Pepducin-mediated GPCR signaling in the cardiovascular system

Abstract

Pepducins are small-lipidated peptides designed from the intracellular loops (iLs) of G protein-coupled receptors (GPCRs) that act in an allosteric manner to modulate the activity of GPCRs. Over the last two decades, pepducins have progressed initially from pharmacologic tools used to manipulate GPCR activity in an orthosteric site-independent manner to compounds with therapeutic potential that have even been used safely in Phase 1 and 2 clinical trials in human subjects. The effect of pepducins at their cognate receptors has been shown to vary between antagonist, partial agonist or biased agonist outcomes in various primary and clonal cell systems, with even small changes in amino acid sequence altering these properties and their receptor selectivity. To date, pepducins designed from numerous GPCRs have been studied for their impact on pathologic conditions, including cardiovascular diseases such as thrombosis, myocardial infarction and atherosclerosis. This review will focus in particular on pepducins designed from protease-activated receptors (PARs), C-X-C motif chemokine receptors (CXCRs), formyl peptide receptors and the β2-adrenergic receptor. We will discuss the historic context of pepducin development for each receptor, as well as the structural, signaling, pathophysiologic consequences and therapeutic potential for each pepducin class.

Pepducins as allosteric modulators of GPCR signaling in the cardiovascular system

GPCRs have long been recognized as easily accessible key regulators of a multitude of cellular functions and remain one of the most highly targeted drug class across a number of disease states^{1, 2}. Within the cardiovascular system, GPCRs have been studied for their roles in the regulation of cardiac and vascular function and remodeling, as well as immune cell regulation, under both physiologic and pathologic conditions such as cardiac injury, hypertension, atherosclerosis and thrombosis³. Strategies to manipulate GPCR signaling toward ameliorating pathologic conditions have traditionally relied upon the use of orthosteric ligands that either activate the receptor similar to its endogenous ligand, or competitively antagonize the binding and action of the endogenous ligand at the receptor. However, allosteric modulation of GPCRs has been studied more frequently as a means to manipulate GPCR signaling independent of the orthosteric ligand binding site^{4, 5}.

With a multitude of potential allosteric sites on GPCRs, there are a several potential signaling outcomes that may manifest for any given GPCR. For instance, some allosteric modulators act as partial agonists independently of the endogenous orthosteric ligand, while others may act to antagonize the action of the orthosteric ligand^{4, 5}. Further, some allosteric modulators may induce structural alterations in their GPCR that preferentially promote or inhibit the engagement of one proximal signaling pathway while maintaining receptor coupling to and activation of other pathways. The recognition that GPCRs can be structurally manipulated to engage distinct proximal signaling mediators to produce different outcomes than the endogenous orthosteric ligand is now recognized as biased signaling^{4, 6}. While much work is underway to determine the benefit of promoting biased GPCR signaling for the treatment of specific cardiovascular pathologies, allosteric modulators represent an area with growing potential for manipulating GPCRs in a signaling pathway-selective manner.

Pepducins, small-lipidated peptides designed from the intracellular loops (iLs) of GPCRs, fall into this category of allosteric modulators. Introduced in the early 2000s by the Kuliopulis group^{7, 8}, a growing body of pepducin-focused work has examined their movement across the membrane, structural requirements for association with their cognate receptors, as well as their antagonist, partial agonist or biased agonist properties in various primary and clonal cell systems^{7, 9–14}. The addition of a long hydrophobic tether, most commonly the 16-carbon fatty acid palmitate, at the N-terminus of the peptide confers cell penetrance by allowing the pepducin to insert into and flip across the plasma membrane where it may then associate with and allosterically modulate its GPCR^{7–9, 11, 15–17}. The palmitoylation-mediated membrane transport of pepducins appears to be crucial for their activity, since non-palmitoylated versions lack cell penetrance and are unable to induce GPCR-dependent responses. Once located on the inner leaflet of the plasma membrane, pepducins have been observed to be membrane-associated for at least 30 minutes and exert either transient (seconds to minutes) or long-lasting (hours to days) effects on cell signaling readouts, depending on the pathway and cell type^{7, 11, 18–23}, although the mechanism(s) by which pepducins undergo intracellular handling to terminate their activity are unclear.

