centrifugation,

and the supernatant is collected. Agarose gel electrophoresis technique can be used for

identification.

The dual-function protein prepared as described herein can be purified to have a

substantially homogeneous property, such as has a single band on SDS-PAGE electrophoresis.

For example, when the recombinant protein is expressed for secretion, a commercially available

ultrafiltration membrane (such as products of Millipore and Pellicon etc.) can be used to separate

said protein, wherein firstly, the expression supernatant is concentrated. The concentrate can be

purified by the method of gel chromatography, or by the method of ion exchange

chromatography, for example, by anion exchange chromatography (DEAE etc.) or cation

exchange chromatography. The gel matrix can be common matrices for protein purification, such

as agarose, glucan, and polyamide etc. Q- or SP-groups is a relatively ideal ion exchange group.

Finally, the above-mentioned purified product can be further refined and purified by the methods

of hydroxyapatite adsorption chromatography, metal chelate chromatography, hydrophobic

interaction chromatography and reversed high performance liquid chromatography (RP-HPLC).

All the above-mentioned purification steps can be used in different combination in order to make

the protein purity substantially homogeneous.

The expressed dual-function protein can be purified using an affinity column containing a

specific antibody, receptor or ligand of said dual function protein. According to the properties of

the affinity column, conventional methods, such as high salt buffer and changing pH etc. can

used to elute the fusion polypeptide binding to the affinity column. Optionally, at the amino

terminus or carboxyl terminus of said dual function protein, one or more polypeptide fragments

also can be contained as protein tags. Any suitable tags can be used in the present disclosure. For

example, said tags can be FLAG, HA, HA1, c-Myc, 6-His or 8-His etc. These tags can be used

for purifying the dual function protein.

EXAMPLES

Example 1: Construction of an expression plasmid of the sVnergistic dual function protein

All gene sequences encoding alpha 1 microglobulin secretion leader signal, GLP-1 analog,

LI, FGF21 mature protein, L2 (comprising a flexible linker unit and rigid linker unit) and human

IgG Fc variants were optimized using CHO preferred codons and the full-length gene sequences

were synthesized. There are a Spel at the 5' and a EcoRI at the 3' for subcloning the target gene

encoding the fusion protein into the expression vector PXY1A1 modified from PCDNA3.1 (Fig.

1 exemplarily set forth the nucleotide sequence of the dual-function protein FP4I-2 and the

translated amino acid sequence). The expression plasmid contained the early promoter of

cytomegalovirus, leading to high expression of exogenous genes in mammalian cells. The

plasmid also contained a selective marker conferring kanamycin resistance in bacteria, and G418

resistance in mammalian cells. Furthermore, the host cell carrying DHFR- mutant, PXY1A1

expression vector contained the gene of mouse dihydrofolate reductase (DHFR) could amplify

the fusion gene and DHFR gene in the absence of methotrexate (MTX) (see U.S. Patent No.

4,399,216).

Various dual-function proteins comprising GLP-1 and FGF21 were constructed. Here, three

are exemplified: FP4I-1, FP4I-2 and FP4I-3. The amino acid composition is shown in Table 1

(Li and L2 were underlined, and mutated amino acids in Fc variants were boxed).

Table 1. Amino acid composition of each synergistic dualfunction protein FP4I-1 HGEGTFTSDVSSYLEGQAAKEFIAWLVKGRGGGGSGGGGSGGGGSH PIPDSSPLLQFGGQVRQRYLYTDDAQQTEAHLEIREDGTVGGAADQ SPESLLQLKALKPGVIQILGVKTSRFLCQRPDGALYGSLHFDPEAC SFRELLLEDGYNVYQSEAHGLPLHLPGNKSPHRDPAPRGPARFLPL PGLPPAPPEPPGILAPQPPDVGSSDPLSMVGPSQGRSPSYASGGGG SGGGGSGGGGSGGGGSGGGGSSSSSKAPPPSSSSSKAPPPSESKYG PPCPPCPAPEF GGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQ EDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWL NGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTK NQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFL YSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLSLSLGK

