(Received for publication, November 18, 1996, and in revised form, March 5, 1997)
From the Institute for Physiological Chemistry and
Pathobiochemistry, Johannes Gutenberg University, Duesbergweg 6, D-55099 Mainz, Germany and the ¶ Institute of Molecular Pathology,
Dr. Bohr-Gasse 7, A-1030 Vienna, Austria
Kinins are proinflammatory peptides that dilate
vessels, increase vascular permeability, contract smooth muscles, and
provoke pain. The known mammalian kinin receptors are classified as two subtypes, i.e. the B1 receptor triggered by
[des-Arg9]bradykinin and inhibited by
[des-Arg9,Leu8]bradykinin, and the B2
receptor stimulated by bradykinin and antagonized by HOE140. Here we
report the cloning of a non-mammalian kinin receptor gene amplified
from genomic chicken DNA. The protein predicted from the open reading
frame shows 31 and 49% sequence identity to the human B1 and B2
receptors, respectively, suggesting that it represents a G
protein-coupled receptor of the kinin receptor family. The
recombinantly expressed chicken receptor had IC50 values of
4.7 nM for the authentic ligand, ornithokinin
([Thr6,Leu8]bradykinin), 3.8 nM
for HOE140, and 10 µM for bradykinin,
[des-Arg9]bradykinin, and
[des-Arg9,Leu8]bradykinin. Ornithokinin and
HOE140 at nanomolar concentrations stimulated intracellular inositol
phosphate accumulation and induced a significant transient rise in
intracelluar free Ca2+, whereas bradykinin was ineffective
even at 100 nM. Hence the principal B2 receptor antagonist
HOE140 is a potent agonist of the chicken kinin receptor. This unique
pharmacological profile classifies the ornithokinin receptor as a novel
subtype among kinin receptors and will facilitate further molecular
studies on ligand binding and receptor activation.
In mammals two principal types of kinin receptors, B1 and B2, with
distinct pharmacological properties have been cloned and characterized
(1-4). Their ligands are derived from bradykinin, Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg. Expression of the B1 receptor is
induced under pathophysiological conditions such as injury, infection,
or chronic inflammation where it mediates hyperalgesia and
sensitization (5, 6). The B2 receptor is constitutively expressed in
fibroblasts and in epithelial, endothelial, neuronal, and smooth muscle
cells (7, 8). Stimulation of the B2 receptor induces local responses
such as vasodilation, bronchoconstriction, edema formation, pain
sensations, and inflammatory reactions (1, 9, 10). The principal B1
agonist is [des-Arg9]bradykinin (1). Potent and selective
B1 antagonists such as [des-Arg9,Leu8]bradykinin derive from the
corresponding agonist by a single substitution, Phe Leu, at
position 8 (5, 11). Bradykinin and [Lys0]bradykinin
("kallidin") are equipotent agonists of the B2 but not of the B1
receptor (5). A Phe8
Leu8 substitution in
bradykinin renders the peptide inactive for the B2 receptor (11, 12).
HOE140, a synthetic peptide derived from the bradykinin sequence, is a
selective and tightly binding B2 receptor antagonist (13, 14) that
lacks partial agonistic activity in most physiological settings
(15).
Kinins are well conserved peptides that have been identified in invertebrates, e.g. in the wasp (16). In fact the non-mammalian kinins are identical to their human counterpart including a 4-hydroxylation of a proline residue at position 3 (16). Kinins are present in all the major vertebrate classes including reptiles, amphibia, fish, and birds (8). The avian kinin, first described in 1967 (17), and aptly named ornithokinin, differs from the human bradykinin in two positions: Thr substitutes for Ser at position 6, and Leu for Phe at position 8. Ornithokinin, i.e. [Ser6,Leu8]bradykinin, induces transient hypotension in the chicken and contracts smooth muscle cells of the oviduct (17, 18), two physiological activities reminiscent of the kinin effects in mammals. To date the receptor underlying ornithokinin's effects has not been defined. Unlike most of the insect kinins ornithokinin has no effect on the human kinin system, and vice versa (18). The differential binding profiles are likely to reflect distinct structures of the ligand-binding sites of the corresponding receptors, however, the lack of non-mammalian kinin receptor sequences has precluded such comparisons.
