Hallberg 2014
Hallberg 2014
Hallberg 2014
Mathias Hallberg
Beijer Laboratory, Department of Pharmaceutical Biosciences, Division of Biological Research on Drug
Dependence, Uppsala University, Biomedical Center, Uppsala, Sweden
Abstract: The proteolytic processing of neuropeptides has an important regulatory function and the
peptide fragments resulting from the enzymatic degradation often exert essential physiological roles. The
proteolytic processing generates, not only biologically inactive fragments, but also bioactive fragments that
modulate or even counteract the response of their parent peptides. Frequently, these peptide fragments
interact with receptors that are not recognized by the parent peptides. This review discusses tachykinins,
opioid peptides, angiotensins, bradykinins, and neuropeptide Y that are present in the central nervous
system and their processing to bioactive degradation products. These well-known neuropeptide systems
have been selected since they provide illustrative examples that proteolytic degradation of parent peptides
can lead to bioactive metabolites with different biological activities as compared to their parent peptides.
For example, substance P, dynorphin A, angiotensin I and II, bradykinin, and neuropeptide Y are all
degraded to bioactive fragments with pharmacological profiles that differ considerably from those of
the parent peptides. The review discusses a selection of the large number of drug-like molecules that
act as agonists or antagonists at receptors of neuropeptides. It focuses in particular on the efforts to
identify selective drug-like agonists and antagonists mimicking the effects of the endogenous peptide
fragments formed. As exemplified in this review, many common neuropeptides are degraded to a variety
of smaller fragments but many of the fragments generated have not yet been examined in detail with
regard to their potential biological activities. Since these bioactive fragments contain a small number of
amino acid residues, they provide an ideal starting point for the development of drug-like substances
with ability to mimic the effects of the degradation products. Thus, these substances could provide a rich
source of new pharmaceuticals. However, as discussed herein relatively few examples have so far been
disclosed of successful attempts to create bioavailable, drug-like agonists or antagonists, starting from the
structure of endogenous peptide fragments and applying procedures relying on stepwise manipulations
and simplifications of the peptide structures. C 2014 Wiley Periodicals, Inc. Med. Res. Rev., 00, No. 0, 1–57, 2014
Correspondence to: Mathias Hallberg, Department of Pharmaceutical Biosciences, Division of Biological Re-
search on Drug Dependence, The Beijer Laboratory, Uppsala University, Biomedical Center, Box 591, 75124
Uppsala, Sweden. E-mail: [email protected].
1. INTRODUCTION
Neuropeptides are derived from protein precursors (prepropeptides) that usually comprise
100–250 amino acid residues. These prepropeptides which are most often biologically inert are
formed at the ribosomes which are located at the endoplasmic reticulum of peptide-producing
neurons.1 A series of sequence-specific and tissue-specific proteolytic steps, frequently also
with other modifications, subsequently deliver smaller bioactive peptides. The neuropeptides
are defined as chains of amino acids of limited length released from neurons to exert an
effect on target cells. The physiological action of these neuropeptides is not terminated by
specific synaptic reuptake mechanisms, as often is the case for classic neurotransmittors, but
through degradation. This process is mediated by extracellular proteases anchored in the cell
membranes. Neuropeptides and their receptors are expressed not only in neurons but also in
other types of cells in the brain (such as glial cells) and various types of cells in the peripheral
tissues (such as endocrine cells). The degradation of neuroactive peptides by endopeptidases,
aminopeptidases, and carboxypeptidases (which may be more or less specific) can lead to the
formation of fragments that have similar or very different biological activities from those of the
parent peptide.2, 3
This review discusses tachykinins, opioid peptides, angiotensins, bradykinins, and neu-
ropeptide Y and their proteoloytic processing. These well-known neuropeptide systems have
been selected since they represent good examples that illustrate that proteolytic degradation
of parent peptides (e.g., substance P, angiotensin II, etc.) can lead to bioactive metabolites
with other biological profiles than their parent peptides. Such bioactive fragments and their
receptors can frequently be of considerable interest in a drug discovery context. All of these
neuropeptides are present in the central nervous system (CNS). Neuropeptides, and bioac-
tive fragments derived from neuropeptides are not in general suitable as drugs since they are
not bioavailable after oral administration and are not metabolically stable. Thus, the pharma-
cokinetic profiles of small peptides are far from optimal. Neuropeptides and their fragments,
however, can sometimes serve as starting points in design processes aimed at identifying drug-
like substances that interact with their receptors. New selective, metabolically stable drug-like
agonists (peptidemimetics) and antagonists that are able to cross the blood–brain barrier (BBB)
are constantly required.
A representative selection of drug-like molecules that act as agonists or antagonists at recep-
tors of the neuropeptides is discussed herein. The review focuses on the relatively few selective
drug-like ligands that have been identified and found to interact with the receptors/binding sites
of the bioactive degradation products formed from the parent neuropeptides subsequently. The
short degradation fragments have in some cases been successfully utilized and stepwise con-
verted into drug-like entities. The effects in peripheral tissues have attracted the most attention
in several cases.
