BK, bradykinin; BKB2R, bradykinin B2 receptor; BKB1R, bradykinin B1 receptor; Ang, angiotensin; AT1aR, angiotensin type 1a receptor; GPCR, G-protein coupled receptor; ARA, arachidonic acid; PLA2, phospholipase A2; PI, phosphoinositide; [Ca2+]I, intracellular calcium; PGE2, prostaglandin E2; NO, nitric oxide; IC, intracellular; TM, transmembrane; DCt, distal C-tail; PTX, pertussis toxin; LPS, lipopolysacharide; SNP, single nucleotide polymorphism; MD, molecular dynamics; KO, knockout.
Bradykinin (BK) is a potent short-lived effector belonging to a class of peptides known as kinins. It participates in inflammatory and vascular regulation and processes including angioedema, tissue permeability, vascular dilation, and smooth muscle contraction. BK exerts its biological effects through the activation of the bradykinin B2 receptor (BKB2R) which is G-protein-coupled and is generally constitutively expressed. Upon binding, the receptor is activated and transduces signal cascades which have become paradigms for the actions of the Gαi and Gαq G-protein subunits. Following activation the receptor is then desensitized, endocytosed, and resensitized. The bradykinin B1 (BKB1R) is a closely related receptor. It is activated by desArg10-kallidin or desArg9-BK, metabolites of kallidin and BK, respectively. This receptor is induced following tissue injury or after treatment with bacterial endotoxins such as lipopolysacharide or cytokines such as interleukin-1 or tumor necrosis factor-α. In this review we will summarize the BKB2R and BKB1R mediated signal transduction pathways. We will then emphasize the relevance of key residues and domains of the intracellular regions of the BKB2R as they relate to modulating its function (signal transduction) and self-maintenance (desensitization, endocytosis, and resensitization). We will examine the features of the BKB1R gene promoter and its mRNA as these operate in the expression and self-maintenance of this inducible receptor. This communication will not cover areas discussed in earlier reviews pertaining to the actions of peptide analogs. For these we refer you to earlier reviews (Regoli and Barabé, 1980, Pharmacol Rev 32:1–46; Regoli et al., 1990, J Cardiovasc Pharmacol 15(Suppl 6):S30–S38; Regoli et al., 1993, Can J Physiol Pharmacol 71:556–557; Marceau, 1995, Immunopharmacology 30:1–26; Regoli et al., 1998, Eur J Pharmacol 348:1–10). J. Cell. Physiol. 193: 275–286, 2002. © 2002 Wiley-Liss, Inc.
HISTORICAL HIGHLIGHTS OF KININ RESEARCH
Research on bradykinin (BK) and related peptides (kinins) began when Abelous and Bardier (1909) reported a transient fall in blood pressure in human subjects after intravenous injection of fractions from human urine. However, it was not until the work of Frey that modern concepts of kinin action and metabolism began to take shape. Frey (1926) observed a hypotensive effect in dogs injected with human urine. Frey and Kraut (1928) attributed the bioactive agent in urine to a thermolabile, non-dialyzable substance. Frey and his coworkers regarded this substance to be a circulating hormone which affected the blood vessels and called it kreislaufhormon (circulating hormone). Kraut et al. (1930) then found high concentrations of this hypotensive agent in the pancreas and named the substance kallikrein from the Greek kallikreas for pancreas. In 1949, Rocha e Silva determined that the vasoactive substance was a peptide. He noticed that in addition to its hypotensive effect it also induced a slow contraction of the guinea pig ileum. He named it bradykinin, from the Greek bradys for slow and kinesia for movement (Rocha e Silva et al., 1949). Andrade and Rocha e Silva (1956) purified BK but the amino acid sequence was not determined until 1960 when Boussonnas et al. (1960) synthesized BK as a nanopeptide. Throughout the 1960s and 1970s research concentrated on the physiological effects of BK and its analogs. Regoli and Barabé (1980) described the physiological response to BK and its analog desArg9-BK in a variety of tissues. They proposed that the responses were mediated by two distinct plasma membrane receptors, named the bradykinin B2 receptor (BKB2R) and the bradykinin B1 receptor (BKB1R), respectively (Regoli and Barabé, 1980).
CLONING AND GENETIC CHARACTERIZATION
Evidence of G-protein interaction
Cumulative evidence from a number of cell types originally suggested G-protein involvement in BK activation (Flavahan and Vanhoutte, 1990; Voyno-Yasenetskaya et al., 1989). In bovine pulmonary artery endothelial cells, Voyno-Yasenetskaya et al. (1989) showed a pertussis toxin (PTX) insensitive G-protein mediated PLC activation by BK. The BKB2R receptor was also reported to couple to a PTX sensitive G-protein which stimulated the release of arachidonic acid (ARA) from lipid stores and led to the production of prostaglandin E2 (PGE2) (Yanaga et al., 1991; Ricupero et al., 1993). Much of the ARA release in response to BK was attributable to the increase in phospholipase A2 (PLA2) activity (Paglin et al., 1993; Ricupero et al., 1993). PTX did not affect BK induced Ca2+ mobilization. However, PTX inhibited BK stimulated ARA release. These results indicated that transduction of the BK signal, for both PLA2 activation and Ca2+ mobilization, are likely due to different Gα subunits, a PTX sensitive and an insensitive subunit. Furthermore, the BK and GTPγS stimulated release of ARA appeared to be only partially dependent on [Ca2+]i (Ricupero et al., 1993).