Here we describe the cloning of an avian kinin receptor using a PCR1-based strategy. Expression of the biologically active chicken ornithokinin receptor revealed the unexpected finding that HOE140, the principal antagonist of the mammalian B2 receptors, is a full agonist of the avian receptor as is ornithokinin, whereas bradykinin and [des-Arg9]bradykinin have no effect. Sequence analyses indicate that avian and mammalian kinin receptors are only distantly related. Our results suggest that the ornithokinin receptor is the first member of a novel type of kinin receptors. More importantly, having available a receptor in which HOE140 acts as a full agonist will make the ornithokinin receptor a unique tool to study ligand binding and receptor function at the molecular level.
The pCDNA3, pVL-1392 vectors, and the Liposome Kit were from Invitrogen; oligonucleotides from MWG-Biotech; [125I]3-(4-hydroxyphenylpropyl)-HOE140 (HPP-HOE140) from Hoechst, myo-[2-3H]inositol from Biotrend; Dowex AG 1-8X (hydroxide form, 200-400 mesh) from Bio-Rad; and cell culture reagents were from Life Technologies. Other reagents were obtained from standard commercial sources.
Cloning of the Ornithokinin ReceptorA 376-bp DNA fragment
was amplified by PCR using chicken genomic DNA prepared from chicken
embryonic fibroblasts (CEF) as the template with the following primers:
H669 (5-ACCATGTCCATGGGCCGGATGCGCGG) and H1110
(5
-GCAGCGT(ACGT)TCCAGGAAGGTGCTGATCTGG). The fragment was subcloned
into pBluescript SK+ (pSK+; Stratagene), and
six isolated clones were sequenced by the dideoxy nucleotide sequencing
technique (19). One of the cloned DNA fragments was
32P-labeled by random priming and used to screen 2 × 106 phages of a chicken genomic library constructed in
FIX-II vector (Stratagene). Using the Church buffer system for high
stringency hybridization 10
phages that strongly hybridized with
the labeled fragment were isolated. Restriction mapping and Southern
blot analysis of the purified
DNA identified a 5.6-kilobase pair BamHI DNA fragment containing the ornithokinin receptor
gene. This fragment was subcloned into pSK+ and sequenced.
Sequence comparisons were done by the blastn program searching the
GenBankTM/EMBL data base.
To amplify the
coding sequence of the ornithokinin receptor two oligonucleotides,
OKR11 (5-GCAAGTCGATGGCAGTTTG) and OKR5 (5[prime-CCTATTGTGATAATGGCAAAAC) were synthesized and used for PCR.
The amplified 1341-bp fragment was directly subcloned into pVL1392 and
pCDNA3 to give the following expression vector constructs: pVL1392-OKR and pCDNA3-OKR.
CHO-K1 were grown in Dulbecco`s modified Eagle`s medium supplemented with Ham's F-12 nutrient mixture, 10% (v/v) fetal calf serum, and 0.5% (w/v) penicillin/streptomycin in a humidified CO2 atmosphere at 37 °C. COS-7 cells were kept under the same conditions except that RPMI medium was used. CEF were prepared from 11 to 13-day-old chicken embryos by standard procedures and cultured in Dulbecco's modified Eagle's medium containing 2% chicken serum and 8% fetal calf serum (for details, see Ref. 20). Insect Sf9 cells were cultured as monolayers in TC100 medium supplemented with 10% fetal calf serum, 0.5% penicillin/streptomycin in a humidified atmosphere at 27 °C.
Transfection of CHO-K1 CellsThe pcDNA3-OKR was used to transfect CHO-K1 cells using the calcium phosphate method (21). Transfectants were isolated by single cell cloning in Dulbecco's modified Eagle's medium for 3 weeks under selection with 500 µg/ml G418 in the same medium and analyzed for their specific binding of [125I]HPP-HOE140 (2100 Ci/mmol). Clones with the highest binding capacity were chosen for further experiments.
Cloning of Recombinant Baculovirus and Infection of Sf9 CellsRecombinant baculoviruses were generated by co-transfection
of Sf9 cells with pVL1392-OKR and linearized AcMVPV by the lipofection method in 60-mm dishes following the instructions of the manufacturer (Invitrogen). Two days after transfection varying dilutions
(103 to 10
6) of the harvested medium were
transferred to fresh Sf9 cells held in 96-well plates, and positive
viral clones were isolated. The subcloned recombinant viruses were
identified by their ability to direct the increase of
[Ca2+]i upon stimulation with ornithokinin or
HOE140 of Sf9 cells infected with the recombinant baculovirus.
