Altered coronary vascular BK channel expression is common in DM (Burnham et al., 2006; McGahon et al., 2007). However, diverse levels of vascular BK channel expression in DM have been observed. In most case, the protein expressions of BK channels are downregulated in coronary arteries (Burnham et al., 2006; Dong et al., 2008; Lu et al., 2008, 2017a; Zhang et al., 2010a; Rueda et al., 2013; Nystoriak et al., 2014; Li et al., 2017), but it was reportedly increased, despite impaired BK channel function in the coronary arteries of Ossabaw miniature swine with metabolic syndrome (Borbouse et al., 2009). Recently, human BK channel expression was examined in coronary arterioles obtained from atrial biopsies of patients who underwent coronary artery bypass grafting surgery. Protein downregulation was found in both BK-α and BK-β1 in patients with T2DM, compared to age-matched non-diabetic subjects (Lu et al., 2019). However, the mRNA levels of BK-β1 were (McGahon et al., 2007) not reduced in the coronary arteries of STZ-induced T1DM rats (Zhang et al., 2010a), db/db T2DM mice (Li et al., 2017) and HFD-induced diabetic mice (Lu et al., 2017a). The varied reports of BK channel expression suggest that a complex assortment of mechanisms exist in the regulation of vascular BK channel expression and function in DM. Reduced BK channel expression leads to impaired Ca²⁺ sparks/STOCs coupling, albeit the Ca²⁺ spark amplitudes and intracellular Ca²⁺ concentrations are known to be elevated in diabetic vascular SMCs.

Ca²⁺-activated K⁺ channel currents (I) are determined by the number of activated channels (N), open probability (Po), and channel unitary conductance (i), where I=N^*Po^*i. BK channel current density is reduced in the coronary arteries of T1DM and T2DM animal models and in humans with DM (Lu et al., 2005, 2008, 2010, 2012, 2016, 2017a, 2019; Pietryga et al., 2005; Burnham et al., 2006; McGahon et al., 2007; Dong et al., 2008; Zhang et al., 2010a; Nystoriak et al., 2014; Yi et al., 2014; Li et al., 2017; Nieves-Cintron et al., 2017; Tang et al., 2017; Zhang et al., 2020). BK channels are activated by intracellular free Ca²⁺ concentration and by membrane depolarization (Cox et al., 1997; Lu et al., 2008), and these are impaired in DM (Lu et al., 2008, 2019). BK channel sensitivity to voltage- and Ca²⁺-mediated activation can be measured by using inside-out patch clamp studies in which the excised cell membrane can be clamped to various voltages and the cytoplasmic surface of the cell membrane directly exposed to bath solutions containing various free Ca²⁺ concentrations. In freshly isolated coronary arterial SMCs of ZDF rats at 8months after the development of hyperglycemia, BK channels had a rightward-shifted Ca²⁺ concentration-dependent curve, with increased EC₅₀ for Ca²⁺ activation and decreased Ca²⁺ cooperativity, compared to those of Lean control rats (Lu et al., 2008). Moreover, BK channel activation by membrane depolarization was also abnormal in coronary arterial SMCs of ZDF rats. The channel open probability–voltage (Po-V) relationships were rightward and downward shifted, with the voltage at 50% maximal Po increased by 40mV. These results indicate that a higher cytoplasmic Ca²⁺ concentration and a more depolarized membrane potential are required to activate BK channels in DM. Changes in the intrinsic free energy of Ca²⁺-binding (ΔΔCa²⁺) that contributes to BK channel activation can be estimated based on the shift of Po-V relationship from 0 to 1 μM free Ca²⁺ in Lean and ZDF rats using the equation: ΔΔCa²⁺=−Δ(zeV_0.5), where z is the number of equivalence charge movement, e is the elementary charge, and V_0.5 is the voltage at half maximal activation (Shi et al., 2002). There was a 62.3% decrease in the ΔΔCa²⁺ in ZDF rats, suggesting a less favorable condition for Ca²⁺ binding to vascular BK channel Ca²⁺ sensors in ZDF rats (Lu et al., 2008). Similar results were also observed in BK channels in freshly isolated coronary microvascular SMCs from the atrial appendages of patients with T2DM. Ca²⁺- and voltage sensitivity were significantly impaired in diabetic patients, with the maximal BK channel activity to free Ca²⁺ and voltage activation reduced by 70 and 50%, respectively (Figure 2; Lu et al., 2019). Such dysregulation contributed to a 27.4% attenuation in shear stress-mediated coronary arteriolar vasorelaxation in diabetic patients compared with non-diabetic controls (Lu et al., 2019). In addition, single BK channel current amplitudes were unaltered in DM, indicating that the conductance property of vascular BK channels is normal in DM.