Cloning of BKB2R and BKB1R genes
Purification of the BKB2R by classic methods proved elusive (Odya et al., 1980; de Vries et al., 1989; Faussner et al., 1991). However, using the knowledge that the BKB2R is G-protein coupled resulted in the cloning of the BK receptor from the rat using a Xenopus oocyte assay (McEachern et al., 1991). Consequently, the cDNA (Hess et al., 1992) and genomic human DNA (Eggerickx et al., 1992; Ma et al., 1994; Kammerer et al., 1995) of the BKB2R were cloned and localized to chromosome 14q32 (Powell et al., 1993; Ma et al., 1994). The rabbit, murine, pig, mouse, rat, and canine BKB2R genes were identified (McIntyre et al., 1993; Hess et al., 1994; Ma et al., 1994; Bachvarov et al., 1995; Farmer et al., 1998; Ni et al., 1998a; Hess et al., 2001). The deduced amino acid sequences of the BKB2Rs from all species have shown extensive homology, ∼80%. The closely related receptor, BKB1R, was first isolated by expression cloning in IL-1 pretreated human embryo lung fibroblasts, IMR90 (Menke et al., 1994). This was followed by the identification of the BKB1R genomic organization and chromosomal localization (14q32 between markers D14S265 and D14S267) (Bachvarov et al., 1996, 1998b; Chai et al., 1996; Yang and Polgar, 1996). The cloned human BKB1R exhibited high affinity binding for desArg10-kallidin and a 2,000 fold lower affinity for desArg9-BK (Menke et al., 1994). The rabbit, mouse, rat, and canine BKB1R genes were also isolated and characterized with a 68–89% homology among BKB1R species (MacNeil et al., 1995; Pesquero et al., 1996; Hess et al., 2001). Both the BKB1R and BKB2R exhibit the seven transmembrane structures typical of G-protein coupled receptors with an amino acid sequence homology of about 36%. The human BKB2R is composed of 359 amino acids with a molecular weight of 41 kDa. The human BKB1R has 353 amino acids while the counterpart receptors in other species have truncated C-termini. Interestingly, the affinity of desArg9-BK and desArg10-kallidin for BKB1R varies among species. The cloned rabbit receptor presented a 150 fold lower affinity for desArg9-BK than desArg10-kallidin. However, the mouse BKB1R which was 73% identical to the rabbit bound desArg9-BK with an affinity threefold higher than desArg10-kallidin. The canine BKB1R bound desArg10-kallidin and desArg9-BK equally while exhibiting an 81% homology with the human BKB1R. Ligand binding determinations obtained with BKB1Rs isolated from various species suggest the existence of a BKB1R which preferably binds either desArg9-BK or desArg10-kallidin or both, depending on the species. So far separate BKB1Rs, preferentially binding desArg9-BK or desArg10-kallidin within the same species, have not been isolated.
Direct evidence of BKB2R coupling to Gαq and Gαi subunits
Immunoprecipitation of photoaffinity-labeled G-proteins demonstrated BKB1R coupling to Gαq/11 and Gαi(1,2) (Austin et al., 1997). A similar approach also showed Gαq and Gαi subunit coupling to the BKB2R (de Weerd and Leeb-Lundberg, 1997). With regard to Gαi, perhaps the most direct evidence of coupling to BKB2R came from knockdown studies in Rat-1 cells (Yang et al., 1999). Knockdown of either Gαi(2) or Gαi(3) protein production did not affect the binding of BK. In the Gαi(2)-depleted cells, BK induced ARA release was reduced by more than 60%. In the Gαi(3)-depleted cells, BK-induced ARA release was decreased by over 50%. Direct evidence also exists that the Gαq protein family interacts with BKB2R and performs an important role in BK-induced total IP formation. This was shown with Gαq-null ES cells (Ricupero et al., 1997). Their results demonstrated that more than one member of the Gαq family is involved in PI turnover. The Gβγ dimer, dissociated from activated Gαi-protein heterotrimer, was also reported as the signal transducer in Gαi-sensitive IP formation (Camps et al., 1992). Clearly, the BKB2R interacts with at least two families of Gα subunits and perhaps also βγ subunits.
Genetic structure and genetic variants of the BK receptors
Using fluorescence in-situ hybridization, Ma et al. (1994) mapped the BKB2R gene to chromosome 14q32. Genomic Southern blot analysis showed that the BKB2R is encoded by a single-copy gene and is expressed in most human tissues. With the same approach, Chai et al. (1996) mapped the BKB1R gene to chromosome 14q32.1-q32.2, in close proximity to the BKB2R gene. Sequencing results confirmed the locations of these genes and placed them within 12 kb (Fig. 1). A three-exon structure for human BKB2R gene has been proposed, with the coding region in exon 2 and 3 (Kammerer et al., 1995). The BKB1R gene also contains three exons, separated by two introns. The first and second exons are non-coding, while the third exon contains the full-length coding region (Yang and Polgar, 1996).
Investigation of promoter polymorphism within the human BKB2R gene provides a unique insight into the genetic association of this gene with essential hypertension (Mukae et al., 1999; Gainer et al., 2000; Wang et al., 2001). A single nucleotide polymorphism (SNP), T/C, is located in the core promoter of the BKB2R gene, 58 base pair (bp) upstream of the transcription start site. The C allele has been shown to be an independent risk factor for essential hypertension in several ethnic groups, including Japanese, Chinese, and African Americans (Mukae et al., 1999; Gainer et al., 2000; Wang et al., 2001). It is not clear if the T to C transition causes an altered expression level of BKB2R.