Beyond in vitro work, pepducins have increasingly been demonstrated to exert functional effects on different cell types and tissues in vivo, with clear beneficial effects in pathological animal models, as well as measurable bioavailability, safety and tolerability in humans^{15, 18, 24}. While only a handful of pepducins have thus far been studied in terms of their pharmacokinetic properties, the Kuliopulos group used a radiolabeled pepducin to determine that intravenous or subcutaneous injection in mice resulted in preferential distribution to highly perfused tissues including the liver, kidneys, lung and spleen at lower doses, with distribution into the heart, blood, muscle and fat achieved at a 10-fold higher dosage¹⁵. Notably, the pepducin was not detected in the brain at either dose, suggesting it was unable to appreciably cross the blood brain barrier. Within a matter of hours, the radiolabel was detected in the urine, indicating that renal excretion plays a role in the systemic elimination of pepducins¹⁵. Overall, intravenous and subcutaneous administration of pepducins have been shown in mice, nonhuman primates and human subjects to result in rapid blood plasma accumulation that peaks within an hour or so, with a half-life of 0.8~1.8 hour in baboons and human subjects, and undergoes clearance to nearly undetectable levels within 24 hours, depending on the dosage and route of administration^{10, 18, 25}.

Pepducins designed from several GPCRs have been reported to exert effects in various processes, including CB1 cannabinoid receptor 1, G-protein-coupled receptors 31 and 35, neurotensin receptor type 1, pituitary adenylate cyclase-activating polypeptide type 1 receptor, sphingosine-1-phosphate receptor 3 and urotensin II receptor^26–34. However, the focus of our current review will center on pepducins designed from GPCRs with obvious application to the cardiovascular system and that have been characterized in several reports, namely those derived from protease-activated receptors (PARs), C-X-C motif chemokine receptors (CXCRs), formyl peptide receptors (FPRs) and the β2-adrenergic receptor (β2AR) (Figure 1).

Figure 1.

Pepducins as allosteric modulators of GPCR signaling in the cardiovascular system

Protease-activated receptors (PARs)

The first examples of the use of pepducins to modulate the activation or inhibition of cellular responses were demonstrated by the Kuliopulos group, whose work focused on pepducins designed from the iLs of PAR1, PAR2 or PAR4^{7, 8, 35}. In an initial screen of iL3-based palmitoylated pepducins for PAR1, the group discovered both activating and inhibiting properties of distinct pepducins (Table 1), with serial truncations of the N-terminal portion of the peptide sequences leading to a shift from PAR1 agonism to antagonism⁷. Functionally, while the longer palmitoylated pepducins (P1pal-19, P1pal-13) promoted biphasic generation of the 2^nd messenger IP₃ in fibroblasts, they increased Ca²⁺ mobilization and aggregation of healthy human platelets to a similar extent as the extracellular PAR1 ligand SFLLRN. Conversely, the N-terminally truncated palmitoylated pepducins (P1pal-12, P1pal-7) lacked these properties and were able to block PAR1 agonist-induced 2^nd messenger generation and platelet aggregation in a concentration-dependent manner. Notably, the selectivity of the activating pepducins also appeared to be reliant on sequence length, as P1pal-19 fully activated both PAR1 and PAR2, as well as partially activated another GPCR, cholecystokinin B receptor, whereas P1pal-13 was only capable of activating PAR1⁷. Additionally, it was shown that not only was the receptor required for pepducin-mediated responses, but that the C-terminal tail was specifically required, suggesting it may provide a point of interaction for iL3-designed pepducins to help coordinate their actions between the receptor and G protein. Indeed, subsequent structural analysis of the association between P1pal-19 and PAR1 demonstrated that the pepducin requires helix 8 in the C-terminus of the receptor to promote the stabilization of GTP-bound G protein, which was not required for extracellular ligand activation of PAR1³⁶.