A. Tachykinins
The most well-known members of the tachykinin family are substance P (SP), neurokinin A
(NKA), and neurokinin B (NKB). The biochemical effects of these substances are mediated by
Medicinal Research Reviews DOI 10.1002/med
METABOLISM TO BIOACTIVE FRAGMENTS r 3
the G-protein-coupled neurokinin receptor subtypes NK1, NK2, and NK3, which are primarily
activated by SP, NKA, and NKB, respectively.4 The neuropeptides, act as neurotransmittors
and neuromodulators and are widely distributed within both the CNS and peripheral tissues.
SP and NKA are the most abundant tachykinins in the brain. The NK1 and NK3 receptors
are expressed at high levels in the adult brain.5 Hemokinin and endokinins represent examples
of new putative tachykinins.6–8 Hemokinin has a similar biological effect to that of SP and is
an agonist primarily for the NK1 receptor subtype.7 The same C-terminal sequence, Phe-X-
Gly-Leu-Met-NH2, is present in SP, neurokinin A, neurokinin B, and hemokinin, where X
is Phe in SP and human hemokinin, Tyr in mouse hemokinin, and Val in neurokinin A and
neurokinin B.
The undecapeptide SP (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2 ), discov-
ered as a neuropeptide by von Euler and Gaddum in 1931,9 is the most studied member of the
tachykinin family. It is characterized by neutral and lipophilic amino acid side chains in the
C-terminal part and two basic amino acid side chains in the N-terminal part.10 Its role in pain
transmission has been studied in greatest depth.11–13 SP and its NK1 receptor are found in such
areas of the brain as the amygdala, septum, hippocampus, hypothalamus, and periaqueductal
gray. These regions of the brain are associated with anxiety and depression.14–24 Furthermore,
the undecapeptide is considered to be involved in the mediation of emesis and SP is present
in the primary sensory afferent fibers and in the dorsal root ganglia. It is found also in the
dorsal horn of the spinal cord, and is a neuromodulator in the primary afferent fibers and
the unmyelinated C-fibers. Notably, it was reported 20 years ago that the undecapeptide does
not mediate pain through the NK1 receptor.25 SP and the other tachykinins are important not
only in various disorders of the brain and in pain processing, but also in the induction and
progression of several inflammatory responses.26–30 It is not surprising, therefore, that antag-
onists to the NK1, NK2, and NK3 receptors have the potential to treat diverse pathological
conditions. These include not only primary pain and CNS disorders such as depression, anx-
iety, schizophrenia, migraine, and emesis, but also disorders that are related to inflammatory
processes such as asthma, arthritis, psoriasis, and inflammatory bowel diseases.15, 31–33
1. Metabolism of Tachykinins
The SP peptide is formed from at least three distinct gene transcripts, the α-, β-, and
γ - preprotachykinin. The peptide is released from these precursors and its C-terminal residue
is subsequently modified by an amidating enzyme.34 The amidated C-terminal is important for
the stability of the peptide. Several different proteases process and degrade SP in the CNS and
CSF. This gives at least six lysine-containing metabolites SP(1–4), SP(1–6), SP(1–7), SP(1–9),
SP(2–11), and SP(3–11).35, 36 The SP(1–4), SP(1–7), SP(1–9), and SP(3–11) peptides have, for
example, been identified in the striatum37 and in the mouse spinal cord, while SP(1–6), SP(1–7),
and SP(1–9) are the major degradation products, together with phenylalanine and two frag-
ments that do not contain lysine, SP(8–9), SP(10–11).36 Angiotensin-converting enzyme (ACE),
neutral endopeptidase (NEP), and substance P endopeptidase (SPE) are all active in the en-
zymatic processing. Prolyl endopeptidase, which cleaves after prolines, cleaves the Pro4 -Gln5
peptide bond. Post proline dipeptidyl aminopeptidase subsequently removes the dipeptides
Arg1 -Pro2 and Lys3 -Pro4 from SP. The major cleaving sites of ACE are the Phe8 -Gly9 and
Gly9 -Leu10 bonds38 but ACE is also acting as a peptidyl dipeptidase, releasing dipeptides
from the remaining N-terminal fragment. Hence ACE can generate, for example, SP(1–7).38, 39
NEP primarily hydrolyzes SP at the Gln6 -Phe7 , Phe7 -Phe8 , and Gly9 -Leu10 bonds.38, 40 Thus,
SP is enzymatically degraded into several fragments, some of which retain their biological
activity.2, 3, 41, 42
Figure 1. Substance P (SP) is not selective but binds preferably to the NK1 receptor. Neurokinin A and B
bind mainly to the NK2 and NK3 receptors, respectively. While the C-terminal fragment of SP activates the NK1
receptor, the N-terminal heptapeptide metabolite substance P (1–7) binds to an unknown receptor, and has often
an opposing and very different biological profile from that of the parent peptide SP.
