Cellular metabolism of substance P produces neurokinin-1 receptor peptide agonists with diminished cyclic AMP signaling

Abstract

Substance P (SP) is released from sensory nerves in the arteries and heart. It activates neurokinin-1 receptors (NK1Rs) causing vasodilation, immune modulation, and adverse cardiac remodeling. The hypothesis was tested: SP and SP metabolites activate different second messenger signaling pathways. Macrophages, endothelial cells, and fibroblasts metabolized SP to N- and C-terminal metabolites to varying extents. SP 5–11 was the most abundant metabolite followed by SP 1–4, SP 7–11, SP 6–11, SP 3–11, and SP 8–11. In NK1R-expressing human embryonic kidney 293 (HEK293) cells, SP and some C-terminal SP metabolites stimulate the NK1R, promoting the dissociation of several Gα proteins, including Gαs and Gαq from their βγ subunits. SP increases intracellular calcium concentrations ([Ca]_i) and cyclic 3′,5′-adenosine monophosphate (cAMP) accumulation with similar −log EC₅₀ values of 8.5 ± 0.3 and 7.8 ± 0.1 M, respectively. N-terminal metabolism of SP by up to five amino acids and C-terminal deamidation of SP produce peptides that retain activity to increase [Ca]_i but not to increase cAMP. C-terminal metabolism results in the loss of both activities. Thus, [Ca]_i and cAMP signaling are differentially affected by SP metabolism. To assess the role of N-terminal metabolism, SP and SP 6–11 were compared with cAMP-mediated activities in NK1R-expressing 3T3 fibroblasts. SP inhibits nuclear factor κB (NF-κB) activity, cell proliferation, and wound healing and stimulates collagen production. SP 6–11 had little or no activity. Cyclooxygenase-2 (COX-2) expression is increased by SP but not by SP 6–11. Thus, metabolism may select the cellular response to SP by inhibiting or redirecting the second messenger signaling pathway activated by the NK1R.

NEW & NOTEWORTHY Endothelial cells, macrophages, and fibroblasts metabolize substance P (SP) to N- and C-terminal metabolites with SP 5–11 as the most abundant metabolite. SP activates neurokinin-1 receptors to increase intracellular calcium and cyclic AMP. In contrast, SP metabolites of N-terminal metabolism and C-terminal deamidation retain the ability to increase calcium but lose the ability to increase cyclic AMP. These new insights indicate that the metabolism of SP directs cellular functions by regulating specific signaling pathways.

INTRODUCTION

Substance P (SP) is an 11-amino acid peptide, generally considered as a sensory nerve neuropeptide, and part of the tachykinin family of peptides. SP has well-established functions in many processes throughout the body including pain, gut function, immune responses, and inflammation (1). In addition, targeting of the neurokinin-1 receptor (NK1R), a cognate receptor for SP, is a potential treatment strategy in cancer (1–3), inflammatory bowel disease (4, 5), kidney disease (6), and cardiac disease (7). In certain situations, SP has adverse effects. For example, upregulated SP activates NK1Rs, promoting adverse cardiac remodeling in rodent models of hypertension (8, 9), myocarditis (10–12), cardiotoxicity (13–15), and hypomagnesemia (16–18). Alternatively, in acute cardiac ischemia, endogenous SP activation of NK1Rs provides protection. Exogenous SP reduces infarct size and improves cardiac function (19–22). Furthermore, serum SP concentrations are decreased in type 2 diabetes and predispose the heart to fibrosis (23, 24). Replacement of SP reduces the fibrosis. It is unknown how SP and the NK1R exert both protective and adverse responses in the same organ in different disease states. Metabolism of SP has not been considered as a pathway modifying the actions of SP in heart disease. In this study, we investigate the cellular metabolism of SP and its impact on activation of the NK1R. This is an essential first step to lay the foundation for future studies exploring the hypothesis that the variable effects of SP in the heart and other diseases are due to differences in its metabolism.

SP is metabolized by a number of enzymes including angiotensin-converting enzyme (ACE), neprilysin (NEP), neurolysin (NLN), thimet oligopeptidase (THOP), endothelin-converting enzyme-1 (ECE-1), matrix metalloproteinase-8 and matrix metalloproteinase-9 (MMP-8/9), fibroblast-activation protein-α (FAP), and dipeptidyl-peptidase IV (DPP-IV) into numerous metabolites (25). These metabolites may invoke differential responses by the NK1R. In addition, the NK1R exists as two isoforms: a full-length (fl) isoform and a truncated (tr) isoform lacking the final 96 amino acids of the carboxyl terminus (3, 26, 27). SP or SP metabolites may have different cellular responses mediated by the two receptor isoforms. Herein, we investigated the metabolism of SP by different cell types and the activity of SP and its metabolites on the human and mouse NK1Rs. These studies indicate that C-terminal SP metabolites retain NK1R activity, whereas N-terminal metabolites are inactive. Furthermore, N-terminal metabolism and C-terminal deamidation of SP change the second messenger pathway activated by the NK1R. Some of the cellular activities of SP are reduced with N-terminal metabolism.

MATERIALS AND METHODS

Chemicals

Substance P, [Tyr⁸]-SP was purchased from Sigma Aldrich; SP 1–7, SP 1–9, SP 6–11, and SP 7–11 were purchased from Bachem (Torrance, CA); SP 2–11, SP 3–11, SP 5–11, and SP 8–11 were purchased from Phoenix Pharmaceuticals (Burlingame, CA); and SP 1–4 was purchased from ABBIOTEC (Escondido, CA). Thioglycollate medium, Brewer modified, and 3-isobutyl-1-methyl-xanthine (IBMX) were purchased from BD Biosciences and Sigma, respectively. Culture medium and additives were purchased from Gibco or Cell Biologics.

Plasmids

Plasmids were obtained from GenScript and expanded according to the manufacturer’s protocol: empty vector [pcDNA3.1/Hygro(+)], human NK1R truncated (tr)-(hNK1R) pcDNA3.1/Hygro(+) [NM_015727.3], full length (fl)-hNK1R pcDNA3.1/Hygro(+) [NM_001058.4], and mouse NK1R (mNK1R) [NM_009313.5]. TRUPATH kit was from Addgene (28).

Animals

Male C57BL/6J mice (10 wk old from Jackson Laboratories) were used in the study. Mice were allowed a 3-day acclimatization period in the new facility upon arrival. Protocols and procedures were approved and performed in accordance with the Institutional Animal Care and Use Committee of the Medical College of Wisconsin (AUA00006086).

Cell Culture

HEK293 cells [American Type Culture Collection (ATCC) CRL-1573] and HEK293T cells (ATCC CRL-3216) were cultured in DMEM media as previously described (29, 30). NIH 3T3 fibroblasts (ATCC CRL-1658) were cultured in DMEM media supplemented with 10% bovine fetal calf serum (FCS) (Hyclone), 100 units/mL penicillin, and 100 µg/mL streptomycin (Invitrogen). C57BL/6 mouse coronary artery endothelial cells (CAECs) (Cell Biologics Inc, Chicago, IL) were grown in gelatin-coated dishes in complete mouse endothelial cell medium (Cell Biologics Inc, Chicago, IL) (31). Chinese hamster ovary (CHO) cells (ATCC CCL-61) were grown in RPMI media supplemented with 10% FBS and 1 mg/mL hygromycin-B. All the cells were incubated at 37°C in a humidified chamber containing 5% CO₂. Experiments were performed on 3 or 4 different stocks of the cell line.