Table 1.

Pepducin	GPCR	Domain	Sequence	Action	References
P1pal-19	PAR1	iL3	pal-RCLSSSAVANRSKKSRALF	Agonist	7
P1pal-13	PAR1	iL3	pal-AVANRSKKSRALF	Agonist	7
P1pal-12	PAR1	iL3	pal-RCLSSSAVANRS	Antagonist	7
P1pal-7 (PZ-218)	PAR1	iL3	pal-KKSRALF	Antagonist	7, 18, 24, 37
x1/2(LCA/pal)-i1	CXCR1/2	iL1	lca- or pal-YSRVGRSVTD	Antagonist	50, 51
x1/2pal-i3	CXCR1/2	iL3	pal-RTLFKAHMGQKHR	Antagonist	50, 52
x4pal-i1 (PZ-218)	CXCR4	iL1	pal-MGYQKKLRSMTD	Antagonist	50, 53
x4pal-i3 (PZ-210)	CXCR4	iL3	pal-HSKGHQKRKALK	Antagonist	50, 53
ATI-2341	CXCR4	iL3	pal-MGYQKKLRSMTDKYRL	Gi protein-biased agonist	19, 54, 59
F2pal-16	FPR2	iL3	pal-KIHKKGMIKSSRPLRV	Agonist	20, 68
F2pal-12	FPR2	iL3	pal-KGMIKSSRPLRV	Antagonist	67
F2pal-10	FPR2	iL3	pal-KIHKKGMIKS	Gi protein-biased agonist	20, 69, 70
ICL3–8	β2AR	iL3	pal-LQKIDKSEGRFHV	Direct Gs protein activation	21
ICL3–9	β2AR	iL3	pal-GRFHVQNLSQVEQDGGRT---IGII	Gs protein-biased agonist	21
ICL1–9	β2AR	iL1	pal-TAIAKFERLQTVTNYFIT	β-arrestin-biased agonist	21, 22, 76, 77

iL, intracellular loop; pal, palmitate; lca, lithocholate

--- in ICL3–9 sequence denotes 20 skipped amino acids in the IL3 of β2AR prior to the first 4 amino acids in transmembrane domain 6

PAR1-designed pepducins have been the most studied to date in terms of their functional effects on cardiovascular parameters such as platelet aggregation. In fact, the most successful example of a pepducin being used translationally is PZ-128 (P1pal-7), which the Kuliopulos group has taken from in vitro studies to clinical trials. After initial identification as a PAR1 antagonist⁷, PZ-128 was shown to decrease PAR1-mediated platelet aggregation and arterial thrombosis under high flow conditions in platelet PAR1-expressing rodents (guinea pigs) and non-human primates (baboons)³⁷. In comparison to the small molecule PAR1 extracellular ligand binding antagonist RWJ-56110, PZ-128 produced a stronger rightward (inhibitory) shift in the thrombin activation curve in human platelets and, in contrast to the direct thrombin inhibitor bivalirudin, PZ-128 had no effect on activated clotting time of human blood samples. Importantly, PZ-128 did not adversely impact hemostatic parameters such as bleeding time or platelet count, and platelet function was recovered within 24 hours of infusion, suggesting that pepducin-mediated PAR1 inhibition provides a safe manner to rapidly and reversibly inhibit platelet activity. A subsequent study in human subjects verified that short-term continuous infusion of PZ-128 was well-tolerated in patients with coronary artery disease, acting to reduce PAR1-mediated platelet aggregation in a dose-dependent manner with no significant hemostatic effects¹⁸. Recently, a randomized placebo-controlled Phase 2 study (Thrombin Receptor Inhibitory Pepducin in Percutaneous Coronary Intervention, TRIP-PCI) was performed to assess the safety and efficacy of PZ-128 infusion, in addition to standard antiplatelet therapy, in patients undergoing percutaneous coronary intervention²⁴. While the overall incidence of major adverse coronary events were not different between the PZ-128 and placebo groups, sub-group analysis showed that patients with elevated baseline cardiac troponin I did have a significant relative risk reduction with PZ-128 at 30 days post-intervention. Overall, these studies suggest that infusion of the PAR1 antagonist pepducin PZ-128 prior to cardiac catheterization may be able to reduce post-procedure thrombotic events and improve long-term patient outcomes.