Vascular large conductance Ca²⁺-activated K⁺ channel gating kinetics contain multiple components of open and closed states and dwell-times (McManus and Magleby, 1988, 1991). In coronary arterial SMCs, the open and closed dwell-time histograms of single BK channels were best fitted with three open-time constants: fast (τo₁), intermediate (τo₂), and slow (τo₃), along with four closed-time constants: fast (τc₁), intermediate (τc₂), slow (τc₃), and very slow (τc₄). DM affects both channel open dwell-times and channel closed dwell-times. The BK channel mean closed-time constant and the individual closed-time constants were significantly prolonged. At the same time, the channel mean open-time constant and individual open-time constants were significantly reduced in DM. These findings were seen in both ZDF rats and in diabetic patients (Lu et al., 2008, 2019). These changes in BK channel gating kinetics suggest that channel openings are abbreviated, and closures prolonged in DM, with reduced channel Po and maximal activation. Hence, diabetes not only affects BK channel expression, but also alters the intrinsic biophysical properties of the channel.

Genome-wide association studies (GWASs) are a powerful tool to find genetic variations associated with diseases. Results from a few studies have shown a strong association between KCNMA1 splicing variants and the incidence of obesity or DM. The results from case–control cohorts involving 4,838 obese and 5,827 control subjects suggested that the KCNMA1 rs2116830*G variant was associated with obesity with a p value of 2.82×10⁻¹⁰ (Jiao et al., 2011). A recent study reported that a de novo missense variant in KCNMA1 (c.1123G>A) was identified in an adult male patient with a plethora of developmental phenotypes including neonatal DM. This loss-of-function polymorphism (p. G375A) of BK channel is located in the S6 transmembrane domain of BK channel (Liang et al., 2019). In addition, it is well known that BK-α and BK-β1 undergo extensive alternative pre-mRNA splicing and that these splice variants have significant changes in BK channel intrinsic properties and surface expression (Poulsen et al., 2009). However, the pathophysiological roles of BK channel variants in the development of BK channelopathy in DM are largely unexplored and warrant further investigation.

Increased reactive oxygen species (ROS) production is a hallmark of diabetic pathophysiology, and the role of ROS on vascular dysfunction has been extensively reviewed (Inoguchi et al., 2003; Konior et al., 2014). ROS is represented by a group of highly reactive molecules that include superoxide anion (O₂^•–), peroxide ion (O₂²⁻), hydrogen peroxide (H₂O₂), and peroxynitrite (ONOO⁻). In vascular SMCs, multiple enzymatic systems such as the NADPH oxidases (NOXs), xanthine oxidase (XO), nitric oxide synthases (NOS), and the mitochondrial electron transport chain are known to produce O₂^•– and H₂O₂ (Taniyama et al., 2004; Byon et al., 2016). The NOXs, in particular NOX1 and NOX4, are the most important because they are commonly expressed in vascular cells and are the major source of ROS generation in vessels (Clempus and Griendling, 2006; Konior et al., 2014; Burtenshaw et al., 2017). O₂^•– is converted to H₂O₂ by superoxide dismutases (SODs) or reacts with nitric oxide (NO) to form ONOO⁻. H₂O₂ is further reduced to H₂O by catalase (CAT) and glutathione peroxidase (GPx; Taniyama and Griendling, 2003). Oxidative stress due to ROS production outweighing their scavenging is implicated in vascular dysfunction associated with T1DM and T2DM. It is well documented that elevated glucose increases the production of intracellular advanced glycation end-products (AGEs), stimulates the protein kinase C (PKC)-dependent activation of NOX1 and NOX4 (Inoguchi et al., 2000; Lu et al., 2006; Deluyker et al., 2017), and reduces the activity and bioavailability of antioxidant enzymes, such as SODs, GSH, CAT, and GPx, which results in higher ROS levels in both vascular ECs and SMCs in DM (Szaleczky et al., 1999; Lu et al., 2012; Tiwari et al., 2013).