Peritoneal Macrophage Isolation

Mice (2) were injected intraperitoneally with 2 mL of sterile, 38.5 mg/mL thioglycolate solution. Four days later, the elicited peritoneal macrophages (PMs) were harvested from isoflurane-euthanized mice by peritoneal lavage. Cells were pelleted at 1,200 rpm for 10 min at 4°C, resuspended in RPMI-1640 media, and seeded into poly-d-lysine-coated tissue culture dishes containing media (31) and cultured overnight. Experiments were performed on three independent sets of PM preparations.

In Vitro SP Metabolism by Cell Types

The exogenous metabolism of SP was studied in mouse 3T3 fibroblasts, CAECs, and PMs. Cells at 80%–90% confluency (∼3 × 10⁶ cells/60-mm dish) were rinsed twice with phosphate-buffered saline (PBS) solution and then incubated with either 1.5 µM SP or vehicle in bicarbonate-buffered phosphate saline solution (PSS) (pH = 7.4). SP was also added to dishes containing no cells to serve as a negative control. After 60 min of incubation at 37°C, 60 µL aliquots of the PSS were removed and mixed with 48 µL of 0.1% formic acid and 12 µL of 10 µM Tyr⁸-SP in 0.1% formic acid as a chromatographic standard. Samples were stored at −20°C until analysis.

Liquid Chromatography-Mass Spectrometry Analysis of SP Metabolites

The amount of SP metabolites synthesized by murine cells was quantified by liquid chromatography-mass spectrometry (LC-MS/MS) as previously described by Chappa et al. (32). Analysis was performed on an Agilent 6460 triple quadrupole mass spectrometer interfaced to an Agilent 1200 HPLC through a Jet Stream interface. Samples were separated on a Kromasil C18 (150 × 2.0 mm, 5 μm) column at a flow rate of 0.2 mL/min. Solvent A was water containing 0.1% formic acid, whereas solvent B was acetonitrile containing 0.1% formic acid. The solvent program was as follows: 0% B for 2 min; increase B from 0% to 40% between 2 and 10 min; increase B to 98% between 10 and 12 min; hold 98% B for 1 min; and return B to 0% in 1 min.

MS/MS scans were performed in multiple reaction monitoring (MRM) mode with the following source parameters: gas temperature 325°C, gas flow 10 L/min, capillary voltage 3,500 V, and cell acceleration 7 V. The collision energy was optimized for each metabolite between 12 and 40 V. Quantitation was based on standard curves of SP metabolite standards obtained from MS under same conditions. Signals from cell-free incubations were subtracted from signals produced by incubations with cells. The precursor and product ions and LC retention times for each of the peptides are listed in Supplemental Table S1.

Immunoblot

Cells were scraped off the surface of the dish, and proteins were isolated using Mem-PER Plus Membrane Protein Extraction kit (Thermo Scientific). After determining the protein concentration, samples were mixed with Lane Marker Reducing Sample Buffer (Thermo Scientific). About 50 µg of total protein was resolved using 10% Criterion SDS-PAGE gel (Bio-Rad). After transfer onto a PVDF membrane, the membrane was blocked with 5% blocking milk (Bio-Rad) and incubated with rabbit polyclonal anti-NK1R antibody (1:1,000) (Novus Biologicals, Cat. No. NB300-119) or mouse monoclonal anti-pan-cadherin (Abcam Cat. No. ab22744) (1:1,000), or mouse anti-cyclooxygenase-2 (COX-2) antibody (1:2,000) (Cayman Chemical Cat. No. 160126) overnight at 4°C. The following day, the membrane was incubated with horseradish peroxidase (HRP)-conjugated donkey anti-mouse IgG secondary antibody (Jackson ImmunoResearch Laboratories, Cat. No. 715-035-150) (1:10,000). HRP-conjugated anti-β-actin antibody (Santa Cruz Biotechnology, Cat. No. sc-47778) was used as a loading control. Blots were developed with ECL substrate (Pierce) and quantified using ImageQuant LAS 4000 imager (GE Healthcare) and ImageJ software.

G-Protein Transducer Pathway (TRUPATH) Assay

To determine which G-proteins transduce the NK1R signal, the TRUPATH assay was performed as described by Olsen et al (28). In brief, HEK293T or CHO cells (2 × 10⁶ cells and 4 × 10⁵ cells/60-mm dish, respectively) were transfected overnight with the NK1R, Gα-RLuc8, Gβ, and Gγ-green fluorescent protein 2 (GFP2) in the ratio 1:1:1:1 (1.25 µg DNA/construct/60-mm dish) using 3 µL/µg TransIT-2020 (Mirus Biosciences, Madison, WI) as the transfecting agent. Transfected cells (5 × 10⁴ cells/well) were plated into white-walled, clear-bottom poly-d-lysine-coated 96-well plates (Corning) using phenol-red-free DMEM media containing 1% dialyzed FBS and incubated overnight. For HEK293T cells, media were replaced with 65 µL of assay buffer/well (1X HBSS + 20 mM HEPES, 0.1% protease-free fraction V BSA, 0.01% ascorbic acid, pH = 7.4) for 1 h at 37°C, followed by incubation with 5 µM coelenterazine 400a (NanoLight Technologies, Pinet, AZ). After a 15-min equilibration period, appropriate amount of SP in 25 µL was added to the wells, and the plate was read immediately in LB940 Mithras plate reader (Berthold Technologies, Oak Ridge, TN) with 410-nm (Rluc8-coelenterazine 400a) and 515-nm (GFP2) emission filters, at 1 s/well integration times. Transfected CHO cells were treated with SP, and after 25 min, coelenterazine 400a was added. The plate was read as described earlier. Bioluminescence resonance energy transfer-2 (BRET-2) signals were calculated as the ratio of GFP2 emission to Rluc8 emission, whereas ligand-induced BRET-2 signals were calculated by subtracting the average vehicle ratio from each ligand concentration ratio. Experiments were performed on 3 or 4 separate cell transfections.

Fluorescence Imaging Plate Reader Assay for [Ca]_i

[Ca]_i mobilization assay using fluorescence imaging plate reader (FLIPR) (Molecular Devices) was performed as previously described (30, 31). In brief, HEK293T cells were transiently transfected with the receptor plasmid using Lipofectamine 2000 (Invitrogen) transfecting agent, according to the manufacturer’s protocol. After 24 h, transfected HEK293T cells were resuspended in 1% dialyzed FBS, phenol-red-free DMEM media and seeded on black-walled, clear-bottom poly-d-lysine-coated plates (Corning) (5 × 10⁴ cells/0.1 mL/well) and incubated overnight. The next day, media were removed, and cells were first incubated with 100 µL Fluo-4 NW (no-wash) dye containing probenecid (2.5 mM) at 37°C for 30 min, followed by 30 min at room temperature while wrapped in foil. SP or its N- and C-terminal metabolites were added, and the signal was recorded for 220 s. The relative intracellular changes in calcium ([Ca]_i) Δ relative fluorescence unit (ΔRFU) were analyzed as previously described (30). Protein concentrations were measured using bicinchoninic acid (BCA) assay (Thermo Scientific) to normalize the signal. Results were expressed as percent change in ΔRFU/µg protein of each metabolite relative to SP. Experiments were performed on 3 or 4 separate cell transfections.

GloSensor Assay for cAMP

GloSensor intracellular biosensor was used to measure cyclic 3′,5′adenosine monophosphate (cAMP) (33). HEK293T cells (4 × 10⁶) or CHO cells (1.2 × 10⁶) stably expressing GloSensor-22F were plated into 100-mm dishes in growth media without antibiotics. The next day, the cells were transiently transfected with empty vector or fl-hNK1R using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol and incubated for 24 h. The transfected cells were plated (50,000/well) into a white-walled clear-bottom poly-d-lysine-coated 96-well plate (Corning) with 1% FBS containing growth media and allowed to adhere overnight. The following day, the cells were rinsed with 300 µL/well of assay buffer (Hanks’ balanced salt solution with 20 mM HEPES, pH 7.4), and 50 µL/well of cAMP reagent (6% GloSensor cAMP Reagent in assay buffer) was added. The cells were incubated at room temperature for 2 h. IBMX (100 µM) was added after 1 h. The plate was equilibrated to room temperature for 15 min. Using a Mithras LB 940 plate reader, baseline luminescence was measured; 10 µL of vehicle, SP, or SP peptides were added; and luminescence was measured every 2 min for 30 min. The baseline value was subtracted from the peak value and reported as relative luminescence unit (ΔRLU). Experiments were performed on 3 or 4 separate cell transfections.