HGEGTFTSDVSSYLEEQAAKEFIAWLVKGGGGGGGSGGGGSGGGGS HPIPDSSPLLQFGGQVRQRYLYTDDAQQTEAHLEIREDGTVGGAAD QSPESLLQLKALKPGVIQILGVKTSRFLCQRPDGALYGSLHFDPEA CSFRELLLEDGYNVYQSEAHGLPLHLPGNKSPHRDPAPRGPARFLP LPGLPPAPPEPPGILAPQPPDVGSSDPLSMVGPSQGRSPSYASGSG FP4I-2 GGGSGGGGSGGGGSGGGGSGGGGSSSSSKAPPPSLPSPSRLPGPSD TPILPQVECPPCPAPPVAGPSVFLFPPKPKDOLMISRTPEVTCVV VDVSHEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVV HQDWLNGKEYKCKVSNKGLPASIEKTISKTKGQPREPQVYTLPPS REEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPMLDS DGSFFLYSKLTVDKSRWQQGNVFSCSV HEALHNHYTQKSLSLSP GK

HGEGTFTSDLSKQMEEEAVRLFIEWLKNGGPSSGAPPPSGGGGGGG SGGGGSGGGGSHPIPDSSPLLQFGGQVRQRYLYTDDAQQTEAHLEI REDGTVGGAADQSPESLLQLKALKPGVIQILGVKTSRFLCQRPDGA LYGSLHFDPEACSFRELLLEDGYNVYQSEAHGLPLHLPGNKSPHRD PAPRGPARFLPLPGLPPAPPEPPGILAPQPPDVGSSDPLSMVGPSQ GRSPSYASGSGGGGSGGGGSGGGGSGGGGSGGGGSSSSSKAPPPSL FP4I-3 PSPSRLPGPSDTPILPQVECPPCPAPPVAGPSVFLFPPKPKDOLM ISRTPEVTCVVVDVSHEDPEVQFNWYVDGVEVHNAKTKPREEQFNS TFRVVSVLTVVHQDWLNGKEYKCKVSNKGLPASIEKTISKTKGQP REPQVYTLPPSREEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPEN NYKTTPPMLDSDGSFFLYSKLTVDKSRWQQGNVFSCSVLHEALHNH YTQKSLSLSPGKVDKSRWQQGNVFSCSV HEALHNHYTQKSLSLS PGK

Example 2: Expression of the dual-function protein in a transfected cell line

A recombinant expression vector plasmid was transfected into a mammalian host cell line to

express the synergistic dual function protein. In order to stabilize the high expression, a preferred

host cell line was DHFR defective CHO-cell (U.S. Patent No. 4,818,679). In the present example,

the host cell was selected from CHO-derived cell line DXB11. A preferred transfection method

was electroporation, but other methods such as calcium phosphate and liposome-induced

transfection also can be used. A Gene Pulser electroporation apparatus (Bio-Rad Laboratories,

Hercules, CA) setting at 300 V of electric field and 1500 Fd of capacitance was used in the

present experiment and 50 g pure expression plasmid was mixed with 5 x 107 CHO cells in the

cuvette. Two days after the transfection, a selection medium containing 0.6 mg/mL G418 was

used. Quantitative ELISA using anti-human IgG Fc was applied to screen the transfectants with

the resistance to G418. Anti-human FGF21 or anti-human GLP-1 by ELISA was used to quantify expression of the dual-function protein. A 96 well culture plate was subjected to the limiting dilution, the well generating a high level of the dual-function protein was subcloned.

In order to achieve a relatively high expression of the dual-function protein, it was

appropriate to use the DHFR gene inhibited by MTX for co-amplification. In another selection

medium containing incremental concentrations of MTX, the gene of the dual-function protein

was co-amplified with the DHFR gene. The subclone with a positive DHFR expression was

subjected to the limiting dilution, the selection pressure was gradually increased and the

transfectant which can grow in a medium with up to 6 M MTX was selected. The secreting rate

of transfectant was determined and the cell line with a high expression of exogenous protein was

screened out. The cell lines with secretory rate higher than about 10 (preferably about 20) pg/106

(i.e. a million) cells/24 hours was subjected to an adaptive suspension culture in serum-free

medium, then the dual-function protein was purified by a specified medium.

Example 3: Purification and qualification of the dual-function protein

This example describes the exemplary purification and qualification methods of FP4I-2.