The BKB2R exon 1 has three alleles, 2G, 3G, and 3T. The alleles 2G and 3G consist of two and three repeat units (GGTGGGGAC), respectively (Braun et al., 1995). The 3T allele consists of three repeats with a G to T transversion (GGTGGTGAC) which occurs in the second repeating unit (Braun et al., 1995). Difference in allele frequency between different ethnic groups has been observed. Caucasians and African Americans have allele frequencies of 0.53 and 0.58 for the 2G allele (Braun et al., 1995; Lung et al., 1997). Lung et al. (1997) reported 100% 2G allele frequency for Asian subjects examined. The 2G allele appears to have a higher receptor expression level and is always present in the most symptomatic cases of C1 inhibitor deficiency, an endogenous inhibitor of plasma kallikrein (Lung et al., 1997). In addition, the (2G/2G) genotype of this polymorphism was associated with increased contractile efficiency of the BKB1R agonist, Sar-[D-Phe8]des-Arg9-BK, but had no effect on BK-induced contractility in human umbilical vein (Houle et al., 2000). This BKB2R polymorphism suggests a linked disequilibrium with a yet unknown, functionally important polymorphism within the adjacent BKB1R gene. In fact, the distance of linkage disequilibrium has been shown to range up to 175 kb for several disease genes and nearby markers (Ott, 2000). Alternatively, an interaction at the receptor level may contribute to this effect.
An interesting polymorphism is the C/T SNP in exon 2 of the BKB2R gene. This C→T SNP determined an Arg→Cys substitution in the extracellular N-terminal domain. The potency of BK (EC50) is selectively higher when the T allele is present (Houle et al., 2000). Furthermore, the T allele has been proposed to play a protective role in the development of end-stage renal failure (Zychma et al., 1999).
One important polymorphism for BKB1R is the G/C SNP in the promoter region 699 bp upstream of the transcription start site of the BKB1R gene. The frequency of the C allele is significantly lower in patients with inflammatory bowel disease and end-stage renal failure (Bachvarov et al., 1998a, 1998b; Zychma et al., 1999). Reporter gene assay showed a 40% increase in expression for the C allele compared to G allele (Bachvarov et al., 1998a), in agreement with the apparent protective effect of the C allele.
Recently, a group at UW-FHCRC Variation Discovery Resource sequenced 24 African-American and 23 Caucasian subjects as part of an effort to identify polymorphisms in a set of inflammatory related genes, including BKB2R (Rieder et al., 2001). A polymorphic map has been generated covering 15 kb genomic sequence including exon 2, exon 3, and a 6 kb segment after exon 3 of BKB2R (Rieder et al., 2001). A total of 77 SNPs were identified for BKB2R (for a complete list, use GeneBank accession number AF378542). Twenty-five of them have a frequency greater than 10%. Although most SNPs discovered are located in the intron and the 6 kb segment after exon 3, there are several other SNPs worthy of notice. One SNP confirmed the previously identified C/T SNP in exon 2 (Fig. 1). Another is located in the 3′ untranslated region of exon 3 with a frequency of 23%. Two relatively highly frequent SNPs with frequencies of 37 and 40% have been located 102 bp upstream and 277 bp downstream of exon 2, respectively. It is yet to be determined if these SNPs alter the function of the BKB2R gene and the underlying physiological responses.
BK, desArg10-kallidin, and desArg9-BK signaling
As discussed above, BK is the primary endogenous agonist of the BKB2R while desArg9-BK and desArg10-kallidin are the favored agonists for the BKB1R. Upon activation, the BKB2R initiates an array of intracellular and intercellular responses which vary with cell and tissue type. Depending upon the cell type, BK induces excitability, contraction, cell division, permeability, and release of a variety of biologically active agents (Goldstein and Wall, 1984; Regoli, 1984; Vincentini and Villereal, 1984; Gaginella and Kachur, 1989; Roberts, 1989). Early post binding events include an increase in cytosolic Ca2+, activation of G-protein, guanylate cyclase, and phospholipases C, D, and A2 (Burch and Axelrod, 1987; Kremer et al., 1988; Voyno-Yasenetskaya et al., 1989; Yanaga et al., 1991; Taylor et al., 1992; Ricupero et al., 1993, 1997; Lee et al., 2000; Zhou et al., 2000; Exton, 2002). The signaling actions of BKB1R in response to desArg9-BK or desArg10-kallidin are very similar to those of BKB2R with respect to the activation of phosphatidyl inositol (PI) turnover, arachidonate (ARA) release, Ca2+ mobilization, and the induction of the immediate early gene, c-fos. (Cahill et al., 1988; Menke et al., 1994; Jong et al., 1996; Zhou et al., 1999; Schaeffer et al., 2001). Like BK, desArg9-BK also induces protein formation and cell division (Goldstein and Wall, 1984; Beny et al., 1987; Churchill and Ward, 1987; Vianna and Calixto, 1998; Prat et al., 1999). DesArg10-kallidin and desArg9-BK activate BKB1R with different specificities depending on the species (Hess et al., 1996; Pesquero et al., 1996).