Transfection and Selection of 3T3 Fibroblasts

NIH 3T3 fibroblasts (1.2 × 10⁶ cells/100-mm dish) were plated in DMEM. After 18 h, they were transfected with Lipofectamine LTX according to the manufacturers’ protocol. The cells were plated in 96-well plates for the individual assays, or a batch selection was carried out by adding 300 µg/mL hygromycin B after 2 days. To select for a stable clone, cells were plated to a density of ½ cell per well in 96-well plates in 300 µg/mL hygromycin B-containing selection media. Wells with single colonies were expanded, tested, and frozen. Experiments were performed on 3 or 4 separate stocks of the NK1R-3T3 fibroblast cell line.

Changes in [Ca]_i and cAMP in 3T3 Fibroblasts

[Ca]_i and cAMP were measured in 3T3 fibroblasts by FLIPR and GloSensor, respectively, using the method described earlier for HEK293 cell (30, 33). Concentration-response curves to SP or SP and SP 6–11 were measured in empty vector transfected- or NK1R-transfected 3T3 cells. Responses to SP were repeated in the presence of 1 µM of inhibitors of ACE (captopril), DPP-IV (linagliptin), and ECE-1/NEP (daglutril).

Measurement of Proliferation and Wound Healing in 3T3 Fibroblasts

Kinetic measurements of proliferation and wound healing used an Incucyte S3 Live Cell Imager (Sartorius) at 37°C in a humidified chamber containing 5% CO₂. 3T3 cells stably expressing hNK1R were maintained in DMEM containing 10% FCS, 100 units/mL penicillin, 100 µg/mL streptomycin, and 300 µg/mL hygromycin B. Proliferation: cells (3,500 cells/well) were plated at in phenol-red-free DMEM with 0.5% FCS in 96-well plates. After 18 h, the media were changed, and the cells were treated with vehicle, SP (100 nM), or SP 6–11 (100 nM). The plate was placed to the Incucyte S3, and images were collected with a ×10 objective (two fields per well) hourly for 12 h and analyzed for % confluency with Incucyte software. Wound healing: cells (35,000 cells/well) were plated in 96-well Essen ImageLock plates. After 18 h, the media were replaced with phenol-red-free DMEM with 0.5% FCS. After 5 h, wounds were made with the Incucyte Woundmaker 96-Tool. The cells were rinsed twice with fresh media and treated with vehicle, SP (100 nM), or SP 6–11 (100 nM). The plate was placed to the Incucyte S3, and images were collected with a ×10 objective (two fields per well) hourly for 12 h and analyzed for % wound confluency with scratch wound acquisition software.

[³H]-Thymidine Incorporation in 3T3 Fibroblasts

[³H]-Thymidine incorporation assay was performed on 3T3 cells transiently transfected with hNK1R (34). Cells were grown to 60% on a 24-well plate and serum starved overnight. Cells were then treated with vehicle, SP (30 nM), or SP 6–11 (30 nM) for 24 h in 1% FCS-DMEM media. During the final 4 h of incubation, [³H]-thymidine (1 µCi/well) was added. The media were removed, and the cells were rinsed twice with ice-cold PBS, twice with 5% trichloroacetic acid, and once with distilled water. The precipitates were solubilized in 0.3 M NaOH, and radioactivity was measured by liquid scintillation spectrometry.

Nuclear factor-κB Activation in 3T3 Fibroblasts

3T3 fibroblasts stably expressing hNK1R were transiently transfected with pNiFty2-luc nuclear factor κB (NF-κB)-inducible reporter plasmid (Invitrogen). After 18 h, the cells (35,000 cells/well) were replated into white-walled, clear-bottom 96-well plates in phenol-red-free DMEM. After 18 h, the media were changed, and the cells were treated with vehicle, tumor necrosis factor α (TNFα) (100 ng/mL), or TNFα after a 10-min pretreatment with SP (100 nM) or SP 6–11 (100 nM). The plates were incubated for 4 h and 30 min in a cell culture incubator. The plates were then placed in BMG VANTAstar plate reader. After an initial 10-s measurement, D-luciferin was added to each well, and subsequent 10-s measurements were made. Delta RLU values were calculated by subtracting the initial from the second measurement.

Collagen Release by 3T3 Fibroblasts

3T3 fibroblasts were transiently transfected with hNK1R, grown to confluency in 96-well plates, and serum starved overnight (24). Cells were stimulated with vehicle, SP (30 nM), or SP 6–11 (30 nM) in 1% FCS-DMEM media. After 24 h, supernatants were collected and assayed using the Mouse Collagen Type 1 ELISA kit (Novus Biologicals, Littleton, CO) as per the manufacturer’s protocol. The concentration of collagen (ng/mL) in different treatment groups was determined based on a standard curve.

COX-2 Expression in 3T3 Fibroblasts

3T3 fibroblasts stably expressing hNK1R (1 × 10⁶⁾ were plated in 35-mm dishes. After 24 h, SP or SP 6–11 (100 nM) was added and incubated for 4 h and 30 min. Cells were washed twice with PBS, scraped into PBS, and pelleted. Pellets were lysed in lysis buffer containing protease inhibitors. Lysates were centrifuged at 16,000 g for 10 min at 4°C, and the concentration of the supernatant protein was determined by BCA assay. Proteins (30 µg) were separated on 7.5% SDS PAGE gel and immunoblotted as described earlier.

Statistical Analysis

The data and statistical analyzes used GraphPad Prism 9. The individual time course and concentration response data points in Figs. 2, 3, 4, 5,6, 8, and 9 represent means ± SD from at least two independent experiments. In Figs. 1, 7, 8, and 9, the boxes are to indicate the median and first and third quartile, and whisker plots to indicate minimum and maximum values. Concentration-response curves were fitted using a four-parameter logistic curve with an estimation of goodness. Statistical differences between fitted nonlinear regression curves were revealed by F test. Statistical differences between groups were represented by box, and whisker plots were revealed by one-way ANOVA with Tukey’s post hoc test for multiple comparisons. Statistical analysis of the groups represented by bar graph was performed using unpaired Student’s t test. *P < 0.05 was considered significant.

Figure 2.

Gα proteins coupled to the NK1R. Dissociation of Gα and Gβγ proteins by SP activation of the full-length (fl)- and truncated (tr)-hNK1R in HEK293T cells as indicated by a decrease in TRUPATH BRET fluorescence. A: Western blot of NK1R-overexpressed cells using pan-cadherin as a membrane marker and β-actin as the loading control. Empty vector (EV), truncated (tr), and full length (fl). BRET changes with Gαq/Gα12 group (B), Gαi group (C), and Gαs group (D) of G proteins in HEK293T cells and (Gαq/Gα12 group and Gαs (E) in CHO cells in fl-NK1R-transfected cells. F: BRET changes with Gαs in tr-NK1R-transfected cells. Results are expressed as the ligand- to vehicle-induced BRET ratio. Each value represents the means ± SD (n = 4–18) (3 or 4 experiment). BRET, bioluminescence resonance energy transfer; CHO, Chinese hamster ovary; NK1R, neurokinin-1 receptor; SP, substance P.

