The present invention relates to a bifunctional protein comprising a fibroblast growth factor 21 (FGF21) protein fused with an antibody directed to proprotein convertase subtilisin/kexin type 9 (PCSK9), pharmaceutical compositions comprising the bifunctional protein, and methods for treating type 2 diabetes, obesity, dyslipidemia, and/or metabolic syndrome using such bifunctional protein.

FGF21 belongs to a family of large polypeptides widely expressed in developing and adult tissues that play crucial roles in multiple physiological functions. FGF21 is a hormone that functions as an important metabolic regulator of glucose and lipid homeostasis. FGF21 promotes glucose uptake in adipocytes by up-regulating GLUT1 expression, a mechanism distinct from that of insulin. FGF21 is a metabolic regulator that has demonstrated an ability to improve multiple metabolic abnormalities in preclinical models of type 2 diabetes mellitus (T2DM). FGF21 as a therapeutic for diabetics has the potential to improve glycemic control, reduce body weight, and decrease low density lipoprotein cholesterol (LDL-C) and triglycerides while increasing high density lipoprotein cholesterol (HDL-C). In diabetic rodents and monkeys, human FGF21 lowered fasting serum concentrations of glucose, and reduced fasting serum

concentrations of triglycerides, insulin and glucagon. Furthermore, in rodent models of diet induced obesity, FGF21 administration led to cumulative body weight loss in a dose dependent manner. Thus, FGF21 has potential utility for the treatment of diabetes, obesity, dyslipidemia, and metabolic syndrome.

PCSK9 is a secreted serine protease, generated primarily in the liver, that regulates plasma concentrations of LDL-C. Rare gain of function mutations in humans cause autosomal dominant hypercholesterolemia, a disorder characterized by LDL-C levels >300 mg/dl and premature atherosclerosis. Loss of function mutations in PCSK9 are associated with reduced LDL-C and reduced risk of cardiovascular diseases. Secreted PCSK9 binds to and internalizes with the LDL receptor (LDLR) located on the surface of hepatocytes. LDLR functions to clear LDL-C from plasma by binding and transporting LDL particles to lysosomes for degradation. Once the LDL particle is delivered for degradation, the LDLR recycles to the hepatocyte cell surface to bind and clear additional LDL-C from the plasma. PCSK9 regulates plasma LDL-C by directing internalized LDLR for degradation rather than recycling to the cell surface, thus reducing LDL-C clearance. Studies in rodents, which lack or over-express PCSK9, have confirmed that PCSK9 controls circulating LDL-C levels by modulating the levels of LDLR. The observation that circulating PCSK9 participates in the degradation of hepatic LDLR suggests that antibody neutralization of PCSK9 is a viable therapeutic approach for lowering of LDL-C. Further, it has been reported that statin drugs, the current standard of care for lowering LDL-C, actually increase the expression and serum levels of PCSK9, which would be expected to counteract the ability to effectively lower LDL-C. Thus, an antibody directed against PCSK9 also has the potential to reduce LDL-C in a manner synergistic with statin therapy.

PCSK9 antibodies have been described in US2009/0246192, US2009/0142352, US2010/0166768, and WO2010/029513. FGF21 proteins have been described in WO2010/042747, WO2010/285131, and WO2009/149171.

Although FGF21 proteins have shown positive effects in treating type 2 diabetes and PCSK9 antibodies have been shown to lower LDL-C, there is still a need for additional beneficial therapeutics for type 2 diabetes with the benefit of also lowering LDL-C.

The present invention provides alternative therapeutics for diabetes. Additionally, the bifunctional protein of the present invention is potentially useful for the treatment of type 2 diabetes, obesity, dyslipidemia, and metabolic syndrome.

The present invention provides a bifunctional protein comprising a polypeptide A and a polypeptide B. Polypeptide A has in order from N-terminus to C-terminus the heavy chain of a specific anti-human PCSK9 antibody, a glycine-rich linker, and a variant of human FGF21 protein. Polypeptide B is the light chain of the anti-human PCSK9 antibody. The bifunctional protein preferably comprises two polypeptide A chains and two polypeptide B chains. The amino acid sequence of polypeptide A (also referred to as polypeptide A chain) is

EVQLVESGGGLVKPGGSLRLSCAASGFPFSKLGMVWVRQAPGKGLEWVSTISSG

GGYTYYPDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAREGISFQGGTY

TYVMDYWGQGTLVTVSSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVT

VSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTKTYTCNVDHKPSNTK

VDKRVESKYGPPCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQE

DPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKC

KVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDI

AVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEA

LHNHYTQKSLSLSLGGGGGSGGGGSGGGGSAHPIPDSSPLLQFGGQVRQRYLYT

DDAQQTECHLEIREDGTVGCAADQSPESLLQLKALKPGVIQILGVKTSRFLCQRP

DGALYGSLHFDPEACSFREDLKEDGYNVYQSEAHGLPLHLPGDKSPHRKPAPRG

PARFLPLPGLPPALPEPPGILAPQPPDVGSSDPLRLVEPSQLRSPSFE

(SEQ ID NO: 1).