BK also elicits many of the intracellular signaling responses that are typically associated with the activation of growth factors. Downstream, BK causes the activation of protein-tyrosine phosphatases (Zhao et al., 1993), Ras-GTPase-activatinG-protein (Tsai et al., 1989), Raf-1 translocation (Rizzo et al., 1999), sphingosine kinase (Melendez et al., 1998), tyrosine kinases (Fleming and Busse, 1997) MEK/MAP kinases pathway (Velarde et al., 1999; Luo et al., 2000; Yang et al., 2001a), and the JAK/STAT signaling proteins (Tyk2, STAT3) which appear to form a complex with the BKB2R (Ju et al., 2000).
Cross regulation of BKB1R and BKB2R has been suggested based on the observation that activation of BKB2R also activates NF-κB, which can then prime the expression of BKB1R (Schanstra et al., 1998; Phagoo et al., 1999; Xie et al., 2000). Additional pathways following BK activation include production of IL-6 and IL-8 in lung fibroblasts (Hayashi et al., 2000), generation of reactive oxygen species in vascular smooth muscle cells (Greene et al., 2000) and transinactivation of EGF receptor in A431 cells (Graness et al., 2000). BK also induces the synthesis of a number of vasoactive, inflammatory agents such as platelet activating factor, endothelium derived hyperpolarizing factor, nitric oxide, and leukotrienes (Regoli and Barabé, 1980; Vane and Botting, 1987; Cahill et al., 1988; Ahluwalia and Perretti, 1999).
Divergence between BKB1R and BKB2R signaling
The signals generated by the BKB1R and the BKB2R at first glance appear to be identical. Both receptors induce an increase of cytosolic Ca2+, activate PLC, and couple to the G-proteins Gαq, Gαi2, and Gαi3 (Yanaga et al., 1991; Liao and Homcy, 1993; Austin et al., 1997; Xie et al., 2000). However, a closer inspection reveals that the kinetics of the increase of [Ca2+]i is quite distinct, suggesting the intracellular signaling that induces the increase of cytosolic Ca2+ are also distinct (Tropea et al., 1993; Marsh and Hill, 1994; Mathis et al., 1996). With regard to [Ca2+]i, the BKB1R utilizes largely extracellular Ca2+ (Zhou et al., 2000) while BKB2R utilizes mostly Ca2+ located within intracellular compartments, likely through IP3-gated Ca2+ channels in the endoplasmic reticulum (Mombouli and Vanhoutte, 1995). These data suggest that different signaling paths are utilized by these two receptors to achieve increases in [Ca2+]i. Activation of the BKB2R leads to processes that can be considered as anti-fibrotic. For example, the activation of the BKB2R does not stimulate collagen or CTGF synthesis in fibroblasts while the BKB1R does (Ricupero et al., 2000; Yu et al., 2002). Also, treatment with angiotensin converting enzyme inhibitors, blocking the production of angiotensin II and promoting the accumulation of BK, has clearly been shown to be anti-fibrotic (Border and Noble, 2001). This is in contrast to BKB1R, which upon activation by desArg10-kallidin stimulates type I collagen synthesis and raises α1(I) collagen mRNA and CTGF mRNA (Ricupero et al., 2000). DesArg10-kallidin, TGFβ, and angiotensin II have also been reported to increase CTGF production in fibroblasts (Ricupero et al., 2000; Yu et al., 2002). Also, IMR90 cells, human embryonic lung fibroblasts, exhibit mitotic and collagen synthetic responses to desArg9-BK (Goldstein and Wall, 1984). Thus, BKB1R plays a distinct, stimulatory role in wound repair and is pro-fibrotic.
With few exceptions (Lung et al., 1998; Schmidlin et al., 1998; Rehbock et al., 1999; Kintsurashvili et al., 2001) the BKB2R is constitutively expressed and its presence on the cell surface is regulated through internalization and resensitization. The main regulatory machinery for BKB2R expression lies at the expression level. Unlike the BKB2R, the BKB1R is an inducible gene. Its expression is upregulated following trauma or injury. However there is some evidence suggesting that the BKB1R also functions when expressed minimally (constitutively). For example, Ricupero et al. (2000) showed that basally expressed BKB1R causes upregulation of collagen synthesis in response to desArg10-kallidin. These observations suggest that, at times, even low gene transcription is sufficient to permit BKB1R to function prior to induction. For the de novo synthesis of BKB1R, specific protein kinases appear to be recruited following tissue injury. Specifically, the activation of the p38 MAP kinase pathway precedes BKB1R activation (Larrivee et al., 1998).
Regulation of BK receptor expression at transcriptional and post-transcriptional levels
The BKB1R gene expression is highly regulated and represents a good model for studies of inducible genes. It is present as a single copy which spans more than 10 kb and includes three exons interrupted by two introns (Fig. 1). While the 5′ untranslated region is distributed on all three exons, the coding region is located entirely on the third exon. Two distinct functional promoters are found in the human BKB1R gene. Exon and intron–exon boundaries, and 5′, 3′ flanking regions were sequenced. While the 5′ untranslated region proved to be distributed on all three exons, the coding region was located entirely in the third exon. The exon–intron junction sequences are highly conserved. Primer extension analysis mapped the transcriptional initiation site 21 bp downstream of a TATA sequence and downstream of numerous transcription factor binding motifs (Fig. 2) (Bachvarov et al., 1996; Yang and Polgar, 1996). From studies by Yang and Polgar (1996), a positive regulatory element (PRE) was located at position −604 to −448 bp upstream of the transcription start site (Fig. 2). This PRE contains a classic powerful enhancer and a negative regulatory element, at position −682 to −604. The negative element ablates the function of the enhancer. The region of the enhancer and silencer was minimized to a 100-bp element and a 78-bp fragment, respectively. Transient transfection of the enhancer construct into a variety of cell types showed that this enhancer is cell-type specific (Yang and Polgar, 1996). In the characterization of the enhancer two motifs were found to be essential for full enhancer activity. Gel shift and antibody supershift assays determined that an AP-1 factor binds one of these motifs. The nuclear protein, which binds the other motif, has yet to be identified. Both factors are the critical regulators for this enhancer activation.
