Neuregulin-4 compounds and methods of use

Abstract

The present invention relates to neuregulin (NRG) 4 compounds and methods of treatment with NRG4 compounds.

Claims

1. A compound comprising the formula: GHEEPCGX.sub.8SHKSFCLNGGLCYX.sub.22IPTX.sub.26PSPFCRCVX.sub.35NYTGARCEX.sub.44VFL wherein: X.sub.8 is P; X.sub.22 is Q; X.sub.26 is I; X.sub.35 is E; and X.sub.44 is K (SEQ ID NO:2) and wherein the compound optionally comprises an N-terminal extension selected from the group consisting of T, PT, MPT, S, GS, GGS, GGGS (SEQ ID NO:20), and (GGGGX.sub.λ).sub.n wherein X.sub.λ is Q, A, E or S and n=1-5 (SEQ ID NO:5).

2. The compound of claim 1 wherein the N-terminal extension is SEQ ID NO:5 wherein X.sub.λ is S and n=1.

3. The compound of claim 1 wherein the compound comprises the amino acid sequence of SEQ ID NO:18.

4. A compound consisting of the amino acid sequence of SEQ ID NO:18.

5. The compound of claim 1, wherein the compound has HER4 binding-related activity which is greater than that of native human NRG4 and at least 70% that of the maximal activity of native human NRG1.

6. The compound of claim 5, wherein the compound has no HER3 binding-related activity.

7. A pharmaceutical composition comprising the compound of claim 1 and a pharmaceutically acceptable carrier, diluent, or excipient.

8. A pump device comprising the compound of claim 1.

Description

PREPARATION OF EXAMPLE COMPOUNDS

(1) Examples of NRG4 compounds of the present invention are described below in Table 2.

(2) TABLE-US-00006 TABLE 2 Components of Example NRG4 compounds. Exam- NRG4 ple Modifications N-terminal AAs/Linker Sequence ID 1 D1G SEQ ID NO: 5 wherein SEQ ID NO: 10 X.sub.λ is S and n is 3 2 D1G/P8E SEQ ID NO: 5 wherein SEQ ID NO: 11 X.sub.λ is S and n is 3 3 D1G/V22Q SEQ ID NO: 5 wherein SEQ ID NO: 12 X.sub.λ is S and n is 3 4 D1G/I26F SEQ ID NO: 5 wherein SEQ ID NO: 13 X.sub.λ is S and n is 3 5 D1G/E35A SEQ ID NO: 5 wherein SEQ ID NO: 14 X.sub.λ is S and n is 3 6 D1G/I26F/E44K SEQ ID NO: 5 wherein SEQ ID NO: 15 X.sub.λ is S and n is 3 7 D1G/I26F/E44H SEQ ID NO: 5 wherein SEQ ID NO: 16 X.sub.λ is S and n is 3 8 D1G/I26F/P8E SEQ ID NO: 5 wherein SEQ ID NO: 17 X.sub.λ is S and n is 3 9 D1G/V22Q/E44K SEQ ID NO: 5 wherein SEQ ID NO: 18 X.sub.λ is S and n is 1 10 D1G/V22Q/E44K SEQ ID NO: 5 wherein SEQ ID NO: 19 X.sub.λ is S and n is 3 11 D1G/V22Q/E44K SEQ ID NO: 5 wherein SEQ ID NO: 19 X.sub.λ is S and n is 3 Complete amino acid sequences for the Example compounds, excluding any IgG Fc region sequences, are provided at the SEQ ID NOs listed in the fourth column. Examples 1-8 and 10-11 further comprise IgG Fc regions. The IgG Fc regions of Examples 1-8 and 10 comprise a dimer of SEQ ID NO: 6 and SEQ ID NO: 7 wherein the N terminal amino acid of the sequence identified in the third column of the table above is fused to the C-terminal amino acid of SEQ ID NO: 7. The IgG Fc region of Example 11 comprises a dimer of SEQ ID NO: 8 and SEQ ID NO: 9 wherein the N terminal amino acid of the sequence identified in the third column of the table above is fused to the C-terminal amino acid of SEQ ID NO: 9.

(3) Biological Expression

(4) Examples of NRG4 compounds comprising Fc regions are produced in a mammalian cell expression system using a CHO GSKO cell line. The GS gene knockout enables tightened selection stringency by eliminating endogenous GS background activity which can allow the survival of low- or non-productive cells under selection conditions. Genes coding for the Fc fusion knob chain and hole chain of the present invention may be sub-cloned into individual glutamine synthetase (GS)-containing expression plasmids for co-transfection or both chains may be sub-cloned into a single GS-containing expression plasmid. Alternatively, different ratios of knob chain and hole chain may be combined into a single GS-containing expression if there is need to alter relative expression levels of either chain. The cDNA sequence encoding the knob chain or hole chain is fused in frame with the coding sequence of a signal peptide, which may be the murine kappa leader sequence, to enhance secretion of the desired product into the cell culture medium. The expression is driven by the viral cytomegalovirus (CMV) promoter.

(5) CHO GSKO cells may be transiently or stably transfected. For stable transfection, CHO GSKO cells are transfected using electroporation and the appropriate amount of recombinant knob chain and hole chain expression plasmids, and the transfected cells are maintained in suspension culture, at the adequate cell density. Selection of the transfected cells is accomplished by growth in glutamine-free, 25 μM methionine sulfoximine (MSX)-containing serum-free medium and incubated at 32-37° C. and 5-7% CO2. Fc fusions are secreted into the media from the CHO cells.

(6) Proteins are purified using either: (1) Protein A affinity chromatography followed by cation exchange or hydrophobic interaction chromatography (or other suitable methods); or (2) multimodal chromatography followed by hydrophobic interaction chromatography (or other suitable methods).