Beyond platelet aggregation, PAR-based pepducins have increasingly been explored for their impact on numerous cell types and disease states in which PARs contribute. For instance, pepducin antagonists of both PAR1 and PAR2 have been shown to inhibit survival signaling, migration or metastasis in several cancer cell types, including ovarian, breast and lung^{25, 38–40}. Notably, long-term in vivo administration of PAR1 antagonist pepducins was not associated with increased immunogenicity or antibody production against the pepducins themselves⁴⁰. Although not directly tested in the context of cardiac disease models, PAR-based pepducins have also been demonstrated to impact processes including neutrophil activity and inflammation^{35, 41–43} and fibrosis^44–46, which are particularly relevant to the response to ischemic cardiac injury and long-term development of heart failure, respectively. Additionally, PAR-based pepducins have been shown to impact parameters of blood vessel regulation normally, in response to injury or during development of disease, including regulation of vascular relaxation⁴⁷, neointimal hyperplasia⁴⁸ and atherosclerosis⁴⁹. Thus, PAR-based pepducins, particularly those with antagonist properties, have tremendous potential for use in numerous cardiovascular pathophysiologic states.

C-X-C motif chemokine receptors (CXCRs)

Aside from PARs, pepducins designed from the iLs of chemokine receptors have also been investigated for their impact on inflammatory processes. Initially, chemokine receptor-designed pepducins were developed based on the iLs of CXCR1, CXCR2 and CXCR4⁵⁰. Since human CXCR1 and CXCR2 have identical iL1 and iL3 amino acid sequences, lipidated pepducins derived from these loops (x1/2LCA-i1, x1/2pal-i3, respectively, Table 1) were each able to antagonize both CXCR1 and CXCR2, blocking ligand (interleukin-8)-induced Ca²⁺ mobilization and migration of primary human neutrophils. Further, with identical iL3 sequence and fairly well-conserved iL1 sequence in mouse CXCR2, the human iL1- and iL3-derived pepducins could also prevent neutrophil chemotaxis in a model of peritonitis in vivo. In particular, both iL1- and iL3-derived CXCR1/CXCR2 pepducins were able to almost completely prevent sepsis-induced mortality when administered either immediately following cecal ligation puncture surgery, or even with a delay of 8 hours⁵⁰. x1/2pal-i1 and x1/2pal-i3 have since been shown to be effective at reducing neutrophil-associated effects and improving outcomes in experimental models of intestinal adenoma and alcoholic steatohepatitis, respectively^{51, 52}.

Pepducins derived from the iLs of CXCR4 also have the translational benefit that the iL1 and iL3 sequences of CXCR4 are identical in mice and humans. Initially, similar to the CXCR1/2 pepducins discussed above, x4pal-i1 (PZ-218) and x4pal-i3 (PZ-210) (Table 1) were shown to antagonize the effects of the endogenous CXCR4 ligand CXCL12/stromal cell-derived factor 1α (SDF-1α) on Ca²⁺ mobilization and chemotaxis in human neutrophils⁵⁰, as well as a variety of clonal and primary human lymphocytic leukemia and lymphoma cells⁵³. However, in contrast to the CXCR1/2 inhibitory pepducins, intraperitoneal administration of x4pal-i1 (PZ-218) enhanced peritoneal transmigration of neutrophils and was unable to prevent sepsis-induced mortality⁵⁰. Highlighting the dependence of pepducin-mediated effects on their specific amino acid composition, addition of the next C-terminal amino acid of iL1 to xpal4-i1 (PZ-218) produced a pepducin (PZ-217) lacking antagonist activity⁵³, while addition of the next four C-terminal amino acids in the iL1 sequence resulted in a pepducin (ATI-2341) (Table 1) with agonist activity⁵⁴.

