Muscarinic receptors are important mediators of smooth muscle contraction in various tissues including airways, gut, and urinary bladder; this largely involves the M3 subtype but M2 receptors can also affect smooth muscle tone (Hegde and Eglen 1999). M3 receptors typically couple to G-proteins of the Gq/11 type leading to activation of a phospholipase C (PLC) (Caulfield and Birdsall 1998). Surprisingly, M3-mediated smooth muscle contraction is not explained by PLC activation, for instance in the bladder (Frazier et al. 2008), but alternative molecular mechanisms to elicit smooth muscle contraction have not been well-defined.

The potency and/or efficacy of β-adrenoceptor agonists to relax smooth muscle is lower when tested against muscarinic agonists such as carbachol than against agonists at other receptors, against passive tension or against receptor-independent contraction elicited by KCl (Dale et al. 2014). Such observations have been made in airways (Russel 1984; Raffestin et al. 1985; Ostrom and Ehlert 1999; Sarria et al. 2002; Naline et al. 2007), ileum (Ostrom and Ehlert 1999), and bladder (Longhurst and Levendusky 1999; Witte et al. 2011; Kanie et al. 2012) of multiple species including rat (Longhurst and Levendusky 1999; Michel and Sand 2009; Witte et al. 2011; Cernecka et al. 2014), guinea pig (Ostrom and Ehlert 1999), dog (Russel 1984), and human (Raffestin et al. 1985; Sarria et al. 2002; Naline et al. 2007; Kanie et al. 2012) and with agonists acting at histamine (Russel 1984; Raffestin et al. 1985; Ostrom and Ehlert 1999; Naline et al. 2007), 5-hydroxytryptamine (5-HT) (Michel and Sand 2009; Cernecka et al. 2014), bradykinin (Michel and Sand 2009; Cernecka et al. 2014), prostanoid receptors (Sarria et al. 2002), passive tone (Naline et al. 2007; Michel and Sand 2009; Cernecka et al. 2014), or KCl (Longhurst and Levendusky 1999; Michel and Sand 2009; Kanie et al. 2012). While most of the above studies have used general β-adrenoceptor agonists such as isoprenaline, similar findings have also been obtained with agonists selective for β2-adrenoceptors such as formoterol, indacaterol, salbutamol, and salmeterol in human bronchi (Naline et al. 2007) or fenoterol in rat bladder (Erdogan et al. 2021) or selective for β3-adrenoceptors such as TRK 380 in human bladder (Kanie et al. 2012) or KUC 7322 in rat bladder (Cernecka et al. 2014). However, it has remained elusive why smooth muscle contraction elicited by muscarinic receptor agonists is more resistant to relaxation by β-adrenoceptor agonists than that elicited by other stimuli. It appears logical that such selective resistance should be related to a signaling pathway activated preferentially by muscarinic receptors as compared to other receptors, but we are not aware of such signaling pathways.

A paper in this issue of the journal demonstrates that NAV2729, an inhibitor of ADP ribosylation factor 6 (ARF6), inhibits coronary vascular smooth muscle contraction elicited by the muscarinic agonists carbachol and methacholine but not that elicited by agonists at other receptors including α1-adrenoceptors, 5-HT, endothelin-1, or prostanoid TP receptors or those elicited by KCl (Huang et al. 2022). The same group had previously shown the existence of ARF6 expression human prostate smooth muscle tissue (Hennenberg et al. 2013) and that NAV2729 inhibited contraction in isolated human prostate strips by α1-adrenoceptor agonists, whereas that elicited by a prostanoid TP receptor agonist, endothelin-1 or by KCl, was not inhibited (Yu et al. 2019). Furthermore, NAV2729 inhibited contraction in human prostate smooth muscle cells, and this was also observed in ARF6 knockout cells (Wang et al. 2021). Inhibition of ARF6 by NAV2729 inhibited contraction elicited by α1-adrenoceptor agonists in human prostate (Yu et al. 2019) but not porcine blood vessels (Huang et al. 2022). Interestingly, the inhibition of contraction in human prostate was accompanied by an inhibition of ARF6 in pull-down assays, while ARF6 was not activated by noradrenaline, phenylephrine, or methoxamine (Yu et al. 2019). The selective inhibition of muscarinic receptor responses in the blood vessels (Huang et al. 2022) raises the possibility that coupling to ARF6 may be a mechanism that distinguishes inhibition by β-adrenoceptor agonists of responses to a muscarinic agonist as compared to those elicited by other means.