[Tyr0]Ornithokinin or HPP-HOE140 was radiolabeled by the IODO-GEN method (Pierce) with previously reported variations (22). The peptides were dissolved in 100 µl of 10 mM sodium phosphate, 150 mM NaCl, pH 7.4 (final concentration 5 µ/ml), applied to a reaction tube coated with 100 µg of IODO-GEN, and 1 mCi of Na125I in 10 µl was added. Unreacted iodine was separated by centrifugation of the reaction mixture over Dowex-1 column equilibrated with 10 mM sodium phosphate, 150 mM NaCl, pH 7.4. The specific activity of [125I-Tyr0]ornithokinin was approximately >500 Ci/mmol.
Radioligand Binding to Intact CellsThe radioreceptor
assays were done as described (23) with minor modifications. Briefly,
clonal CHO-K1 cells recombinantly expressing the ornithokinin receptor
were grown in 96-well plates, rinsed 3 × with 0.15 M
NaCl, 0.1 M phosphate, pH 7.4 (phosphate-buffered saline),
and incubated for 4 h at 0 °C in 100 µl of RPMI 1640, 0.5%
bovine serum albumin, 1.5 mM -D-glucose
(assay buffer) containing 1 nM
[125I]HPP-HOE140 (2100 Ci/mmol) or 1 nM
[125I-Tyr0]ornithokinin (approximately >500
Ci/mmol) and competing peptides at the indicated concentrations. After
3 rapid washes with 200 µl each of ice-cold phosphate-buffered
saline, the cells were solubilized with 0.5 N NaOH, 1%
(w/v) SDS, and the amount of bound radioligand was determined in a
-counter. Sf9 cells infected for 12 h were washed 3 × with phosphate-buffered saline, collected by centrifugation (1,500 × g for 10 min at 20 °C), resuspended in assay buffer,
and incubated for 2 h at 0 °C. Bound ligand was separated from
free ligand by filtration over a Whatman 6F/C filter followed by 3 washes with phosphate-buffered saline and
-counting.
Determination of inositol triphosphate hydrolysis was followed by the original method (24) as modified previously (25). Clonal CHO cells were grown to confluence in 24-well plates and labeled with myo-[2-3H]inositol (1 µCi/ml) for 12 h in serum-free medium. Following incubation with 10 mM LiCl for 15 min the cells were stimulated at 37 °C with various ligands at the indicated concentrations for 10 min. The released inositol phosphate was purified by anion exchange chromatography (Dowex AG 1-8X) and quantified by liquid scintillation counting.
[Ca2+]i Measurement by Fura-2/AMIntracellular Ca2+ concentrations were determined by the fluorescent Ca2+ chelating agent, fura-2/AM according to our previous protocols (3, 26) except for the following. Clonal CHO cells expressing ornithokinin receptor were grown on glass coverslips to confluence loaded with 2 mM fura-2/AM for 30 min at 37 °C. Sf9 cells expressing the ornithokinin receptor were incubated with loading solution for 1.5 h at 27 °C. The cells were challenged with varying ligand concentrations as indicated, and the fluorescence emission was recorded using a HITACHI spectrophotometer. To assess the role of phospholipase C in the [Ca2+]i transients, cells were incubated with 4 µM U73122, a phospholipase C inhibitor, prior to ligand challenge.
Assay of [35S]GTPVarious amounts
of ligand (HOE140 or ornithokinin) were diluted in 50 µl of ice-cold
reaction buffer (50 mM Hepes, pH 8.0, 1 mM
EDTA, 5 mM MgCl2, 1 µM GDP, 1 mM dithiothreitol) containing 20 nM
[35S]GTPS (1300 Ci/mmol) prior to the addition of 50 µl of membranes (10-50 µg of protein) suspended in ice-cold
reaction buffer. The mixed solution was incubated at 37 °C for 10 min. The reaction was stopped by rapid filtration using a Scatron cell
harvester operated with ice-cold washing buffer. The amount of
protein-bound [35S]GTP
S was determined by incubating
the filters in scintillation solution prior to scintillation counting
(total binding). Unspecific binding of [35S]GTP
S was
determined in the presence of 100 µM GTP
S. Specific binding was calculated as the difference of total and unspecific binding.
Competition binding experiments with
125I-[Tyr0]ornithokinin using primary CEF
revealed a single high affinity binding site for ornithokinin with an
apparent IC50 value of 5.5 nM (Fig.