The cell culture supernatant was subjected to clarifying treatments, such as high speed

refrigerated centrifugation and 0.22 [m sterile filtration etc., then purified by three

chromatograph steps including protein A, anion exchange and hydrophobic chromatography, the

specific method was as follows: In the first step, protein A was used for capture, wherein the

equilibrium solution was PBS buffer, the eluant was a citrate buffer at pH 3.5, then the eluted

protein was neutralized by 1 M Tris solution. In the intermediate purification process, high

resolution anion exchange packing material Q Sepharose HP (GE company) was selected to

remove residual impurity proteins. A combined mode was used, that is, 20 mM Tris-HCl, 0.2 M

NaCl, pH 7.5 solution was used for rinsing, and 20 mM Tris-HCl, 0.3 M NaCl, pH 7.5 solution

was used for elution. In the fine purification step, Butyl Sepharose FF (GE) was selected to

remove polymers; due to different hydrophobic properties of FP4I-2 monomer and polymer, the monomer with weak hydrophobic property flowed through directly, but the polymer with high hydrophobic property bound to the medium; hydrophobic chromatography was selected as flow through mode, and the equilibrium solution was PBS buffer.

The qualitative analysis result is shown in Figs. 2a and 2b. The theoretical molecular weight

of single stranded FP4I-2 was about 53 KD, due to the absence of glycosylation sites, under the

reducing condition, SDS-PAGE electrophoresis showed that the actual mass of the single

stranded FP4I-2 molecule was about 70 KD. FP4I-1 and -3 were prepared by the same method.

Example 4: Effect of a single injection of the dual-function protein on glucose utilization in

C57BL/6 mice

8 weeks aged male C57BL/6J mice at SPF grade (purchased from Beijing HFK Bioscience

Ltd.) were selected. Housing conditions: temperature 22-25°C, relative humidity 45-65%, and

12h-light/dark cycle. After acclimation for 1 week, mice were randomly divided into control

group, Dulaglutide 120 nmol/kg group, FP4I-2 120 nmol/kg group and FP4I-1 120 nmol/kg

group (n = 7) according to body weight. The mice in the treatment groups were injected

subcutaneously with corresponding drug solutions, while the mice in the control group were

injected subcutaneously with PBS buffer. After the injection, mice in each group were fasted for

16 h, and then glucose tolerance test was performed. The fasting blood glucose values of the

mice were determined followed by an intraperitoneal injection of a 2g/kg glucose solution, the

blood glucose values were determined at 15 min, 30 min, 60 min, 90 min and 120 min after

glucose injection, and the increased area below the curve and above the baseline (iAUC) was

calculated by the trapezoidal method. The glucose tolerance test was further performed on the

mice of each group at 96 h and 144 h after the administration, and the method was the same as

above. The data were represented as means ±SEM, and analyzed using SPSS18.0 statistical

software. For the Gaussian distribution data, statistical comparison of the means among the

groups was performed using one-way ANOVA, followed by LSD test for the homogeneity of variance or Dunnet T3 test for the heterogeneity of variance; non-parametric test was used for the Non-Gaussian distribution data. P < 0.05 represented a significant statistical difference.

As shown in Figs.3a and 3b, FP4I-1 and FP4I-2 significantly improved glucose utilization

in the mice at 16 h after administration when compared with the control group (P< 0.01). It can

be known from Figs.4a and 4b that FP4I-1 and FP4I-2 also can significantly improve the glucose

utilization level in mice at 96 h after administration as well (P < 0.01). It can be known from

Figs.5a and 5b that FP4I-1 and FP4I-2 significantly improved the glucose utilization level in

mice even at 144 h after administration (P < 0.05). The results showed that in the case of a

suddenly increased glucose level in vivo, the GLP-1-FGF21 dual-function protein had a rapid

response to the glucose level and normalized it to the physiological level by promoting release

and secretion of insulin with a long-acting activity, and therefore it could be used for treating

diabetes and the complications induced by the absolute or relative deficiency of insulin. When

the mice were fasted for 16 h after FP4I-1 or FP4I-2 administration, no shock or death due to

hypoglycemia were noted in any mouse, indicating that the dual-function protein would not

result in hypoglycemic symptoms as insulin.

In addition, the activity of Exendin4-FGF21 dual-function protein FP4I-3 on the glucose

utilization was determined by above-mentioned method as well. C57BL/6 mice were divided

into the control group, Dulaglutide group and FP4I-3 group. Corresponding drug solutions

(120nmol/kg) were administrated subcutaneously to the mice in Dulaglutide group and FP4I-3

group, respectively, and PBS buffer was administrated to the mice in the control group. The

glucose tolerance test was performed at 16 h, 96 h, and 144 h after the injection. As shown in

Figs. 6a and 6b, 16 h after the administration, compared with the control group, FP4I-3

significantly improved the glucose utilization (P < 0.01), but the activity was significantly

weaker than that of Dulaglutide (P < 0.01). As shown in Figs. 7a and 7b, 96 h after the

administration, FP4I-3 significantly improved the glucose utilization in mice (P < 0.01) though

the efficacy was significantly lower than Dulaglutide as well (P < 0.01). As shown in Figs. 8a and 8b, 144 h after the administration, the ability of FP4I-3 to improve glucose utilization in mice was not observed (P > 0.05).