The presence of an IL-1β responsive element has proved elusive. Studies by Yang and Polgar (1996); Yang et al. (1998); Bachvarov et al. (1996) have not been able to identify domains in the human BKB1R gene promoter region involved in the induction of expression by IL-1 or LPS. However, two other groups (Ni et al., 1998a, 1998b; Schanstra et al., 1998) reported the upregulation of the human BKB1R gene by IL-1β at the transcriptional level and strongly correlated with the activation of transcription factor NF-kB. Differences in the localization of the putative NF-kB-binding domain have been reported. One study identifies this motif at −1172 to −1162 (Schanstra et al., 1998) while another study points to −67 to −57 position (Ni et al., 1998b). Other regulatory elements also appear to be involved in the upregulation and downregulation of the BKB1R gene at the promoter level. For instance, in vivo studies using ultraviolet and dimethylsulfate footprinting analyses have shown that BKB1R promoter putatively bound to a number of sequence-specific DNA binding proteins (GATA-1, PEA3, AP-1, CAAT, Sp1, Pit-1a, Oct-1, CREB) (Angers et al., 2000). Yang et al. (2000a), reported construction of a human BKB1R minigene which contained a 1.8 kb promoter, the entire exon 1, 1.5 kb of intron I, the entire exon II, intron II, and the luciferase gene as a reporter. Transient transfection of the minigene into SV40-transformed IMR90 cells (IMRSV) resulted in a promoter activity which was activated by LPS and desArg10-kallidin. In contrast, these mediators did not induce the activity of the 1.8 kb promoter construct alone. Thus, motifs exclusive of the promoter such as 5′-UTR and/or intron regions appear to be required for mediator-induced expression of this gene. Promoter activities of both the minigene and the 1.8 kb promoter construct were enhanced in a dose-dependent manner with c-Jun. Furthermore, cotransfecting c-Jun with the minigene achieved the maximal promoter activity with no further increase in response to mediators. Conversely, the induction of the minigene promoter activity by mediators was abolished upon cotransfection with a dominant negative mutant of c-Jun. Other experiments suggest that multiple AP-1 sites are interactive with c-Jun. These results point to c-Jun as a key intermediary in the activation and expression of this gene. However, participation of motifs outside of the promoter appears necessary to obtain this inducible expression. The promoter has also been identified as a target of the tumor suppressor, p53 (Yang et al., 2001b). The suppression by p53 was not mediated by competition with the TATA-binding protein and was not through interaction with the putative p53-binding site.
The nuclear factors involved in regulating the BKB2R gene expression are largely unknown. Early studies of the rat BKB2 gene revealed lack of a TATA box but identified potential transcription factors such as cyclic AMP response element-binding protein, AP-1, NF-B, SP-1, and Egr-1 (Pesquero et al., 1994). More recently, studies of the promoter region of the rat BKB2R gene identified a p53-binding site (Saifudeen et al., 2000). In this study, two p53-like binding sites were identified in the 5′-flanking region of the rat BKB2R gene. AP1 site, a sequence at 70 bp that is conserved in the murine and human BKB2R genes; and P2 site, a sequence that is not conserved in any of the species (Saifudeen et al., 2000). Although p53 has been implicated to be induced and activated during cellular injury and chronic inflammation (Arai et al., 1999), it has also been shown to play a role in the regulation of BKB2R gene expression in the developing kidney (El-Dahr et al., 2000a). The biological role of the p53-mediated upregulation of BKB2R gene expression still remains unclear.
RNA destabilization as a further control of BKB1R expression
The BKB1R gene expression has also been shown to be regulated through mRNA stabilization (Zhou et al., 1999). The short 3′-untranslated region (3′-UTR) of the BKB1R contains only 14 bases with an alternative polyadenylation signal (AUUAAA) which overlaps the stop codon (Zhou et al., 1999). A luciferase gene construct containing the BKB1R 3′-UTR displayed a half-life of ∼1 h, which is considerably shorter than the 6-h half-life of the wild type luciferase construct. Thus this data provided evidence that the 3′-UTR is participating in the regulation of BKB1R mRNA stability and its ultimate expression (Zhou et al., 1999). In addition, induction of BKB1R by immunostimulants, such as IL-1, has been reported to involve not only transcriptional activation but also increased mRNA stability. The IL-1 treatment has been shown to double the mRNA half-life of the BKB1R from 1 to 2 h in IMR90 cells (Zhou et al., 1998). Furthermore, treatment with protein synthesis inhibitors dramatically increases BKB1R mRNA stability as well as the BKB1R mediated cellular responses (Deblois et al., 1991; Zhou et al., 1998). These results strongly suggested a post-transcriptional mRNA stability regulator, which most likely is a short-lived protein factor that is sensitive to protein synthesis inhibitors. This protein factor may interact with the 3′-UTR of the BKB1R gene and selectively regulate the mRNA stability of this gene.