Reactive oxygen species triggers many signaling pathways and promotes redox-mediated protein posttranslational modification. We found that redox modification is involved in BK channel dysfunction through hyperglycemia. High glucose culture of HEK293 cells stably expressing BK-α resulted in altered BK-α activity and channel kinetics that were mimicked by the effects of exogenously applied H₂O₂ in BK-α expressing cells cultured in normal glucose (Lu et al., 2006). A 1-week culture with 22mM glucose markedly downregulated the protein expression of CAT and CuZn-SOD in HEK293 cells, leading to a 3.3-fold increase of H₂O₂ concentration to the 10⁻³M range. Consequently, high glucose culture produced a 50% reduction of BK-α current density, prolonged the channel activation and deactivation time constants (τ_A and τ_D), and upward shifted the τ-V curve, indicating that BK-α activation is suppressed in high glucose conditions (Lu et al., 2006). The effects of high glucose on BK-α voltage-dependent activation were mimicked by acute exposure to 2mM H₂O₂. Furthermore, the cysteine residue at 911 (C911) in BK-α is particularly vulnerable to H₂O₂-mediated regulation (Tang et al., 2001), and a single substitution of C911 by alanine (C911A) eliminated most of the inhibitory effects of BK-α under high glucose conditions and to exogenously applied H₂O₂ (Lu et al., 2006). In addition, acute exposure to ONOO⁻ (5–100μM) significantly suppressed BK channel activity in vascular SMCs (Brzezinska et al., 2000; Liu et al., 2002), but did not alter BK-α voltage-dependent activation (Lu et al., 2006), suggesting that the molecular mechanisms underlying BK channel regulation by H₂O₂ and ONOO⁻ are different. Further studies revealed a 3- to 4-fold increase of 3-nitrotyrosine levels on BK-α protein in freshly isolated aortas from STZ-induced T1DM rats compared to non-diabetic controls, suggesting that ONOO⁻-induced modification of BK-α may be mediated through protein tyrosine nitration rather than protein oxidation (Lu et al., 2010). The precise amino acid residue(s) in BK-α modified by ONOO⁻ has not been identified. Nevertheless, an increase of ROS accumulation is the culprit for the development of BK channel dysfunction in DM.

Angiotensin II (Ang II) is an oligopeptide hormone, exerting its physiological and pathophysiological effects through binding to Ang II type 1 (AT1R) and type 2 (AT2R) receptors and activating their downstream signaling pathways (Dasgupta and Zhang, 2011). In vascular SMCs, where AT1R is predominantly expressed, Ang II causes vasoconstriction and promotes vascular wall remodeling (Ribeiro-Oliveira et al., 2008). In contrast, activation of AT2R produces vasodilatation and impairs vascular remodeling, effects opposite to those of AT1R (Danyel et al., 2013). AT1R is a G-protein-coupled receptor, which is coupled to Gαq, Gβγ, Gαi, and β-arrestin (Kawai et al., 2017; Wang et al., 2018). Binding of Ang II to AT1R in vascular SMCs activates Gαq which in turn activates the phospholipase C (PLC)-dependent inositol-1,4,5-triphosphate (IP₃)/diacylglycerol (DAG)-mediated Ca²⁺ signaling cascades, causing an increase in protein kinase C (PKC) activity (De Gasparo et al., 2000; Touyz and Schiffrin, 2000). Activation of PKCβ stimulates NOXs with ROS overproduction under hyperglycemic conditions (Inoguchi et al., 2000; Evcimen and King, 2007) and is a cause of impaired vascular BK channel function in diabetic vessels (Figure 3; Zhou et al., 2006; Lu et al., 2012; Zhang et al., 2020). In addition to redox-mediated modification of BK-α, it has been shown that PKC-induced serine phosphorylation at 695 (S695) and 1151 (S1151) in the C-terminus of BK-α inhibits BK channel current density by 50%, and S1151 phosphorylation by PKC also abolishes BK-α activation by protein kinase A (PKA) and protein kinase G (PKG; Zhou et al., 2001, 2010). On the other hand, the activity of tyrosine-protein kinase is regulated by Gαi and β-arrestin upon AT1R stimulation, causing BK channel dysfunction (Ma et al., 2000; Alioua et al., 2002; Fessart et al., 2005; Tian et al., 2007). Another study reported that the C-terminus of AT1R physically interacts with the C-terminus of BK-α in heterologous expression system, and such protein–protein interaction between AT1R and BK-α directly inhibits BK-α activity, independent of G-protein mediated processes (Zhang et al., 2014).

However, AT1R expression, Ang II bioavailability, and tissue sensitivity to Ang II are upregulated in diabetic vessels (Arun et al., 2004; Kawai et al., 2017). The pathophysiological importance of Ang II-mediated BK channel regulation in diabetic coronary arteries is supported by the evidence that cardiac infarct size induced by experimental ischemia/reperfusion in STZ-induced T1DM mice was twice as large as non-diabetic mice (Lu et al., 2016). The effects of DM on myocardial ischemia/reperfusion injury can be reproduced by infusion of 2μM Ang II or 0.1μM membrane impermeable BK channel inhibitor, IBTX, but attenuated by the BK channel activator, NS-1619 (Lu et al., 2016). Similar results were observed in Akita T1DM mice with exacerbated cardiovascular complications and cardiac and vascular dysfunction, from an imbalance of Ang II/AT1R signaling in DM (Patel et al., 2012). Most importantly, the pathological roles of Ang II signaling are supported by clinical outcomes showing that treatment with AT1R blockers and ACE inhibitors reduced cardiovascular complications and cardiovascular death in patients with DM by 25–30% (Niklason et al., 2004; Abuissa et al., 2005; Cheng et al., 2014; Lv et al., 2018).

