`Vol. 23, No. 1, pp. 31–42
`© 2005 Neva Press, Branford, Connecticut
`
`The Preclinical Pharmacology
`of BIBN4096BS,
`a CGRP Antagonist
`
`Debbie L. Hay and *David Poyner
`
`School of Biological Sciences, University of Auckland, Auckland, New Zealand and
`*School of Life and Health Sciences, Aston University, Birmingham, UK
`
`Keywords: Adrenomedullin — BIBN4096BS — Calcitonin receptor —
`CGRP — CGRP antagonist — CGRP receptor — CGRP8–37 — Migraine.
`
`ABSTRACT
`
`CGRP is an important neuropeptide found throughout the cardiovascular system.
`However, until recently it has been difficult to define its pharmacology or physiological
`role because of the lack of suitable antagonists. BIBN4096BS is a high-affinity, non-
`peptide antagonist that shows much greater selectivity for human CGRP1 receptors com-
`pared to any other drug. Its pharmacology has been defined with studies on transfected
`cells or cell lines endogenously expressing receptors of known composition. These have
`allowed confirmation that in many human blood vessels, CGRP is working via CGRP1 re-
`ceptors. However, it also interacts with other CGRP-activated receptors, of unknown com-
`position. In vivo, clinical studies have shown that BIBN4096BS is likely to be useful in
`the treatment of migraine. It has also been used to define the role of CGRP in phenomena
`such as plasma extravasation and cardioprotection following ischemia.
`
`INTRODUCTION
`
`BIBN4096BS is a new calcitonin gene-related peptide (CGRP)-selective antagonist.
`As a stable, non-peptide antagonist of high affinity and selectivity it represents a marked
`improvement on the previous CGRP antagonist, CGRP8–37 and as such, it may be ex-
`pected to greatly facilitate the study of this peptide. It is also starting to leave its mark in
`the clinic, for the treatment of migraine. This article reviews its pharmacology.
`
`CHEMISTRY
`
`The structure of BIBN4096BS (1-piperidinecarboxamide, N-[2–[[5-amino-1-[[4-(4-
`pyridinyl)-1-piperazinyl]carbonyl]pentyl]amino]-1-[(3,5-dibromo-4-hydroxyphenyl)me-
`thyl]-2-oxoethyl]-4-(1,4-dihydro-2-oxo-3(2H)-quinazolinyl)) is shown in Fig. 1. This
`
`Address correspondence and reprint requests to: Dr. David Poyner, School of Life and Health Sciences, Aston
`University, Birmingham, B4 7ET, UK.
`Tel: +41 (121) 204-3997, Fax: +41 (121) 359-5142, E-mail: D.R.Poyner@aston.ac.uk
`
`31
`
`EX2068
`Eli Lilly & Co. v. Teva Pharms. Int'l GMBH
`IPR2018-01425
`
`1
`
`
`
`32
`
`D. L. HAY AND D. R. POYNER
`
`Br
`
`OH
`
`Br
`
`O
`
`N
`
`N
`
`N
`
`H N
`
`O
`
`O
`
`N H
`
`N
`
`N
`
`O
`
`N
`
`H
`
`FIG. 1. Structure of BIBN4096BS.
`
`NH2
`
`drug was developed at Boehringer Ingelheim Pharmaceuticals, Inc. from a dipeptide lead.
`It is a white solid substance with a molecular weight of 867. Boehringer Ingelheim scien-
`tists developed also a radiolabeled version of the drug, [3H]BIBN4096BS. BIBN4096BS
`is best dissolved in 1 M HCl, thereafter it may be diluted with buffer and then adjusted to
`pH 6.5–7.0 with 1 M NaOH. It is soluble in aqueous solution at concentrations well in
`excess of 1 mg/mL. The solution may be conveniently stored in frozen aliquots.
`
`PHARMACOKINETICS
`
`The pharmacokinetics of BIBN4096BS following i.v. infusion in humans has been de-
`scribed in several reports (22,33,49). Its distribution is best described by a three-com-
`partment model, where a central compartment with a volume of distribution of 8.4 L is in
`equilibrium with shallow and deep compartments of 4.4 and 15.6 L (49). The non-com-
`partmental studies estimated the volume of distribution as about 20 L (22). Overall plasma
`clearance was approximately 12 L/h. As renal clearance was only 2 L/h, the authors con-
`cluded that the kidney played only a minor role in the removal of the unmetabolized drug
`(22). The pharmacokinetics did not appear to depend on the dose of drug (22,33).