(7) For purification starting with protein A chromatography, proteins from harvested media are captured onto Mab Select SuRe Protein A resin (GE Healthcare). The resin is then briefly washed with a running buffer, such as a phosphate buffered saline (PBS), pH 7.4 or a running buffer containing Tris, to remove non-specifically bound material. The protein is then eluted from the resin with a low pH solution, such as 20 mM acetic acid/5 mM citric acid. Fractions containing Fc fusion are pooled. The pH can be increased as needed by adding a base such as 0.1 M Tris pH 8.0. At this stage Fc fusions may be used to screen binding/activity, or if desired may be further purified by hydrophobic interaction chromatography using resins such as Phenyl Sepharose HP. Fc fusions can be eluted from the Phenyl Sepharose HP column using a 500 mM to 0 mM gradient of sodium sulfate in 10 mM sodium phosphate, pH 7 over 10 column volumes. The Fc fusions may be further purified by size exclusion chromatography by using a Superdex 200 column (GE Healthcare) with isocratic elution in PBS, pH 7.4 or buffer exchanged into the desired buffer. Examples 1-8 and 10 are purified in this manner.

(8) For purification starting with multimodal chromatography, proteins from harvested media are captured onto CaptoMMC resin (GE Healthcare). The resin is then briefly washed with 100 mM citrate, pH 5.0 (“A” buffer) prior to an 80%/20% wash of “A” buffer and 25 mM sodium phosphate, 1 M sodium chloride, pH 7.5 (“B” buffer). The desired protein is then eluted from the resin at 70% “B” buffer. Fc fusions may be further purified by hydrophobic interaction chromatography using resins such as Phenyl Sepharose HP (GE Healthcare). Fc fusions can be eluted from the Phenyl Sepharose HP column using a 800 mM to 0 mM linear gradient of sodium sulfate in 20 mM Tris, pH 8. The Fc fusions may be further purified by ion exchange chromatography by using a Q Sepharose column (GE Healthcare). Fc fusions can be eluted from the column using a 0 M to 1 M sodium chloride gradient in 20 mM Tris, pH 8. Fractions are pooled and buffer exchanged into the desired buffer for storage. Example 11 is purified in this manner.

(9) Chemical Synthesis

(10) NRG4 compounds of the present invention which are peptides that do not comprise IgG Fc regions (e.g., Example 9 in Table 2 above) may also be generated by solid-phase peptide synthesis using Fmoc/t-Bu strategy. The peptides are synthesized on a SymphonyX automated peptide synthesizer (PTI Protein Technologies Inc.).

(11) Fmoc-L-Leu-Wang resin (0.3-0.8 mmole/gram, 200-400 mesh, Chem-Impex) is used for synthesizing the peptide. Fmoc deprotection is carried out using a 20% v/v solution of piperidine in DMF. Amino acid couplings are performed using 10 equivalents of Fmoc-amino acid, 0.9 M diisopropylcarbodiimide (DIC) and 0.9 M Oxyma (1:1:1 molar ratio) in DMF for 2 h at 25° C. Washing steps are performed with DMF and are included after every coupling and deprotection step. Additional details are provided below in Table 3.

(12) TABLE-US-00007 TABLE 3 Fmoc-deprotection and coupling protocol. Step Solvent/Operation Mixing Time Repetitions 1 DMF 00:01:00 2 2 25% Piperidine/DMF 00:10:00 3 3 DMF 00:01:00 1 4 DMF 00:01:00 5 5 Methylene Chloride 00:01:00 1 6 Methylene Chloride 00:00:02 1 7 Reagent (Amino Acid) 00:00:01 1 8 0.9M Oxyma in DMF 00:00:01 1 9 0.9M DIC in DCM 03:00:00 1 10 DMF 00:01:00 3

(13) All amino acids used in the main sequence are L amino acids: Fmoc-Ala-OH, Fmoc-Cys(Trt)-OH, Fmoc-Asp(Otbu)-OH, Fmoc-Glu(Otbu)-OH, Fmoc-Phe-OH, Fmoc-Gly-OH, Fmoc-His(Trt)-OH, Fmoc-Ile-OH, Fmoc-Lys(Boc)-OH, Fmoc-Leu-OH, Fmoc-Met-OH, Fmoc-Asn(Trt)-OH, Fmoc-Pro-OH, Fmoc-Gln(Trt)-OH, Fmoc-Arg(Pbf)-OH, Fmoc-Ser(tBu)-OH, Fmoc-Thr(tBu)-OH, Fmoc-Tyr(tBu)-OH, Fmoc-Val-OH, Fmoc-Trp(Boc)-OH.

(14) After finishing the elongation of the peptide-resin described above, the final Fmoc group is removed and resin is washed with methylene chloride before subjecting the peptides to cleavage. 1.5 mL solution of water, triisopropylsilane, thioanisole, 1,2 ethanedithiol (1:1:1) added into 13 mL of trifluoroacetic acid (TFA) and the resultant cleavage cocktail is added to the resin and mixed for 2.5 h in a reaction syringe. TFA solution containing cleaved peptide is transferred into a 50 mL conical vial containing cold diethyl ether. Precipitated peptide solution is spun down to a thick pellet and ether is decanted. Ether precipitation is repeated twice to wash the residual cocktail before proceeding to refolding.

(15) The crude peptide is dissolved in water and acetonitrile is added if necessary. The crude peptide concentration is adjusted to 2.5-3 mg/mL (200 mL) in acetonitrile and water. The sample is then subjected to refolding directly. Refolding buffer is prepared by adding oxidized glutathione and reduced glutathione at the ratios, GSSG:GSH (2 mM:1 mM) in 0.1 M tris buffer, pH 8. Crude peptide solution is added into refolding buffer to the concentration of 0.13-0.15 mg/mL (diluted 20 fold) and the refolding mixture was kept at 4° C. without stirring. After 1 day, the refolding mixture is quenched by adding TFA up to 0.2% in order to bring the pH to 3.0. The quenched solution is then filtered and purified.