Caveolae, which are nonclathrin-coated, flask-shaped invaginations of plasma membrane lipid raft subdomains, are characterized by their signature structural protein caveolin, with caveolin-1 (Cav-1) predominantly expressed in the vasculature (Gratton et al., 2004; Krajewska and Maslowska, 2004). Caveolae have emerged as a central platform for signal transduction in many tissues through the interaction between the Cav scaffolding domain and protein partners that contain a Cav-binding motif (ΦxΦxxxxΦ or ΦxxxxΦxxΦ, where Φ is an aromatic amino acid, and x is any amino acid; Okamoto et al., 1998). Many signaling molecules that are associated with BK channel regulation, such as the β-adrenergic receptors (Bucci et al., 2004), AT1R (Ushio-Fukai and Alexander, 2006; Basset et al., 2009), NOX1 (Hilenski et al., 2004; Wolin, 2004), cellular tyrosin protein kinase Src (c-Src; Zundel et al., 2000; Lee et al., 2001), guanylyl cyclase (Linder et al., 2005; Vellecco et al., 2016), PKA (Heijnen et al., 2004; Linder et al., 2005), protein kinase B (PKB or AKT; Sedding et al., 2005), PKC (Zeydanli et al., 2011; Ringvold and Khalil, 2017), PKG (Linder et al., 2005), NOS (Garcia-Cardena et al., 1996; Vellecco et al., 2016), and prostacyclin (PGI₂) synthase (PGIS; Spisni et al., 2001), are found in the low buoyant density, caveolae-rich membrane fractions of vascular ECs and SMCs. The significance of Cav-1 on vascular physiology is demonstrated by findings in Cav-1 knockout (KO) mice that show constitutively activated eNOS with elevated NO production as well as a failure to maintain a constant vasocontractile tone, resulting in the development of cardiovascular pathologies (Drab et al., 2001; Razani et al., 2001). Overgeneration of NO facilitates the production of ONOO⁻ and contributes to vascular dysfunction with excessive H₂O₂ accumulation (Pacher et al., 2007).

The consensus sequence of the Cav-binding motif is present in BK-α, but not in BK-β1. Indeed, only BK-α but not BK-β1 is detected in the caveolae-rich fractions of SMCs (Lu et al., 2016). Moreover, BK-α is colocalized in the caveolae with other ion channels (Wang et al., 2005; Riddle et al., 2011; Howitt et al., 2012; Lu et al., 2016), especially those associated with Ca²⁺ spark/sparklet generation, such as L-type Ca²⁺ channels (Suzuki et al., 2013; Saeki et al., 2019), T-type Ca²⁺ channels (Hashad et al., 2018), TRPV4 (Goedicke-Fritz et al., 2015; Lu et al., 2017b), TRPC1, TRPC3, and TRPC6 (Bergdahl et al., 2003; Adebiyi et al., 2011; Grayson et al., 2017) in vascular ECs and SMCs. The close proximity of BK channels with Ca²⁺ entry molecules leads to Ca²⁺ spark-coupled STOCs. However, it has been reported that Cav-1 interacts with BK channels and inhibits BK channel activities in coronary ECs (Wang et al., 2005; Riddle et al., 2011). Cholesterol depletion by methyl-β-cyclodextrin and silencing of Cav-1 by small interference RNA enhance BK currents, while exposure to the scaffolding domain peptide of Cav-1 (AP-CAV) inhibits BK currents (Wang et al., 2005; Riddle et al., 2011). Hence, the presence of caveolae may exert an inhibitory effect on BK channel activity.