`
`PHARMACOLOGY OF CGRP RECEPTORS
`
`Introduction
`
`To understand the pharmacology of BIBN4096BS, it is necessary to briefly provide
`some background information on the role of CGRP and allied peptides in the cardiovas-
`cular system and their pharmacological profiles.
`
`CGRP and Related Peptides in the Cardiovascular System
`
`CGRP is a 37 amino acid peptide; it is the main neurotransmitter released by capsai-
`cin-sensitive sensory nerve fibers (C-fibers). It is found in most branches of the cardiovas-
`cular system where it exerts a wide range of effects (4). Its receptors are found throughout
`the cardiovascular system (4,16,29). It has positive chronotropic and inotropic effects on
`the heart and is a very potent vasodilator. Activation of C-fibers during myocardial
`ischemia releases it, causing local vasodilation and reducing the effects of any infarction
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`2
`
`
`
`BIBN4096BS PHARMACOLOGY
`
`33
`
`(16). CGRP release also plays a major part in the cardioprotective effect of ischemic pre-
`conditioning (28). In congestive heart failure, CGRP infusion increases cardiac output,
`mainly due to vasodilation (44). In chronic stable angina, it increases exercise tolerance
`(5,50). CGRP is also of potential use in peripheral circulatory diseases. Beneficial effects
`have been reported in Raynaud’s disease where it produces a long-lasting increase in hand
`skin blood flow and promotes healing of ulcers (6,42). Early clinical trials involving
`CGRP to treat the vasospasm that follows subarachnoid hemorrhage gave promising re-
`sults but were not followed up as the peptide frequently caused hypotension. However,
`recent gene transfer experiments in animal models have given positive results, suggesting
`that the peptide will be useful if it can be targeted appropriately (15,48). Gene transfer of
`CGRP in animal models has also been used successfully to treat pulmonary hypertension
`(7). CGRP overproduction can be significant in a number of pathological conditions. In
`septic shock, CGRP release is triggered by endotoxin and this results in inappropriate
`vasodilation (1). CGRP is also released in neurogenic inflammation; the resulting ex-
`cessive vasodilation may be important in conditions such as migraine and chronic pelvic
`pain (27,45). Studies on CGRP knockout mice have not always given consistent results;
`there are reports that some animals have elevated blood pressure but its importance in
`physiological (as opposed to pathophysiological) conditions is at best unclear (34).
`CGRP is related to the peptide adrenomedullin (AM) (20). Human (h) AM is a 52
`amino acid peptide; however, there is about 25% homology with CGRP over residues
`13–52 and this has essentially the same biological activity as full length AM. There is
`cross-reactivity between CGRP and AM at their receptors (8). AM functions as an auto-
`crine/paracrine factor and is released by vascular endothelial cells. It has similar pharma-
`cological properties to CGRP on the heart and vasculature. Data from AM knockout mice
`suggest that it is important in the physiological control of blood pressure, blood flow and
`vascular development (21). In severe heart failure in man, there is a reduction in coronary
`and pulmonary AM release (20). Components of the receptors for both CGRP and AM are
`also reduced in models of heart failure (20).
`CGRP and AM are also related to two other peptides, calcitonin (CT) and amylin
`(AMY). Both of them have distinct receptors although they are of minor significance in
`the cardiovascular system. However, CGRP can activate AMY receptors, at least pharma-
`cologically. An additional recently discovered member of this family is intermedin (IMD)
`or adrenomedullin 2 (AM2) (39,47). It can activate CGRP, AM and AMY receptors.
`
`Is the CGRP Receptor Heterogenic?
`
`The serious study of CGRP receptor pharmacology began with the discovery that N-
`terminally truncated fragments of CGRP, particularly CGRP8–37, could act as antagonists
`(9,11,12). Quirion and co-workers observed that some receptors (typified by those on the
`rat atria) were antagonized more potently by these fragments than those on the tissues
`such as the guinea pig vas deferens. By contrast, ring-opened derivatives of CGRP such as
`(Cys[ACM])2,7-CGRP and (Cys[Et])2,7-CGRP were more potent agonists at the vas defe-
`rens than at the atria. The antagonist sensitive receptors were designated the CGRP1-sub-
`type, as opposed to the CGRP2-subtype typified by the vas deferens (11,23).