(16) The peptide solution is loaded onto a HPLC column and then equilibrated with buffer A before triggering the gradient. Buffer A: 0.1% TFA in water; Buffer B: 100% acetonitrile. Gradient: 10% B to 25% B in 100 min. Flow rate: 18 mL/min; detection at 220 nm (Waters 2489 Detector). Column: Waters Symmetry Prep C18 column, 7 μm, 19×300 mm (Waters Part #WAT066245). Fractions auto triggered by UV and collected using Waters Fraction Collector III.

(17) The pooled fractions containing the desired peptide are combined and lyophilized.

In Vitro Studies

(18) HER4/NRG-Fc Binding Assay for Screening of NRG4 Variants for Improved Binding

(19) A binding assay is utilized to perform high-throughput screening of NRG4 variant Fc fusion proteins to HER4 receptor, to identify affinity-driven variants with the potential to increase receptor activation and signaling. ELISA plates are coated with an anti-His tag antibody at 2 μg/ml in PBS to facilitate the standardized capture of a C-terminal His-tagged HER4 receptor extracellular domain (ECD) construct after blocking the plate with 1% BSA/PBS/0.1% Tween 20 (HER4 ECD at 5 μg/ml). Saturating amounts of unpurified NRG4-Fc variants in expression media (Starting at 10 μg/ml with serial dilutions in 1% BSA/PBS/0.1% Tween 20) are incubated with the captured receptor for 1 hour. After washing, bound variant NRG4-Fc is detected colorimetrically with an anti-human Fc-specific alkaline phosphatase secondary detection reagent to quantitate binding. Incubation times of variant binding to receptor or durations of the subsequent wash steps are varied to further identify binding variants driven by affinity on-rate and off-rate, respectively. This screening process identifies the D1G modification, as included in Example 1, as having high affinity and potential to increase receptor activation and signaling, as indicated in the data below in Table 4.

(20) TABLE-US-00008 TABLE 4 Binding to HER4 extracellular domain. Dose Example 1 WT NRG1 WT NRG4 # dose (μg/mL) O.D. 405 dose (μg/mL) O.D. 405 dose (μg/mL) O.D. 405 1 10.65 1.485 10 1.501 17 0.651 2 5.325 1.438 5 1.439 8.5 0.597 3 2.6625 1.4 2.5 1.399 4.25 0.485 4 1.33125 1.332 1.25 1.31 2.125 0.347 5 0.665625 1.217 0.625 1.27 1.0625 0.224 6 0.332813 1.081 0.3125 1.022 0.53125 0.154 7 0.166406 0.865 0.15625 0.8 0.265625 0.111 8 0.083203 0.607 0.078125 0.514 0.132813 0.088 9 0.041602 0.404 0.039063 0.316 0.066406 0.071 10 0.020801 0.235 0.019531 0.179 0.033203 0.07 11 0.0104 0.136 0.009766 0.114 0.016602 0.066 12 0.0052 0.096 0.004883 0.084 0.008301 0.066

(21) As seen in Table 4, Example 1, which includes the D1G modification, has similar HER4 binding to WT NRG1 at similar doses, and significantly greater HER4 binding than WT NRG4 at similar doses in this study.

(22) Generation of CHO Human and Rat HER2-4 and HER2-3 Cell Lines

(23) A cell based assay utilizing CHO-K1 cells overexpressing human or rat HER receptors is used to determine activity of NRG4 compounds. CHO-K1 cells (ATCC) are grown in DMEM-F12 3:1 with 5% FBS with 20 mM HEPES, 40 μg/mL L-Proline, lx antibiotics and were split 1:5 every 2-3 days with TripLE Express (Gibco). Cells are transfected with plasmid DNA encoding HER receptors in pairs (hHER2-3, hHER2-4, rHER2-3 and rHER2-4) with Fugene 6 (Promega) according to the manufacturer's instructions. HER2 is included, because this receptor does not bind ligand but is the preferred dimerization partner of the HER receptors for downstream signaling. Transfected cells are selected with Puromycin (12 ug/ml) and Hygromycin (1 mg/ml) for 3-4 weeks and clonal lines are obtained by limited dilution cloning into 96 well plates. Clonal lines are selected by appropriate gene expression of HER2, HER3 or HER4 by Taqman method. Lines with appropriate receptor expression are confirmed by NRG1 induced phospho-ERK 1/2 responses. Clones are grown up, harvested, aliquoted into cryovials and then frozen under liquid nitrogen for long-term storage.

(24) Screening of Molecules by Human HER2-HER4 Mediated ERK 1/2 Phosphorylation In Vitro

(25) A phospho-ERK1/2 activity assay is used to determine potencies of NRG4 compounds. The potency of Examples 1-8 and 10 as compared to WT human NRG1 and/or human NRG4 are analyzed in studies measuring stimulation of human HER2-HER4 mediated ERK 1/2 phosphorylation. This assay comprises a CHO-K1 stable line expressing human HER2 and HER4. CHO-hHER2-hHER4 (clone 1H6) cells are routinely cultured in DMEM: F12(3:1), 5% FBS, 40 μg/mL L-Proline, 10 μg/mL puromycin and 1 mg/mL hygromycin at 37° C., 5% CO.sub.2. Cells are washed twice with 1×PBS, dissociated with 0.05% trypsin/0.53 μM EDTA and collected by centrifugation at 300×g for 10 m. Cells are re-suspended in maintenance media. 20,000 cells/well, in 30 μL, are seeded into 384 well Poly-D-Lysine cell culture plates (Greiner, Cat No. 781946). Plates are covered with a manufacturer's lid and moved to a 37° C. tissue culture incubator at 5% CO.sup.2, 80%+ humidity and allowed to adhere. After 24 h, media are removed from the plates and replaced with 30 μL serum-free starvation media (Low glucose DMEM with 0.1% BSA) using a Biomek FX liquid handler. Plates are returned to the incubator for 24 h.