The C-terminal fragment SP(6–11) has anxiogenic effects after administration to the dorsal
periaqueductal gray in rats.23, 24 The effect is mediated through the NK1 receptor. In contrast,
the N-terminal fragment SP(1–7) did result in the opposite reaction,24, 43 Fig. 1 and Table I.
Notably, the latter effect was reported not to be mediated through the NK1 receptor.44 Thus,
the C-terminal part mimics the effect of the parent neuropeptide SP, while the N-terminal part
does not. The N-terminal heptapeptide SP(1–7) is the major metabolite of substance P in the
rat and has been studied in depth.42, 45, 46 SP can be metabolized to SP(1–7) by SPE, ACE as
well as NEP. SPE-like activity is present in the CSF45 in spinal cord tissue,47 and in various
other areas of the brain.48 Furthermore, the metabolic biotransformation of substance P in
liver microsomes from mouse, rat, and human was recently studied and the data compared.
Five major substance P metabolites, SP(3–11), SP(5–11), SP(6–11), SP(8–11), and SP(1–7) were
identified and quantified.49
Substance P (1–7) administered spinally attenuates thermal hyperalgesia in diabetic mice50
and has several effects that are opposite to those of SP. Thus, SP(1–7) has antinociceptive,51 anti-
inflammatory,52 antihyperalgesic,50 and anxiolytic effects.24, 43 It also attenuates several with-
drawal signs in morphine-dependent rodents53, 54 and the development of morphine tolerance.53
These effects are mediated through a specific receptor for SP(1–7)55–57 that is distinct from any
of the known opioid and tachykinin receptors. It is possible that SP(1–7) acts by interfering
with allosteric sites present on other neuropeptide receptors.
The SP(1–7) receptor has not been cloned nor studied in detail, but considerable efforts
have been devoted to making drug-like compounds with the same attractive pharmacological
profile as SP(1–7), and hence serve as SP(1–7) mimics. Structure-activity relationship studies,
involving an alanine-scan and truncations as well as C- and N-terminal modification of the
Table I. Continued
Peptide Sequence Receptor target Activity type Biological activity
Angiotensin (1–7) Asp-Arg-Val-Tyr-Ile-His-Pro Mas receptor Vasodilation, cardioprotection, decreases hypertropy, and
r HALLBERG
fibrosis.281–283
Angiotensin III Arg-Val-Tyr-Ile-His-Pro-Phe AT1 and AT2 receptor Agonist AT1R and AT2R stimulation: Natriuretic effect, see angiotensin
II.285–287
Angiotensin IV Val-Tyr-Ile-His-Pro-Phe AT4 receptor/ Inhibitor Improves memory and learning,288, 291, 293, 297–299 protection
Insulin-regulated against ischemic stroke294, 295 and hyperglycemia.296
aminopeptidase (IRAP)
Kallidin Lys-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe- Mainly B2 receptor Agonist The receptor constitutively expressed and agonists are
Arg hypotensive, potential cardioprotective, and
proinflammatory.358, 363
Bradykinin Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg Mainly B2 receptor Agonist See kallidin.358, 363
Figure 2. Small ligands with high affinity to the substance P(1–7) binding site. The compounds were identified
after stepwise modifications and truncations of the heptapeptide SP(1–7).
heptapeptide led to improvements of affinity for the SP(1–7) binding sites58 and stronger effects
in behavioural tests than those of SP(1–7).59, 60 The C-terminal part of the heptapeptide was
found most essential for its binding and a primary amide rather than a carboxylic group in the
C-terminal was preferable. By systematic modifications, a series of high affinity binding ligands
were identified, exemplified by the amides 1 and 2,61, 62 Fig. 2. The amide 1 strongly attenuates
diabetic mechanical allodynia och thermal hyperalgesia in mice.63 About 60% of all diabetic
patients experience diabetic neuropathy and the pain of this condition is poorly relieved by
opiates.64 There is a great need for new strategies for the treatment of diabetic neuropathy, and
drug-like SP(1–7) mimetics that are active in the CNS may be a new therapeutic alternative.
Figure 3. Examples of selective and nonselective NK1, NK2, and NK3 receptor antagonists. Aprepitant pre-
vents chemotherapy-induced nausea and vomiting. Selective and dual NK1/NK3 antagonists as well as NK2
antagonists have attracted considerable interest as potential agents against CNS disorders as depression and
schizophrenia. NK2 antagonists and triple antagonists, such as compound 9, could serve as useful pharmaceu-
ticals for the treatment of GI disorders. A fluorophenyl or chlorophenyl group attached to the central part of the
scaffold is often a characteristic feature of NK antagonists.
In 2004, the first selective and very potent nonpeptide NK1 receptor agonists were
disclosed.83 Among those, the quinoline amide 11, derived from a series of structurally re-
lated NK1 antagonists such as 10 that had been disclosed by Takeda,84 exhibited subpicomolar
affinity to the NK1 receptor, Fig. 4. Compound 11 comprises structural elements frequently rec-
ognized in the NK antagonists. It does not resemble the endogenous agonist, the undecapeptide
SP.