Figure 3.

Effect of SP peptides on Gα protein activation. Dissociation of Gα and Gβγ proteins by SP and SP metabolite activation of the fl-hNK1R in HEK293T and CHO cells as indicated by a decrease in TRUPATH BRET fluorescence. N- and C-terminal peptides were tested on Gαq (n = 8–12) (A), Gα11 (n = 8–12) (B), and Gα13 (n = 3–12) (C) in HEK293T cells and GαsS (n = 6–10) (D) in CHO cells. Each value represents the means ± SD (n = 3–18) (3 or 4 experiments). CHO, Chinese hamster ovary; fl-hNK1R, full-length human neurokinin-1 receptor; SP, substance P.

Figure 4.

Effect of SP on [Ca]_i in NK1R-transfected HEK293T cells. Increase in [Ca]_i elicited by SP and SP peptides in HEK293T cells overexpressing tr-hNK1R, fl-hNK1R, or fl- mouse (m)NK1R. A: [Ca]_i concentration-response curves to SP in cells expressing empty vector (EV), fl-human, fl-mouse (m), or tr-human NK1R. B: [Ca]_i response elicited by C- and N-terminal SP metabolites in HEK293T cells expressing empty vector. Cells overexpressed either hNK1R (C and D) or mNK1R (E and F) were treated with N-terminal (C and E) and C-terminal SP metabolites (D and F). SP was used as a reference. Values represent means ± SD (n = 3–29) (4 or 5 experiments). [Ca]_i, intracellular calcium concentration; fl-hNK1R, full-length human neurokinin-1 receptor; SP, substance P; tr, truncated.

Figure 5.

Effect of SP on cAMP in NK1R-transfected HEK293T and CHO cells. Stimulation of cAMP accumulation by SP and its N- and C-terminal SP metabolites in HEK293T (B and E) and CHO (A, C, D, and F) cells expressing hNK1R (B, C, E, and F) or empty vector (A and D). SP was used as a reference. Values represent means ± SD (n = 3–18) (4 or 5 experiments). CHO, Chinese hamster ovary; EV, empty vector; hNK1R, human neurokinin-1 receptor; SP, substance P.

Figure 6.

Effect of SP and SP 6–11 on [Ca]_i and cAMP in 3T3 fibroblasts. A and B: [Ca]_i and cAMP concentration-response curves with SP in 3T3 cells transfected with empty vector (EV) or hNK1R. C and D: [Ca]_i and cAMP concentration-response curves with SP and SP 6–11 in 3T3 cells transfected with hNK1R. E and F: [Ca]_i and cAMP concentration-response curve with SP in 3T3 cells transfected with hNK1R in the presence and absence of inhibitors of ACE, NEP, DPP-IV, and ECE-1. Values represent means ± SD (n = 4-25) (3-4 experiments). ACE, angiotensin-converting enzyme; DPP-IV, dipeptidyl-peptidase IV; ECE-1, endothelin-converting enzyme-1; [Ca]_i, intracellular calcium concentration; hNK1R, human neurokinin-1 receptor; NEP, neprilysin; SP, substance P.

Figure 8.

Effect of SP and SP 6–11 (100 nM) on wound healing and proliferation in NK1R-transfected 3T3 fibroblasts. A: micrographs of wound healing at 12 h with vehicle, SP, and SP 6–11. The light gray area represents the area of the wound. Wound healing measured with vehicle, SP, and SP 6–11 measured a % wound confluence over 0–12 h (B) (n = 12–14) and summarized for 12 h (C) (n = 12–14). Proliferation in response of vehicle, SP, and SP 6–11 measured as % confluency over 0–12 h for 30 nM (D) (n = 10–12) and 100 nM (F) (n = 17) and summarized data for 12 h at 30 nM (E) and 100 nM (G) (n = 17). Values represent means ± SD for B, D, and F and median, first and third quartile boxes, and minimum and maximum values for C, E, and G. *P < 0.05; **P < 0.01; ****P < 0.0001 (ANOVA and Tukey’s tests) (4 or 5 experiments). NK1R, neurokinin-1 receptor; SP, substance P.

Figure 9.

Involvement of cAMP and NF-κB in SP- and SP 6–11-induced proliferation in NK1R-transfected 3T3 fibroblasts. Proliferation in response of vehicle, forskolin (1 µM), to increase cAMP or the NF-κB inhibitor andrographolide (10 µM) was measured as % confluency over 0–12 h (A) (n = 12–24) and summarized data for 12 h (B) (n = 12–24). The effect of vehicle, SP, and SP 6–11 (100 nM) on NF-κB activation was measured after 4 h as change in luminescence in NK1R-3T3 cells with TNFα (100 nM) stimulation (C) (n = 12). Proliferation in response to vehicle, TNFα, SP + TNFα, and SP 6–11 + TNFα was measured as % confluency over 0–12 h (D) (n = 16–36) and summarized for the 4 h (E) (n = 16–36) and 12 h (F) (n = 16–36). Values represent means ± SD for A and D and median, first and third quartile boxes, and minimum and maximum values for B, C, E, and F. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001(ANOVA and Tukey’s tests) (4 or 5 experiments). NK1R, neurokinin-1 receptor; SP, substance P; TNFα, tumor necrosis factor α.

Figure 1.

Cellular metabolism of SP. Formation of peptide metabolites from exogenously added SP (1.5 µM) in cultures of 3T3 fibroblasts, primary mouse CAECs, and murine elicited PMs after 1 h of incubation. Amounts of SP 2–11 (A), SP 3–11 (B), SP 5–11 (C), SP 6–11 (D), SP 7–11 (E), SP 8–11 (F), SP 1–4 (G), SP 1–7 (H), and SP 1–9 (I) produced by the three cell types (n = 8 for each treatment). Data are presented as median, first and third quartile boxes, and minimum and maximum values. CAEC, coronary artery endothelial cells; PM, peritoneal macrophage; SP, substance P. *P < 0.05; **P < 0.01; ***P < 0.001 (ANOVA and Tukey’s tests; 3 experiments).

Figure 7.

Effect of SP and SP 6–11 on collagen release and COX-2 expression in NK1R-transfected 3T3 fibroblasts. A: collagen release with vehicle, SP, and SP 6–11 (30 nM) was measured after 24 h in the media of NK1R-3T3 cells by ELISA (n = 24–28). B: immunoblot of COX-2 from cells treated with vehicle, SP, and SP 6–11 (100 nM). Summarized data of band density for COX-2:β-actin ratio (n = 8). Data are presented as median, first and third quartile boxes, and minimum and maximum values. *P < 0.05; ****P < 0.0001 (ANOVA and Tukey’s tests) (3 or 4 experiments). COX-2, cyclooxygenase-2; NK1R, neurokinin-1 receptor; SP, substance P.

RESULTS

In Vitro Metabolism of SP by Cells of Importance in Heart Disease

All three cardiovascular cell types—fibroblasts, CAECs, and PMs—metabolized SP. After incubating cells for 1 h with 1,500 pmol/mL of SP, the cell media were analyzed by LC-MS/MS for 9 C- and N-terminally truncated SP metabolites. The most active cell type was the fibroblast with 840 pmol/mL of total metabolites formed, followed by CAECs with 660 pmol/mL, whereas PMs produced the least amount with 470 pmol/mL. SP 5–11 was the most abundant metabolite, followed by SP 1–4, SP 7–11, SP 1–7, SP 6–11, SP 8–11, SP 3–11, SP 2–11, and SP 1–9. The formation pattern of individual metabolites also revealed some differences. Out of nine metabolites, five showed significant or substantial differences between cell types (Fig. 1). Despite the PMs’ relatively low overall activity, these cells produced the highest amount of SP 3–11, SP 6–11, and SP 1–9. Fibroblasts and CAECs produced the greatest amount of SP 5–11, SP 1–4, and SP 7–11, respectively. SP 1–9, SP 2–11, and SP 8–11 were detected at low concentrations (range 0.5–10 pmol/mL) and with no significant differences between cell types.