(26) hFcNRG4 variants, transiently expressed as 293F supernatants, are prepared in starvation media in a clear 384W plate (Greiner, cat no. 781185) as a 4pt 1:8 serial dilution (dose range averaged between 250 pM to 140000 pM). Media are removed from the cell culture plates and the 30 uL/well prepared variants are stamped into the plates using a Biomek FX liquid handler. Plates are sealed with a foil seal and stimulated for 15 minutes at room temperature on an orbital shaker. Cells are then washed once with 50 uL cold PBS. 30 uL lysis buffer is added to the cells and oscillated for 10 m at RT on an orbital shaker.

(27) Detection of phosphorylation of ERR 1/2 is performed with either an AlphaScreen SureFire phospho-ERK1/2 kit (T202/Y204) (Perkin Elmer, catalog no. TGRES) or an AlphaLISA® SureFire® Ultra™ p ERK1/2 assay kit together with the AlphaScreen Protein. A IgG Detection Kit (Protein-A) (PerkinElmer, catalogue no. 6760617). After lysis of the CHO-K1 cells. 4 μL lysate is transferred to a 384W Alpha Proxiplate (Perkin Elmer, cat no, 6008280) with a 384W Biomek FX. 4 uL positive and negative lysates (Perkin Elmer, TGRES-L) are added to available wells as controls. 5 uL/well acceptor mix (containing anti-phoso-Thr202/Tyr204 antibodies and Protein A conjugated acceptor beads) is added to the plate with a Forumlatrix Mantis. The plate is sealed, spun down for 1 min at 300×g and incubated for 2 hr at room temperature on an orbital shaker. 2 μL/well donor mix (containing strepavidin coated donor beads and biotinylated antibodies to a distal ERK1/2 epitope) is dispensed by Mantis. The plate is sealed, spun down for 1 min at 300×g and incubated for 2 hr at room temperature on an orbital shaker before reading on the Perkin Elmer Envision 2103 Multilabel Reader (HIS Alpha mode, excitation time: 40 ms, total measurement time: 130 ms).

(28) Each sample is plotted against the parental hFcNRG4 dose response, included on every plate in replicate in XLFit. This data is plotted in GraphPad Prism with a four parameter fit, and EC50 potency values are extracted. Results are provided in Tables 5 and 6.

(29) TABLE-US-00009 TABLE 5 Activity of Examples 1-5 as compared to WT NRG4. Data for Examples 1-5 represent an average of two replicates with standard deviations indicated. Sample EC50 (nM) WT NRG4 2.53 Example 1 0.77 ± 0.16 Example 2 2.77 ± 0.39 Example 3 0.29 ± 0.01 Example 4 0.53 ± 0.05 Example 5 0.84 ± 0.27

(30) As seen in the data in Table 5, consistent with its increased affinity to HER4, in this study the D1G modification also significantly increases activity at HER4 and HER2 as compared to wild type NRG4, including in examples which include additional amino acid modifications.

(31) TABLE-US-00010 TABLE 6 Activity of Examples 6-8 and 10 as compared to WT NRG1 and NRG4. Sample EC50 (nM) WT NRG1 0.3192 WT NRG4 1.2980 Example 6 0.3947 Example 7 0.5629 Example 8 0.4033 Example 10 0.2448

(32) As seen in the data in Table 6, in this study Examples 6-8 and 10, each of which includes the D1G modification and additional modifications, show increased activity at HER4 and HER2 as compared to WT NRG4 and similar activity as compared to WT NRG1.

(33) Human and Rat HER2-4 or HER2-3 Phospho-ERK 1/2 (Thr202/Tyr204) Assays for Testing of Purified Proteins

(34) A phospho-ERK1/2 activity assay is utilized with CHO-K1 cells overexpressing human or rat HER receptors to determine potencies and selectivity of purified NRG4 compounds. CHO-K1 cell lines expressing the human or rat HER receptors are cultured with selection medium (DMEM-F12 3:1 with 5% FBS with 20 mM HEPES, 40 μg/mL L-Proline, 1× antibiotics, 12 μg/mL puromycin, 1 mg/mL Hygromycin B). On Day −1 (the day before phospho-ERK1/2 assay), cells are washed once with PBS, detached with enzyme free cell dissociation solution (GIBCO cat #13151-014), and resuspended in plating medium (DMEM-F12 3:1 with 20 mM HEPES, lx antibiotics, 0.2% FBS). Cells are plated in a 96-well Poly-D-Lysine coated plate (BD cat #354640) at 10,000 cells per 0.1 mL per well. Cells are cultured in a tissue culture incubator at 37° C. 5% CO2 overnight. On Day 1 (the day of pERK1/2 assay), plates are incubated at room temperature for 30 min. The culture media are removed followed by addition of 504, ligand at various concentrations diluted in PBS-20 mM HEPES-0.005% BSA. Native human NRG1 and NRG4 peptides are purchased from Reprokine (Tampa, Fla.). The stimulation time is 30 min (except for hHER2-4 1H6 line in which 15 min is used for stimulation). At the end of stimulation, ligands are removed and 704, Lysis buffer (made according to manufacturer's recipe, Perkin Elmer cat #ALSU-PERK-A10K) is added. Plates are incubated at room temperature for 10 min on plate shaker with 350 rpm agitation, then on ice for 1 hour without shaking. 304, of cell lysis is transferred to a 96-well white Optiplate (Perkin Elmer cat #6002290) for phospho-ERK1/2 assay. Briefly 7.5 μL acceptor beads are added to 30 μL lysis in Optiplate. The plate is covered with aluminum foil and incubated at room temperature for 1 hour on plate shaker with 350 rpm agitation. Then 7.5 μL donor beads are added and the plate is covered with aluminum foil and incubated at room temperature for 2 hours on plate shaker with 350 rpm agitation or at 4° C. for overnight (warm up plate at room temperature for 2 h before reading). Alpha signal data are obtained on an Envision instrument (Alpha Technology-compatible plate reader). Donor beads and acceptor beads are made according to assay protocol. The data shown are generated with either the AlphaScreen SureFire p-ERK1/2 (Thr202/Tyr204) assay kit (Cat #TGRES50k) or the AlphaLISA® SureFire® Ultra™ p-ERK 1/2 (Thr202/Tyr204) assay kit (Perkin Elmer cat #ALSU-PERK-A10K). The activity of the NRG1 peptide was similar in both types of assays. Raw data obtained from the Envision instrument is imported into GraphPad Prism software (version 7). The EC50 value is generated by a variable slope-four parameter dose response curve. The % of maximal NRG1 activity data are generated by expressing the average maximal raw data for the NRG4 compound over the average maximal raw data for the native human NRG1 per plate and multiplying by 100. Data are provided below in Tables 7-9. N values reflect number of independent assays run.