Increased Cav-1 expression has been found in most diabetic vessels (Hillman et al., 2001; Bucci et al., 2004; Pascariu et al., 2004; Elcioglu et al., 2010; Uyy et al., 2010; Li et al., 2014). Cav-1 expression is directly upregulated by the Forkhead Box O (FOXO) transcription factor (Sandri et al., 2004; Van Den Heuvel et al., 2005). The FOXO-3a phosphorylation levels are significantly reduced in STZ-induced T1DM rat arteries and in cultured human coronary arterial SMCs (Zhang et al., 2010a). This explains the underlying mechanism that leads to Cav-1 upregulation in DM (Figure 3). Furthermore, in STZ-induced T1DM rats, our results in co-immunoprecipitation experiments show that AT1R, c-Src, and BK-α are enriched in the low buoyant density, caveolae-rich membrane fractions of aortas, compared to non-diabetic rats (Lu et al., 2010). Infusion with Ang II (0.05μg/kg) results in markedly enhanced AT1R protein translocation to the low buoyant density fractions of aortas after 1h (83.4% of total membrane AT1R in STZ-induced T1DM rats vs. 28.5% in controls), suggesting enhanced AT1R translocation into caveolae-rich lipid rafts upon agonist activation in diabetic vessels, consistent with previous report in cultured vascular SMCs (Ishizaka et al., 1998). However, the precise mechanism underlying AT1R translocation is currently unclear. The levels of vascular BK-α protein oxidation, tyrosine phosphorylation, and tyrosine nitration are significantly increased in STZ-induced T1DM rats, likely due to the co-localization of NOS, NOX1 and c-Src in the caveolae. Since BK-α but not BK-β1 is present in caveolae, BK-α translocation into the caveolae of arteries in STZ-induced T1DM mice may promote the physical dissociation of BK-α and BK-β1 (Lu et al., 2016), which may explain the uncoupling of BK-α and BK-β1 in diabetic vessels. A working framework has emerged in caveolae targeting of BK channel regulation, in which caveolae compartmentalize BK-α with AT1R, NOS, NOXs, and c-Src to form BK-α-receptor-enzyme microdomain complexes in vascular SMCs (Figure 3). Such caveolae compartmentation is enhanced in diabetic vessels, which facilitates the redox modification of BK-α. Of note, because BK-β1 does not translocate into caveolae, such subcellular distribution of BK-α and BK-β1 may contribute to BK-α and BK-β1 functional uncoupling, thereby exacerbating BK channelopathy in diabetic vessels (Figure 3). Additionally, caveolae take part in endosomal trafficking and regulating surface expression of many membrane proteins (Elkin et al., 2016). Taking into account the consequences of upregulation of caveolae formation in the vascular SMCs in DM, BK-α caveolae translocation may have important pathophysiological implications for vascular BK channel dysfunction in DM.

Protein homeostasis with a balanced regulation between synthesis and degradation is essential for the maintenance of normal cellular function. Cellular proteins are degraded mainly through the lysosomes and the ubiquitin proteasome system (UPS; Ciechanover, 2005). Lysosomal protein degradation occurs through fusion with endocytotic vesicles. This mechanism of protein degradation is non-specific, and all proteins are digested indiscriminately at the same rate. UPS-mediated protein degradation accounts for 80–90% of protein degradation in mammalian cells and it is substrate-specific (Powell, 2006; Schapira et al., 2019). This process is facilitated by three distinct enzymatic steps that involve an ubiquitin-activating enzyme (E1), a ubiquitin-conjugating enzyme (E2), and a ubiquitin ligase (E3). E1 interacts with ubiquitin through an E1-ubiquitin thioester bond in an ATP-dependent manner. It transfers the activated ubiquitin molecule to a cysteine residue on the E2 enzyme to form an E2-ubiquitin thioester-linked intermediate. The E3 ligase facilitates transfer of the E2-ubiquitin moiety to the substrate protein via an amide bond between the carboxy terminus of ubiquitin and a lysine side chain of the substrate protein. The E3 ligase is substrate-specific, allowing repeated positioning of the distal end of ubiquitin molecule for ubiquitin chain assembly with high precision. The poly-ubiquitinated protein is then recognized for enzymatic degradation in the 26S proteasome (Powell, 2006; Schapira et al., 2019). Hence, the E3 reaction is critical for determining the turnover of specific proteins. There are 617 E3 ligases functionally annotated in the human genome (Li et al., 2008). It is known that F-box (FBXO) proteins are a key component of the Skp1-Cullin-F-box (SCF)-type ubiquitin ligase complex (SCF^FBXO) and serve as sites for enzyme-substrate interaction (Kipreos and Pagano, 2000). FBXO proteins contain several functional domains such as the F-box domain, the LRRs, and the WD40 repeats for protein-protein interaction. Two muscle-specific FBXO proteins, FBXO-9 and FBXO-32 (also known as atrogin-1), have been found to be upregulated in diabetic vessels. They mediate BK-β1 protein ubiquitination in coronary arterial SMCs (Zhang et al., 2010a). The molecular basis of FBXO-32 and BK-β1 interaction was identified using site-directed mutagenesis and co-immunoprecipitation approaches, which showed that the PDZ-binding motif (ETSV) on BK-β1 is critical for FBXO-32-dependent ubiquitination (Zhang et al., 2010a). Deletion of the consensus sequence of the PDZ-binding motif in BK-β1 significantly decreases BK-β1 protein ubiquitination (Figure 4; Zhang et al., 2010a). Activation of FBXO proteins reduces BK-β1 expression, while knockdown of FBXO and proteasomal inhibition enhances BK-β1 levels, suggesting that accelerated UPS-mediated degradation of BK-β1 is an important mechanism of BK channel regulation in DM.