1A). To further characterize this binding
site we applied various kinin peptides (Table I). The
IC50 values were 206 nM for
[Ser6]ornithokinin, an ornithokinin-bradykinin hybrid
which corresponds to [Leu8]bradykinin, >10
µM for the B2 receptor agonist, bradykinin, >10 µM for the B1 receptor agonist
[des-Arg9]bradykinin, and >10 µM for the
B1 receptor antagonist,
[des-Arg9,Leu8]kallidin (Fig. 1A).
The principal B2 receptor antagonist, HOE140, displaced radiolabeled
ornithokinin most effectively with an IC50 of 0.9 nM (Fig. 1A). To further address this unexpected
binding we employed 125I-labeled HPP-HOE140 (Fig.
1A). The IC50 values were 1.5 nM for HOE140, 44 nM for ornithokinin, 1.0 µM for
[Ser6]ornithokinin, and >10 µM for
bradykinin, [des-Arg9]bradykinin and
[des-Arg9,Leu8]kallidin (Fig. 1A).
These results indicate that CEF expose high affinity binding site(s)
for ornithokinin and HOE140. Our results do not allow discrimination
whether ornithokinin and HOE140 are binding to an identical binding
site or to two overlapping binding sites within the same receptor
molecule.
|
To analyze the ability of the various ligands to induce an
intracellular signal typical of kinin receptors we studied the ligand-induced stimulation of inositol phosphate hydrolysis in CEF. The
apparent EC50 values were 0.2 nM for HOE140,
0.3 nM for ornithokinin, 14.7 nM for
[Ser6]ornithokinin with a maximum stimulation of about
5.5-fold above baseline. Bradykinin was much less effective, evoking a
half-maximal response at 187 nM with a significantly
smaller total increase (<2-fold, Fig. 1B). Hence the
putative ornithokinin receptor is functionally coupled to the
phospholipase C pathway of CEF, and both ornithokinin and HOE140 are
efficient ligands for inositol triphosphate mobilization. Application
of 100 nM bradykinin to CEF did not change the
concentration of intracellular free calcium ([Ca2+]i) whereas the subsequent stimulation with
100 pM ornithokinin or with 100 pM HOE140
triggered a [Ca2+]i transient (Fig.
2). Preincubation with 4 µM of the PLC
inhibitor U73122 completely abolished cytoplasmic calcium increase
(data not shown). Thus CEF expose a high affinity kinin-binding site,
putative ornithokinin receptor, which is coupled to the phospholipase
C/inositol triphosphate/Ca2+ pathway. Ornithokinin and
HOE140 are efficient agonists and [Ser6]ornithokinin is a
weak agonist, whereas bradykinin and truncated versions thereof do not
have (ant)agonistic activity on the chicken receptor.
Cloning and Sequencing of the Ornithokinin Receptor
To
isolate the gene encoding the ornithokinin/HOE140-binding site, two
degenerated oligonucleotide primers were devised, based on two well
conserved segments corresponding to the second intracellular domain,
ID2 (primer H669) and the sixth transmembrane region, TM6 (primer
H1110), of the human, murine, and rat B2 receptor cDNA sequences.
Using this pair of primers we PCR-amplified a 376-bp fragment from
chicken genomic DNA. Northern blot analysis using the radiolabeled PCR
fragment revealed a single transcript with an apparent size of about
3.3 kilobases in poly(A)+ RNA from CEF (Fig.
3A). No such signal was detectable in the poly(A)+ RNA fraction of MC29 transformed
chicken macrophages (Fig. 3A) that lack ornithokinin-binding
sites. Concomitant with a decrease in the ornithokinin-binding sites
the mRNA signal was lost when CEF went through eight passages (Fig.
3A). Southern blot analysis indicated that the 376-bp
fragment forms part of a single copy gene (Fig. 3B).
We then used the 32P-labeled PCR fragment to search a
genomic chicken DNA library for the complete ornithokinin receptor
gene. Ten individual clones that hybridized under high stringency
conditions were isolated and purified. The inserts of the isolated
phage DNAs varied in length between 15 and 24 kilobases. Four clones contained a 5.6-kilobase BamHI fragment that strongly
hybridized to the 376-bp PCR fragment; a BamHI fragment of
identical size was also present in the Southern blot (Fig.