`/CGRP2 receptor classification has been very fruitful in furthering re-
`The CGRP1
`search. There is abundant evidence that CGRP can interact with several receptors, at least
`pharmacologically if not physiologically. However, it is questionable whether a single
`CGRP2 receptor exists. A wide range of pA2 estimates for CGRP8–37 against different
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`3
`
`
`
`34
`
`D. L. HAY AND D. R. POYNER
`
`tissues has been reported (see ref. 38 for review). (Cys[ACM])2,7-CGRP is a partial ag-
`onist, which complicates its use in receptor classification (53). Whilst the CGRP1 receptor
`has been defined molecularly, the CGRP2 receptor remains elusive. Accordingly, it has
`been suggested that the “CGRP2 receptor” may represent the action of CGRP at certain
`subtypes of AM or AMY receptor which have a high affinity for CGRP but a low affinity
`for CGRP8–37 (38). This is an area of controversy.
`
`Molecular Pharmacology of CGRP and AM
`
`The CGRP1 receptor is a G-protein coupled receptor (GPCR), but most unusually it is a
`heterodimer. The 7-transmembrane component is termed calcitonin receptor-like receptor
`(CRLR or CL). This is a secretin-like GPCR. However, by itself it will not respond to
`CGRP. It interacts with a single transmembrane accessory protein called receptor activity
`modifying protein 1 (RAMP1). This gives a complex that corresponds to the CGRP1 re-
`ceptor (29).
`There are two proteins homologous to RAMP1: RAMP2 and RAMP3. RAMP2/CL
`gives the AM1 receptor; this has high selectivity for AM over CGRP. RAMP3/CL gives
`the AM2 receptor; this shows less discrimination between AM and CGRP, particularly
`â-CGRP (23,29). Both of these receptors have a low affinity for CGRP8–37. RAMPs will
`also associate with other GPCRs, particularly the CT receptor. Here they produce three
`AMY receptors, depending on which RAMP is involved. The AMY1 (RAMP1/CT-re-
`ceptor) and to some extent, the AMY3 (RAMP3/CT-receptor) have significant affinities
`for CGRP. These appear to have low affinity for CGRP8–37 but can be activated by analogs
`of (Cys[ACM])2,7-CGRP. Thus they could contribute to “CGRP2” receptor pharmacology,
`although other factors may also be at work (38).
`
`Receptor Pharmacology of BIBN4096BS
`
`The interaction of BIBN4096BS with CGRP receptors
`
`An interesting property of BIBN4096BS is the great selectivity it shows for human (or
`more strictly, primate) receptors over rodent receptors (see below) (13). This property has
`been used to study its interaction with CGRP1 receptors. Receptors formed from mixed
`human/rat RAMP1 and CL demonstrated that the species selectivity was determined by
`the RAMP. Site directed mutagenesis has shown that it can be traced to a single amino
`acid at position 74; this is tryptophan in humans and lysine in rats. Presumably the trypto-
`phan stabilizes a hydrophobic interaction required for high affinity BIBN4096BS binding
`(30). It is not known whether there is a direct interaction between RAMP1 and the antag-
`onist; this is possible but equally it might promote a fold in CL necessary for binding. Re-
`gardless of the role of RAMPs, it has been reported that BIBN4096BS interacts with CL
`(40). It is usual to find that there is at best only partial overlap between the binding sites
`of peptide and non-peptide ligands at GPCRs and this would be expected to be true of
`BIBN4096BS. As will be seen below, there are some reports of non-competitive anta-
`gonism with BIBN4096BS; until the binding site is mapped in more detail it is difficult to
`know if they are consequences of examining the antagonist under non-equilibrium condi-
`tions or whether there is true non-competitive antagonism.
`The role of RAMP1 in promoting high affinity binding of BIBN4096BS raises the pos-
`sibility that the AMY1 receptor, a complex between the CT-receptor and RAMP1, might
`also have some affinity for the antagonist. This remains to be tested.