The amino acid sequence of polypeptide B (also referred to as polypeptide B chain) is

DIVMTQSPLSLPVTPGEPASISCRSSKSLLHRNGITYSYWYLQKPGQSPQLLIYQLS NL AS G VPDRFS GS GS GTDFTLKISR VE AED VG V Y YC YQNLELPLTFGQGTKVEIK RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQES VTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

(SEQ ID NO: 2).

The present invention includes polynucleotides comprising DNA or RNA that encodes polypeptide A, polypeptide B, and both polypeptide A and polypeptide B. The polynucleotides encoding the above-described bifunctional protein may be in the form of RNA or in the form of DNA, which DNA includes cDNA, and synthetic DNA. The DNA may be double-stranded or single-stranded. The coding sequences that encode the bifunctional protein of the present invention may vary as a result of the redundancy or degeneracy of the genetic code. The polynucleotides that encode for the bifunctional protein of the present invention may include the following: only the coding sequence for the protein, the coding sequence for the protein and additional coding sequence such as a leader or secretory sequence or a pro-protein sequence; the coding sequence for the protein and non-coding sequence, such as introns or non-coding sequence 5' and/or 3' of the coding sequence for the protein. Thus the term "polynucleotide encoding a protein" encompasses a polynucleotide that may include not only coding sequence for the protein but also a polynucleotide that includes additional coding and/or non-coding sequence.

The polypeptides of the present invention will be expressed from the

polynucleotides in a host cell after the polynucleotide sequences have been operably linked to an expression control sequence. The expression vectors are typically replicable in the host organisms either as episomes or as an integral part of the host chromosomal DNA. Commonly, expression vectors will contain selection markers, e.g., tetracycline, neomycin, and dihydrofolate reductase, to permit detection of those cells transformed with the desired DNA sequences.

The bifunctional protein of the present invention may readily be produced in mammalian cells such as CHO, NSO, HEK293 or COS cells; in bacterial cells such as E. coli, Bacillus subtilis, or Pseudomonas fluorescence; or in fungal or yeast cells. The host cells are transformed and cultured so as to express the polypeptides using techniques well known in the art. The preferred mammalian host cell is CHO, particularly the CHOKISV cell line containing a glutamine synthetase (GS) selection marker (see US 5,122,464).

Various methods of protein purification may be employed and such methods are known in the art and described, for example, in Deutscher, Methods in Enzymology 182: 83-89 (1990) and Scopes, Protein Purification: Principles and Practice, 3rd Edition, Springer, NY (1994).

The present invention also provides a process for producing a bifunctional protein that comprises polypeptide A and polypeptide B, wherein the amino acid sequence of polypeptide A is SEQ ID NO: 1 and wherein the amino acid sequence of polypeptide B is SEQ ID NO: 2, said process comprising the steps of: i) cultivating in a suitable medium a recombinant mammalian host cell that has been transfected to express polypeptide A and polypeptide B and to secrete the bifunctional protein in the medium under conditions such that each polypeptide is expressed and the bifunctional protein is secreted into the medium; and

ii) recovering the bifunctional protein from the medium.

The present invention also provides the bifunctional protein produced by the above- described processes.

The present invention also provides a pharmaceutical composition comprising a bifunctional protein of the present invention and at least one pharmaceutically acceptable carrier, diluent, or excipient.

The present invention also provides a method of treating type 2 diabetes, obesity, dyslipidemia, and/or metabolic syndrome in a patient comprising administering to the patient a bifunctional protein of the present invention.

The present invention also provides a method of treating type 2 diabetes, obesity, dyslipidemia, and/or metabolic syndrome in a patient comprising administering to the patient a pharmaceutical composition of the present invention.

Furthermore, the present invention provides a bifunctional protein of the present invention for use in therapy. Preferably, the present invention provides a bifunctional protein of the present invention for use in the treatment of type 2 diabetes, obesity, dyslipidemia, and/or metabolic syndrome.