RECEPTOR LEVEL REGULATION
Unlike the BKB2R, the BKB1R does not desensitize as measured by PI turnover, does not internalize or resensitize (Zhou et al., 2000). Unlike the BKB2R, BKB1R once induced under pathological conditions or expressed stably in cultured cells its activation results in a persistent PI turnover suggesting a lack of desensitization and absence of internalization and resensitization (Austin et al., 1997; Zhou et al., 2000; Phagoo et al., 2001). Moreover, long term stimulation of this receptor actually leads to increased expression. Another unique feature of the BKB1R is its apparent selective desensitization. The BKB1R activated PI turnover proceeds unimpeded for at least 60 min while BKB2R mediated PI turnover is desensitized within 30 min (Zhou et al., 2000). However, BKB1R mediated ARA release is short-lived, with a very similar pattern as that seen with BKB2R. This selective desensitization phenomenon has also been observed with the G-protein coupled gonadotropin-releasing hormone receptor, GnRHR (Anderson et al., 1995; McArdle et al., 1996). This differential desensitization would enable BKB1R, and perhaps other inducible receptors, to manifest both transient and longer lasting signaling events linked to both transient and longer lasting biological functions (Regoli et al., 1993; Zhou et al., 2000). Interestingly, unlike ligand activated BKB2R, which internalizes rapidly, the activated BKB1R is internalized minimally with no significant receptor uptake observed within 60 min of exposure to desArg10-kallidin (Austin et al., 1997; Zhou et al., 2000; Sabourin et al., 2002). Thus, in case of the BKB1R, two desensitization processes appear to be taking place. A rapid desensitization, seen with arachidonic release and a slow desensitization seen with PI turnover. The rapid process has been linked to receptor phosphorylation (Blaukat et al., 1996; Prado et al., 1998; Smith et al., 1998). Phosphorylation may either directly inhibit the receptor function or attract the binding of a protein such as α-arrestin which causes the dissociation of G-protein from GPCRs (Hausdorff et al., 1990; Zhang et al., 1997). The slow desensitization process could be receptor uptake linked.
Receptor desensitization, resensitization, and internalization
BKB2R expression is maintained at the protein level with motifs within the seven transmembrane sequence regulating receptor desensitization, uptake, and resensitization, parameters which are responsible for the receptor self maintenance (Prado et al., 1996, 1997, 2001). BK activation of its receptor is followed by receptor desensitization followed by its internalization, and resensitization (Prado et al., 1997, 1998, 2001). Interestingly, the route of BKB2R internalization differs from most GPCRs. The major mechanism of agonist-induced internalization of GPCR is by attachment to β-arrestin and dynamin-dependent clathrin-coated vesicles (Luttrell and Lefkowitz, 2002; Laporte et al., 2002). The sequestration of the BKB2R has been associated with the caveolae (Austin et al., 1997; de Weerd and Leeb-Lundberg, 1997) while little if any uptake is observed with the BKB1R. These distinct receptor-specific patterns of self-destruction imply different roles for BKB1R and BKB2R with respect to signal function. More recently, the initial sequestration of the GPCR has been reported to be accompanied by signaling cascades separate from those of the G-protein (Claing et al., 2002; Luttrell and Lefkowitz, 2002). Thus, although receptors, BKB1R and BKB2R classically link to Gαi and Gαq other as yet undetermined signal pathways may exist for each resulting in disparate functions.
Active motifs of the BKB2R
As studies have shown, the BKB2R undergoes rapid and pronounced internalization while the BKB1R displays slow and very limited internalization (Faussner et al., 1998, 1999; Zhou et al., 2000). Studies using C-terminus exchanges between the BKB1R and BKB2R have shown that the BKB2R C-terminus contains sequences which are involved in receptor uptake, motifs which the BKB1R does not appear to possess (Faussner et al., 1998). This suggested that the BKB2R C-terminus participate in its endocytosis. However, at this time it is not clear as to all the motifs within the C-terminus involved in receptor uptake. It is clear that one of these is the S/T cluster located in the distal portion of the BKB2R C-terminus (Fig. 3). Several laboratories have investigated the actions of these hydroxyl-possessing residues. In the rat BKB2R, elimination of the last 34 residues within the distal C-terminus which includes the S/T cluster results in defective signal transmission and uptake, suggesting that this region plays a role in both receptor function and maintenance (Prado et al., 1997). With regard to receptor maintenance, the presence of S348 is crucial for normal BKB2R uptake. Pinpoint replacement of any of the other three serines (335, 341, or 350) to alanine and threonine (347) to valine resulted in defective internalization as compared to WT. However, close examination of this S/T cluster illustrated that the presence of S348 promotes internalization while the presence of S341 dampens it (Prado et al., 2001). Other groups have done similar studies. One study showed that in the human BKB2R, truncations after positions Y320 and L334, deletions within the segment covering positions 335–351, substitutions with alanine at position S339, S346, S348, T342, T345 (S341, S348, S350, T344, and T347 in the rat) led to diminished receptor internalization (Pizard et al., 1999). [The numbering system of the human BKB2R sequence used by Hess et al. (1992) is two bases off in comparison with the numbering system used for the rat BKB2R studies by Prado et al. (1997)]. These internalization-impaired mutants also failed to produce BK-induced receptor phosphorylation (Pizard et al., 1999) suggesting that phosphorylation is involved in internalization. Further studies from the same group using in vivo studies showed that in response to BK, S339, and S346/S348 are phosphorylated and that S348 by itself is constitutively phosphorylated (Blaukat et al., 2001). An earlier study using mass spectrometric analysis of the rat BKB2R showed phosphorylation of all four serines and one threonine occurring in a pairwise fashion (Soskic et al., 1999). Thus all four serines and two threonines in the C-terminus may be integral components for internalization. Although these residues appear to undergo phosphorylation, it is not clear at this time whether phosphorylation of all or any is necessary for receptor maintenance or function.