The glucose tolerance test of FP4I-3 in animals demonstrated that Exendin-4 did not display

a synergistic effect with FGF21. The hypoglycemic effect of FP4I-3 was significantly weaker

than that of Dulaglutide indicated that the circulating half-life of FP4I-3 was shorter than

Dulaglutide. In contrast, the preferred GLP-1-FGF21 dual-function proteins FP4I-2 and FP4I-1

had a relatively strong stability in vivo, and were not easily degraded and inactivated, and

maintained a longer in vivo pharmacodynamic activity relative to Exendin4-FGF21 dual-function

protein FP4I-3. Above results indicated that the combination modes of three functional

components, GLP-1 analogs, FGF21 and Fc fragment in the dual-function protein were not

random and arbitrarily, wherein the selection of GLP-1 analogs, the structure of linker peptide,

the fusion sequence, even the difference of glycosylation pattern would affect accuracy and

stability of the dual-function protein conformation to varying degrees, and it determined whether

the active molecules were functionally synergetic and the half-life was prolonged or not.

Example 5: Hypoglycemic effect of dual-function protein in db/db mice

8 weeks aged male db/db mice were purchased from Shanghai SLAC Laboratory Animal

Ltd. Housing conditions: temperature 22-25°C, relative humidity 45-65%, and 12 h-light/dark

cycle. After housed individually for 1 week as acclimation, the mice were divided into 4 groups

according to body weight, blood glucose and food intake: control group, Dulaglutide group,

FP4I-1 group and FP4I-2 group (n = 7). Mice in the control group were injected subcutaneously

with PBS buffer, and mice in other groups were injected subcutaneously with 120 nmol/kg

corresponding drug solutions (twice per week, totally 8 times). Daily food intake of each mouse

was recorded. At the end of the dosing period, mice were fasted for 16 hours, 5 L whole blood

sample was collected from the eye socket to measure glycosylated hemoglobin. The data were

represented as means standard error ( ± s), and were analyzed using SPSS 18.0 statistical software. For the data follow Gaussian distribution, one-way analysis of variance was used for comparing mean difference among groups, followed by LSD test for the homogeneity of variance or Dunnet T3 test for the heterogeneity of variance; non-parametric test was used for the data follow non-normal distribution. P < 0.05 represented a significant statistical difference.

As shown in Fig. 9, compared with Dulaglutide group, after the first administration, the

food intakes within 24 hour of the mice in FP4I-1 and FP4I-2 groups were significantly elevated

(P < 0.01). The results showed that GLP-1-FGF21 dual-function protein could significantly

relieve symptoms of severe gastrointestinal adverse effects induced by the first administration of

long-acting GLP-1 receptor agonist drugs. As shown in Fig. 10, FP4I-1, FP4I-2, Dulaglutide

group can significantly lower the glycosylated hemoglobin values of db/db mice (P < 0.01), and

the glycosylated hemoglobin values of mice in FP4I-1 and FP4I-2 groups were significantly

lower than that in Dulaglutide group (P< 0.05). db/db mouse is a spontaneously hyperglycemic

animal model with a severe insulin tolerance. The GLP-1-FGF21 dual-function protein exhibited

a better property than Dulaglutide in the long-term glycemic control. Based on the data in

Example 4, the insulinotropic activity of GLP-1-FGF21 dual-function protein was not

significantly better than Dulaglutide. Wild type FGF21 exhibited a good insulin sensitization

effect in the hypeinsulinemic-euglycemic clamp test (Xu J et al., Diabetes, 2009, 58:250-259),

but there was no direct evidence showing that Dulaglutide had an insulin sensitization effect in

vivo. In conclusion, the superiority in blood glucose control exhibited by GLP-1-FGF21

dual-function protein should be the result of synergistic effect of GLP-1 analog promoting

release and secretion of insulin and FGF21 enhancing insulin sensitivity. As shown in Fig. 11, in

the experimental period, the cumulative food intake of mice in FP4I-1 and FP4I-2 groups was

significantly higher than that in Dulaglutide group (P < 0.05), the results showed that in the

condition of excluding factors intervening food intake, the blood glucose control activity of

FP4I-1 and FP4I-2 groups on type 2 diabetes should be higher than that of Dulaglutide.