Furthermore, the present invention provides the use of a bifunctional protein of the present invention in the manufacture of a medicament for the treatment of type 2 diabetes, obesity, dyslipidemia, and/or metabolic syndrome.

The bifunctional protein of the present invention is a tetrameric protein composed of two identical polypeptide A chains and two identical polypeptide B chains. Each half of the tetramer consists of one polypeptide A chain associated with one polypeptide B chain to form an AB heterodimer. Two AB heterodimers associate through polypeptide A - polypeptide A interactions to form the tetramer consisting of two polypeptide A chains and two polypeptide B chains. Polypeptides in the tetramer are associated through non- covalent interactions and intermolecular disulfide bonds. More specifically, within each AB heterodimer, the polypeptide A forms inter-chain disulfide bond(s) with the polypeptide B chain at cysteines that are well-known for antibodies. Two AB

heterodimers bond to each other by disulfide bonds at well-known positions in the hinge area. Polypeptides A and B each have intra-chain disulfides well-known for antibodies and polypeptide A also has intra-chain disulfide bonds within the FGF21 portion.

Mammalian cell expression of such polypeptides may result in formation of these disulfide bonds and also in glycosylation sites at a highly conserved N-glycosylation site, for example at a well-known position in the heavy chain portion of polypeptide A.

The pharmaceutical compositions of the bifunctional protein of the present invention may be administered by any means known in the art that achieve the generally intended purpose to treat type 2 diabetes, dyslipidemia, and/or metabolic syndrome. The preferred route of administration is parenteral, in particular intravenous or subcutaneous administration. The dosage administered will be dependent upon the age, health, and weight of the recipient, kind of concurrent treatment, if any, frequency of treatment, and the nature of the effect desired. Typical dosage levels can be optimized using standard clinical techniques and will be dependent on the mode of administration and the condition of the patient and can be determined by a person having ordinary skill in the art.

The bifunctional protein of the present invention is formulated according to known methods to prepare pharmaceutically useful compositions. A desired formulation is a stable lyophilized product that is reconstituted with an appropriate diluent or an aqueous solution of high purity with optional pharmaceutically acceptable carriers, preservatives, excipients or stabilizers [Remington, The Science and Practice of

Pharmacy, 19th edition, Gennaro, ed., Mack Publishing Co., Easton, PA 1995].

The bifunctional protein of the present invention may be formulated with a pharmaceutically acceptable buffer, and the pH adjusted to provide acceptable stability, and a pH acceptable for administration. Moreover, the compositions of the present invention may be placed into a container such as a vial, a cartridge, a pen delivery device, a syringe, intravenous administration tubing or an intravenous administration bag, wherein the container is a unit dose container. "Dyslipidemia" means a disorder of lipoprotein metabolism, including lipoprotein overproduction or deficiency. Dyslipidemia may be manifested by elevation of the total cholesterol, LDL-C and the triglyceride concentrations, and/or a decrease in HDL-C concentration in the blood.

"Metabolic syndrome" is characterized by a group of metabolic risk factors in one person. They include: abdominal fat— in most men, a 40-inch waist or greater; high blood sugar— at least 110 milligrams per deciliter (mg/dl) after fasting; high triglycerides— at least 150 mg/dL in the bloodstream; low HDL-C— less than 40 mg/dl; and/or, blood pressure of 130/85 or higher.

"Obesity" is a condition in which there is an excess of subcutaneous fat in proportion to lean body mass (Stedman' s Medical Dictionary 28th edition, 2006, Lippincott Williams & Wilkins).

A "patient" is a mammal, preferably a human.

The term "treating" (or "treat" or "treatment") means slowing, reducing, or reversing the progression or severity of a symptom, disorder, condition, or disease.

The term "therapeutically effective amount" refers to the amount or dose of a bifunctional protein of the present invention which, upon single or multiple dose administration to a patient, provides the desired treatment.

"Type 2 diabetes" is characterized by excess glucose production in spite of the availability of insulin, and circulating glucose levels remain excessively high as a result of inadequate glucose clearance.

The present invention may be practiced by referencing the following examples. However, this is not to be interpreted as limiting the scope of the present invention.

Example 1

Expression of the bifunctional protein in CHOKISV Cells The bifunctional protein of the present invention is produced in a mammalian cell expression system using CHOKISV cells. Polynucleotides containing cDNAs coding for the two polypeptides of the bifunctional protein are sequenced to verify that the correct sequences will be expressed and once the sequences are verified they are sub-cloned into a Glutamine Synthetase (GS)-containing expression pEE12.4-based plasmids. The cDNA sequence encoding the bifunctional protein is ligated in frame with the coding sequence of signal peptide sequences to enhance secretion of the desired product into the cell culture medium.