The possibility that distinct receptor structures dictate different signaling mechanisms was investigated by using rat angiotensin type I a receptor (AT1aR) as a motif donor. As BKB2R, AT1aR possess a similar serine/threonine complex within the (326–342) portion of its C-terminus. The chimeric exchange of rat BKB2R (333–351) with the corresponding rat AT1aR fragment (326–342) C-terminus, resulted in a receptor with a retarded internalization but a normal BK activated PI turnover (Prado et al., 2001). However, the exchange of last 34 residues of BKB2R with AT1aR resulted in normal uptake and normal signaling (Prado et al., 2001). These results illustrated that the precise S/T sequence is not necessary and that another motif within the distal portion of the C-terminus, outside of the S/T complex, is also involved the regulation of BKB2R uptake.
Motifs in the proximal tail and other intracellular loops have also been investigated. The proximal tail contains a YXXL motif and the intracellular (IC) loop 2 (IC2) contains a DRY motif. Mass spectroscopic studies showed that in the rat BKB2R, Y322, and Y131 are phosphorylated (Soskic et al., 1999). Prado et al. (1997) showed that Y131 and Y322 are critical for receptor uptake and signal transduction. Threonine at position 137 with potential for phosphorylation also proved important for receptor signal function and maintenance (Prado et al., 1998). In fact, introduction of a negatively charged residue, aspartate, in this position resulted in a mutant with poor signal function and normal receptor uptake. As shown in Figure 3, other hydroxyl possessing residues T137 and S139 in IC2 and T239 in IC3 showed no role in rBKB2R signal capacity or receptor internalization (Pizard et al., 1999; Prado et al., 2001). These results illustrate the complex role of the IC motifs and OH phosphorylation with respect to both receptor maintenance and its signal function. Clearly, within the IC face of the BKB2R, at this time, there is no clear separation between receptor maintenance and function.
Formation of BKB2R/AT1aR hybrids
The AT1aR is very similar to the BKB2R in length of the IC loops. It also shares some homology with the BKB2R within the IC face (Prado et al., 2001). However, the physiologic roles of this receptor such as its hemodynamic actions are most often opposed to those of the BKB2R (de Gasparo et al., 2000). As pointed out in the previous section, the replacement of the last 34 amino acids of the C-terminus of BKB2R with those from the C-terminus of AT1aR resulted in a receptor which functioned, internalized and resensitized as BKB2R WT (Prado et al., 2001). Yu et al. (2002) made single as well as multiple global replacements within the IC of BKB2R with the corresponding regions of the AT1aR. When stably transfected into Rat-1 cells, this hybrid bound BK with high affinity. Single replacement of the IC2 or the distal 34 residues of the C-terminus (DCt) with the corresponding regions of AT1aR resulted in chimeras, which continued to function in response to BK to turnover PI and release ARA as WT BKB2R. In fact, the simultaneous exchange of IC2 and IC3 loops of BKB2R with AT1aR resulted in a receptor responding to BK with PI turnover and ARA release approximately fourfold greater than WT BKB2R. Likewise, the simultaneous replacement of the IC2 and DCt resulted in a 2.8-fold and 1.6-fold increase in PI turnover and ARA release, respectively. Replacement of all three IC domains (IC2, IC3, and DCt) resulted in PI closer to that of AT1aR than BKB2R. The rate of uptake of the receptor chimeras was similar to that of WT BKB2R. When transfected into Rat-1 cells, the AT1aR markedly increased the expression of connective tissue growth factor (CTGF) mRNA while BK slightly decreased it. The dual IC2/DCt and triple IC2/IC3/DCt hybrids both up-regulated CTGF mRNA in response to BK. These results show that the IC face of the BKB2R can be exchanged with that of another receptor such as AT1aR which produce hybrid receptors taking on the functional characteristics of the donor receptor. The creation of chimera with stepwise replacement of the IC domains should allow for assignment of specific roles to the individual loops and C-terminus in the signaling and internalization of the BKB2R. This approach could also facilitate the generation of receptors with BKB2R binding and selected functions of the donor receptor (Yu et al., 2002).