3. In summary
The degradation of the tachykinin substance P is an example of a metabolic process that
produces fragments with very different pharmacological profiles than the parent neuropeptide.
Drug-like molecules mimicking the effect of substance P, such as 11, and the substance P
degradation product, substance P (1–7), such as 2, have been developed. A large number of
selective and nonselective drug-like neurokinin receptor antagonists are known.
Figure 4. The development of an NK1 receptor antagonist to the first series of drug-like NK1 receptor agonists
(2004). The two classes of compounds are structurally very similar.
B. Opioid peptides
1. Opioid neuropeptides
The opioid peptide family contains several members, which bind to three types of opioid
receptor: μ (MOP), δ (DOP), and κ (KOP). These G-protein-coupled receptors (GPCRs)
are well established by pharmacological and molecular biological methods.85 Subtypes of
the various types have been defined, where κ1, κ2, and κ3 are examples of subtypes of the
κ receptor, μ1 and μ2 are subtypes of the μ opioid receptor, and δ1 and δ2 are subtypes of the
δ opioid receptor. The alkaloid morphine, the archetypical nonpeptide opioid is not selective
and activates all three opioid receptors, but activates preferentially the μ receptor. In fact,
morphine and the related nonpeptide opioids were the only known drug-like agonists to pep-
tide receptors until 1995, when Perlman et al. disclosed the second example; the nonpeptide
angiotensin receptor agonists.86, 87 The opioids have been used for hundreds of years, and it is re-
markable that the first endogenous peptides that activate the opioid receptors, the enkephalins,
were not identified and reported until 1975.88
The enkephalins are pentapeptides with a tyrosine residue at their N-terminal (Tyr-Gly-Gly-
Phe-X; X = Leu or Met). More recently, tetrapeptides with opioid activity have been identified,
and have been given the name “endomorphins” (Tyr-Pro-X-Phe-NH2 ; X = Trp or Phe). These
have a C-terminal amide function.89 The enkephalin sequence resides in the N-terminal of most
opioid peptides, including β-endorphin, dynorphin A and B, and α-neoendorphin. The opioid
peptides β-endorphin, the dynorphins and the enkephalins are widely distributed in the brain,
whereas in the spinal cord dynorphins are mainly present in interneurons. Spinal enkephalins
are found primarily in long descending pathways from midbrain to the dorsal horn.90 The
opioid peptides have been attributed to a variety of behavioural processes, such as reward,
dependency, sedation, and stress response but their modulatory actions in pain processing
have probably attracted the greatest interest.91 It is worth noting that the opioid peptides are
produced also in nonneuronal cells, such as endocrine cells and cells of the immune system.
The enkephalins, activate mainly the δ-opioid receptors, while the dynorphins activate
mainly the κ opioid receptors. The selectivity of the dynorphins for the κ-opioid receptor is
mediated by the C-terminal, where Arg7 and Lys11 are important residues. The N-terminal of
the peptide accounts for the opioid activity. Regarding β-endorphin, this peptide seems less
selective than other members of the family and can produce a response through all three re-
ceptors. This response is somewhat stronger for the μ opioid and δ receptor than it is for the
κ opioid receptor. It is still not clear which peptide that is the endogenous ligand of the μ
receptor: the endomorphins label this site with high affinity and specificity and are considered
to be good candidates.89 Stimulation of the μ-opioid receptor leads to euphoria, respiratory
depression, constipation, sedation, dependence, and strong analgesic effects.92 All of the en-
dogenous opioid peptides have very similar N-terminals that act as the “message” part of the
neuropeptides, while their C-terminal act as the “address” part and account for the rather
moderate receptor selectivity observed. This property and their low capacity to cross BBB, and
the fact that the endogenous opioid peptides are readily degraded in vivo by various peptidases
make these neuropeptides not useful in clinic to treat pain.93, 94
Figure 6. Examples of a selective κ opioid receptor agonist, a selective δ opioid receptor agonist, and a
selective μ opioid receptor agonist. These opioids are often used as research tools—all are amines but they
have very different structures. Nalfurafine is used as a nonaddictive antipruritic drug. Fentanyl and its more
recently developed congeners have largely replaced morphine for anesthesia.
system, where κ agonists produce dysphoria, while stimulation of δ/μ opioid receptors results
in euphoria.3, 99, 100
The enzymatic removal of the N-terminal tyrosine residue from dynorphin A, that induces
anxiety-like responses, sedation, and analgesic effects,101, 102 yields a peptide with 16 amino acid
residues that is bioactive but that does not have the opioid effect of the parent peptide.103 Thus,
the deletion of single residue can have a dramatic effect on the biological response. This frag-
ment was reported to interact with specific sites located on the N-methyl-D-aspartate receptor.