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`4
`
`
`
`BIBN4096BS PHARMACOLOGY
`
`35
`
`Pharmacology on recombinant receptors and cell lines
`
`The binding of BIBN4096BS has been studied either directly using [3H]BIBN4096BS
`or indirectly, in competition with [125I]iodohistidyl-háCGRP (Table 2). For
`hCL/hRAMP1 complexes expressed in Cos 7 cells, [3H]BIBN4096BS has a pKd of
`10.05 ± 0.03 (n = 3) (10). This is in fair agreement with a pKi of 10.74 estimated on the
`same complex expressed in 293 EBNA cells in a competition study (30). For rat (r)
`CL/rRAMP, the pKi in competition studies was 8.67 (30).
`Several studies have looked at the binding of BIBN4096BS to SK-N-MC cells. These
`express CL and RAMP1 and provide an endogenous system to study hCGRP1 receptors.
`In direct binding, BIBN4096BS had a pKd of 10.35. This also illustrated that the ligand
`has slow kinetics; at 50 pM it required about 2 h to reach equilibrium and the t1/2 for dis-
`sociation was 357 min (i.e., koff = 0.0018 min–1) (41). The slow dissociation would be ex-
`pected for a high affinity ligand. In competition studies at SK-N-MC cells, pKi values
`cover a range from 11.4 to 10.8, with a mean of 11.2 (Table 2). Pooling all direct and in-
`direct binding data on exogenous or endogenous expressed human CL /RAMP1 com-
`/pKd estimate of 10.8 ± 0.22 (n = 6).
`plexes gives a mean pKi
`/pKd has also been estimated in a number of functional studies on SK-N-MC
`The pA2
`cells (Table 1). The mean value is 10.9 ± 0.22 (n = 3), in excellent agreement with the
`radioligand binding data. In one study (18), the Schild slope was significantly greater than
`one, indicating that the binding was not strictly competitive; the pA2 derived from the x
`intercept on the Schild plot was 9.95. The apparent non-competitive behavior was con-
`sidered to be an artifact due to the slow kinetics of the antagonist; the slope of the plot was
`influenced by the two concentrations at the extreme ends of the range used, where the ki-
`netic effects would be most pronounced. If the slope was constrained to unity, an apparent
`pKb of 10.47 could be calculated.
`/pKb values in two other cell lines (18). Rat L6 cells are
`The same study measured pA2
`a model of the rCGRP1 receptor, expressing CL and RAMP1 (albeit with RAMP2 as
`well). On these cells, the Schild slope (0.89) was just significantly less than unity, giving a
`pA2 of 9.25. As with the SK-N-MC cells, this was considered to be the result of the slow
`kinetics of the antagonist (again, the slope of the Schild plot was largely skewed by the
`very highest and lowest antagonist concentrations) and when constrained to unity, an
`apparent pKb of 9.1 was calculated. Human Col 29 cells have been reported to express a
`CGRP2-like receptor, although the molecular nature of this is unknown. Here, the Schild
`plot had a slope that was not different from unity, allowing a pKb of 9.75 to be calcu-
`lated. This was significantly different from the pKb for BIBN4096BS on SK-N-MC cells.
`BIBN4096BS has also been examined for activity against Rat 2 cells, which are a model
`for rAM1 receptors (CL/RAMP2). It had no antagonist effects at concentrations up to
`10 ìM (18). It was similarly inactive on Cos 7 cells expressing hAM1 and AM2
`(CL/RAMP3) or rAM2 receptors (19).
`In summary, a clear picture has emerged of the pharmacology of BIBN4096BS against
`molecularly defined receptors expressed exogenously or endogenously in cell lines. It is a
`/pKb of 10 to 11) antagonist of hCGRP1 receptors. As might be predicted
`very potent (pKd
`from this affinity, its kinetics are correspondingly slow. It is much less potent (100–1000
`fold) against rCGRP1 receptors. It seems to be able to discriminate between CGRP-acti-
`vated receptors, as evidenced by its lower affinity for the CGRP-responsive receptor on
`Col 29 cells. However it has virtually no affinity for AM1 or AM2 receptors. As will be
`seen below, this profile can also be seen in tissues.