G Protein Transducer Pathways Mediating SP Activation of the hNK1R

G protein activation was measured as a dissociation of Gαs from Gβγs by a reduction in BRET (28). Both the fl- and 96-amino acid C-terminal-deleted tr-NK1R were expressed in HEK293T cells. By immunoblotting, a 46-kDa protein band was detected in the fl-NK1R-expressing cells but not in the empty vector or the tr-NK1R-expressing cells (Fig. 2A). A 34-kDa protein was detected with cells expressing the tr-NK1R but not the empty vector or the fl-NK1R. Expression of the NK1R was confirmed by real-time polymerase chain reaction (RT-PCR) using specific primers (data not shown). In fl-NK1R-expressing HEK293T cells, concentration-related G-protein dissociations to SP were observed with both Gαq and Gα12 groups. The log IC₅₀ values were similar (−8.7 ± 0.08, −8.9 ± 0.12, −8.1 ± 0.16, and −8.5 ± 0.08 M for Gαq, Gα11, Gα12, and Gα13, respectively), but the maximal effects (E_max) varied significantly (Fig. 2B and Supplemental Table S2). In fl-NK1R cells transfected with members of the Gαi group, significant concentration-related decreases were observed with Gαz, GαoB, GαoA, Gαi1, and Gαi2, but not with Gαi3 (Fig. 2C). The E_max responses in this group were less than the Gαq and Gα12 groups (Fig. 2C and Supplemental Table S2). The log IC₅₀ values were −8.32 ± 0.15, −7.62 ± 0.18, and −7.19 ± 0.29 M for Gαz, GαoB, and GαoA, respectively (Supplemental Table S2). Changes were not observed by SP with the short (S) or long (L) forms of Gαs proteins in HEK293T cells (Fig. 2D); however, concentration-related dissociation of GαsS occurred with SP in NK1R-transfected CHO cells (log IC₅₀ of −9.2 ± 0.18 M) (Fig. 2E and Supplemental Table S2). SP had no effect on the dissociation of GαsL in CHO cells. Unlike GαsS, SP induced similar changes in Gαq and Gα12 groups in both CHO (Fig. 2E) and HEK293T cells (Fig. 2B). In contrast to cells transfected with the fl-NK1R, SP was without effect in cells transfected instead with the empty vector (Fig. 2, B–D). In tr-NK1R-expressing cells transfected with the Gα proteins, SP was without effect (Fig. 2F).

We tested the effect of C- and N-terminal metabolites of SP on activation of Gα11, Gαq, and Gα13 in HEK293T cells and GαsS in CHO cells expressing the NK1R (Fig. 3). Activation of Gαq, Gα11, and Gα13 and GαsS was observed with C-terminal metabolites (Fig. 3) (Supplemental Table S3). However, the extent of the activation became less with each amino acid loss from SP 2–11 to SP 7–11 with a log IC₅₀ value (Supplemental Table S3). In contrast, the N-terminal peptide SP 1–7 was without effect. Removal of the C-terminal amide, SP-free acid, reduced activity. The pattern of reductions in activation by shorter peptides was similar with Gαq, Gα11, and GαsS and more pronounced with N-terminal shortening with Gα13.

[Ca]_i Increases in hNK1R- and mNK1R-Expressing HEK293 Cells by SP Peptides

NK1R activating Gαq and Gα11 with SP implicates [Ca]_i as a second messenger. In both fl-mNK1R- and fl-hNK1R-expressing cells, SP stimulated a concentration-related increase in [Ca]_i (Fig. 4A). The [Ca]_i increases were transient, reaching a maximum at 50 s and then declining to baseline over 200 s (data not shown). There was no difference between [Ca]_i transient signals with hNK1R and mNK1R (data not shown). The SP concentration-response curves for mNK1R and hNK1R showed a small, but not statistically significant, difference (Fig. 4A) with log EC₅₀s of −8.20 ± 0.29 M and −8.53 ± 0.30 M, respectively (Supplemental Table S4). SP did not change [Ca]_i in cells expressing the tr-hNK1R or empty vector (Fig. 4, A and B).

Both N- and C-terminal peptides of SP were tested for their ability to activate fl-mNK1R and fl-hNK1R. [Ca]_i was not changed by the peptides in empty vector-expressing cells (Fig. 4B). Removal of the SP C-terminal amide to form SP-free acid reduced activity with the hNK1R (Fig. 4C) and mNK1R (Fig. 4E). No response was detected with the N-terminal SP 1–4, SP 1–7, or SP 1–9 peptides in mNK1R cells (Fig. 4E). In hNK1R cells, only SP 1–9 at largest concentration induced some moderate [Ca]_i increase (Fig. 4C). In contrast, C-terminal peptides were active on hNK1R and mNK1R (Fig. 4, D and F). The active SP peptides all produced transient increases in [Ca]_i (data not shown). Peptide length did not alter the shape or duration of these transients. Metabolites shortened by up to five amino acids (SP 2–11, SP 3–11, SP 5–11, and SP 6–11) either do not differ from SP or caused slightly higher [Ca]_i increases (Fig. 4, D and F) (Supplemental Table S4). Further shortening of the peptide significantly reduced activity. The concentration-response curves to SP 7–11 and SP 8–11 were shifted to the right. It is important to note that both SP and its metabolites have similar activities in the hNK1R and mNK1R. Log ED₅₀s and statistics are listed in Supplemental Table S4. Shortening of the N-terminal end of SP by up to five amino acids was associated with the gain of activity; however, further shortening, removal of the amide, or shortening of the C-terminal end reduced activity.

cAMP Increases in hNK1R-Expressing HEK293T and CHO Cells by SP Peptides

The NK1R also couples to Gαs and increases cAMP production (35, 36). SP increased intracellular cAMP in fl-hNK1R-expressing HEK293T cells (Fig. 5, B and E). The increase was concentration related with a log ED₅₀ of −7.81 ± 0.07 M (Supplemental Table S5). SP 1–7 was inactive, and SP-free acid only increased cAMP at the highest concentration (Fig. 5B). SP 2–11 and SP 3–11 had significantly reduced activity (log ED₅₀s of −7.4 ± 0.08 and −7.14 ± 0.06 M, respectively) compared with SP (Fig. 5E). The activities of SP 5–11 and SP 6–11 were reduced further with log ED₅₀s of −6.2 ± 05 and −5.7 ± 0.09 M, respectively. There was a 1,000-fold difference in the log ED₅₀s for these peptides between [Ca]_i and cAMP. SP 7–11 and SP 8–11 were without activity on cAMP but had significant activity on [Ca]_i (Fig. 5E). The peptides had no effect on cAMP in HEK293 cells transfected with the empty vector (Fig. 5A). N-terminal metabolism reduced the cAMP responses of SP. The reductions in activity with SP 5–11, SP 6–11, SP 7–11, and SP 8–11 were disproportionately greater with cAMP than [Ca]_i.

Measurements of cAMP were also carried out in CHO cells transfected with empty vector or fl-hNK1R (Fig. 5, A, C, D, and F, respectively) and confirmed the results in HEK293 cells. SP increased cAMP with a log ED₅₀ of −7.6 ± 0.14 in CHO cells. SP and SP 3–11 did not significantly differ in activity in CHO cells, but the activities of SP 5–11, SP 6–11, SP 7–11, and SP 8–11 were reduced similarly in CHO as in HEK293 cells (Fig. 5F). The peptides were without effect in CHO cells transfected with the empty vector (Fig. 5D). Interestingly, SP produced similar increases in cAMP in HEK293 and CHO cells (Fig. 5), despite the cells differing in SP activation of GαsS (Fig. 2).