The biotransformation of dynorphin A in striatum of freely moving rats after direct infusion
of the peptide has been investigated. Microdialysis-mass spectrometry technique revealed that
proteolysis of, for example, the Arg7 -Ile8 bond to provide dynorphin A(1–7) and dynorphin
A(8–17) occurred.104 Several different enzymes are involved in the enzymatic processing of the
opioid peptides and as an example dynorphin A is a substrate a) for PC2 to cleave Arg9 -Pro10
where after carboxypeptidase E cleaves the Ile8 -Arg9 bond,105 b) for a thiol-dependent dynor-
phin A converting enzyme with Arg6 -Arg7 as major cleavage site,106 and c) for a thiol-sensitive
metalloprotease, called dynorphin A-17 processing enzyme with Ile8 -Arg9 as the cleavage site.107
Hence, at least three different proteases are engaged in the selective cleavage of peptide bonds
in the central part of dynorphin A.
Inhibition of neprilysin and aminopeptidase N, that are involved in the degradation of
the enkephalins, increases the levels of the neuropeptides and induces analgesic effects in
various animal models of inflammatory and neurophatic pain. Notably, dual enkephalinase
inhibitors and fatty acid hydrolase inhibitors, the latter preventing degradation of endogenous
cannabinols, are in clinical trials.108
Figure 7. Structural simplification of morphine produces the opioid receptor agonist pethidine, which comprises
the 4-phenylpiperidine unit of morphine. The N-terminal part of Leu-enkephalin, essential for its activity, is shown
for comparison.
receptor to which morphine and morphine analogues preferentially bind. These drugs, however,
are also associated with a series of side effects including sedation, constipation, and respiratory
depression. The drugs create euphoria and opioid dependence and withdrawal syndrome can
develop. Thus, μ opioid receptor agonists as morphine provide the best analgesics but also most
side effects. As a consequence, considerable efforts have been devoted to the development of new
better selective chemical entities that do not have undesired side effects or that have fewer. The
biological response to a special opioid is a function of to which receptor or receptor subtypes
it binds, the affinity to the various receptors and the pharmacokinetics and metabolism of the
drug. The in vivo duration of drug action depends not only on macroscopic pharmacokinetic
properties such as plasma half-life and the time needed to equilibrate between the plasma and
the effect compartments, but also on long-lasting target binding and rebinding.109–111
Pethidine (16) is a structurally simplified morphine analogue encompassing the character-
istic pharmacophore scaffold of the morphine opioids, Fig. 7. It was synthesized in 1933 and
was the first synthetic opioid to be made. The morphine opioids can be anticipated to bind to
the receptor sites that are occupied of the N-terminal tyrosine or tyrosine-glycine residues of
the endogenous peptides, for example, 17. It is somewhat remarkable that no drug-like low-
molecular-weight opioid agonist seems to have been developed by stepwise modifications of an
endogenous enkephalin. In contrast, hundreds of bioactive analogues, such as pethidine, have
been prepared derived from morphine as structural template. In fact, the majority of the opioid
analgesics used in clinical practise are μ opioid receptor agonists derived from morphine.112, 113
Following the discovery of pethidine and after the development of a large number of other
related analogues, fentanyl (14) was synthesized by Paul Jansen in 1960. Notably, fentanyl that
is lacking the Ar-C-C-C-N element found in morphine/pethidine analogues is 100 times more
potent than morphine. It is used in clinic and is a good example of powerful and selective
μ opioid receptor agonist.
To make peptidemimetics starting from endogenous peptides is a tremendous challenge.
It is possible that the detailed structural information obtained from the recently reported
high-resolution crystal structure of the μ opioid receptor bound to a morphinan antagonist
(revealing a large solvent-exposed pocket into which the morphinan ligand binds deeply)114 will
improve the understanding of the essential binding interactions and aid the discovery of better
selective opioid drugs.
Considerable efforts have been directed toward developing opioids that do not interfere
with the μ opioid receptor, due to the side effects, such as addiction, that are associated
with this receptor. Activation of the κ opioid receptor results in a weaker analgesic response
and in addition, receptor activation is associated with miosis, sedation, psychotomimesis, and
dysphoria. Nevertheless, several κ selective compounds have now been disclosed, for example,
U-50,488H (18), CI-977,115, 116 HZ2117 and salvinorin A isolated from Salvia divinorum. The
Medicinal Research Reviews DOI 10.1002/med
METABOLISM TO BIOACTIVE FRAGMENTS r 13
Figure 8. The arylacetamides 18 and 19 are subtype selective and serve as selective κ1 agonists, while
pentazocine (20) and bremazocine (21) are κ2 receptor selective. The latter two compounds include the ben-
zomorphane skeleton found in nalfurafine (Fig. 6), but nalfurafine is κ3 receptor selective.