The expression is driven by the viral cytomegalovirus (CMV) promoter.

CHOK1SV cells are stably transfected using electroporation and an appropriate amount of recombinant expression plasmid, and the transfected cells are maintained in suspension culture at an adequate cell density. Selection of transfected cells is accomplished by growth in serum-free medium containing methionine sulfoximine (MSX) at 35-37 °C and 5-7 % C0 ₂ .

Clonally-derived cell lines are generated by use of a flow cytometer. Bifunctional proteins secreted into the media from the CHO cells are purified by Protein A affinity chromatography followed by preparative size exclusion chromatography following standard chromatographic techniques. Briefly, bifunctional proteins from harvested media are captured onto Mab Select Protein A (GE, Piscataway, NJ) with PBS pH 7.4 running buffer; briefly washed with running buffer to remove non-specifically bound material; and eluted with 10 mM citrate pH 3.5. Fractions containing bifunctional proteins are pooled and pH is neutralized by adding 1/10 volume of 1M Tris pH 8.0. The neutralized pool is concentrated and loaded onto a Superdex 200 size exclusion chromatography column (GE, Piscataway, NJ) with PBS pH 7.4 mobile phase. Fractions containing the bifunctional protein are pooled, concentrated, and stored.

Alternatively, the cell free media containing bifunctional protein is treated with detergent (Triton X-100) for viral inactivation. The pH of media is adjusted to 6.0 and applied to a Capto MMC column, that is equilibrated in 10 mM citrate, 150 mM NaCl, pH 6. After sample application the resin is washed with equilibration buffer to remove non-specifically bound materials. The bifunctional protein is eluted from the column with pH gradient in 50 mM Tris, pH 8. The Capto MMC mainstream is heated to 55 °C for two hours. Precipitates that form are removed by depth filtration (Millipore). The bifunctional protein is further purified on POROS 50 HQ anion exchange column equilibrated in 50 mM Tris pH 8. Bound proteins are eluted with salt gradient in 20 mM Tris 300 mM NaCl pH 8. Eluted bifunctional protein is further purified by hydrophobic interaction chromatography. The POROS 50 HQ mainstream pool is adjusted to 1 M sodium sulfate and applied to a Phenyl Sepharose HP column equilibrated with 1 M sodium sulfate in 20 mM Tris pH 7. Bifunctional protein is eluted in a reversed salt gradient in 20 mM Tris pH 7. Purified bifunctional protein can be passed through a viral retention filter such as Planova 20N (Asahi Kasei Medical) followed by

concentration/diafiltration into 10 mM citrate, 150 mM NaCl pH 7 using tangential flow ultrafiltration on a regenerated cellulose membrane (Millipore).

Binding Kinetics and Affinity

A surface plasmon resonance (SPR) assay as well known in the art is used to assess the binding kinetics and affinity of the bifunctional protein of the present invention to human and cynomolgus PCSK9. Under physiological buffer conditions (ionic strength and pH) at 25 °C, the bifunctional protein of the present invention binds to human PCSK9 with an average association rate (k _on ) of 9.1 x 10 ⁴ M ^' V ¹ and an average dissociation rate (koff) of 1.5 x lO^ s ^"1 . The average ¾ for human PCSK9 binding for the bifunctional protein of the present invention was determined to be about 1.7 nM. The bifunctional protein of the present invention binds to cynomolgus PCSK9 with an average association rate (k _on ) of 13 x 10 ⁴ M ^' V ¹ and an average dissociation rate (k ₀ ff) of 3.8 x 10 ^" s ^" , resulting in a ¾ for cynomolgus PCSK9 binding of about 3.0 nM. Table 1 shows a summary of results obtained with the bifunctional protein of the present invention using human and cynomolgus PCSK9. These data indicate that the bifunctional protein of the present invention binds with nanomolar affinity to both human and cynomolgus PCSK9 under physiological conditions of pH and ionic strength at 25 °C. Table 1 : Binding Kinetics and Affinity of the bifunctional protein of the present invention to Human and Cynomolgus PCSK9

Assay performed at 25 °C.