As an agonist pepducin, ATI-2341 was demonstrated to increase Ca²⁺ mobilization and ERK1/2 phosphorylation in a Gi protein-sensitive manner, wherein absence of CXCR4 or expression of a mutated CXCR4 deficient in G protein binding prevented ATI-2341-mediated signaling effects^{19, 54}. Further, the pepducin-mediated responses appeared specific to CXCR4 as ATI-2341 was unable to induce Gi protein association with other Gi protein-coupled receptors, α2AR and CCR2. While another group later reported that ATI-2341 can activate neutrophils in a formyl peptide receptor 2 (FPR2)-sensitive manner, it was unclear whether this involved a direct effect on FPR2⁵⁵. Subsequent structural analysis revealed the direct interaction of a photo-activatable analogue of ATI-2341 with CXCR4⁵⁶ and computer modeling highly predicted ATI-2341 binds to the intracellular domains of CXCR4⁵⁷, congruent with the concept that pepducins bind the intracellular domains of their cognate receptors to exert their actions.

In vivo, ATI-2341 was demonstrated to promote the mobilization of neutrophils in both mice and monkeys to a similar extent as the FDA-approved CXCR4 antagonist AMD3100, consistent with the observation that CXCR4 antagonists and agonists can each act to mobilize neutrophils from the bone marrow via direct inhibition of CXCR4 or disruption of endogenous CXCR4 gradients, respectively⁵⁸. Unlike neutrophils, however, ATI-2341 was not able to mobilize lymphocytes in vivo, which was later suggested to reflect the possibility that AT1–2341 exhibits biased, or selective, agonist properties that may lead to distinct chemotactic responses in a cell type-specific manner depending on the intracellular signaling machinery present. Indeed, AT1–2341 was subsequently demonstrated to exert partial agonism for Gi proteins as compared to SDF-1α, was unable to induce CXCR4 association with G13 protein and displayed reduced G protein-coupled receptor kinase (GRK)-mediated β-arrestin recruitment responses compared to SDF-1¹⁹. Notably, despite mediating relatively less β-arrestin recruitment to CXCR4 than SDF-1α, AT1–2341 was still capable of promoting receptor internalization⁵⁴, which may be attributed to enhanced PKC-mediated phosphorylation of distal C-terminal serine residues as compared to SDF-1α⁵⁹. Whether these signaling properties relay specific effects that may be advantageous over canonical orthosteric CXCR4 ligands remains to be tested.

The majority of studies to date have focused on application of CXCR-derived pepducins toward autologous bone marrow transplantation, sepsis and cancer. However, since neutrophils and other CXCR4-expressing immune cells play essential roles in a number of cardiovascular disease states, including myocardial infarction and atherosclerosis^60–64, agonist or antagonist CXCR-derived pepducins could prove to be effective, and perhaps more selective, modulators of inflammatory responses and disease progression.

Formyl Peptide Receptors

Similar to chemokine receptors, formyl peptide receptors (FPRs) are Gi protein-coupled receptors that regulate immune cell responsiveness to injury, and selective activation of FPR2 in particular has been shown to promote resolution of myocardial infarction in mice⁶⁵. Pepducins designed from the iLs of FPR1–3 have been tested over the last decade by the Bae and Forsman groups for their impact on inflammatory and chemotactic responses primarily in neutrophils, as well as monocytes^{20, 66–68}. Initially, F2pal-16 (Table 1), the 16 amino acid palmitoylated sequence of FPR2 iL3, was shown to increase Ca²⁺ mobilization, superoxide production, inflammatory cytokine expression and migration of human monocytes in a Gi protein-dependent manner⁶⁸, with subsequent confirmation of Ca²⁺ mobilization and superoxide production in human neutrophils²⁰. Notably, the extracellular-binding FPR2 inhibitor WRWWWW (WRW4) was unable to block the effects of F2pal-16, whereas FPR2 pepducin altered the binding of the orthosteric agonist WKYMVM²⁰, consistent with an intracellular allosteric mode of action of pepducins.