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`5
`
`
`
`36
`
`D. L. HAY AND D. R. POYNER
`
`Tissue/Cell
`
`TABLE 1. pA2/pKb determinations for BIBN4096BS
`/pKb
`
`Comments
`
`pA2
`
`Human
`SK-N-MC cells (CL/RAMP1)
`
`Col 29 cells
`Temporal artery
`
`Cerebral artery
`Coronary artery
`Subcutaneous artery
`Pial artery
`
`Rat
`L6 cells (CL/RAMP1)
`Atrium
`Vas deferens
`
`11.0
`11.2
`10.5
`9.75
`10.1
`10.4
`10.1
`10.4
`11.25
`14.5
`
`9.1
`8.5
`7.2
`8.4
`
`Single dose-ratio
`
`Schild slope > 1
`CGRP2 receptor
`Before KCl contracture
`During KCl contracture
`
`Only competitive at 10 pM
`Biphasic Schild plot slope < 1
`
`Schild slope < 1
`
`vs. (Cys[Et])2,7-CGRP
`
`References
`
`13
`14
`18
`
`50
`
`14
`
`43
`31
`
`18
`55
`
`Other species
`Bovine cerebral artery
`Porcine coronary artery
`
`Porcine cerebral artery
`
`Schild slope < 1
`7.9
`vs. háCGRP
`8.0
`vs. hâCGRP
`6.6
`vs. háCGRP
`7.9
`vs. hâCGRP
`6.7
`Values are all against háCGRP unless otherwise stated. BIBN4096BS was inactive at 10 ìM on Rat 2
`cells (rat CL/RAMP2) and Cos 7 cells transfected with human CL/RAMP2 and rat and human
`CL/RAMP3 (18,19).
`
`31
`54
`
`Source
`
`TABLE 2. pKi/pKd values for BIBN4096BS
`pKi/pKd
`
`
`
`Comments
`
`Human
`CL/RAMP1 (293 EBNA cells)
`CL/RAMP1 (Cos 7 cells)
`SK-N-MC (CL/RAMP1)
`
`Rat
`CL/RAMP1 (293 EBNA cells)
`Brain
`Spleen
`
`10.74
`10.05
`10.35
`10.8
`11.4
`
`8.8
`8.8
`8.5
`
`pKi
`pKd
`pKd
`pKi
`pKi
`
`pKi
`pKi
`pKi
`
`References
`
`30
`10
`41
`13
`14
`
`30
`
`13
`
`Marmoset
`Cortex
`
`Total brain
`
`Dura mater
`Spleen
`
`10.2
`pKd
`10.2
`pKi
`9.9
`pKd
`10.4
`pKi
`10.3
`pKd
`9.7
`pKd
`10.0
`pKi
`pKd values obtained with [3H]BIBN4096BS. pKi values obtained by competition experiments against
`[125I]iodohistidyl-háCGRP.
`
`41
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`6
`
`
`
`BIBN4096BS PHARMACOLOGY
`
`37
`
`The pharmacology of BIBN4096BS in isolated tissues
`
`The direct binding of BIBN4096BS has been measured in marmoset cerebral cortex,
`total brain, dura mater and spleen (Table 2), where pKd values vary from 9.7 to 10.5. There
`is excellent agreement with pKi values measured against [125I]iodohistidyl-háCGRP. The
`authors of the study noted that the lowest pKd and pKi values were both obtained for
`spleen, where, in parallel studies, CGRP8–37 also showed its lowest pKi (41). However, in
`view of the small difference between spleen and other tissues, it would be unwise to over
`interpret this. The marmoset is a primate and in keeping with the data on cell lines and
`transfected cells, the pKi of BIBN4096BS on rat spleen (8.5) was over an order of mag-
`nitude lower than the value for the marmoset. BIBN4096BS has also been used in compe-
`tition studies against [125I]hAM, which labels AM receptors (24). It showed some ability
`to displace the radioligand from rat brain homogenates above 100 nM, but was inactive at
`1 ìM on rat lung and vas deferens membranes. Whatever the explanation for the slight ac-
`tivity on rat brain, it is clear that there is again good agreement between these studies and
`those on cell lines, confirming that it has virtually no affinity for AM receptors examined
`in these studies (although, as discussed below, there is a report that it can selectively an-
`tagonize adrenomedullin in the rat vas deferens) (55).
`
`An important series of experiments have been done on human blood vessels to de-
`termine the pA2 of BIBN4096BS. In temporal, cerebral and coronary arteries, pA2 values
`vary from 10.1 to 10.4 (Table 1). These match the pA2 values found against CL/RAMP1
`complexes. These studies indicate the advantage of using BIBN4096BS over CGRP8–37.