widely abused hallucinogen salvinorin A has a very different chemical structure from those of
other opioids, in that it does not contain nitrogen.118, 119 In fact, it was the first nonnitrogenous
opioid agonist reported. Nalfurafine (TRK-820) (12), with the characteristic Ar-C-C-C-N
scaffold and encompassing a tyrosine element provides a good example of κ agonist that is
very selective, Fig. 6. Nalfurafine is used as an antipruritic drug.120–122 It has been claimed that
this naltrexone-like derivative is the first opioid that does not cause addiction.123 Hence, while
nalfurafine neither exhibits aversive nor additive effects,124 many arylacetamide derivatives such
as U-50488H (18) and U-69,593 (19) cause psychomimetic and aversive reactions.125, 126 These
differences in psychological effects have been postulated to be attributed to affinity preferences
versus various κ receptor subtypes, where arylacetamide derivatives prefer the κ1 receptor
subtype127, 128 and nalfurafine the κ3 receptor subtype.129–133 The benzomorphans pentazocine
(20) and bremazocine (21) are proposed to show high affinity to the κ2 subtype of the κ opioid
receptor,128, 134 Fig. 8. Thus, very structurally diverse compounds activate the three subtypes of
the κ opioid receptor. The recently reported crystal structure of the human κ opioid receptor
in complex with a κ opioid receptor selective antagonist,135 in combination with modeling
of selective agonists related to salvinorin A as well as morphinan-derived antagonists, have
provided essential insights on ligand-receptor interactions useful in future design of even more
selective κ opioid receptor ligands. A large binding cavity with several potential anchoring
points to ligands was revealed, which can explain the acceptance of broad structural diversity
of the κ opioid receptor ligands.135 Furthermore, access to 3D structures of selective ligands to
the subtypes of the κ receptor will be very helpful in future discovery processes.136
Many selective agonists, not only to the μ and κ opioid receptors but also to the δ re-
ceptor, have recently been reported.137–141 The δ opioid receptor agonists have less powerful
antinociceptive activity but have a potential for use as analgesics since they are associated with
relatively few severe side effects and are anticipated in general to be less addictive.142 Among
such agonists, SNC 80 (13) is highly selective for the δ receptor and is 2000-fold selective over the
μ opioid receptor,137, 139 Fig. 6. Another selective δ agonist, TAN-67 (23)143, 144 is structurally
Figure 9. From a selective δ opioid receptor antagonist to a subtype selective δ1 opioid receptor agonist.
Figure 10. Opioid receptor antagonists used after overdoses of opioids and to treat alcohol dependence. These
compounds, which are often used as research tools, are characterized by the benzomorphane skeleton, the
vinyl, or cyclopropyl group linked to the basic nitrogen and the tertiary hydroxyl group. These elements are also
present in the δ opioid receptor antagonist naltrindole (Fig. 9).
very different. It has a rotable bond between the two ring systems and contains a quinoline nu-
cleus. TAN-67 has been derived from the rigid and selective δ antagonist, naltrindole (22),141, 145
Fig. 9. The “message (efficacy) – address (selectivity)” concept introduced by Portoghese has
been useful in the design process of opioid agonists and antagonists of this class.146 The crystal
structure of the δ receptor from mouse bound to the subtype-selective antagonist has recently
been reported.147 The lower part of the binding pocket, which is divided into two distinct
regions is highly conserved among the opioid receptors. The upper part, in contrast, contains
divergent residues and confers subtype selectivity. Thus, a structural rationale for the useful
“message-address” concept has been obtained.147
Naloxone (24), often referred to as an opioid antagonist is an inverse agonist and is used
in the clinic to counter the effects overdoses of opioids. In addition, naloxone is added to
the partial opioid agonist buprenorphine, in a mixture called Suboxone, to prevent abuse of
buprenorphine, which is probably the major current use of naloxone. Naltrexone (25) is mainly
used to treat alcohol dependence,148 Fig. 10. In addition, it is added to buprenorphin (in the
mixture called suboxone) to prevent abuse of buprenorphin. This is probably the major current
use of naltrexone.
The compounds described in this section so far activate receptors of opioid peptides
and opioid peptide fragments and are drug-like. In addition, there are a large number of
compounds disclosed in the literature that retain the peptidic character of the endogenous
ligands. These pseudopeptides have served and serve as important research tools and have
improved the understanding of potential bioactive conformations and binding modes of the
opioid peptides to their receptors. Various modifications such as introduction of D-amino
acids, of β-amino acids, of peptide bond bioisosters, and of conformational constraints by
Figure 11. The ring size of the macrocycles is critical for opioid receptor selectivity.
cyclizations, for example, have been applied. By these procedures, very potent and selective
agonists were provided, exemplified here with the selective μ opioid receptor agonist JOM-6
(26) and the structurally very similar selective δ opioid receptor agonist JOM-13 (27), Fig. 11.
The C-terminal carboxylate has been replaced with a carboxamide in JOM-6, and this favors
μ binding over δ opioid receptor binding.149
The tyrosine residue constitutes the pharmacophore associated with binding and
activation,150–152 while the position of the aromatic ring of phenylalanine is proposed to be
responsible for the observed subtype receptor selectivity149, 153 (for a review, see154 ).