Akin to results observed with PAR pepducins (described above), C-terminal truncation of the iL3 sequence of F2pal-16 resulted in shorter pepducins (F2pal-8, F2pal-7) lacking agonist effects, while N-terminal truncation of F2pal-16 resulted in a pepducin that antagonized the actions of conventional FPR2 agonists (F2pal-12, Table 1)^{20, 67, 68}. Notably, an intermediary C-terminal truncation, F2pal-10, was shown to be a FPR2-selective partial agonist for superoxide production in the absence of β-arrestin2 recruitment or neutrophil migration, despite a reduction in FPR2 surface expression equivalent to the full agonist WKYMVM over time^{20, 69}. These observations suggest that while β-arrestin2 is dispensable for FPR2 internalization, FPR2 stimulation may regulate neutrophil chemotaxis in a β-arrestin2-dependent manner, and that F2pal-10 may exhibit biased signaling properties that could provide cell type-specific effects⁶⁹, though studies directly defining signal bias of FPR pepducins is lacking.

Studies focused on F2pal-10 also highlight the potential importance of the FPR2 iL3 mid-sequence lysine and serine residues for relaying the pepducin activity since truncation of the C-terminal serine residue of F2pal-10 resulted in a pepducin (F2pal-9) with reduced agonist activity^{20, 70}. However, the lysine residue in position 5 of F2pal-10 was also shown to be essential in mediating its activity, as its substitution to the conserved iL3 glutamine of the closely related FPR1 resulted in an inactive pepducin²⁰, and a pepducin encoding the full length iL3 of FPR1 (F1pal-16) was shown to antagonize FPR2 activation⁷¹. Notably, an iL3-mutated FPR2 with substitutions of 2 conserved FPR1 amino acids, including the aforementioned glutamine, was still able to be activated by F2pal-10²⁰, demonstrating that although the iL3 sequence of F2pal pepducins is essential for activation of FPR2, the iL3 within FPR2 itself does not alone dictate pepducin sensitivity.

The work of the Forsman group on FPR2 pepducin development and characterization has also highlighted key concepts for consideration in such studies, including both receptor family- and species-selectivity. For instance, pepducins derived from iL2 and iL3 of the P2Y12 receptor were shown to promote FPR2-mediated Ca²⁺ mobilization and superoxide generation in human neutrophils, with disruption of FPR2 agonist binding, suggesting a direct effect of the P2Y12-derived pepducins on FPR2⁷². In terms of species selectivity, the group demonstrated that pepducins designed from the full iL3 sequence of either human or murine FPR1 or FPR2 could each act at both human and murine FPR2 but with different outcomes⁷³. Whereas both human and murine FPR1 iL3 pepducins (F1pal-16 and mF1pal-16, respectively) antagonized human FPR2 activity, they activated mouse FPR2. Conversely, human FPR2-derived F2pal-16 activated, while murine FPR2 pepducin mF2pal-16 inhibited, both human and mouse FPR2. In all, these data show the necessity of properly attributing the actions of a pepducin on its intended GPCR in a cell type- and species-specific manner, where differences in GPCR expression profiles could lead to off-target effects, especially when the amino acids sequences used to design the pepducin may not be conserved between receptor isoforms or species.