The muscle RING-finger protein 1 (MuRF1) is another E3 ligase involved in UPS-dependent vascular BK-β1 degradation (Yi et al., 2014). Nuclear factor-κB (NF-κB) sites in the MuRF1 promoter are required for transcriptional activation, while FOXO sites are not (Wu et al., 2014). Overexpression of MuRF1 downregulates BK-β1 expression, impairs BK-β1-mediated BK channel activity, and reduces BK channel-induced vasodilation in mouse coronary arteries. We found that the N-terminus of BK-β1 and the coiled-coil region of MuRF1 are necessary for BK-β1 and MuRF1 interaction (Yi et al., 2014). Importantly, the protein expressions of FBXO-9, FBXO-32, and MuRF1 are unregulated in the arteries of STZ-induced T1DM animals and in primary human coronary arterial SMCs cultured with high glucose (Zhang et al., 2010a, 2020; Lu et al., 2012; Yi et al., 2014). Such upregulation of FBXO expression is mediated through the suppression of PI3K/AKT-dependent phosphorylation in FOXO-3a, thereby promoting FOXO-3a nuclear translocation and binding to the consensus sequence [GTAAA(C/T)A] in the promoter of Fbxo gene, activating its transcription (Furuyama et al., 2000). However, activation of MuRF1 is due to an increase of NF-κB-mediated Trim63 (encoding MuRF1) transcription (Wu et al., 2014). In DM or hyperglycemia, the activity of AKT is reduced (Okon et al., 2005), while that of NF-κB is augmented (Narayanan et al., 2014), thereby promoting FBXO and MuRF1 expression (Figure 4). Indeed, inhibition of PKCβ activity by ruboxistaurin, NF-κB activity by TPCA-1, and proteasomal activity by MG132 downregulates BK-β1 ubiquitination, preserves BK-β1 expression, and improves BK channel function in coronary arterial SMCs (Zhang et al., 2010a; Lu et al., 2012; Yi et al., 2014).

BK-α protein expression is also regulated by lysosome and UPS degradation (Wang et al., 2013; Liu et al., 2014; Leo et al., 2015; Song et al., 2018). It has been found that the CRL4A and its substrate cereblon (CRBN) complex (CRL4A^CRBN) serves as the ubiquitin ligase that interacts with the C-terminus of BK-α and induces BK-α protein degradation in neurons (Liu et al., 2014). A recent study reported that both CRBN and BK-α proteins were targeted by SCF^FBXO-7 ubiquitin ligase complex for ubiquitination and proteolysis, controlling BK-α function and regulating the learning and memory processes in the brain (Song et al., 2018). However, the specific E3 ligase(s) responsible for BK-α protein ubiquitination in blood vessels is unknown, and how the BK-α-specific E3s are regulated in DM remains to be determined.

Nuclear factor erythroid-2-related factor 2 (Nrf2) plays a critical role in the maintenance of intracellular redox homeostasis by regulating multiple downstream antioxidant enzymes and phase II detoxifying enzymes, which include NADPH dehydrogenase quinone 1 (NQO1), glutathione-disulfide reductase (GSR), glutathione translocase (GSTA), thioredoxin (TXN), thioredoxin reductase 1 (TXNRD1), heme oxygenase-1 (HO-1), SODs, CAT, and GPx (Gao and Mann, 2009; Chen et al., 2014). In addition, Nrf2 negatively regulates the expression of NOXs (McSweeney et al., 2016). The function of Nrf2 is principally regulated by the kelch-like ECH-association protein 1 (Keap1), which mediates Nrf2 ubiquitination and subsequent proteasomal degradation (Canning et al., 2015; Suzuki and Yamamoto, 2015). In the nuclei, Nrf2 binds to the promoters of antioxidant response elements (AREs) and electrophile response elements (EpREs) through interaction with the Nrf2-binding motif [TGA(G/C)xxxGC], where x represents any amino acid (Chorley et al., 2012). Both the KCNMA1 and KCNMB1 genes contain the consensus sequences of Nrf2-binding motifs in their promoter regions. Using promoter luciferase reporter assays, we confirmed that Nrf2 binds to the ARE of the KCNMA1 promoter, but not to that of KCNMB1 promoter. Mutation of the Nrf2-binding motif in the KCNMA1 promoter abolishes the transcription response to Nrf2 (Sun et al., 2020). In addition, adenoviral expression of Nrf2 significantly augmented the mRNA levels of BK-α and BK-β1 in coronary arterial SMCs (Lu et al., 2017a; Sun et al., 2020). These results suggest that Nrf2 facilitates BK-α mRNA expression through activation of KCNMA1 transcription, whereas the stimulatory effect of Nrf2 on BK-β1 mRNA expression is indirect and may be achieved by activating other transcription factor(s) or signaling mechanisms that upregulate KCNMB1 transcription and expression in vascular SMCs.