3B). This BamHI fragment was subcloned and
sequenced using H669 and H1110 for priming. In this way a DNA sequence
with an uninterrupted open reading frame of 1341 bp was identified
(Fig. 4). Within the 201 bp 5 to this reading frame,
three in-frame ATG codons are present that could serve as translation
initiation site(s) for proteins of 447 (OKR447), 421 (OKR421), and 381 (OKR381) residues, respectively. This situation is reminiscent of the
mammalian B2 receptor where the NH2-terminal domain does
not influence the ligand binding and signaling properties of the
receptor (26). Because we have failed to isolate sufficient amounts of
the purified receptor we presently cannot define the precise
translation site(s).
Hydropathy analysis for OKR381 predicts seven putative transmembrane
domains consistent with a G protein-coupled receptor (Fig. 4). The
predicted protein has five potential N-linked glycosylation sites, and four of them are located in the first extracellular domain,
ED1 (NH2 terminus), and a single one in the third
extracellular domain, ED3 (second extracellular loop). Consensus
phosphorylation sites for cAMP-dependent protein kinases
and protein kinase C are present in all of the intracellular domains,
ID1-ID4 (Fig. 4). The longer variants, OKR421 and OKR447, are
characterized by NH2-terminal extensions of 40 and 66 residues as compared with OKR381, and contain another hydrophobic
stretch at their extreme NH2-terminal regions (not shown).
Comparison of the amino acid sequences of the chicken OKR381 receptor
with other G protein-coupled receptors reveals the highest sequence
identity with mammalian B1 (27-31%) and B2 receptors (46-49%); all
other known G protein-coupled receptors share less sequence identity
(<27%). These relationships are compatible with a phylogenetic tree
that sets the chicken receptor apart from both the mammalian B1 and B2
receptors, respectively (Fig. 5).
Functional Expression of the Chicken Ornithokinin Receptor in CHO and Sf9 Cells
To analyze the pharmacological properties of the chicken receptor we amplified the 1341-bp coding sequence by PCR and subcloned it into the pCDNA3 vector for stable expression in CHO cells; which do not possess endogenous ornithokinin-binding sites (not shown). Of 12 stably transfected CHO clones showing similar high affinity binding for ornithokinin, clone pCDNA3-OKR1 was chosen for detailed analyses. The exogenously expressed OKR447 was almost indistinguishable in the ligand binding features from the receptor expressed endogenously in CEF, as revealed by competition binding experiments using 125I-labeled ornithokinin. Apparent IC50 values of 1.8 nM were determined for HOE140, 4.7 nM for ornithokinin, 147 nM for [Ser6]ornithokinin, and >10 µM for bradykinin (Fig. 1C). Likewise [des-Arg9]bradykinin and [des-Arg9]ornithokinin did not compete, IC50 > 10 µM (Table II). Application of 125I-labeled HOE140 yielded IC50 values of 3.8 nM for HOE140, 38 nM for ornithokinin, >10 µM for [Ser6]ornithokinin and bradykinin (Fig. 1), [des-Arg9]bradykinin, and [des-Arg9]ornithokinin (Table II). The recombinantly expressed ornithokinin receptor was coupled to the phospholipase C/inositol triphosphate: the corresponding EC50 values were 0.8 nM for HOE140, 0.6 nM for ornithokinin, and 21.6 nM for [Ser6]ornithokinin with a maximum stimulation of 14-fold above baseline (Fig. 1D). In the same setting bradykinin induced only a 2-fold increase with an apparent EC50 of 138 nM. Finally, incubation of pCDNA3-OKR1 with 100 nM bradykinin failed to induce a [Ca2+]i transient whereas the subsequent stimulation with 100 pM HOE140 or ornithokinin induced a significant cytoplasmic Ca2+ increase (Fig. 2). From this data we conclude that the cloned chicken receptor expressed in CHO cells has a pharmacological profile that is strikingly similar to that of the endogenous ornithokinin receptor from CEF. Both the full-length cDNA form including three AUG codons and a truncated version thereof holding only the third AUG site gave rise to functionally equivalent receptors when recombinantly expressed in CHO cells (data not shown). This situation is reminiscent of the mammalian B2 receptor where the NH2-terminal domain does not influence the ligand binding and signaling properties of the receptor (26).
|
To further prove that the pharmacological properties are an intrinsic function of the cloned receptor protein independent of its cellular background we expressed the cloned receptor in an insect cell line, Sf9. Two baculovirus clones containing the 1341-bp fragment containing the entire open reading frame of the receptor were constructed. Both clones directed the expression of high affinity binding sites for 125I-labeled HOE140 in Sf9 cells (data not shown). Using one of these clones (pVL1392-OKR) we found that both ornithokinin and HOE140 at 100 pM but not bradykinin at 100 nM were able to induce a transient increase in [Ca2+]i (Fig. 2). Hence we conclude that we have cloned, expressed, and characterized the authentic, bioactive kinin receptor of the chicken.