4. Nociceptin
Nociceptin155 also known as “orphanin FQ”,156 is not a classical opioid but is presented here
due to its similarities with the opioid peptides. This neuropeptide of length 17 amino acid
residues includes the sequence Phe-Gly-Gly-Phe in the N-terminal part rather than the Tyr-
Gly-Gly-Phe present in the enkephalins and dynorphins. Thus, a hydroxyl group is the difference
at the N-terminal end. The nociceptin opioid receptor (NOP) sometimes referred to as opiate
receptor-like 1, ORL-1, LC132, OP4 , or NOP1 was cloned in 1994 and identified using a human
cDNA library on the basis of high homology with the classic μ, κ, and δ opioid receptors.157, 158
This GPCR remained an orphan receptor for a year after it was cloned, until the endogenous
neuropeptide agonist nociceptin or orphanin FQ (N/OFQ) was isolated from brain extracts.
The neuropeptide was found to bind with high affinity to the NOP site, but not to the μ, κ,
or δ opioid receptors.155, 156 Nociceptin, its precursor prepronociceptin and the NOP receptor
are localized to the corticolimibic regions and like the μ opioid receptor, δ opioid receptor and
κ opioid receptor, the NOP receptor, with around 60% sequence homology with the opioid
receptors, is negatively coupled to adenylate cyclase, activates potassium channels, and inhibits
calcium channels.159 The peptide is associated with processes that are relevant to pain,160, 161
learning and memory162, 163 as well as stress and anxiety.164, 165 The peptide is present in several
areas known to be involved in pain perception.166 Activation of the receptor produces spinal
analgesia but it also seems to antagonize the effects of opioids.
5. Metabolism of nociceptin
Nociceptin is metabolized by several proteases of different specificity. Aminopeptidase
N cleaves the Phe1 -Gly2 bond delivering phenylalanine and nociceptin(2–17) in
mouse brain slices, whereas endopeptidase 24.15 acts by cleaving nociceptin at the
Ala7 -Arg8 , at the Ala11 -Arg12 , and at the Arg12 -Lys13 bonds.167 The neutral endopeptidase
was reported not to be involved in the metabolism. In mouse spinal cord on the other hand,
Figure 12. Nociceptin is degraded to peptide fragments that have similar biological profiles to that of nociceptin,
i.e., hyperalgesia and antinociception, and to bioactive fragments that have opposite or different biological
profiles to those of nociceptin. Such fragments may, for example, interact with an unknown receptor that leads
only to antinociception.
synaptic membranes endopeptidase 24.15 is not involved but here the neutral endopeptidase
delivers nociceptin(1–13), nociceptin(14–17), and phenylalanine as the major metabolites from
nociceptin.168 Furthermore, small amounts of nociceptin(1–7), (1–9), (1–10), (1–11), (2–9), (2–
11), (2–13), (2–17), and (13–17) were detected in the synaptic membranes. Nociceptin is first me-
tabolized to nociceptin(1–13) and (14–17) in rat hippocampus. The nociceptin(1–13) fragment
is subsequently metabolized into nociceptin(1–9) and nociceptin(10–13).169 Nociceptin(2–17)
that is the major metabolite in human plasma undergoes further proteolytical processing at
its N-terminal to produce nociceptin(3–17), (4–17), and (5–17).170 The metabolism pattern of
nociceptin is highly specific for tissue type.
Some of the nociceptin fragments retain biological activity. The N-terminal nociceptin
(1–7), (1–9), and (1–13) fragments modulate nociceptin-induced scratching, biting, and lick-
ing behavior in mice.171 Some nociceptin fragments have similar biological effects as those of
nociceptin, while other fragments have opposite or different effects. It is probable that at least
some of the biological effects attributed to nociceptin, originate from its bioactive fragments.
For example, since nociceptin is known to produce both antinociception and hyperalgesia, it
was suggested that the antinociceptive activity arises after the conversion of the parent pep-
tide. This hypothesis is supported by studies that show that i.c.v. injection of the nociceptin
fragments (1–7) and (1–11) elicits antinociception without causing hyperalgesia,172 Fig. 12
and Table I. Furthermore, studies on nociceptin C-terminal fragments, such as nociceptin
(13–17), demonstrated an induction of nociceptive response in mice spinal cord, whereas
N-terminal fragments of nociceptin exhibited no effect.173, 174 It appears that effects of
N-terminal nociceptin fragments such as (1–7) and (1–11) are mediated through receptors
that are not recognized by the parent compound nociceptin.
Figure 13. Two nociceptin receptor agonists and one antagonist. The agonists 28 and 29 exhibit anxiolytic-like
activity in rat.