`When the latter was used, pA2 values as low as 6.6 were recorded on the temporal artery; a
`value that would normally be considered as indicating a CGRP2 receptor (50). However,
`RT-PCR studies of human vessels confirmed the presence of CL and RAMP1; and
`BIBN4096BS confirms that CGRP is acting against a CGRP1 receptor. The low affinity of
`CGRP8–37 may be due to problems with peptidases or poor accessibility.
`Numerous studies on human arteries are worth special mention. Against relaxation of
`subcutaneous arteries, competitive behavior was only apparent at 10 pM BIBN4096BS
`(giving a pA2 estimate of 11.26). At higher concentrations, the antagonism appeared non-
`competitive (43). In similar experiments where the CGRP-induced reduction in intracellu-
`lar calcium was measured, non-competitive behavior was observed at all concentrations.
`RT-PCR showed that CL and RAMP1 were present in this artery type, similar to other
`human blood vessels and so it is difficult to imagine that the CGRP receptor is molecular-
`ly distinct. Furthermore, the pA2 is in line with a normal CGRP1 receptor. The authors
`suggest that the slow kinetics of BIBN4096BS may be one explanation of their results;
`they used an equilibrium time of 30 min (43). Whilst the same equilibtium time was used
`in another of their studies where BIBN4096BS showed apparent competitive behavior
`(14), it is an attractive hypothesis to account for their results. The data illustrate the need
`for care in using the compound.
`
`BIBN4096BS has also been investigated on human primary cultures. These experi-
`ments are difficult to interpret as the authors did not measure pA2 or pKb values but in-
`stead observed pIC50 values for blocking the effects of 50 nM CGRP (a supramaximal
`concentration) on cAMP stimulation (32). It is difficult to convert these values into ab-
`solute measurements of receptor affinity. The experiments also measured pIC50 values for
`CGRP8–37, so allowing an approximate BIBN4096BS/CGRP8–37 potency ratio to be cal-
`culated. On smooth muscle, this was around 3000, at endothelial cells it was 30,000 fold
`and at astrocytes two sites were revealed by BIBN4096BS, one where it was 100-fold
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`7
`
`
`
`38
`
`D. L. HAY AND D. R. POYNER
`
`more potent than CGRP8–37 and the other where it was over 106-fold more potent. For
`comparison, at CGRP1 receptors in cultured cells, the expected potency ratio is typically
`around 1000-fold but where the CGRP8–37 affinity is depressed (e.g., temporal artery)
`it can reach 10,000-fold. In the light of these observations, it may be suggested that
`BIBN4096BS was acting at a CGRP1 receptor on smooth muscle and endothelial cells.
`The situation at the astrocytes is more complicated. Interestingly, the same group investi-
`gated the effects of BIBN4096BS on human pial and meningeal arteries (31), where they
`also found evidence for a biphasic action of BIBN4096BS, with a population of sites
`showing sub-picomolar responsiveness. The pial artery data were subjected to Schild
`analysis; this produced a biphasic plot. Although the authors fitted the high-affinity com-
`ponent to obtain a pA2 of 14.5, the Schild slope was still less than unity (0.73) with a
`failure of the antagonist to block the higher concentrations of CGRP. In contrast, on
`bovine cerebral arteries, BIBN4096BS behaved much more normally, with a pA2 of 7.9
`(albeit with a Schild slope slightly less than unity), compared to 6.3 for CGRP8–37. The ar-
`teries almost certainly express CGRP1 receptors; it is difficult to fit these observations of
`ultra-high affinity into the known pharmacology of either these receptors or BIBN4096BS
`and more work is needed to explain them.