ARF6 is a small GTP-binding protein that contributes to several cellular processes including G-protein-coupled receptor (GPCR) trafficking, actin organization, and contractile response through diverse downstream component interaction (Fig. 1, Table 1). ARF6 function is modulated by two distinct components. Guanine nucleotide exchange factors (GEFs) mediate the activation of ARF6, whereas GTPase-activating proteins (GAPs) mediate inhibition. NAV2729 interferes in the formation of ARF6-GEF complex to inhibit the ARF6 activation (Yamauchi et al. 2017).

Fig. 1
figure 1

Schematic drawing of ARF6-mediated cellular effects in GPCRs agonist stimulation. Dashed red arrow, negative regulatory effect; black arrow, positive regulatory effect; GAP, GTPase-activating proteins; GDP, guanosine diphosphate; GEF, guanine nucleotide exchange factors; GTP, guanosine triphosphate; ↑, increase; ↓, decrease; ↔ , no effect

Full size image
Table 1 ARF6-mediated responses through G-protein isoform/G-protein-coupled receptors
Full size table

ARF6 was shown to be a prerequisite component further promoting either clathrin- or caveolin-mediated pathway in agonist-induced endocytosis of several GPCRs such as β-adrenoceptor in adipocytes (Liu et al. 2010) and β2-adrenoceptor in HEK293 cells (Claing et al. 2001; Lawrence et al. 2005; Macia et al. 2012); M2 muscarinic receptors in HeLa cells (Delaney et al. 2002), in JEG-3 human choriocarcinoma cells (Reiner and Nathanson 2008), and in HEK293 cells (Houndolo et al. 2005); the luteinizing hormone chorionic gonadotropin receptor in HEK 293 cells (Kanamarlapudi et al. 2012); angiotensin type 1 receptor (Houndolo et al. 2005; Cotton et al. 2007); μ-opioid receptor (Rankovic et al. 2009); the vasopressin V2 receptor; and endothelin type B receptor (Houndolo et al. 2005) in HEK293 cells. Upon agonist stimulation, GPCR endocytosis was found mostly activated via ARF6-dependent pathway with some exceptions which show inhibitor regulatory effect of ARF6 in M2 muscarinic receptor (Delaney et al. 2002; Reiner and Nathanson 2008) and in β2-adrenoceptor (Macia et al. 2012) internalization. Furthermore, VIP receptor internalization was not affected by ARF6 depletion (Houndolo et al. 2005). ARF6 involvement in trafficking did not exist for some receptors such as M4 muscarinic receptor (Reiner and Nathanson 2008) and vasopressin V2 receptor (Madziva and Birnbaumer 2006). ARF6 requirement in endocytosis was mostly demonstrated in agonist-induced settings, which may not reflect the ARF6 function for basal condition for the same receptor (Cotton et al. 2007). Moreover, ARF6 involvement of muscarinic receptor internalization has mostly been studied with the M2 subtype because of the well-defined, clathrin-dependent pathway-mediated internalization of M1, M3, and M4 receptors (Reiner and Nathanson 2008).