Induction of [35S]GTPThe ornithokinin receptor
indirectly stimulates accumulation of inositol phosphates, probably by
activation of PLC. To analyze the direct activation of G proteins by
the ornithokinin receptor we measured [35S]GTPS
binding in membranes prepared from CEF and CHO-OKR477 cells.
[35S]GTP
S binding was induced in a
dose-dependent fashion by HOE140 and ornithokinin showing a
half-maximum stimulation at a ligand concentration of about 10 nM (Fig. 6). The differences observed between CEF and CHO-OKR477 [35S]GTP
S binding are
probably due to the differences in the receptor copy number expressed
in these cells. In conclusion the ability of the ornithokinin receptor
to stimulate G protein activation classifies the ornithokinin receptor
as a member of the family of G protein-coupled receptors.
In this study we have cloned and characterized a hitherto unknown avian kinin receptor that is a member of the family of G protein-coupled receptors; we refer to it as ornithokinin (OK) receptor. The OK receptor is functionally coupled to the phospholipase C pathway, and mediates the transient increase of free cytoplasmic Ca2+ in various cell types. This signal transduction activity of the OK receptor is pertussis toxin-insensitive since preincubation with up to 100 ng/ml pertussis toxin had no effect on PLC activation (data not shown). These results are in good agreement with other reports on bradykinin induced signal transduction pathways. Goodemote et al. (27) showed a pertussis toxin-independent Ca2+ release in Swiss 3T3 cells and Wilk Blaszczak et al. (28) reported that B2 receptor coupling to ion channels is mediated by pertussis toxin-insensitive Gq and/or G11 G proteins in neuroblastoma cells. Whereas Liebmann et al. (29) reported a B2 receptor coupled to pertussis toxin-sensitive G proteins of the Gi family in rat myometrium membranes. In conclusion kinin receptors are obviously coupled to different G proteins depending on the signal transduction pathway, the type of cell, and the species analyzed.
The pharmacological profile of the OK receptor is exceptional among kinin receptors: [Ser6,Leu8]bradykinin, i.e. ornithokinin, has high affinity and efficacy, and [Leu8]bradykinin ([Ser6]ornithokinin) has a moderate affinity whereas the "classical" B1 and B2 agonists, [des-Arg9]bradykinin and bradykinin show low if any affinity for the OK receptor. Most remarkably, the principal antagonist of mammalian B2 receptor, HOE140, is a full agonist of the OK receptor. Our preliminary experiments showing a transient dilatation of chick aortic rings by ornithokinin and HOE140 but not by bradykinin (data not shown) are in line with our in vitro data. We therefore propose that this receptor represents a third kinin receptor subtype, operationally defined here as the "B0" type ("0" for ornithokinin). Table III lists the typical rank orders of agonist potency for this B0 receptor subtype in comparison to the mammalian kinin receptor subtypes B1 and B2.
|
A particular striking feature of the OK receptor is its full activation
by the principal B2 receptor antagonist HOE140. Few other G
protein-coupled receptor families have been identified so far where
ligands serve such a dual role, i.e. as full agonist of one
subtype and as full antagonist of another. For instance, the compound
ICI-215,001 is an antagonist for
1-/
2-adrenoreceptors but an agonist for
3-adrenoreceptor (30). Substance PD-136,450 is an
antagonist for type B but an agonist for type A cholecystokinin receptor (31). Both substances have a considerably lower agonistic efficacy than the authentic ligand, rendering the OK receptor a truly
exceptional case since on this receptor HOE140 is at least equipotent
with the authentic ligand, ornithokinin. Furthermore, ICI-215,001 and
PD-136,450 are organic compounds likely to bind to a receptor site
distinct from the agonist-binding site. In contrast, HOE140 is a
peptide derivative (13) that binds to a site overlapping or even
congruent with the kinin-binding site (26). For these reasons, the OK
receptor may be a unique tool to construct chimeric receptors between
the various kinin receptor subtypes, functional analysis of which will
provide important clues to the identification of structural elements
that convey the remarkable ligand selectivity of kinin receptors. By a
similar chimeric approach, one may also be able to define the
structural changes, which the receptors undergoes in response to either
agonists or antagonists.