Motif interactions within the IC face of BKB2R
Accumulating data obtained from mutagenesis of the IC face of BKB2R cDNA are suggestive of cooperative interactions among the intracellular regions of the BKB2R (Prado et al., 1998, 2001; Piserchio et al., 2001; Yu et al., 2002). For instance, while a single mutation Y131A and Y131S in IC2 results in receptors with poor signal capacity and an altered receptor internalization rate, a double mutant involving simultaneous mutation of (Y131A/Y322A, Y131S/Y322S), resulted in normal signal capacity. Receptor modeling indicated that the C-terminus and the distal IC2 are in close proximity (Fig. 4a,b). This could account for the functional synergy (Prado et al., 1997, 1998). Another example of this synergy is the mutant T137D which has poor PI turnover (Prado et al., 1998) and the C-terminus S335A, S341A mutant which also has poor signal capacity (Prado et al., 2001). When the two mutations were combined, the new mutant receptor returned to normal signal function (Piserchio et al., 2001). Molecular modeling suggested that the altered signaling of the T137D mutant is due to the introduction of a negative charge, indicating that phosphorylation of this residue takes place and participates in the life cycle of this receptor (Piserchio et al., 2001). These results point to a functional interaction between the IC2 and the C-terminus (Piserchio et al., 2001). Other recent data supporting interactions between intracellular regions involved the exchange of the IC3 between AT1aR and the BKB2R (Yu et al., 2002). When IC3 was exchanged but the IC2 was retained the hybrid did not function. Double exchange of IC2 and IC3 of BKB2R with the corresponding AT1aR regions resulted in a fully functional hybrid (Yu et al., 2002). Molecular modeling determined that K237 and R239 of the IC3 of the AT1aR were in close proximity to K136 of IC2 of BKB2R in the chimera containing the IC3 of AT1aR. The consequent repulsion between the IC2 and IC3 would account for the lack of function. This interaction is not present in the WT or in the double chimera receptor containing both IC2 and IC3 of AT1aR. It is of interest to note that when the IC2 and IC3 of BKB2R are both replaced with AT1aR, the function of the hybrid is that of the AT1aR with respect to the upregulation of CTGF expression (Yu et al., 2002).
PHYSIOLOGY OF BKB2R AND BKB1R KNOCKOUTS
Targeted disruption (“knockout”) of the BKB2R gene by homologous recombination has been produced in a number of laboratories. An early study showed that knockout (KO) mice developed normally without any hemodynamic changes (Borkowski et al., 1995). However, another study showed that when KO mice were overloaded with high salt, severe hypertension occurred (Alfie et al., 1996). Also El-Dahr et al. (2000a,b) demonstrated in KO mice that the renin–angiotensin system was suppressed and abnormal renal development occurred. Other studies established that BK interacts with the renin–angiotensin system with BK apparently stimulating renin gene expression (Yosipiv et al., 2001) and neutralizing vasoconstricting activities of angiotensin II (Cervenka et al., 2001). KO mice also showed a chronically elevated heart rate with an impaired baroreflex control which contributed to the development of cardiomyopathy (Madeddu et al., 1999). The vasodilatory responses to BK are also absent in BKB2R KO mice showing the importance of the receptor in the action of BK (Berthiaume et al., 1997). In fact, BKB2R disruption can lead to hypertension and cardiac impairments such as left ventricular remodeling (Emanueli et al., 1999). Together with the genetic results discussed in Section 3 (Receptor Signaling), these results point to a role for BKB2R in the regulation of blood pressure and cardiovascular function. These damaging effects in the heart also implicate the AT1aR receptor and the lack of BKB2R to counterbalance its actions in KO mice (Madeddu et al., 2000). Thus an important interaction between Ang II and BK may be taking place and may be essential for the development of the normal heart. The BKB1R has been implicated in nociception and the accumulation of leukocytes in inflamed tissues (Rupniak et al., 1997; Perron et al., 1999). BKB1R KO mice developed normally, were normotensive but displayed a drastic reduction in the accumulation of polymorphonuclear leukocytes at sites of inflammation and hypoalgesia (Pesquero et al., 2000). Recent studies have also shown that apoptosis of neutrophils is impaired in mice lacking the BKB1R (Araujo et al., 2001).
As this report shows, the work on the mechanisms regulating the action and the ultimate physiologic roles of BK and its associated effectors, desArg9-BK and desArg10-kallidin, are only in their infancy. Clearly, a lot more elucidation is yet to come. The BKB1 and BKB2 receptors present an ideal opportunity to understand the signaling functions and self-maintenance mechanisms of two very closely allied GPCR, located adjacently on chromosome 14q32 but with very different mechanisms of expression. The BKB1R gene is induced rapidly. The mechanisms operating in its induction are only understood at a very rudimentary level. Participation of such trans factors as NFkB, c-Jun, and p53, has been reported. However, how these elements fit together to regulate gene expression is totally unclear. Also, what is the role of the potent enhancer in the induction of the BKB1R gene? Since BKB1R is not endocytosed and only partially desensitized, at best, its regulation through mRNA destabilization and gene expression are paramount and in need of understanding.
Considerable progress has been made toward understanding the motifs operating in BKB2R signaling function and self-maintenance. Motifs within the IC2, IC3, and the C-terminus interact to attain both signal transduction and self-maintenance of the receptor. Signaling, internalization, and receptor resensitization appear to share common motifs. It is also becoming clear that global exchanges within the BKB2R IC face, with receptors with similar sequence length such as the AT1, can be achieved to produce viable hybrids.
Another area in need of better understanding is that of signal transduction by these receptors. This question is exacerbated when the angiotensin type 1 receptor is also considered. Clearly, all three receptors, the BKB2, BKB1 and AT1, link to Gαi and Gαq but the differences in their physiologic functions are often divergent and opposing. These signal capacities seek explanation well beyond the actions of the G-protein subunits that these receptors are known to link. Progress in this area is being made and results suggest that indeed signaling is taking place outside of the classic G-protein interactions.