Effects of SP and SP 6–11 in 3T3 and NK1R-Transfected 3T3 Fibroblasts

SP did not alter [Ca]_i or cAMP in 3T3 fibroblasts transfected with empty vector (Fig. 6, A and B). Also, the NK1R was not detected by RT-PCR in 3T3 cells (data not shown). However, when 3T3 cells were transfected with the hNK1R, SP elicited a concentration-related increase in both [Ca]_i and cAMP with log EC₅₀ of −8.53 ± 0.27 and −8.04 ± 0.18 M, respectively (Fig. 6, A–D). SP 6–11 also increased [Ca]_i and cAMP with log EC₅₀ of −8.07 ± 0.27 and −6.78 ± 0.27 M, respectively (Fig. 6, C and D). Of note, SP 6–11 was more potent, with a smaller ED₅₀, in increasing cAMP in NK1R-3T3 cells than NK1R-HEK293 cells (Fig. 6D and Supplemental Table S6). Under these experimental conditions, the [Ca]_i and cAMP responses to SP were not altered by pretreatment of the fibroblasts with inhibitors of ACE, DPP-IV, ECE-1, and NEP (Fig. 6, E and F). Based on these results, NK1R-3T3 fibroblasts were selected as an experimental model to determine whether SP N-terminal metabolism reduces cAMP-mediated cell functions. Since SP 6–11 shows the greatest reduction in cAMP accumulation but an identical increase in [Ca]_i as SP, it was chosen for these studies.

Collagen secretion was measured in NK1R-transfected 3T3 fibroblasts (Fig. 7A). SP significantly increased collagen secretion when compared with the vehicle. SP 6–11 also stimulated collagen secretion when compared with the vehicle but was less active than SP. A 72-kDa COX-2 immunoreactive protein was detected in NK1R-3T3 cells (Fig. 7B). SP significantly increases the expression of COX-2 when compared with the vehicle. However, its expression was not increased by SP 6–11.

The effects of SP and SP 6–11 on wound healing were determined with the fibroblasts (Fig. 8). Micrographs of wound healing with the three experimental treatments are illustrated in Fig. 8A and indicate more cells in the light gray wound area with vehicle than SP peptides. Wound healing was measured as % wound area confluence and increased over the 12 h with vehicle (Fig. 8B). SP and SP 6–11 deceased the rate of wound area confluence with the effect of SP being greater than SP 6–11. Figure 8C summarizes the wound area confluence at 12 h. Both SP and SP 6–11 significantly reduced wound healing; however, the inhibition by SP was significantly greater than SP 6–11.

The effects of the peptides were tested on cell proliferation in NK1R-3T3 fibroblasts using two methods. First,³H-thymidine incorporation was measured after 24-h incubation in serum-starved cells. Compared with the vehicle, SP (30 nM) significantly inhibited the incorporation of ³H-thymidine from 16,487 ± 3,704 counts/min (n = 27) to 12,947 ± 2,152 counts/min (n = 29) (P < 0.001), whereas SP 6–11 (30 nM) was without effect (16,489 ± 2,794 counts/min, n = 21). Second, the proliferation was also determined by measuring the cell number (Fig. 8). The % confluence increased linearly over 12 h with vehicle treatment (Fig. 8, D and F). As with thymidine incorporation, SP (30 nM) decreased the rate of confluence, whereas SP 6–11 (30 nM) did not decrease the rate of confluence (Fig. 8D). However, at the 100-nM concentration, both SP and SP 6–11 inhibited the rate of confluence when compared with vehicle (Fig. 8F). SP had a greater effect than SP 6–11. The % confluence at 12 h is summarized for 30 nM (Fig. 8E) and 100 nM (Fig. 8G) concentrations. At 30 nM, only SP significantly decreased proliferation. At the higher concentration, SP and SP 6–11 significantly reduced proliferation, but the reduction by SP was significantly greater than SP 6–11.

To gain an insight into the mechanism of SP inhibition of proliferation, NK1R-3T3 cells were treated with forskolin (1 µM) to increase cAMP and with the NF-κB inhibitor andrographolide (10 µM). Like SP, both forskolin and andrographolide decrease the rate of confluence when compared with the vehicle (Fig. 9A). The % confluence at 12 h was significantly decrease by both compounds (Fig. 9B). The effects of SP and SP 6–11 were tested on NF-κB activity after 4.5 h in NK1R-3T3 fibroblasts. SP, but not SP 6–11, significantly decreased basal NF-κB activity (P < 0.01) (data not shown). TNFα significantly increased NF-κB activity (Fig. 9C). This increase by TNFα was blocked by SP and significantly inhibited by SP 6–11. SP was significantly more active than SP 6–11. Over the same 0–12 h, TNFα increased the rate of proliferation compared with vehicle, and the increase with TNFα was blocked by SP (Fig. 9D). SP 6–11 was without effect. Figure 9, E and F, shows the summarized data for the 4-h and 12-h time points, respectively. These studies raise the possibility that SP increases cAMP that inhibits NF-κB activity, resulting in reduced proliferation.

Summary of SP and SP Peptide Effects on [Ca]_i and cAMP

Figure 10A compares the [Ca]_i and cAMP responses to 100 nM SP and SP peptides in HEK293 cells based on data in Figs. 4 and 5. It emphasizes that the N-terminal metabolism of SP has little effect on [Ca]_i responses, but cAMP responses decline stepwise with the removal of each N-terminal amino acid. SP 5–11 and SP 6–11 are as active as SP on [Ca]_i but have little to no cAMP response. Removal of the C-terminal amide also reduces the cAMP responses more than [Ca]_i. Thus, depending on the concentration, SP metabolism can select the second messenger pathway mediating the actions of SP.

Figure 10.

Summary of the [Ca]_i and cAMP responses to SP and SP peptides in HEK293 cells expressing the hNK1R. A: comparison of [Ca]_i (dark bars) and cAMP (light bars) responses to 100 nM SP and each of its N- and C-terminal metabolites. Peptidases producing the SP metabolites are indicated below the x-axis of the graph. Values represent means ± SD. *P < 0.05 (Student’s t test). B: diagram of SP metabolism selectively activating the NK1R to increase [Ca]_i rather than cAMP and effects on cell proliferation, wound healing, COX-2 expression, and collagen release in the NK1R-3T3 fibroblast experimental model. The image was created with BioRender.com. [Ca]_i, intracellular calcium concentration; hNK1R, human neurokinin-1 receptor. Figure created with BioRender.com.