Our notion that the OK receptor represents the first member of a novel class of kinin receptors is supported by the fact that the sequence identity between the B0 subtype and the members of the B1 and B2 families is moderate to low. The sequence identity between the OK receptor and the mammalian B2 receptors averages 43% at the nucleotide level and 48% at the protein level. This is in sharp contrast to the amino acid sequence conservation between other mammalian and chicken receptors, e.g. the muscarinic receptor M2 with 85% (32), the purinergic receptor P2YR with 84% (33, 34), and the melatonin receptor 1a with 79% (35, 36) sequence identity. Notably the intra-species sequence identity between human B1 and B2 receptors is also low, i.e. 36% (DNA) and 54% (protein) (3). Hence it appears that the kinin receptor subtypes have considerably diverged during evolution, a fact that has hampered the efforts to expand the spectrum of cloned kinin receptors by homology screening. The low homology of the OK receptor gene to mammalian B1/B2 receptor may also explain our initial failure to isolate the OK receptor gene by low stringency hybridization using a human B2 receptor cDNA probe (not shown), requiring to employ primers from those receptor portions that are best conserved among the various subtypes (>80% identity at the nucleotide level). Additional reasons for the failure to clone the OK receptor cDNA were probably in fact the initially low mRNA expression rate in CEF which showed a dramatic decrease during repeated cell culture passages; a loss in receptor expression during primary cell culture has been described for various receptor types (37, 38).
Identification of a novel kinin receptor in chicken raises the questions of whether (i) there is a mammalian B0 counterpart, and (ii) whether the chicken genome bears B1 and/or B2 counterparts? Low stringency hybridization of human genomic DNA using the OK receptor DNA failed to reveal a specific gene fragment (data not shown). By the same procedure and using a B2 receptor probe we have been unable to identify a B2-like gene in the chicken genome (see above); so far we did not screen for B1-like genes. These findings are not unexpected because man and other mammals lack the ornithokinin sequence; rather they express the canonical bradykinin sequences with variable NH2 termini. Conversely the chicken has only a single kininogen type, i.e. ornithokininogen, that holds the ornithokinin sequence (18) and lacks the bradykinin sequence.
The exquisite ligand specificity is a hallmark of kinin receptors. A
striking example of this is the conservative exchange of Thr Ser at
position 6 of ornithokinin, i.e. the deletion of a single
methylene group, which lowers the potency of
[Ser6]ornithokinin by a factor of 25 compared with that
of the parental peptide, ornithokinin. The reverse exchange, Ser
Thr, occurs naturally, e.g. [Thr6]bradykinin
is found in wasp, frog, or turtle; this Ser
Thr exchange at
position 6 leaves the activity of the variant kinin almost unchanged
(39). Hence it appears that the B0 subtype is much more sensitive to
changes at position 6 than the B2 subtype. A unifying feature of all
kinin receptor subtypes is their remarkable sensitivity to changes in
position 8 where a Phe
Leu exchange transforms the B1 agonist
[des-Arg9]bradykinin into a potent B1 antagonist,
[des-Arg9,Leu8]bradykinin. The same exchange
converts the B2 agonist [Ser6]bradykinin into a B0
agonist, [Ser6]ornithokinin. It will be interesting to
see whether a Phe
Oic ([3aS,7aS]octahydroindol-2-carboxyl)
substitution at position 8 which is present in HOE140 (see Table I)
will convert bradykinin into an OK receptor agonist. In addition, the
above mentioned experiments using chimeric chicken/mammalian receptors
may help to unravel the structural basis of the exquisite specificity
of kinin receptors.
We thank Drs. B. Schölkens and H. G. Eckert, Hoechst AG, Frankfurt, for HOE140 and [125I]HPP-HOE140, and Dr. Gotthold Schaffner and Roland Kurzbauer, Institute of Molecular Pathology, Vienna, for DNA sequencing, and L. Gibson for critically reading the manuscript.