Nuclear factor erythroid-2-related factor 2 deficiency has been implicated in diabetic complications including those associated with the heart (Tan et al., 2011; Bai et al., 2013), blood vessels (Ungvari et al., 2011; Miao et al., 2012; Li et al., 2017; Lu et al., 2017a), kidneys (Zheng et al., 2011; Cui et al., 2012), and the brain (Pu et al., 2018; Tarantini et al., 2018). The expression of Nrf2 and its downstream genes is slightly increased in the cardiovascular systems of STZ-induced T1DM mice at 2–3months after the onset of hyperglycemia, but then becomes significantly downregulated at 5–6months after the development of hyperglycemia (Tan et al., 2011; Miao et al., 2012; Bai et al., 2013), suggesting the burnout of an important redox protective mechanism in the advanced stages of DM. In db/db and HFD-induced diabetic mice 6months after the development of hyperglycemia, BK channel activity and BK channel-mediated vasodilation in coronary arteries are impaired, accompanied by a remarkable reduction in Nrf2 and its associated antioxidant enzymes (Li et al., 2017; Lu et al., 2017a). Nrf2 KO mice show excessive ROS production, as well as diminished BK channel expression and function in vascular SMCs (Ashino et al., 2013; Sun et al., 2020). Both mRNA and protein expression of BK-α are downregulated, whereas BK-β1 proteins but not mRNA levels are decreased in the arterial SMCs of Nrf2 KO mice, consistent with the notion that Nrf2 regulates BK-α via transcription, and BK-β1 through posttranscriptional mechanisms (Figure 4; Sun et al., 2020). Administration of dimethyl formamide (DMF, an FDA-approved Nrf2 activator) preserves BK channel protein expression, BK channel activity, and BK channel-mediated vasodilation in the coronary arteries of db/db and HDF-induced diabetic mice (Li et al., 2017; Lu et al., 2017a). Currently, Nrf2 activators such as DMF and sulforaphane (SFN) are being used in clinical trials for cardiovascular diseases and metabolic disorder (Yagishita et al., 2020), but it has not been administered for diabetic patients with coronary heart disease (Houghton, 2019). Whether the beneficial effects of Nrf2 activators observed in animal studies would translate into better outcomes in diabetic patients with cardiovascular complications needs to be determined.

Nuclear factor of activated T cells cytoplasmic 3 isoform (NFATc3) belongs to the nuclear factor of activated T cells (NFAT) family of transcription factors that were originally discovered in resting T cells and is important in immune response (Rao et al., 1997). NFATc3 is also involved in the development of skeletal muscle and of the cardiovascular systems (Crabtree and Olson, 2002). The activity of NFATc3 is modulated by the Ca²⁺/calmodulin-dependent phosphatase, calcineurin. Elevation of the intracellular Ca²⁺ concentration activates calmodulin and promotes its binding to calcineurin, leading to calcineurin activation. Activated calcineurin dephosphorylates NFATc3, which in turn induces NFATc3 nuclear translocation. Calcineurin binds to the scaffolding protein A-kinase anchoring protein 150 (AKAP150), corresponding to AKAP79 in humans, which also anchors PKA and L-type Ca²⁺ channel to form a dynamic Ca²⁺ signaling complex (Oliveria et al., 2007). AKAP79/150 strongly suppresses PKA-mediated L-type Ca²⁺ channel phosphorylation and is required for the activation of NFAT by local Ca²⁺ influx through L-type channels (Oliveria et al., 2007).