DISCUSSION

SP is modified by both N-terminal and C-terminal metabolism by a number of enzymes (Fig. 10A). DPP-IV and FAP remove the N-terminal dipeptides from SP, giving rise to SP 3–11 and SP 5–11 (37, 38). C-terminal metabolism is accomplished by dipeptide or tripeptide elimination by ACE (39). The resulting metabolites include SP 1–9 and SP 1–8. The endopeptidases NEP and NLN cleave SP to produce two peptides. NEP produces SP 1–6 and SP 7–11 or SP 1–7 and SP 8–11 as well as SP 1–9, and NLN cleaves SP to SP 1–5 and SP 6–11 and SP 1–7 and SP 8–11 as well as SP 1–8 (39, 40). ECE-1 also metabolizes SP with a specificity like NEP (41), and THOP has the same specificity as NLN (40). MMP-9 and MMP-8 produce mainly SP 7–11 and SP 1–9 as a minor product (42, 43). SP has a C-terminal amide group that is removed by deamidases producing SP-free acid (44–46). These enzymes may hydrolyze SP alone or act in combination, resulting in the formation of a variety of SP peptides of varying length. DPP-IV, FAP, ACE, NLN, THOP, ECE-1, and NEP are widely distributed in tissues and cells including cells of the cardiovascular system (47–50). They are plasma membrane-associated enzymes that are capable of extracellular metabolism of charged, hydrophilic peptides such as SP (47, 51). Specifically, we studied the metabolism of SP by fibroblasts, endothelial cells (ECs), and PMs. All three cell types metabolized SP, but the patterns of metabolites differed among the cells. Fibroblasts degraded SP to a greater extent than CAECs or PMs. N-terminal metabolism tended to occur to a greater extent than C-terminal metabolism. SP 5–11 was the most common metabolite formed by the three cell types as a likely result of sequential N-terminal metabolism by DPP-IV and FAP. This possibility is supported by our detection of SP 3–11 formation by the cells. The NEP, ECE-1, or MMP-9 metabolite SP 7–11 was produced to a lesser extent than SP 5–11. The major metabolite that was produced by C-terminal metabolism was SP 1–4 with lesser amounts of SP 1–7 and SP 1–9. The NLN and THOP metabolites SP 1–8 and SP 6–11 were minor metabolites. ECs are considered a major source of ACE so it was surprising that CAECs produced very little SP 1–8 and SP 1–9. However, ECs produced the greatest amount of the MPP-9, ECE-1, and NEP metabolite SP 7–11 among the cells studied. The brain and umbilical ECs also produced SP 5–11 as a major metabolite and SP 1–9 was a minor product (32, 50); however, these studies did not analyze for SP 7–11. Interestingly, human smooth muscle cells were more active than EC in metabolizing SP (50). PMs and fibroblasts also produced only small amounts of SP 1–7 and SP 1–9. By comparison, the liver produced SP 5–11, SP 6–11, and SP 1–7 as major metabolites, whereas the spinal cord produced predominantly SP 8–11 and SP 1–7 (52, 53).

The NK1R mediates most of the actions of SP including pain, vasodilation, cardioprotection, cardiac remodeling, and immune modulation and shows specificity for SP when compared with other kinins and neurokinins (1, 54). Understanding the activity of the major SP metabolites on NK1Rs has broad implications in physiology and disease. It will provide important insights into the role of metabolism in limiting or altering the actions of SP in the cardiovascular, gastrointestinal, and central nervous systems and indicate the therapeutic potential of inhibitors of SP metabolism. We tested the effects of SP metabolites on the fl- and tr-NK1Rs (1, 3). The activities of SP peptides were compared between the human and mouse NK1Rs. SP peptides have not previously been studied on the mouse NK1R. Some SP metabolites have been tested on the NK1R (54).

The fl-NK1R and tr-NK1R, an isoform missing the 96 C-terminal amino acids (3), both retain the binding sites for SP and SP antagonists (26, 27). The fl-NK1R is coupled to Gαq and Gαs and increases [Ca]_i, ERK phosphorylation, and cAMP in cells. The tr-NK1R is not thought to be G protein-coupled based on the crucial role of the C-terminal region of the receptor for G protein activation (26, 27). SP increases ERK phosphorylation in cells expressing the tr-NK1R with a time course that is much slower than the fl-NK1R. We directly tested coupling of the fl- and tr-NK1Rs to a series of G proteins using BRET detection of G protein subunit dissociation. SP stimulation of the fl-NK1R promoted a concentration-related dissociation of the Gα subunits from the βγ subunits with members of the Gαq/12 groups showing the greatest maximal efficacy. Smaller maximal effects were observed with Gαs, Gαz, and Gαo and little Gαi activity. Using GTP photoaffinity labeling and immunoprecipitation with Gα subtype-specific antibodies, Gαq/11, Gαs, and Gαo, but not Gαi, were activated by SP in NK1R-expressing CHO cells (35). In the current study, the tr-NK1R did not initiate dissociation of any of the G proteins studied, confirming the importance of the C-terminal tail of the receptor in G protein interaction. Similarly, SP did not increase [Ca]_i in tr-NK1F cells. These data also indicate that other mechanisms must mediate ERK phosphorylation that is attributed to the tr-NK1R. We tested a series of SP metabolites on activation of Gαq, Gα11, Gαs, and Gα13 by the fl-NK1R. The activation of the four Gαs was less with SP metabolites than with SP. Thus, changes in Gα protein activation occurred with SP metabolites. A similar pattern and order of activity for the SP peptides were observed in SP ligand-binding assays in NK1R-expressing cells (54–56). Several G proteins that were not previously known to couple to the fl-NK1R were identified, laying the foundation for further studies on SP and NK1R signaling.

Since the NK1R is coupled to Gαq, we examined in a more comprehensive manner the activity of SP metabolites on NK1R-mediated changes in [Ca]_i in HEK293 cells. SP increased [Ca]_i in cells expressing the fl-hNK1R, consistent with the activation Gαq and/or Gα11 by SP. Thus, N-terminal metabolism of SP by DPP-IV results in SP peptides with full agonist activity and potency similar to SP. This may explain why the activity of SP is not altered by DPP-IV inhibitors in humans (57). In contrast, C-terminal metabolism by ACE, ECE-1, MMP-9, NLN, and/or NEP to SP 1–9, SP 1–7, or SP 1–4 will reduce or eliminate the NK1R activity of SP, and inhibitors of these enzymes should enhance or prolong the activity of SP. Removal of the C-terminal amide of SP with an amidase such as calpain or cathepsin A reduces activity by only 12-fold. These peptidases are likely to act in combination in degrading SP, and cells may differ in the peptidase(s) responsible for the metabolism of SP. Thus, a single peptidase inhibitor will not adequately protect SP from inactivation.

The NK1R also couples to Gαs that signals through cAMP, and SP increases cAMP in cells expressing the NK1R (35, 36). SP metabolites have not been extensively studied on cAMP. As with [Ca]_i, SP increased cAMP in a concentration-related manner. The cAMP-stimulating activity of the SP peptides decreased with the removal of N-terminal amino acids and decreased to a much greater extent than was observed when [Ca]_i was measured. This difference was most pronounced with SP 5–11, SP 6–11, and SP 7–11. Similarly, SP-free acid had little effect on cAMP but was a potent agonist on [Ca]_i. The N-terminal metabolite SP 1–7 failed to increase cAMP and [Ca]_i. At physiological concentrations, the differences in the ability of the SP peptides to stimulate [Ca]_i and cAMP are strikingly apparent. This is illustrated with 100-nM peptides in Fig. 10A. The N-terminal metabolism of SP reduces cAMP accumulation with the removal of each amino acid without altering calcium stimulation. SP 6–11 and SP-free acid selectively stimulate calcium at this concentration.

Although there has not been a comprehensive evaluation of SP metabolites in the past, SP and SP 6–11 have been compared on Gαq-mediated inositol phosphate and [Ca]_i increase and Gαs-mediated cAMP accumulation (36, 55, 58, 59). SP increased both [Ca]_i and/or inositol phosphate and increased cAMP accumulation. In contrast, SP 6–11 had similar activity on [Ca]_i and inositol phosphate as SP, but its ability to increase cAMP was greatly reduced. We found similar results with SP 6–11 and expanded these findings to include other SP metabolites such as SP 7–11 and SP-free acid. These differences in G protein activation by SP and SP 6–11 were attributed to the activation states of NK1R. Mutation of specific residues of the receptor regulated the Gαq and Gαs coupling (55, 58). Combining NK1R structural analysis with G-protein coupling, Harris et al. showed the N-terminal end of SP interacts with the second extracellular loop of the NK1R and confers Gαq and Gαs coupling (59). SP 6–11 with the shorten N-terminal end does not interact with the extracellular loop, so Gαs coupling is reduced and Gαq coupling is retained. The consequences of this signaling shift with SP 6–11 on cell function have not been studied. Similar to SP 6–11, we found that SP-free acid was a potent agonist in increasing [Ca]_i and had little effect on cAMP. This was not previously recognized, and the structural basis was not studied.