Inhibition of PCSK9-induced Reduction of LDL Uptake To determine the effect of a bifunctional protein of the present invention on LDL uptake, HepG2 cells are seeded at 4,000 cells per well in 100 ul of DMEM / F-12 (3: 1) medium supplemented with 5 % human LPDS on poly-D-lysine coated 96-well black plates and incubated at 37 °C in an atmosphere of 5 % C0 ₂ for 18 hours. Human PCSK9 (69 nM) is added to the cells with or without the bifunctional protein of the present invention (3.9 nM to 2000 nM) or, in a separate experiment, a human IgG4 control (2.6 nM to 1333 nM) and pre-incubated with cells for 2 hr at 37 °C. Following the addition of 100 ng/well of fluorescently-labeled LDL (BODIPY-LDL, Invitrogen), the cells are incubated for 4 hr at 37 °C. Cells are fixed in a formalin-free fixative (Prefer;

ANATECH, Ltd.) for 20 min at room temperature. After washing cells twice with PBS, cells are permeablized with PBS buffer containing 0.01 % Triton X- 100 for 15 min at room temperature and stained with 10 ug/mL of propidium iodide to determine total cell number. LDL uptake is determined using an Acumen Explorer™ laser-scanning fluorescence microplate cytometer and expressed as a percentage of fluorescent cells relative to total cells. The response to the bifunctional protein of the present invention or control IgG is expressed as percentage inhibition of PCSK9, i.e., the percent return to maximum LDL uptake in the absence of PCSK9 relative to baseline LDL-C uptake in the presence of PCSK9 alone. Corresponding IC5 ₀ values for inhibition of PCSK9-induced reduction of LDL uptake are also calculated. Following procedures substantially as described above, human PCSK9 caused a concentration-related reduction of LDL uptake in HepG2 cells with an EC5 ₀ of 32 nM. The bifunctional protein of the present invention reversed the PCSK9-induced inhibition, reflected as increased LDL uptake, whereas the control IgG did not reverse the inhibition. Specifically, bifunctional protein of the present invention demonstrated a mean maximum percent inhibition of PCSK9 of 96 % and an average IC ₅₀ of 115 ± 20 nM (mean + SEM, n=4). These data demonstrate that the bifunctional protein of the present invention inhibits PCSK9- induced reduction of LDL uptake.

Human 293 cell- Klotho-SRE luciferase FGF21 Activity Assay

First, 293- Klotho-SRE luciferase (luc) reporter cells are constructed as follows.

HEK-293 cells (human embryonic kidney cells) are cultured at 37 °C, 5 % C02 in growth medium containing 10 % fetal bovine serum in Dulbecco's modified Eagle's medium.

Cells are co-transfected with a plasmid containing a CMV promoter driven human βΚΙοώο expression cassette and a plasmid containing a Serum Response Element (SRE) driven luciferase expression cassette. The βΚΙοώο expression plasmid also contains an

SV40 promoter driven neomycin phosphotransferase expression cassette to confer resistance to the aminoglycoside antibiotic G418. Transfected HEK-293 cells are selected with 600 μg/mL of G418 to select for cells where the transfected plasmids have been integrated into the genome. Selected cells are cloned by dilution and tested for an increase in luciferase production at 24 hours post addition of FGF21. The clone demonstrating the largest FGF21 dependant increase in luciferase is chosen as the cell line used to measure relative FGF21 proteins activity.

The FGF21 activity assay is carried out as follows. 293- Klotho-SRE luc cells are rinsed and placed into CD 293 suspension culture media (Invitrogen). Cells are grown in suspension overnight at 37 °C, 6 % C02, 125 rpm. Cells are counted, pelleted by centrifugation, and re-suspended in CD 293 media containing 0.1 % BSA. Cells are placed in white 96 well plates at 25,000 cells per well. A four-fold serial dilution in CD

293/0.1 % BSA is prepared for each test bifunctional protein to generate eight dilutions with final concentrations from 100 nM to 0.006 nM. Dilutions are added to cells in triplicate and incubated for 16-20 hours at 37 °C, 5 % C02. Lucif erase level is determined by the addition of an equal volume of OneGlo™ luciferase substrate

(Promega) and measuring relative luminescence. Data is analyzed using a four parameter logistic model (XLfit version 5.1) to fit the curves and determine EC ₅₀

Following the protocol essentially as described above, the average in vitro potency (EC5 ₀ ) of the bifunctional protein of the present invention, based on triplicate

determinations on each of two assay plates, is determined to be 0.25 nM.

3T3-Ll- Klotho Fibroblast Glucose Uptake Assay

3T3-Ll- Klotho fibroblasts are generated from 3T3-L1 fibroblasts by retroviral transduction of a CMV-driven mammalian expression vector containing the coding sequence of wild type mouse βΚΙοώο and a Blasticidin resistance marker. Blasticidin- resistant cells are selected after growth for 14 days in the presence of 15 uM Blasticidin, and βΚΙοώο protein expression is verified by immunoblot with an anti- Klotho antibody. The 3T3-Ll- Klotho fibroblasts are maintained in Dulbecco's Modified Eagle Medium (DMEM) with 10 % calf serum, and 15 uM Blasticidin until plated for experimental use.