(35) TABLE-US-00011 TABLE 7 Human HER2/HER4 Assay. Sample EC50 (nM) % Maximal NRG1 activity WT NRG1 0.67 ± 0.07 (N = 9) 100 WT NRG4 90.09 ± 33.21 (N = 8) 75 ± 4 (N = 8) Example 9 0.20 ± 0.01 (N = 11)  91 ± 3 (N = 11) Example 10 0.16 ± 0.02 (N = 8) 82 ± 4 (N = 8) Example 11 0.55 ± 0.11 (N = 5) 79 ± 2 (N = 5)

(36) TABLE-US-00012 TABLE 8 Rat HER2/HER4 Assay. Sample EC50 (nM) % Maximal NRG1 activity WT NRG1 0.29 ± 0.05 (N = 4) 100 Example 9 0.17 ± 0.02 (N = 4) 87 ± 4 (N = 4) Example 10 0.12 ± 0.04 (N = 2) 88 ± 7 (N = 2) Example 11 0.17 ± 0.06 (N = 2) 96 ± 2 (N = 2)

(37) As seen in Table 7, wild type human NRG1 peptide exhibits potent activity at the human HER4 and HER2 receptors while wild type human NRG4 peptide behaves as a weak partial agonist. Examples 9-11 exhibit potent activity with EC50 values slightly more potent than wild type human NRG1 and % maximal activities greater than wild type human NRG4 peptide and approaching that of human NRG1. As seen in Table 8, similar results are obtained at the rat HER4 HER2 receptors with slight increases in potency relative to human NRG1 and maximal activity approaching that of human NRG1. In conclusion, the examples shown of NRG4 compounds exhibit greater phospho-ERK1/2 activity at both human and rat HER4 HER2 receptors relative to native human NRG4 and approaching that of native human NRG1.

(38) TABLE-US-00013 TABLE 9 Human and Rat HER2/HER3 Assay Human HER2/HER3 Assay Rat HER2/HER3 Assay Sample EC50 (nM) EC50 (nM) WT NRG1 0.83 ± 0.20 (N = 6) 32 ± 14 (N = 2) WT NRG4 No activity up to nd 1000 nM (N = 1) Example 9 No activity up to No activity up to 3000 nM (N = 6) 3000 nM (N = 4) Example 10 No activity up to No activity up to 1000 nM (N = 3) 100 nM (N = 2) Example 11 No activity up to No activity up to 1000 nM (N = 1) 1000 nM (N = 2)

(39) As seen in Table 9, wild type human NRG1 peptide exhibits potent activity at the human HER3 HER2 receptors while wild type human NRG4 peptide and all the Examples show no activity at this receptor pair demonstrating differential selectivity to the HER4 HER2 receptor pair. In conclusion, the examples of NRG4 compounds demonstrate no phospho-ERK1/2 activity at HER3 HER2 receptors thus showing maintenance of the HER4 and HER2 receptor selectivity of the wild type NRG4 peptide.

In Vivo Studies

(40) Effects of Example 9 on Cardiac Function and Structure in Rat MI Model

(41) Two nonclinical efficacy pharmacology studies are conducted in a rat model of heart failure with reduced ejection fraction (HFrEF) to study plasma concentration and duration of exposure for Example 9. Both studies measure effects on cardiac function and structure in male Sprague Dawley rats with surgically induced myocardial infarction.

(42) Similar methodologies are used for both studies. Male Sprague Dawley rats with surgically induced myocardial infarction are purchased and anesthetized and positively ventilated throughout the procedure. An incision is made between the fourth and fifth ribs, revealing the heart. The left coronary artery is permanently ligated. The sham-operated animals undergo the same procedure except the silk suture is placed around the left coronary artery without being tied. All rats are individually housed in a temperature and humidity controlled room and maintained on a 12 hour light/dark cycle. Two to three weeks after surgery, rats undergo transthoracic echocardiography for determination of ejection fraction (EF %) and left ventricle dimensions (LVD) using a Vevo 2100 ultrasound system. Rats are randomized across treatment groups according to EF % and LVD.

(43) Measurements of EF % and LVD are expressed as mean values±standard error (SE). Statistical analysis is performed with JMP® 13 software (SAS Institute, Inc.; Cary, N.C.) and Dunnett's Test is used for statistical comparisons across treatment groups. Statistical significance is accepted at P<0.05.

(44) Study Design for Determination of Example 9 Plasma Concentration for Nonclinical Efficacy.