Example 6. Therapeutic effects of the dual-function protein on weight loss, hepatic steatosis

and lipid metabolism disorder in obese mice induced by high-fat diet

8 weeks aged C57BL/6 mice were purchased from Shanghai SLAC Laboratory Animal Ltd.

Housing conditions: temperature 22-25°C, relative humidity 45-65%, and lighting time 12 h/d.

After acclimation for 1 week, 7 mice were selected and fed with low-fat diet (D12450B,

Research Diets), and other mice were fed with high-fat diet (D12451, Research Diets). 40 weeks

later, obese mice were subjected to adaptive feeding with single animal/cage for 1 week, then the

obese mice were divided into five groups according to body weight and weekly food intake:

obese control group, Dulaglutide group, high fat diet pair-fed group, FP4I-1 group and FP4I-2

group (n = 7). In the experiment, the amounts of daily diet per mouse in high fat diet pair-fed,

FP4I-land FP4I-2 groups were consistent with daily food intake per mouse in Dulaglutide group.

Mice in the obese control group and high fat diet pair-fed group were injected subcutaneously

with PBS buffer solution, and mice in other groups were injected subcutaneously with 120

nmol/kg corresponding drug solutions, once every 6 days, and totally 2 times. The body weight

of each mouse was recorded before and after the dosing period. At the end of the dosing period,

mice in each group were fasted for 16 hours, whole blood was collected from the eye socket, and

centrifuged at 2000xg for 15 min to obtain serum. Serum lipid profiles were determined by an

automatic biochemical analyzer. Liver tissue was excised, washed with normal saline, then

removed residual liquid with filter paper and weighed. About 50 mg liver tissue at the same part

of each live was taken, and the triglyceride content was determined using the Folch method. The

results were represented in the form of triglyceride content per mg liver tissue. The data were

represented as means ±SEM, and analyzed using SPSS18.0 statistical software. For the Gaussian

distribution data, statistical comparison of the means among the groups was performed using

one-way ANOVA, followed by LSD test for the homogeneity of variance or Dunnet T3 test for

the heterogeneity of variance; non-parametric test was used for the Non-Gaussian distribution

data. P < 0.05 represented a significant statistical difference.

As shown in Figs. 12 to 17, after administrated with Dulaglutide, body weight, liver mass,

liver triglyceride, triglycerides, total cholesterol and low density lipoprotein-cholesterol contents

in serum were significantly lowered (P<0.01) in the obese mice induced by high-fat diet.

Dulaglutide could cause severe gastrointestinal adverse effects and suppressed appetite by

regulating central nervous system, resulted in reduction in food intake. In this example, daily

supplied the same amount of diet to the mice in the high-fat diet pair-fed group as the

corresponding mice in Dulaglutide group, despite the parameters mentioned above were

significantly lowered when compared with that in the obese control group, there was no

significant statistical difference from that of Dulaglutide group (P > 0.05). The results showed

that the effects of Dulaglutide on weight loss, hepatic steatosis and lipid metabolic disorder

substantially depended on inhibition of appetite without any other mechanisms. The obese mice

in FP4I-1 group and FP4I-2 group were given the same amount of diet as the corresponding mice

in Dulaglutide group, compared with Dulaglutide group, body weight and serum triglyceride

level of the mice in FP4I-2 group were significantly decreased (P < 0.01), which demonstrated

that FP4I-2 had additional functions of reducing fatty acid synthesis and promoting fatty acid

metabolism and utilization in vivo. The results indicated that FP4I-2 could be used for treating

obesity and obesity-induced metabolic syndrome.

Compared with Dulaglutide group, liver mass and liver triglyceride content of mice in

FP4I-2 group were significantly decreased (P < 0.01 or P < 0.05), which demonstrated that

FP4I-2 effectively reduced the excessive accumulation of triglyceride in liver, improved liver

function. The results indicated that FP4I-2 could be used for treating various liver diseases

induced by hepatic steatosis, such as nonalcoholic fatty liver, nonalcoholic steatohepatitis, liver

fibrosis and liver cirrhosis.