For glucose uptake, 3T3-Ll- Klotho fibroblasts are plated at 20,000 cells/well in 96-well plates and incubated for 48 hours in DMEM with 10 % calf serum. On the day of assay, the cells are incubated for 3 hours in serum-free DMEM with 0.1 % bovine serum albumin (BSA) with or without the bifunctional protein of the present invention, followed by 1 hour incubation in Krebs-Ringer phosphate buffer (15 mM Hepes, pH 7.4, 118 mM NaCl, 4.8 mM KCl, 1.2 mM MgS04, 1.3 mM CaC12, 1.2 mM KH2P04, 0.1 % BSA)

14

containing 2uCi/ml of 2-deoxy-D-( C) glucose. Non-specific binding is determined by incubation of selected wells in Krebs-Ringer bicarbonate/Hepes buffer containing 20mM of un-labeled 2-deoxy-D-glucose and 2 u Ci/ml of the 2-deoxy-D-( ¹⁴ C) glucose. The reaction is terminated by addition of 200 uM of cytochalasin B to the cells and glucose uptake is measured using a liquid scintillation counter.

The in vitro potency (EC50) of the bifunctional protein of the present invention is 0.35 nM with 95% confidence interval of 0.19 to 0.62. Glucose Lowering in Ob/ob Mice

Male ob/ob mice and age-matched ob/m lean controls are 7 weeks of age upon arrival and 8-9 weeks of age at initiation of treatment. On the day of treatment initiation (day 0), the mice are sorted into groups based on the pretreatment body weight and blood glucose. Blood samples are collected via tail bleed and glucose levels are measured using an Accu-Check Avivia blood glucose meter (Roche). On day 0, mice are dosed with a subcutaneous injection of 10 or 30 nmol/kg of the bifunctional protein of the present invention, in a volume of 10 mL/kg. Dosing vehicle is sterile phosphate buffered saline (HyClone DPBS/Modified - Calcium - Magnesium) containing 0.03 % mouse serum albumin (Sigma A3139). Blood glucose is measured daily for 10 days.

Vehicle treated ob/ob mice were hyperglycemic with mean blood glucose levels measured at 338 + 41 mg/dl (mean + SEM) , while ob/m lean control mice had blood glucose levels of 132 + 4.6 mg/dl (mean + SEM). The bifunctional protein of the present invention lowered blood glucose dose-dependently on day 3 after dosing, reaching a level comparable to the ob/m lean controls. Blood glucose returned to the level observed in vehicle control by day 7 or 8 after a single dose.

LDL-C Lowering Efficacy in Mice Expressing Human PCSK9 The expression of human PCSK9 in mice was induced by injecting an adeno- associated virus (AAV) vector containing the native human PCSK9 cDNA driven by a liver specific promoter (AAV2/8.TBG.PI.hPCSK9.RBG; ReGenX Biosciences). Male C57BL/6 mice, 7 - 8 wk of age, were obtained from Taconic Farms, Germantown, MD. After acclimation for approximately 2 weeks, the mice were anesthetized with isoflurane and given a single injection into the retro-orbital sinus of 2 x 10 ¹⁰ genome copies of the AAV-PCSK9 vector. LDL-C levels increase to 70 to 120 mg/dl, compared to approximately 20 mg/dl in mice injected with the same titre of a LacZ control vector. At 7 - 8 weeks after injection of AAV, the mice were given a single subcutaneous injection of the bifunctional protein of the present invention at a dose of 100 nmol/kg. An additional group of mice were given a single dose of a control human IgG4 at 100 nmol/kg. Blood samples were collected 24 hr after antibody treatment by cardiac puncture following C0 ₂ asphyxiation for assessment of serum LDL-C by HPLC, as described by Kieft et al. (J. Lipid. Res. 32: 859-866, 1991).

Serum LDL-C in the PCSK9-expressing mice treated with the control IgG was 78 + 3 mg/dl (mean + SE), compared to 53 + 5 mg/dl in PCS K9 -expressing mice treated with the bifunctional protein of the present invention, representing a 32 % reduction of LDL-C at 24 hr after dosing.