(45) At an infusion rate of 10 μl/h, Example 9 or vehicle (1× Dulbecco's phosphate-buffered saline [DPBS]) is administered subcutaneously (SC) continuously for 4 days through Alzet® pumps (Model 2ML1; Alzet® Osmotic Pumps; Cupertino, Calif.). The dose levels administered for Example 9 are 0.22, 0.73, 2.18, and 7.28 mg/kg/day using pump infusion. On Day 4, blood samples are collected to determine plasma concentration and the infusion pump is removed. Seven days after dosing started, cardiac function (EF %) and structure (LVD) are evaluated for all animals. Fourteen days after dosing started, EF % and LVD are evaluated for vehicle-treated and two highest dose groups.

(46) Results.

(47) Example 9 treatment demonstrates a dose-dependent improvement in cardiac function (EF %) when administered for 96 hours using the osmotic pump (Table 10). High-dose groups administered 2.18 and 7.28 mg/kg/day, exhibit improved cardiac function for 2 weeks post-infusion started (Table 10). These doses achieve steady state plasma concentrations of 29.62 and 72.24 nM, respectively (Table 11). All treated MI animals exhibit less LVD dilatation compared to the vehicle group (Table 10).

(48) TABLE-US-00014 TABLE 10 Effect of 96-Hour Infusion on EF in a Rat Model of HFrEF. 7 days after 14 days after Baseline dosing started dosing started Dose (mg/kg/day) MEAN s.e. MEAN s.e. MEAN s.e. Ejection Fraction Sham, N = 3 70.2 6.42 68.1 4.64 ND ND Vehicle, N = 7 41.7 1.97 39.6 1.92 39.6 1.73 0.22, N = 5 41.6 2.57 43.9 2.04 ND ND 0.73, N = 7 42.3 1.42 45.8 2.09 ND ND 2.18, N = 7 41.4 1.90 48.9* 1.72 45.5 1.61 7.28, N = 7 41.1 1.06 49.8* 2.10 48.0* 2.08 Left Ventricle Diameter (end diastole) Sham, N = 3 7.21 0.51 7.49 0.76 ND ND Vehicle, N = 7 7.56 0.35 7.94 0.26 7.90 0.25 0.22, N = 5 7.51 0.42 7.53 0.57 ND ND 0.73, N = 7 7.55 0.27 7.83 0.30 ND ND 2.18, N = 7 7.50 0.34 7.68 0.24 7.75 0.19 7.28, N = 7 7.60 0.29 7.52 0.31 7.67 0.19 Abbreviations: N = number of animals; ND = not determined; SE = standard error. *p < 0.05 vs vehicle.

(49) TABLE-US-00015 TABLE 11 Exposure in Rat Plasma Following 96 Hours of Infusion. Dose (mg/kg/day) Mean (nM) Standard Deviation 0.22 2.59 0.71 0.73 9.94 1.59 2.18 29.62 4.86 7.28 72.24 13.66

(50) Study Design for Determination of Infusion Duration for Nonclinical Efficacy.

(51) At an infusion rate of 10 μl/h, 2.18 mg/kg/day of Example 9 or vehicle (DPBS) is administered SC for 48, 72, or 96 h through Alzet® pumps. At the end of designated infusion phase, blood samples are collected to determine plasma concentration and the infusion pump is removed. Seven days after dosing started, cardiac function (EF %) and structure (LVD) are evaluated for all animals.

(52) Results.

(53) As seen in Table 12, in this study Example 9 treatment demonstrates an infusion duration-dependent improvement in cardiac function when administered for 48 to 96 h using the Alzet® pump. Example 9 infusion for 72 h and 96 h results in significant EF improvements on Day 7 after dosing started. All treated MI animals show less LVD dilatation compared to the vehicle group.

(54) TABLE-US-00016 TABLE 12 Effect of Example 9 Infusion for Multiple Durations on Cardiac Function. Baseline 7 days after dosing started Duration Time (hours) MEAN s.e. MEAN s.e. Ejection Fraction Vehicle, N = 8 40.8 2.80 39.2 2.78 48, N = 8 40.6 1.26 42.6 1.35 72, N = 9 41.0 1.96 44.6* 1.69 96, N = 9 41.8 1.77 46.2* 2.19 Left Ventricle Diameter (end diastole) Vehicle, N = 8 7.49 0.20 8.10 0.24 48, N = 8 7.52 0.29 7.68 0.30 72, N = 9 7.55 0.21 7.87 0.20 96, N = 9 7.44 0.16 7.70 0.22 Abbreviations: N = number of animals; ND = not determined; SE = standard error. *p < 0.05 vs vehicle.

(55) TABLE-US-00017 TABLE 13 Exposure in Rat Model of Myocardial Infarction Plasma Following Example 9 Infusion. Infusion Time (hours) Mean (nM) Standard Deviation 48 40.13 6.37 72 47.28 17.85 96 31.42 9.86

(56) Nonclinical data showed that LVEF improved in the rat model of HFrEF when plasma concentrations of Example 9 were maintained at 3 to 100 nM for a duration of 72 to 96 hours.

(57) Effects of Example 10 on Cardiac Function and Structure in Rat MI Model

(58) A nonclinical efficacy pharmacology study is conducted in a rat model of heart failure with reduced ejection fraction (HFrEF) to evaluate cardiac function and structure changes after captopril treatment, Example 10 treatment or combination treatment.

(59) Methods.

(60) Male Sprague Dawley rats with surgically induced myocardial infarction are purchased and anesthetized and positively ventilated throughout the procedure. An incision is made between the fourth and fifth ribs, revealing the heart. The left coronary artery is permanently ligated. All rats are individually housed in a temperature and humidity controlled room and maintained on a 12 hour light/dark cycle. Two to three weeks after surgery, rats undergo transthoracic echocardiography for determination of ejection fraction (EF %) and left ventricle dimensions (LVD) using a Vevo 2100 ultrasound system. Rats are randomized across treatment groups according to EF % and LVD.