Other new aspects of SP-NK1R pharmacology were uncovered and are in need of further study. We found that the SP-activated NK1R couples to a variety of Gα proteins, and activation of some of these Gα proteins also differ between SP and SP metabolites. Thus, SP peptides may differentially affect NK1R coupling to other Gα proteins and enlist other second messengers as described for Gαq and Gαs. This adds to the complexity of SP metabolism on the cellular response.

SP and SP 6–11 have been compared in detail on the NK1R as discussed earlier. Since SP 6–11 selectively stimulates [Ca]_i and SP stimulates both [Ca]_i and cAMP, we compared the cellular actions of the two peptides. Cardiac fibroblasts in primary cultures have NK1Rs and respond to SP (8, 24, 60, 61); however, 3T3 fibroblasts do not. Transfection of the NK1R into 3T3 cells bestowed consistent and reliable responsiveness to SP peptides. For example, SP and SP 6–11 increased [Ca]_i in NK1R-3T3 fibroblasts with similar ED₅₀s, whereas SP was more potent than SP 6–11 in increasing cAMP. Thus, N-terminal metabolism of SP does not change calcium signaling but reduces cAMP signaling. The NK1R-3T3 fibroblasts provided a reliable experimental model to investigate the impact of N-terminal metabolism on cells.

The actions of SP and SP 6–11 were compared on some functions in NK1R-3T3 fibroblasts. Overall, SP and SP 6–11 differed in their cellular effects. SP inhibits proliferation, wound healing, and NF-κB activity to a greater extent than SP 6–11. Elevations in cAMP are known to decrease proliferation, wound healing, and NF-κB activity (62–66). In contrast, increases in [Ca]_i stimulate these effects (67, 68). We found that forskolin that increases cAMP inhibited 3T3 cell proliferation like SP. These results are consistent with cAMP mediating the inhibition of these cellular effects by SP. SP 6–11 at 30 nM did not increase cAMP; however, at 100 nM, it has 25% of the activity of SP in increasing cAMP in NK1R-3T3 fibroblasts. Thus, the diminished inhibition of cellular activities by SP 6–11 compared with SP is explained by its reduced ability to elevate cAMP. NF-κB is a mediator of cell growth through cyclin D1 induction (69), and inhibition of NF-κB activity by a NF-κB inhibitor reduced NK1R-3T3 fibroblast proliferation. Thus, SP, and to a lesser extent SP 6–11, inhibits proliferation and migration through cAMP inhibition of NF-κB activity. There may be other activities of cAMP that affect proliferation. SP, but not SP 6–11, inhibits thymidine incorporation. Along these lines, cAMP is known to inhibit thymidine kinase (70, 71). Both SP and SP 6–11 increase collagen release; however, SP is significantly more active than SP 6–11. Collagen release by SP peptides may also be mediated by cAMP since increases in [Ca]_i decrease collagen release (72). Physiological elevations in cAMP increase collagen release, whereas high cAMP levels from pharmacological agents decrease collagen by degradation of newly synthesized collagen (73–75). Increases in cAMP also decrease transforming growth factor β (TGFβ)-mediated collagen release (66, 76). Thus, the effects of SP on 3T3 fibroblasts are numerous. However, the N-terminal metabolism may result in decreases in some pro- and antifibrotic effects of SP.

NF-κB is also a major regulator of inflammation, promoting the expression of inflammatory proteins such as cytokines, immune cell adhesion molecules, and COX-2 (77). SP inhibition of NF-κB activity would be expected to reduce inflammation. However, SP also increased the expression of COX-2, which is likely mediated by cAMP (78, 79). Any pro- and anti-inflammatory activities of SP may depend on NF-κB activity, its regulation by cAMP, and other cellular activities of cAMP (80, 81).

These in vitro studies of SP metabolism, signaling, and cell function provide a framework for future research into the beneficial and detrimental effects of SP and the utility of pharmacological interventions that inhibit SP metabolism. This research has implications to cardiovascular disease as well as diseases affecting the central nervous and gastrointestinal systems. For example, in acute cardiac ischemia, SP promotes vasodilation, increases perfusion, reduces infarct size, and improves cardiac function (19–22). However, SP also promotes adverse cardiac remodeling in chronic models of hypertension and heart failure (8, 9). The extent of SP metabolism in the heart and the contribution of SP metabolism to these beneficial and/or detrimental cardiac effects are not known. However, inhibiting SP inactivation and prolonging its duration of action may not be a useful therapeutic approach. In fact, DPP-IV inhibition and ACE inhibition failed to alter the vasodilator response to SP in human subjects (57) and increased plasma norepinephrine in hypertensives (82). Since calcium and cAMP often have opposite effects in cells comprising the vasculature, heart, and sympathetic and immune systems, the response to DPP-IV inhibitors in cardiovascular disease may be complicated by N-terminal metabolism dialing back the cAMP pathway while maintaining the calcium pathway. Also, in small human resistance arteries, NEP and ACE inhibitors enhanced the dilation to bradykinin but not SP (83). These studies suggest that these inhibitors of SP inactivation may not enhance the beneficial effects of SP in myocardial ischemia and complicate heart failure. Although NEP inhibitors do not show clinical benefit in heart failure or hypertension, the combination of NEP and ACE inhibition is more effective than an ACE inhibitor alone (84, 85). ACE inhibitors alone are beneficial in heart failure (86), and plasma SP concentrations are increased in patients with heart failure treated with ACE inhibitors (87). SP was not studied in patients with heart failure treated with NEP and ACE inhibitors. It is possible that the beneficial effects of ACE and NEP and ACE inhibition in heart failure are opposed by the drugs’ ability to increase in SP and SP-mediated adverse cardiac remodeling.

In summary, N- and C-terminal metabolism of SP occurs in fibroblasts, macrophages, and endothelial cells (Fig. 10B). N-terminal metabolism predominates, and SP 5–11 is the major metabolite of these cells. SP 5–11 is as active as SP in increasing [Ca]_i but less active than SP in increasing cAMP. Thus, N-terminal metabolism selectively downregulates the cAMP pathway, and the cAMP-mediated cellular responses of SP are reduced, whereas [Ca]_i-mediated cellular responses are unaltered. In contrast, N-terminal metabolites such as SP 1–9 have no NK1R activity. Interestingly, C-terminal deamidation of SP also diminishes the NK1R signaling pathway mediated by Gαs/cAMP. Thus, some SP-metabolizing enzymes have the ability to tune the cellular response to SP by producing signaling pathway-selective NK1R agonists. Future studies will determine whether metabolism explains some of the different effects of SP in various cell types and pathological conditions.

DATA AVAILABILITY

All data will be made available upon request.

SUPPLEMENTAL DATA

Supplemental Tables S1–S5.pdf: https://doi.org/10.6084/m9.figshare.25669074.v1.

GRANTS

These studies were supported by a grant from the National Heart, Lung and Blood Institute (HL-132908).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.-K.P., A.W., S.P.L., and W.B.C. conceived and designed research; T.K., J.N., A.H., and M.J.T. performed experiments; T.K., J.N., A.H., and M.J.T. analyzed data; T.K., S.L.P., and W.B.C. interpreted results of experiments; T.K. prepared figures; W.B.C. drafted manuscript; T.K., S.-K.P., S.L.P., A.W., S.P.L., and W.B.C. edited and revised manuscript; S.P.L. and W.B.C. approved final version of manuscript.