LDL-C Lowering Efficacy in Cynomolgus Monkeys Four healthy, male cynomolgus monkeys were dosed intravenously with a single injection of 33.3 nmol/kg of the bifunctional protein of the present invention. Serum was obtained from each monkey prior to dosing on days -13, -8,-4 and pre-dose on day 1 to establish a baseline of LDL-C, and over the course of 56 days after dosing of the bifunctional protein of the present invention. The efficacy was determined by comparing pre- and post-dose LDL-C levels of each monkey, and by comparing with LDL-C levels in monkeys receiving a single intravenous dose of control IgG. Serum concentrations of LDL-C were measured by autoanalyzer (Direct LDL-C Plus, 2 ^nd Gen., Roche

Diagnostics).

The average LDL-C on the four pre-dose days was 64 + 3 mg/dl (mean + SEM). The LDL-C lowering efficacy of the bifunctional protein of the present invention was observed within 12 hours of dosing, with the maximum LDL-C reduction between 2 and 5 days post-dose. At maximum, LDL-C was decreased to 37 + 3 mg/dl (mean + SEM) by the bifunctional protein of the present invention, representing a 42 % reduction from the pre-dose level. LDL-C returned to baseline and control IgG levels on day 21.

Sequences

SEP ID NO: 1 - Polypeptide A

EVQLVESGGGLVKPGGSLRLSCAASGFPFSKLGMVWVRQAPGKGLEWVSTISSG

GGYTYYPDSVKGRFTISRDNAKNSLYLQMNSLRAEDTAVYYCAREGISFQGGTY

TYVMDYWGQGTLVTVSSASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVT

VSWNSGALTSGVHTFPAVLQSSGLYSLSSVVTVPSSSLGTKTYTCNVDHKPSNTK

VDKRVESKYGPPCPPCPAPEAAGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSQE

DPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTVLHQDWLNGKEYKC

KVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQEEMTKNQVSLTCLVKGFYPSDI

AVEWESNGQPENNYKTTPPVLDSDGSFFLYSRLTVDKSRWQEGNVFSCSVMHEA

LHNHYTQKSLSLSLGGGGGSGGGGSGGGGSAHPIPDSSPLLQFGGQVRQRYLYT

DDAQQTECHLEIREDGTVGCAADQSPESLLQLKALKPGVIQILGVKTSRFLCQRP

DGALYGSLHFDPEACSFREDLKEDGYNVYQSEAHGLPLHLPGDKSPHRKPAPRG

PARFLPLPGLPPALPEPPGILAPQPPDVGSSDPLRLVEPSQLRSPSFE

SEP ID NO: 2 - Polypeptide B

DIVMTQSPLSLPVTPGEPASISCRSSKSLLHRNGITYSYWYLQKPGQSPQLLIYQLS NLAS G VPDRFS GS GS GTDFTLKISR VE AED VG V Y YC YQNLELPLTFGQGTKVEIK RTVAAPSVFIFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQES VTEQDSKDSTYSLSSTLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC

SEP ID NP: 3 - DNA of Polypeptide A

GAGGTGCAGCTGGTGGAGTCTGGGGGAGGCTTGGTAAAGCCTGGGGGGTCCC

TGAGACTCTCCTGTGCAGCCTCTGGATTCCCGTTCAGTAAGCTCGGCATGGTT

TGGGTCCGCCAGGCTCCAGGGAAGGGGCTGGAGTGGGTCTCAACCATTAGTA

GTGGTGGTGGTTACACATACTATCCAGACAGTGTGAAGGGGCGGTTCACCAT

CTCCAGAGACAATGCCAAGAACTCACTGTATCTGCAAATGAACAGCCTGAGA

GCCGAGGACACGGCCGTATATTACTGTGCGAGAGAAGGAATTAGCTTTCAGG

GTGGCACCTACACTTATGTTATGGACTACTGGGGCCAGGGCACCCTGGTCACC GTCTCCTCAGCCTCCACCAAGGGCCCATCGGTCTTCCCGCTAGCGCCCTGCTC

CAGGAGCACCTCCGAGAGCACAGCCGCCCTGGGCTGCCTGGTCAAGGACTAC

TTCCCCGAACCGGTGACGGTGTCGTGGAACTCAGGCGCCCTGACCAGCGGC

GTGCACACCTTCCCGGCTGTCCTACAGTCCTCAGGACTCTACTCCCTCAG

CAGCGTGGTGACCGTGCCCTCCAGCAGCTTGGGCACGAAGACCTACACCT

GCAACGTAGATCACAAGCCCAGCAACACCAAGGTGGACAAGAGAGTTGAG

TCCAAATATGGTCCCCCATGCCCACCCTGCCCAGCACCTGAGGCCGCCGG

GGGACCATCAGTCTTCCTGTTCCCCCCAAAACCCAAGGACACTCTCATGA

TCTCCCGGACCCCTGAGGTCACGTGCGTGGTGGTGGACGTGAGCCAGGAA

GACCCCGAGGTCCAGTTCAACTGGTACGTGGATGGCGTGGAGGTGCATAA

TGCCAAGACAAAGCCGCGGGAGGAGCAGTTCAACAGCACGTACCGTGTGG

TCAGCGTCCTCACCGTCCTGCACCAGGACTGGCTGAACGGCAAGGAGTAC

AAGTGCAAGGTCTCCAACAAAGGCCTCCCGTCCTCCATCGAGAAAACCAT

CTCCAAAGCCAAAGGGCAGCCCCGAGAGCCACAGGTGTACACCCTGCCCC

CATCCCAGGAGGAGATGACCAAGAACCAGGTCAGCCTGACCTGCCTGGTC

AAAGGCTTCTACCCCAGCGACATCGCCGTGGAGTGGGAAAGCAATGGGCA

GCCGGAGAACAACTACAAGACCACGCCTCCCGTGCTGGACTCCGACGGCT

CCTTCTTCCTCTACAGCAGGCTAACCGTGGACAAGAGCAGGTGGCAGGAG

GGGAATGTCTTCTCATGCTCCGTGATGCATGAGGCTCTGCACAACCACTA

CACACAGAAGAGCCTCTCCCTGTCTCTGGGTGGTGGTGGTGGCTCCGGAG

GCGGCGGCTCTGGTGGCGGTGGCAGCGCTCACCCCATCCCTGACTCCAGT

CCTCTCCTGCAATTCGGGGGCCAAGTCCGGCAGCGGTACCTGTACACCGA

CGACGCCCAGCAGACCGAGTGCCACCTGGAAATCCGGGAGGACGGCACCG

TGGGCTGTGCCGCCGACCAGTCCCCTGAGTCCCTGCTGCAGCTGAAGGCC

CTGAAGCCTGGCGTGATCCAGATCCTGGGCGTGAAAACCTCCCGGTTCCT

GTGCCAGAGGCCTGATGGCGCCCTGTACGGCTCCCTGCACTTCGACCCTG

AGGCCTGCTCCTTCCGGGAGGACCTGAAGGAAGATGGCTACAACGTGTAC

CAGTCCGAGGCTCACGGCCTGCCTCTGCATCTGCCTGGCGACAAGTCCCC

CCACCGGAAGCCTGCTCCTAGGGGCCCTGCCAGATTCCTGCCACTGCCTG

GCCTGCCTCCAGCTCTGCCTGAGCCTCCTGGCATCCTGGCCCCTCAGCCT CCAGACGTGGGCTCCTCCGACCCTCTGCGGCTGGTCGAGCCTTCCCAGCT GCGGAGCCCTAGCTTCGAG

SEP ID NO: 4 - DNA of Polypeptide B

GATATTGTGATGACTCAGTCTCCACTCTCCCTGCCCGTCACCCCTGGAGAGCC

GGCCTCCATCTCCTGCAGGTCTAGTAAGAGTCTCTTACATCGTAATGGCATCA

CTTATTCGTATTGGTACCTGCAGAAGCCAGGGCAGTCTCCACAGCTCCTGATC

TATCAGCTGTCCAACCTTGCCTCAGGAGTCCCAGACAGGTTCAGTGGCAGTGG

GTCAGGCACTGATTTCACACTGAAAATCAGCAGGGTGGAGGCTGAGGATGTT

GGAGTTTATTACTGCTATCAAAATCTAGAACTTCCGCTCACGTTCGGCCAGGG

CACCAAGGTGGAAATCAAACGGACTGTGGCTGCACCATCTGTCTTCATCTTCC

CGCCATCTGATGAGCAGTTGAAATCTGGAACTGCCTCTGTTGTGTGCCTGCT

GAATAACTTCTATCCCAGAGAGGCCAAAGTACAGTGGAAGGTGGATAACG

CCCTCCAATCGGGTAACTCCCAGGAGAGTGTCACAGAGCAGGACAGCAAG

GACAGCACCTACAGCCTCAGCAGCACCCTGACGCTGAGCAAAGCAGACTA

CGAGAAACACAAAGTCTACGCCTGCGAAGTCACCCATCAGGGCCTGAGCT

CGCCCGTCACAAAGAGCTTCAACAGGGGAGAGTGC