(61) Measurements of EF % and LVD are expressed as mean values±standard error (SE). Statistical analysis is performed with JMP® 13 software (SAS Institute, Inc.; Cary, N.C.) and Dunnett's Test is used for statistical comparisons across treatment groups. Statistical significance is accepted at P<0.05.

(62) Study Design.

(63) Three weeks after MI surgery, animals are randomized to 2 treatment groups (Placebo or captopril, 2 g/L in drinking water, ad lib) according to EF % and LVD. Three weeks and four weeks after dosing starts, cardiac function (EF %) and structure (LVD) are evaluated by transthoracic echocardiography. Animals are further randomized to additional groups based on cardiac function (EF %) and structure (LVD) to the following groups: Placebo, Example 10, captopril and Example 10+captopril. Animals in the Example 10 groups receive a single injection on week 4 after the 2.sup.nd randomization. Cardiac function (EF %) and structure (LVD) are evaluated on week 5 and 6.

(64) Results.

(65) Results are provided in Table 14.

(66) TABLE-US-00018 TABLE 14 Effects of captopril, a single injection of Example 10 or combination therapy on EF in a Rat Model of Heart Failure with reduced Ejection Fraction. Baseline Week 3 Week 4 Week 5 Week 6 AVG s.e. AVG s.e. AVG s.e. AVG s.e. AVG s.e. Ejection Fraction Vehicle 38.2 1.42 32.8 1.55 Vehicle 33.4 2.49 30.0 2.70 29.9 2.43 Captopril 39.1 1.30 38.4 2.43 Captopril 36.1 2.66 36.2 3.23 34.6 3.09 Example 10 33.3 2.06 46.5* 3.68 46.0* 2.46 Example 35.8 1.96 47.3* 1.97 46.7* 2.45 10 + Captopril Left Ventricle Diameter (end diastole) Vehicle 8.75 0.25 9.00 0.18 Vehicle 9.25 0.44 10.16 0.46 9.92 0.40 Captopril 8.66 0.16 8.57 0.14 Captopril 8.64 0.29 8.64* 0.24 8.71* 0.21 Example 10 9.33 0.24 9.35 0.37 9.01 0.25 Example 8.82 0.12 8.51* 0.23 8.41* 0.28 10 + Captopril Abbreviations: AVG = mean. s.e. = standard error. *p < 0.05 vs vehicle.

(67) As seen in Table 14, while captopril treatment slowed down cardiac function decline over the course of 6 weeks of treatment (not statistically significant), Example 10 treatment significantly improved cardiac function (EF %) and reduced LV dilation after a single treatment.

(68) Effects of Example 11 on Cardiac Function and Structure in Rat MI Model

(69) A nonclinical efficacy pharmacology study is conducted in a rat model of heart failure with reduced ejection fraction (HFrEF) to identify dose-response relationship for Example 11.

(70) Method.

(71) Male Sprague Dawley rats with surgically induced myocardial infarction are purchased from Charles River. Briefly, the rats are anesthetized and positively ventilated throughout the procedure. An incision is made between the fourth and fifth ribs, revealing the heart. The left coronary artery is permanently ligated. The sham-operated animals undergo the same procedure except the silk suture is placed around the left coronary artery without being tied. All rats are individually housed in a temperature and humidity controlled room and maintained on a 12 hour light/dark cycle. Three weeks after surgery, rats undergo transthoracic echocardiography for determination of ejection fraction (EF %) and left ventricle dimensions (LVD) using a Vevo 2100 ultrasound system. Rats are randomized across treatment groups according to EF % and LVD.

(72) Measurements of EF % and LVD are expressed as mean values±standard error (SE). Statistical analysis is performed with JMP® 13 software (SAS Institute, Inc.; Cary, N.C.) and Dunnett's Test is used for statistical comparisons across treatment groups. Statistical significance is accepted at P<0.05.

(73) Study Design.

(74) Two to three weeks after MI surgery, animals are allocated to treatment groups and treated with a single injection of vehicle (DPBS) or one of the 3 dose levels (0.3, 1, 3 mg/kg) of Example 11. Seven days, 14 days and 21 days after dosing, cardiac function (EF %) and structure (LVD) are evaluated for all animals.

(75) Results. Results are provided in Table 15.

(76) TABLE-US-00019 TABLE 15 Effect of Example 11 on EF in a Rat Model of Heart Failure with reduced Ejection Fraction. Ejection Fraction Baseline 7 days after dosing 14 days after dosing 21 days after dosing Dose MEAN s.e. MEAN s.e. MEAN s.e. MEAN s.e. Sham, N = 5 73.7 2.24 72.5  1.76 71.8 0.81 71.9 3.60 Vehicle, N = 8 45.0 1.50 41.2  2.90 38.4 1.98 37.8 1.98 0.3 mg/kg, N = 8 45.4 2.46 50.1* 3.22 42.6 1.43 36.4 1.43 1 mg/kg, N = 8 46.6 1.86 52.4* 2.61 42.2 2.77 39.7 2.77 3 mg/kg, N = 8 45.8 1.48 54.7* 2.11  54.0* 2.49 39.6 2.49 Abbreviations: N = number of animals; s.e. = standard error. *p < 0.05 vs vehicle.

(77) As seen in Table 15, Example 11 treatment demonstrates a dose-dependent improvement in cardiac function (EF %). The high-dose group exhibits improved cardiac function for 2 weeks post-treatment. No differences in LVD dilation are observed across groups during the course of the study (data not shown).