β2-Adrenergic Receptor

Directly stemming from the work on CXCR4 pepducins (described above) that had suggested ATI-2341 may act as a biased agonist by favoring activation of Gi protein signaling versus G13 protein signaling or GRK/β-arrestin recruitment¹⁹, the Benovic group sought to determine whether a similar paradigm may exist for pepducins designed from the iLs of β2AR. After screening over 50 pepducins designed from iL1, iL2 and iL3, several iL1- and iL3-designed pepducins were revealed to be partial agonists for β-arrestin recruitment or Gs protein-dependent cAMP generation, respectively²¹. In particular, ICL3–8 and ICL3–9 (Table 1) were highlighted as ICL3–8 induced Gs protein activation directly, in the absence of receptor, while ICL3–9 promoted Gs protein activation and cAMP accumulation in a β2AR-dependent manner. ICL3–8 contains more of the N-terminal iL3 amino acids than ICL3–9, which could mediate the direct actions on Gs protein versus requiring the receptor itself. Notably, ICL3–9 also activated β1AR, despite iL3 sequence differences between β1AR and β2AR (but not the unrelated prostaglandin E2 receptor), and its effects were sensitive to progressive N- or C-terminal truncations as well as alanine substitutions throughout its sequence, making it difficult to pinpoint a single essential sequence that relays it activity²¹. Remarkably, ICL3–9 promoted cAMP accumulation in the complete absence of GRK-mediated β2AR phosphorylation, β-arrestin recruitment and β2AR internalization. Indeed, a steady state of cAMP generation was produced by ICL3–9 in human airway smooth muscle cells that was sustained at a higher level than the agonist salbutamol, which highlighted the potential of Gs protein-biased pepducins as new therapeutic tools for asthma research and development, where β2AR desensitization is an issue with non-biased agonists²¹.

Converse to iL3-based pepducins, iL1-based β2AR pepducins were shown to promote the recruitment of β-arrestins to the receptor in the absence of Gs protein activation²¹. A subsequent study by the Benovic group in collaboration with our group investigated the properties of ICL1–9 (Table 1), which was the most potent β-arrestin-biased β2AR-based iL1 pepducin identified²². ICL1–9 was shown to be a partial agonist for both phosphorylation of GRK-sensitive phospho-serine sites in the C-terminus of β2AR and receptor internalization. Interestingly, ICL1–9 induced the same peak level of ERK1/2 phosphorylation as the full agonist isoproterenol, but with delayed kinetics, presumably due to independence from Gs protein activation, and did so in a β-arrestin-dependent manner. Since β-arrestin-biased GPCR signaling was previously shown to enhance cardiomyocyte contractility⁷⁴, we tested whether ICL1–9 impacted the contractility of primary isolated left ventricular murine cardiomyocytes. Although ICL1–9 did not alter canonical contractility-promoting Ca²⁺ mobilization, congruent with a lack of Gs protein engagement, it did enhance cardiomyocyte contractility in a β2AR- and β-arrestin-dependent manner, with β-arrestin1 playing more of a role due to its higher expression than β-arrestin2 in cardiomyocytes²². While the mechanism by which ICL1–9 engaged β-arrestin-dependent contractile signaling in the absence of Gs protein engagement was not entirely clear, others demonstrated that the orthosteric β-blocker carvedilol can promote β1AR-Gi protein coupling to enable β-arrestin-mediated signaling⁷⁵. Indeed, we subsequently determined that the contractile effect of ICL1–9 is sensitive to inhibition of Gi protein activity, suggesting that ICL1–9 allosterically promotes β2AR coupling with Gi protein to engage β-arrestin-dependent contractile signaling⁷⁶. Further, the effect of ICL1–9 was sensitive to inhibition of Rho-associated protein kinase (ROCK), protein kinase D (PKD) and myosin light chain kinase (MLCK) activity, suggesting that, in the absence of Gs protein-dependent Ca²⁺ mobilization, ICL1–9 may promote the post-translational modification of myofilament proteins to enhance contractility.