`Studies on other species with BIBN4096BS also show how this compound does
`not always give simple results. The rat atrium and vas deferens are classic examples of
`CGRP1 and CGRP2-receptor expressing tissues. The receptors on the atrium have a
`10-fold higher affinity than those on the vas deferens (Table 1), mirroring the affinity dif-
`ference seen with CGRP8–37 (55). The atrial receptor probably represents CL/RAMP1, as
`judged by the pA2 for this complex when expressed on L6 cells. A curious feature re-
`vealed by this study was a second population of receptors on the vas deferens, activated
`by (Cys[Et])2,7-CGRP and AM, at which BIBN4096BS had a higher affinity than it did
`against CGRP. The molecular nature of these receptors is unclear. On porcine arteries a
`different pattern of receptor heterogeneity was apparent (Table 1). BIBN4096BS antago-
`nized háCGRP, (Cys[Et])2,7-CGRP and AM with very similar pA2 values (7.8 to 8.4) but
`was much weaker against hâCGRP (54). Differential antagonism of á and âCGRP has
`also been noted in the human temporal artery (52). In addition, the studies in the atrium
`and vas deferens suggested that both BIBN4096BS and CGRP8–37 had higher affinities
`against háCGRP compared to hâCGRP but the differences were generally small (<0.5 log
`units) (55). It may be relevant that hâCGRP has been reported to show a higher potency
`than háCGRP on complexes such as CL/RAMP3, albeit in a species-dependent manner
`(19). Thus, one explanation for á vs. âCGRP differences may be preferential activation of
`non CL/RAMP1 complexes by bCGRP. This needs to be addressed by construction of
`Schild plots to determine whether the antagonism by BIBN4096BS is likely to involve
`multiple receptors.
`The work on isolated tissues with BIBN4096BS is starting to clarify issues in CGRP
`pharmacology. There is generally good agreement between the pA2 of the antagonist
`found on human arteries and on CGRP1 receptors expressed on cell lines. This agrees with
`the presence of CL and RAMP1 in these tissues and suggests that the low pA2 values
`found for CGRP8–37 may be due to problems of accessibility or stability for this peptide
`antagonist. However, fresh questions are also arising from use of the compound. It has
`slow kinetics. These may explain why in some studies it appears to show non-competitive
`behavior. However, it is also possible that its binding epitope on the receptor may not
`entirely overlap with that of CGRP. It also seems to interact with more than one receptor.
`What is required are more studies on molecularly defined receptors, taking account of
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`8
`
`
`
`BIBN4096BS PHARMACOLOGY
`
`39
`
`species variation. Studies of CT-receptor/RAMP complexes might be particularly infor-
`mative. With this information, it should be possible to make further inroads into defining
`the complex pharmacology of CGRP.
`
`The pharmacology of BIBN4096BS in vivo
`
`BIBN4096BS has been used successfully to define CGRP pharmacology in several
`in vivo models. In the case of migraine, this has lead to successful clinical trials for the
`drug (32).
`In the very first report on BIBN4096BS, it was shown that when injected i.v. into mar-
`mosets, it inhibited the increase in facial blood flow caused by stimulation of the trige-
`minal nerve with an IC50 of 3 ìg/kg (13). Overactivity of the trigeminovascular system
`leading to release of CGRP is believed to be an important factor in the pathogenesis of mi-
`graine. In rats, i.v. BIBN4096BS at 333 ìg/kg can block middle meningeal artery dilation
`induced by transcranial electrical stimulation or systemic á- orâ-CGRP. It has no action
`on pial arteries in the same preparation (35). The authors suggest that this is because it
`cannot cross the blood brain barrier (pial arteries, unlike meningeal arteries, have a
`blood-brain barrier). There is currently no direct data on whether BIBN4096BS can cross
`the blood-brain barrier, although it is predicted to be charged at physiological pH, sug-
`gesting it probably will not penetrate this structure. Vasodilation may not be the primary
`mechanism of action of CGRP in migraine. BIBN4096BS has been used to demonstrate
`the role of CGRP receptors within the trigeminal pathway (46). It is possible to record
`electrical activity from neurons in the cat trigeminocervical complex (i.e., trigeminal
`neurons extending to the caudal brainstem and upper cervical spinal cord). These neurons
`can be activated by stimulation of the superior saggital sinus; the activation is blocked by
`i.v. BIBN4096BS with an ED50 of 31 ìg/kg. If the neurons are stimulated with glutamic
`acid, iontophoretic BIBN4096BS will block the stimulation, suggesting a post-synaptic
`locus of action for the drug. Thus, in migraine the significant action of CGRP may be in
`sensory transmission along the trigeminal nerve.
`The role of CGRP in morphine tolerance has been investigated with BIBN4096BS.