Compared with Dulaglutide group, both total serum cholesterol and low density

lipoprotein-cholesterol content of mice in FP4I-2 group were significantly reduced (P< 0.01 or

P < 0.05), indicating that FP4I-2 can be used for treating hypercholesteremia and relevant cardiovascular and cerebrovascular diseases, such as hypertension, coronary heart disease, chronic heart failure, cerebral infarction and atherosclerosis. Compared with Dulaglutide group, the body weight, liver mass, liver triglyceride content, serum triglycerides, total cholesterol and low density lipoprotein cholesterol levels in FP4I-1 group were mildly decreased but no significant differences were observed.

The present study demonstrated that FP4I-1 and FP4I-2 could treat obesity, fatty liver

disease and lipid metabolic disorder via the physiological activity of FGF21, and was not

completely dependent on the food intake regulation effect of GLP-1 analogs; the therapeutic

effect of FP4I-2 in the obese mice was superior to Dulaglutide, which indicated that it could

compensate for the deficiency of Dulaglutide in the clinic. In conclusion, the therapeutic

mechanisms of FP4I-2 are more abundant than that of Dulaglutide, which is more suitable for the

requirement of diversified clinical therapy.

This disclosure provides merely exemplary embodiments of the disclosure. One skilled in

the art will readily recognize from the disclosure and claims, that various changes, modifications

and variations can be made therein without departing from the spirit and scope of the disclosure

as defined in the following claims.

All documents mentioned in this application are hereby incorporated by reference as if each

document were individually incorporated by reference. In addition, it should be understood that

after reading the above teachings of the invention, those skilled in the art can make various

changes or modifications to the invention, and these equivalent forms also fall within the scope

defined by the appended claims of this application.

Claims

1. A dual-function protein comprising sequentially human GLP-1 analog, linker peptide

1, human FGF21, linker peptide 2 and human immunoglobulin Fc fragment from the N to

C-terminus;

wherein:

the linker peptide 1 comprises a flexible peptide; the linker peptide 2 comprises a flexible peptide and a rigid peptide;

the rigid peptide comprises at least 1 rigid unit; and

the rigid unit comprises a full-length or truncated sequence consisting of carboxyl terminal

amino acids 113 to 145 of human chorionic gonadotropin p-subunit.

2. The dual-function protein of claim 1, wherein said dual-function protein is

glycosylated.

3. The dual-function protein of claim 1, wherein said human GLP-1 analog is an analog,

fusion peptide, or derivative thereof which is obtained by substituting, deleting or adding one or

more amino acid residues on the amino acid sequence of SEQ ID NO: 1 and can maintain human

GLP-1 activity.

4. The dual-function protein of claim 3, wherein the GLP-1 analog comprises an amino

acid sequence of SEQ ID NO: 2, 3, 4 or 5.

5. The dual-function protein of claim 1, wherein said linker peptide 1 comprises a

flexible peptide consisting of 2 or more amino acids.

6. The dual-function protein of claim 5, wherein the amino acids are selected from G, S,

A and T.

7. The dual-function protein of claim 6, wherein the amino acid sequence of the flexible

peptide is GGGGGGGSGGGGSGGGGS.

8. The dual-function protein of claim 1, wherein said human FGF21 comprises the

sequence of SEQ ID NO: 6 wherein the leader peptide of amino acid position 1-28 is deleted.

9. The dual-function protein of claim 1, wherein said human FGF21 comprises the

sequence of SEQ ID NO: 6 wherein the leader peptide of amino acid position 1-28 is deleted and

which has G141S or LI74P substitution.

10. The dual-function protein of claim 1, wherein the flexible peptide constituting said

linker peptide 2 comprises 2 or more amino acids selected from G, S, A and T.

11. The dual-function protein of claim 10, wherein the general structural formula of the

amino acid composition of said flexible peptide is (GS)a(GGS)b(GGGS)c(GGGGS)d, wherein a,

b, c and d are integers greater than or equal to 0, and a + b + c + d > 1;

12. The dual-function protein of claim 10, wherein the amino acid composition of said

flexible peptide is selected from:

(i) GGGGS;

(ii) GSGGGSGGGGSGGGGS;

(iii)GSGGGGSGGGGSGGGGSGGGGSGGGGS;

(iv)GGGGSGGGGSGGGGSGGGGSGGGGSGGGGSGGGGS;

(v) GGGSGGGSGGGSGGGSGGGS; and

(vi) GGSGGSGGSGGS.