β-arrestin-dependent signaling has also been associated with the promotion of cell survival³, thus we also investigated the potential utility of using ICL1–9 to increase cardiomyocyte survival in a murine model of ischemia-reperfusion (I/R) injury⁷⁷. Indeed, direct intramyocardial injection of ICL1–9 at the time of I/R injury was able to decrease acute cardiomyocyte death and infarct size in a β2AR- and β-arrestin-dependent manner, while better preserving cardiac function and decreasing fibrotic remodeling over time. Notably, at a cellular level the pro-survival signaling effects of ICL1–9 were shown not to involve canonical kinases such as ERK1/2, but rather stable activation of RhoA/ROCK-dependent signaling, wherein RhoA activation could be detected up to 24 hours following cardiomyocyte treatment in vitro or intramyocardial injection in vivo without significant β2AR downregulation⁷⁷. Such long-lasting effects of ICL1–9 on RhoA activity may therefore have the potential to relay chronic benefits in the failing heart in terms of promoting both cardiomyocyte survival and contractility.

Beyond cardiomyocytes, β2AR are almost ubiquitously expressed throughout the body, including in immune cells. We have shown that immune cell-specific β2AR plays a key role in regulating their responsiveness to acute cardiac injury, wherein either deletion of β2AR in all cells of hematopoietic origin or prior β-blockade led to reduced recruitment of neutrophils and monocytes to the heart following myocardial infarction, preventing stable scar formation^{23, 78}. One of the mechanisms responsible for this effect was shown to be a reduction in chemokine receptor 2 (CCR2) expression specifically in the absence of GRK/β-arrestin-dependent β2AR signaling, which reduced the ability of the cells to migrate toward CCL2/MCP-1⁷⁹. Notably, exogenous expression of either wild-type or a G protein coupling-deficient β2AR in immune cells was able to restore CCR2 levels and recruitment to MI-induced injury, whereas a GRK phosphorylation site-deficient β2AR was unable to rescue the effects, suggesting that GRK/β-arrestin-dependent β2AR signaling specifically promotes CCR2 expression in leukocytes⁷⁹. Thus, pepducin-mediated regulation of GRK/β-arrestin-biased β2AR activity in immune cells acutely following injury, when their levels in the peripheral circulation are elevated and easily accessible to pepducins, may offer a novel approach to fine-tune the injury response. Whether peripheral infusion of β2AR-selective pepducins would be able to acutely alter immune cell responsiveness to myocardial injury in vivo, and whether such a strategy would also impact cardiomyocyte survival and function locally within the heart, remains to be tested.

Summary and Perspectives

Since their introduction as pharmacologic tools, pepducins have been immensely useful in defining facets of GPCR signaling, including pathway-selective biased responses and structural requirements for these outcomes. They have also become to be viewed as potentially valuable and novel therapeutics since they have been shown to be bioavailable, well-tolerated and safe in human subjects. The most obvious application for pepducins would be for the regulation of GPCRs expressed on cells within the peripheral circulation, such as neutrophils and monocytes, or within highly perfused organs such as the liver⁴⁴, where they would have ready access to their targets and would not have to cross multiple membrane barriers. Since pepducins administered in vivo have been shown to provide stable, long-lasting effects that are reversible, in this respect they may provide an advantage over other therapeutics requiring more frequent dosing or that act irreversibly. However, as peptides, they would require parenteral administration, which could limit their applicability to acute clinical situations. A major consideration that must be taken into account when developing pepducins as therapeutics, similar to other drug classes, is receptor selectivity. Numerous studies, including those described above and others⁸⁰, highlight the importance of testing pepducins designed from the iLs of GPCRs for receptor selectivity across both related and unrelated receptor families, as well as defining whether the pepducins indeed act in a receptor-dependent manner. As allosteric modulators that can be designed to act as partial agonists or antagonists with or without proximal pathway bias, pepducins have enormous potential to alter GPCR signaling responses to improve cardiovascular outcomes.