Nuclear factor of activated T cells share a conserved DNA-binding domain that specifically binds to the DNA core sequence [(A/T)GGAAA] at the promoter region of target genes, activating gene transcription (Rao et al., 1997). Human and mouse KCNMA1 and KCNMB1 contain at least one NFAT-binding motif in their promoters. Inhibition of vascular BK channels by NFATc3 has been reported, while upregulation of NFATc3 expression by Ang II results in decreased BK channel activity in mouse arteries due to the downregulation of BK-β1 mRNA expression (Nieves-Cintron et al., 2007). The effects of NFATc3 on BK channel activity and BK-β1 mRNA expression are abolished by calcineurin inhibitors, FK506 and cyclosporin A, in the presence of Ang II, a finding that has been confirmed in NFATc3 KO mice (Nieves-Cintron et al., 2007). AKAP150 also participates in NFATc3-mediated BK channel downregulation in HFD-induced diabetic mice (Figure 5; Nystoriak et al., 2014). In HFD-induced diabetic mice, the activity of the AKAP150-NFATc3 signaling pathway is upregulated, contributing to impaired BK channel function with reduced BK-β1 expression and increased vascular tone in the mesenteric arteries. However, in AKAP150 KO mice with HFD consumption, the deleterious effects of HFD on BK channels are not observed (Nystoriak et al., 2014). Recently, in vivo administration of a NFATc3 inhibitor (A285222, Abbott Labs) in Akita T1DM mice is found to improve vascular endothelial function, enhance eNOS activity and NO production, reduce endothelin-1 secretion, lower blood pressure, and improve survival (Garcia-Vaz et al., 2020). The beneficial effects of NFATc3 inhibitors on coronary BK channel function in DM warrant further investigation.

Arachidonic acid (AA), a polyunsaturated omega-6 fatty acid, is abundant in normal human diet and in membrane phospholipids. It is an important precursor to a wide range of bioactive mediators and eicosanoids that regulate a multitude of essential functions in the body (Tallima and El Ridi, 2018). AA is metabolized by three major enzyme systems: It is converted by 12-lipoxygenase (12-LOX) into leukotrienes and 12-hydroxyeicosatetraenoic acid (12-HETE), by cytochrome P-450 (CYP-450) epoxygenase into epoxyeicosatrienoic acids (EETs), and by cyclooxygenases (COX) into prostaglandins, including PGI₂ and thromboxane A2 (TXA₂; Brash, 2001; Vila, 2004). Additionally, AA can be metabolized by CYP-450 omega-hydroxylase to produce 20-hydroxyeicosatetraenoic acid (20-HETE).

Arachidonic acid (Lu et al., 2005; Kur et al., 2014; Martin et al., 2014, 2021) and its metabolites (EETs, PGI₂, 12-HETE, and 20-HETE; Li and Campbell, 1997; Yamaki et al., 2001; Zhang et al., 2001; Zink et al., 2001; Lauterbach et al., 2002; Morin et al., 2007) are known to activate vascular BK channels and promote vasodilation through endothelium-dependent hyperpolarization mechanisms. Direct exposure to 10μM AA robustly increases BK channel activity in inside-out excised patches from human umbilical arterial SMCs, suggesting activation of BK channels directly by AA (Martin et al., 2021). Extracellular application of AA results in BK channel activation and hyperpolarization of resting membrane potentials in vascular SMCs (Kur et al., 2014; Martin et al., 2021). These changes can be blocked by LOX, CYP, and COX inhibitors, suggesting that AA metabolites affect BK channels. The effects of AA on BK channels require the presence of BK-β1 (Sun et al., 2007; Martin et al., 2021).

The activation of vascular BK channels by PGI₂ is associated with cAMP-dependent, PKA-mediated phosphorylation. EETs and their metabolites dihydroxyeicosatrienoic acids (DHETs) are also potent BK channel activators and vasodilators, including the human coronary microvessels and internal mammary arteries (Quilley et al., 1997; Archer et al., 2003; Feletou and Vanhoutte, 2006; Larsen et al., 2006). Several different mechanisms of EET- and DHET-mediated BK channel activation have been proposed, including direct activation (Wu et al., 2000; Lu et al., 2001), ADP-ribosylation of Gsα (Fukao et al., 2001; Li et al., 2002), and stimulation of PKA-mediated phosphorylation (Dimitropoulou et al., 2007; Imig et al., 2008). However, AA-induced vasodilation of coronary arterioles via BK channel activity is impaired in high glucose conditions and DM (Lu et al., 2005; Zhou et al., 2005, 2006; Yousif and Benter, 2007; Tsai et al., 2011). PGI₂ and EET levels are decreased in patients with cardiovascular diseases (Theken et al., 2012; Mokhtar et al., 2013; Schuck et al., 2013) and DM (Lane et al., 1982; Kazama et al., 1987; Migdalis et al., 2001; Duflot et al., 2019). As a result of these findings, AA metabolites and analogues have been developed as potential therapeutic agents for cardiovascular diseases and diabetic vascular complications (Campbell et al., 2017; Wang et al., 2021).