13. The dual-function protein of claim 1, wherein the rigid units constituting said linker

peptide 2 are selected from SEQ ID NO: 7 and the truncated amino acid sequences thereof;

wherein said truncated amino acid sequences comprise at least 2 glycosylation sites.

14. The dual-function protein of claim 13, wherein the rigid units comprise one of the

following amino acid sequences:

(i) SSSSKAPPPSLPSPSRLPGPSDTPILPQ;

(ii) PRFQDSSSSKAPPPSLPSPSRLPGPSDTPILPQ;

(iii)SSSSKAPPPS;

(iv) SRLPGPSDTPILPQ; or

(v) SSSSKAPPPSLPSPSR.

15. The dual-function protein of claim 13, wherein said rigid units comprise an amino

acid sequence that has at least 90% or 95% amino acid identity with SEQ ID NO: 7 or the

truncated amino acid sequences of claim 13.

16. The dual-function protein of claim 1, wherein said rigid peptide comprises 1, 2, 3, 4

or 5 rigid units.

17. The dual-function protein of claim 1, wherein said human immunoglobulin Fc

fragment is a variant having a reduced ADCC effect and/or CDC effect and/or enhanced binding

affinity with FcRn receptor.

18. The dual-function protein of claim 17, wherein said Fc variant is selected from:

(i) hinge, CH2 and CH3 regions of human IgG1 containing Leu234Val, Leu235Ala and

Pro331Ser mutations;

(ii) hinge, CH2 and CH3 regions of human IgG2 containing Pro331Ser mutation;

(iii) hinge, CH2 and CH3 regions of human IgG2 containing Thr250Gln and Met428Leu

mutations;

(iv) hinge, CH2 and CH3 regions of human IgG2 containing Pro331Ser, Thr250Gln and

Met428Leu mutations; and

(v) hinge, CH2 and CH3 regions of human IgG4 containing Ser228Pro and Leu235Ala

mutations.

19. The dual-function protein of claim 1, wherein the amino acid sequence of said

dual-function protein is of SEQ ID NO: 13 or 15.

20. A DNA molecule encoding the dual-function protein of claim 1.

21. The DNA molecule of claim 20, wherein said DNA molecule comprises the sequence

as shown in SEQ ID NO: 14.

22. A vector comprising the DNA molecule of claim 20.

23. A host cell, wherein the host cell comprises the vector of claim 22, or is transfected

with the vector comprising the DNA molecule of claim 20.

24. A pharmaceutical composition, wherein the pharmaceutical composition comprises a

pharmaceutically acceptable carrier, excipient or diluent, and an effective dose of the

dual-function protein of claim 1.

25. A method for preparing the dual-function protein, comprising:

(a) introducing the DNA sequence encoding the dual-function protein of claim 20 into a

mammalian cell;

(b) screening a high-yield cell strain expressing more than 20 [g/10 6 (million) cells within a

period of every 24 hours in the growth medium thereof from step (a);

(c) culturing the screened cell strain in step (b), and expressing the dual-function protein;

(d) harvesting fermentation supernatant obtained from step (c), and purifying the

dual-function protein; preferably, said mammalian cell in step (a) is a CHO cell; and more

preferably, said mammalian cell is CHO-derived cell line DXB-11.

26. A method of treatment of one or more FGF21 related diseases and GLP-1 related

diseases, and other metabolic, endocrine and cardiovascular diseases; comprising administering

to a person suffering from at least one of said diseases an effective amount of the dual-function

protein of claim 1.

27. The method of treatment of claim 26, wherein the disease is obesity, type 1 diabetes,

type 2 diabetes, pancreatitis, dyslipidemia, nonalcoholic fatty liver disease, nonalcoholic

steatohepatitis, insulin resistance, hyperinsulinemia, glucose intolerance, hyperglycemia,

metabolic syndrome, acute myocardial infarction, hypertension, cardiovascular disease,

atherosclerosis, peripheral arterial disease, stroke, heart failure, coronary heart disease,

nephropathy, diabetic complication, neuropathy, gastroparesis, or symptoms associated with the

severe inactivation mutations of insulin receptor.

28. A method of treatment of one or more FGF21 related diseases and GLP-1 related

to a person suffering from at least one of said diseases an effective amount of the pharmaceutical

composition of claim 24.

29. The method of treatment of claim 28, wherein the disease is obesity, type 1 diabetes,

severe inactivation mutations of insulin receptor.