(78) Effects of Examples 9-11 on Cardiac Toxicity

(79) Toxicology studies are conducted on Examples 9-11 in rats and/or cynomolgus monkeys. Example 10 is administered to rats in a single SC injection of 0.3 mg/kg, which results in detectable plasma concentrations for longer than 168 hours. Rats are sacrificed, and inspections of tissue reveal cardiac degeneration and necrosis. Example 11 is administered to rats in a single SC injection of 30 mg/kg; plasma concentration was undetectable at 96 hours. Rats are sacrificed, and inspections of tissue reveal no evidence of cardiac degeneration and/or necroses. Example 11 is administered in a single 3 mg/kg dose to male and female monkeys by SC bolus injection. In contrast to its pharmacokinetics in rats, the single injections in monkeys results in detectable plasma concentrations that take longer than 168 hours to be cleared. Monkeys are sacrificed, and inspections of tissue reveal cardiac degeneration and necrosis. Example 9, which shows a half-life after SC administration of less than an hour in rats, is administered in doses resulting in plasma concentrations ranging from 3-300 nM for 96-168 hours by subcutaneously (SC) implanted pumps. Rats are sacrificed, and inspections of tissue reveal no evidence of cardiac degeneration and/or necroses. Example 9, which has a half life after SC administration of less than 5 hours in monkeys, is administered in doses resulting in plasma concentrations ranging from 30-1000 nM for 96 hours to male and female monkeys by surgically placed SC or intravenous (IV) catheter. Monkeys are sacrificed, and inspections of tissue reveal no evidence of cardiac toxicity and/or necroses. These data indicate that NRG4 compounds may be administered without causing cardiac toxicity by limiting the duration of exposure to the NRG4 compound.

(80) TABLE-US-00020 Sequences Human NRG4 SEQ ID NO: 1 DHEEPCGPSH KSFCLNGGLC YVIPTIPSPF CRCVENYTGA RCEEVFL NRG4 Compound SEQ ID NO: 2 GHEEPCGX.sub.8SHKSFCLNGGLCYX.sub.22IPTX.sub.26PSPFCRCVX.sub.35 NYTGARCEX.sub.44VFL wherein:  X.sub.8 is E or P; X.sub.22 is Q or V; X.sub.26 is F or I; X.sub.35 is E or A; and X.sub.44 is H, K or E. Human NRG4 Full Sequence As Expressed SEQ ID NO: 3 MPTDHEEPCG PSHKSFCLNG GLCYVIPTIP SPFCRCVENY TGARCEEVFL PGSSIQTKSN LFEAFVALAV LVTLIIGAFY FLCRKGHFQR ASSVQYDINL VETSSTSAHH SHEQH NRG4 Compound SEQ ID NO: 4 GHEEPCGPSH KSFCLNGGLC YQIPTIPSPF CRCVENYTGA RCEKVFL Glycine rich peptide or linker SEQ ID NO: 5 (GGGGX.sub.λ).sub.n wherein: X.sub.λ is Q, A, E or S; and n is 1-5. IgG Fc region SEQ ID NO: 6 ECPPCPAPPVAGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVQF NWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSN KGLPAPIEKTISKTKGQPREPQVYTLPPSREEMTKNQVSLWCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSC SVMHEALHNHYTQKSLSLSPG IgG Fc region SEQ ID NO: 7 ECPPCPAPPVAGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVQF NWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSN KGLPAPIEKTISKTKGQPREPQVYTLPPSREEMTKNQVSLSCAVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLVSKLTVDKSRWQQGNVFSC SVMHEALHNHYTQKSLSLSPG IgG Fc region SEQ ID NO: 8 ECPPCPAPPVAGPSVFLFPPKPKDTLMASRTPEVTCVVVDVSHEDPEVQF NWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSN KGLPAPIEKTISKTKGQPREPQVYTLPPSREEMTKNQVSLWCLVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLTVDKSRWQQGNVFSC SVMHEALHNAYTQKSLSLSPG IgG Fc region SEQ ID NO: 9 ECPPCPAPPVAGPSVFLFPPKPKDTLMASRTPEVTCVVVDVSHEDPEVQF NWYVDGVEVHNAKTKPREEQFNSTFRVVSVLTVVHQDWLNGKEYKCKVSN KGLPAPIEKTISKTKGQPREPQVYTLPPSREEMTKNQVSLSCAVKGFYPS DIAVEWESNGQPENNYKTTPPVLDSDGSFFLVSKLTVDKSRWQQGNVFSC SVMHEALHNAYTQKSLSLSPG NRG4 Compound SEQ ID NO: 10 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYVIPTIPSPFCRCVE NYTGARCEEVFL NRG4 Compound SEQ ID NO: 11 GGGGSGGGGSGGGGSGHEEPCGESHKSFCLNGGLCYVIPTIPSPFCRCVE NYTGARCEEVFL NRG4 Compound SEQ ID NO: 12 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYQIPTIPSPFCRCVE NYTGARCEEVFL NRG4 Compound SEQ ID NO: 13 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYVIPTFPSPFCRCVE NYTGARCEEVFL NRG4 Compound SEQ ID NO: 14 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYVIPTIPSPFCRCVA NYTGA RCEEVFL NRG4 Compound SEQ ID NO: 15 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYVIPTFPSPFCRCVE NYTGARCEKVFL NRG4 Compound SEQ ID NO: 16 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYVIPTFPSPFCRCVE NYTGARCEHVFL NRG4 Compound SEQ ID NO: 17 GGGGSGGGGSGGGGSGHEEPCGESHKSFCLNGGLCYVIPTFPSPFCRCVE NYTGARCEEVFL NRG4 Compound SEQ ID NO: 18 GGGGSGHEEPCGPSHKSFCLNGGLCYQIPTIPSPFCRCVENYTGARCEKV FL NRG4 Compound SEQ ID NO: 19 GGGGSGGGGSGGGGSGHEEPCGPSHKSFCLNGGLCYQIPTIPSPFCRCVE NYTGARCEKVFL SEQ ID NO: 20 GGGS