Phospholipase D (PLD) is known to be involved in smooth muscle contraction through PKC activation but its contribution in urinary bladder contraction was proposed to be minor (Frazier et al. 2008). ARF6-mediated PLD activation was reported by several researchers in in vivo animal (Le Stunff et al. 2000) and in vitro cultured cell line studies (Mitchell et al. 2003; Johnson et al. 2006; Rankovic et al. 2009; Chakraborti et al. 2017; Charles et al. 2018). In human pulmonary artery smooth muscle cells (HPASMCs), stimulation of prostanoid TP receptor stimulates cytohesin-1 coupling to ARF6 which further leads to PLD2 isoform and subsequent NADPH oxidase activation (Chakraborti et al. 2017). In HPASMCs, involvement of ARF6 in endothelin-1-induced PLD and NADPH oxidase activation was shown by same study group (Chakraborti et al. 2018). ARF6 was shown to be involved in PLD activation in N376D mutant 5-HT2A-stimulated PLD activation but not in mediated via WT 5-HT2A stimulation (Johnson et al. 2006). This study also showed the sensitivity of other class A GPCRs which contain DPxxY motif such as purinergic P2u, thrombin PAR, and gonadotropin-releasing hormone receptor to ARF6 for further PLD activation in COS7 cells (Johnson et al. 2006). In the same cell line, agonist-stimulated M3 activation induced both PLD1/2 activation through ARF6-mediated pathway, whereas PLD2 activation was found linked to PKC and ARF6 in purinergic P2U receptor and to only PKC in N376D mutant 5-HT2a receptor (Mitchell et al. 2003). ARF6 is involved in the promotion of prostate smooth muscle contraction. Inhibition of ARF6 activation by cytohesin (a GEF) inhibitor resulted in reduced noradrenaline, phenylephrine-, thromboxane A2-, and endothelin-1- and endothelin-3-induced contraction (Herlemann et al. 2018). In vascular smooth muscle cell, both ARF1 and ARF6 are involved in actin polymerization which subsequently migrate and proliferate but only ARF1 affected contractile responses (Charles et al. 2018). However, in a latter study, ARF6 was found to promote contraction and proliferation in human prostate stromal cells (WPMY-1) (Wang et al. 2021).

ARF6 activation interrupts the recycling of β2-adrenoceptors and lead desensitization of receptor in HEK293 cells (Macia et al. 2012); similarly, it is postulated that ARF6 may involve in α1-adrenoceptor desensitization in human prostate tissue (Hennenberg et al. 2013). In contrast, ARF6-mediated endocytosis was suggested beneficial in development of opioid tolerance through preventing receptor desensitization HEK293 cells (Rankovic et al. 2009).

Other effects of ARF6 include the calcium-sensing receptor-mediated plasma membrane ruffling which is required for chemotaxis in HEK cells (Bouschet et al. 2007). In endothelial cells, endothelin ETB receptor stimulation by endothelin 1 results in ARF6 activation which facilitates cell migration via actin reorganization. Moreover, endothelin 1 stimulation did not promote capillary tube formation in ARF6 depleted cells which indicates ARF6 involvement in angiogenesis (Daher et al. 2008). ARF6 also has regulatory effects on metabolic pathway. Depletion of ARF6 resulted in inhibition of isoproterenol-induced lipolysis in 3T3-L1 adipocytes. ARF6 mRNA and protein level was found higher in WAT tissue of ob/ob mice compared to obesity resistance mice (Liu et al. 2010). In 3T3-L1 adipocytes, endothelin ETA receptor-mediated lipolysis was found to be dependent downstream ARF6-ERK1/2 signaling (Davies et al. 2014). Additionally, endothelin 1 stimulated GLUT4 translocation through heterotrimetric Gq protein signaling pathway activated by ARF6 in 3T3-L1 adipocytes (Bose et al. 2001; Lawrence and Birnbaum 2001) and specifically Gα11 isoform of Gαq family.

Taken together, the above data demonstrate that ARF6 is involved in cellular processes and is activated by cellular effects stimulated by various GPCR, including smooth muscle contraction. Within a given tissue, the involvement of ARF6 in pathways leading to smooth muscle contraction, e.g., in arteries or prostate, appears to be preferential for some GPCR (muscarinic receptors and α1-adrenoceptors, respectively) over others apparently coupling to the same G-proteins. Some of these data are in line with the hypothesis that coupling to ARF6 may explain the relative resistance of smooth muscle contraction elicited by muscarinic receptors as compared to other GPCR or receptor-independent contraction such as receptor desensitization to relaxation by β-adrenoceptor agonists. However, not all data support this hypothesis. Thus, the role of ARF6 in this phenomenon cannot be considered proven, but it remains as a reasonable molecular candidate to explain the resistance of muscarinic receptor-mediated smooth muscle contraction against relaxation. Further studies at the molecular level are required to further explore this, specifically studies in which the role of agonists at various GPCR is compared quantitatively.