`Repeat intrathecal administration of morphine in rats results in tolerance to its antinoci-
`ceptive effects in the paw pressure (mechanical nociception) and tailflick (thermal noci-
`ception) tests (36). Co-administration of BIBN4096BS reduced the development of
`tolerance in both of these tests. Its effects were broadly similar to those of CGRP8–37, al-
`though it was applied at doses ranging from 0.001 to 0.1 ìg, as opposed to 4 ìg for
`CGRP8–37. Unlike CGRP8–37, it had a small effect to block directly the antinociceptive ef-
`fects of morphine. Morphine increases CGRP expression in cultured adult dorsal root gan-
`glion neurons and this effect is blocked by CGRP8–37 and BIBN4096BS (37).
`CGRP is important in neurogenic inflammation where its most significant action is
`vasodilation, leading to edema. AM is also a potent vasodilator and can lead to edema. It
`is not clear if this action of AM is mediated by its own receptors or CGRP receptors. This
`has been addressed in a murine skin model, where the ability of both peptides to induce
`plasma extravasation was blocked by BIBN4096BS, demonstrating an action at the
`CGRP1 receptor (17). This illustrates how the selectivity of BIBN can be used to inves-
`tigate receptor mechanisms, a topic that was previously very difficult to study in vivo due
`to lack of selective antagonists. The same study also used BIBN4096BS to investigate
`why CGRP produces such a long lasting vasodilation. This could be due to long-term
`changes in vascular reactivity produced by the initial activation of CGRP receptors.
`
`Cardiovascular Drug Reviews, Vol. 23, No. 1, 2005
`
`9
`
`
`
`40
`
`D. L. HAY AND D. R. POYNER
`
`However, BIBN4096BS is able to terminate rapidly established vasodilation, indicating
`that continuous CGRP receptor activation is needed.
`A number of studies have used BIBN4096BS to address the role of CGRP in the car-
`diovascular system. Studies with knockout mice suggest that CGRP plays only a minor
`role in normal physiological control of the circulation. Studies with BIBN4096BS confirm
`this. In the anesthetized pig, BIBN4096BS, at doses of 1 mg/kg, i.v., did not affect heart
`rate, mean arterial blood pressure, systemic vascular conductance, or cardiac output. It
`caused a small decrease in vascular conductance in the lungs, kidneys, spleen and adrenals
`but not the brain, heart, gastrointestinal system, skin or skeletal muscles. By contrast, it
`antagonized the effects of exogenously added CGRP and some actions of capsaicin (which
`releases CGRP from sensory neurons) (25,26). Interestingly, it increased capsaicin-in-
`duced release of CGRP into the plasma, suggesting it might block an autoreceptor that
`normally decreases CGRP release or a reuptake mechanism (25). A similar picture has
`been noted in rats; BIBN4096BS, administered i.v. at 3 mg/kg blocked the cardiovascular
`effects of exogenous CGRP but had little effect by itself (2). Whilst CGRP may play little
`role in the normal physiology of the cardiovascular system, it is probably more important
`in pathophysiology, either as a protective factor or in some cases, as a pathological factor.
`BIBN4096BS has been used in these investigations. Exogenous CGRP reduces the size of
`myocardial infarcts in rats following 60 min of coronary artery occlusion. This effect is re-
`duced by i.v. BIBN4096BS (20 nmol/kg/h) (56). BIBN4096BS by itself had no signif-
`icant effect, suggesting that endogenous CGRP does not effectively protect from this
`degree of insult.
`BIBN4096BS has proved to be a useful drug for in vivo investigations of CGRP. Its
`greater stability and selectivity mean that it is much more useful than the existing peptide
`fragments for probing the role of CGRP and its receptors in physiology and patho-
`physiology. As more becomes known about its receptor selectivity in different species, its
`utility is likely to increase.
`
`SUMMARY
`
`The pharmacology of BIBN4096BS against CGRP1 receptors is now well defined. It
`has been shown to be very useful experimentally, both to define CGRP pharmacology and
`to investigate its role in pathophysiology. It is well tolerated and shows promise in the
`treatment of migraine. For the future, more work is needed to understand how it interacts
`with other CGRP-activated receptors, particularly CT-receptor/RAMP complexes.
`
`Acknowledgments. This work was supported by project grants from the University of Auckland
`Staff Research Fund, the Auckland Medical Research Foundation, Lottery Health Commission (New
`Zealand), the Wellcome Trust, the Biotechnology and Biological Sciences Research Council
`(C20090) and the British Heart Foun