SCH-221510 (29) that is orally active have been disclosed.181 It has been demonstrated that
RO64–6198 exhibits an anxiolytic-like activity in rats182 and more recently a similar effect
was observed also with SCH 221510 that was qualitatively comparable with that achieved
with benzodiazepine.181 Thus, it was suggested that NOP agonists may represent a new class
of anxiolytic agents181 and in addition to that, based on new experimental data, it has been
proposed that mixed MOP/NOP receptor agonists might provide good analgesics since co-
activation of MOP and NOP receptors produce synergistic antinociception with a minimum
of side effects.183, 184 Furthermore, a mixed-action profile of NOP/opioid activity is suggested
to provide a valuable therapy to treat addiction to various abused substances and/or polydrug
addiction.185 Notably, SCH 486757, an orally available NOP receptor agonist whose structure
is very similar to that of SCH 221510, has recently entered human clinical trials for cough.186, 187
Potent selective drug-like NOP receptor antagonists such as Trap-101 (30) have also
been developed.188 Such antagonists may relieve the symptoms of patients with Parkinson’s
disease.189 The crystal structure of the nociceptin/orphanin FQ receptor (NOP receptor) in
complex with a selective antagonist, encompassing a spiropiperidine scaffold and that mimics
the first four N-terminal amino acid residues of a close analogue of nociceptin was recently
reported. Substantial conformational differences between NOP and the classical κ and μ opioid
receptors were observed.190
8. In summary
Members of the opioid peptide family, such as the precursor peptide dynorphin A, activate
one type of receptor, the κ opioid receptor (KOP) while one of its degradation products, Leu-
enkephalin activates other receptor types, the δ and μ opioid receptors (DOP and MOP). Thus,
Figure 14. Captopril, the first ACE inhibitor in clinic (1978), and aliskiren, the first renin inhibitor in clinic (2007).
the rate of proteolytic degradation in various structures in brain and the pattern of degradation
of opioid peptides are of outmost importance for the biological outcome. Several different
types of proteases operate, and these proteases may be targets for future pharmaceuticals.
The opioid peptidemimetics, such as morphine (isolated 1804), pethidine (1933), and fentanyl
(1960), existed long before the endogenous ligands to the opioid receptors were identified in the
early 1970-ties. No drug-like opioid receptor agonist or antagonist have been developed and
reached clinical trials, where the endogenous opioids have been utilized as structural prototypes
and starting points, although the N-terminal tyrosine element can be recognized in many of
the drug-like opioid compounds. Newly available structural information about the detailed
molecular interactions of selective ligands binding to the three classical opioid receptors and
the nociceptin/orphanin FQ peptide receptor will be useful in the discovery processes aimed at
identifying new subtype selective pharmaceuticals that target the opioid system. Thus, a large
number of receptor-selective drug-like molecules mimicking the effects of opioid fragments at
their receptors have been developed. Opioid receptor antagonists are known, such as 25.
C. Angiotensins
Angiotensin II (Asp1 -Arg2 -Val3 -Tyr4 -Ile5 -His6 -Pro7 -Phe8 ), Ang II, elicits a pronounced hy-
pertensive effect and is a powerful modulator of a variety of cardiovascular functions. It was
observed in the early 1970s that minor modifications of the amino acid residue sequence of
Ang II could produce peptides that block the action of Ang II. These peptides (e.g., sarile and
saralasin) helped to confirm that the renin-angiotensin system (RAS) was a suitable target for
drugs aimed at combating hypertension, in particular.194–198
Figure 15. Losartan, the first AT1R antagonist in clinic; L-163,491, the first selective drug-like AT1R agonist;
M024/C21, the first selective drug-like AT2R agonist; and EMA401, a selective AT2R antagonist undergoing
clinical trials.
The first nonpeptide angiotensin II receptor blocker (ARB) losartan (33), acting as a se-
lective antagonist to the Ang II AT1 receptor (AT1R) was introduced onto the market in 1995,
Fig. 15. Several effective ARBs, collectively known as “sartans”, have subsequently been dis-
closed and are now widely used in clinical practice.205, 206 Efforts to develop single compounds
that can block both the AT1R and the receptor of the potent vasoconstrictor endothelin A are
now in progress.207 Furthermore, LCZ696, a dual antagonist of AT1R and inhibitor of neutral
endopeptidase is in late clinical trials.208 Drug-like compounds able to activate the AT1R with
high selectivity are known; one such (the first to be reported) is L-163,491 (34).209 The other
major Ang II receptor, the AT2 receptor (AT2R) has recently emerged as a new target for drug
therapy.210–212 The AT2R is abundant in fetal tissues but in adults this GPCR remains abun-
dant only in certain tissues such as vascular endothelium213 and the brain.214, 215 Activation of
the AT2 receptor affects neuronal cell differentiation and nerve regeneration.216–218 The AT2R
exerts vasodilatory, antiproliferative and anti-inflammatory effects,219–221 neuroprotection,222
and several peripheral effects mediated through the AT2R oppose those mediated through
the AT1R.223, 224 It is worth noting that the AT2R is re-expressed in some disease condi-
tions such as heart failure, renal failure, myocardial infarction, hypertension, and some brain
disorders.215, 225–230 The first receptor–selective and drug-like AT2R agonists M024/C21 (35),
that exhibits a pronounced anti-inflammatory effect231 was reported in 2004.232 It was recently
suggested that the combination of drug-like selective AT2R agonists with antihypertensive
treatment might lead to vasculoprotective effects even beyond the blood-pressure-reducing
effect.210, 233–235 No selective AT2R agonists, however, have still entered clinical trials. In con-
trast, EMA401 (36), which does not have significant CNS distribution after oral dosing236 and
which is an analogue of the commonly used research tool, the AT2R antagonist PD123319
is in clinical trials and is aimed for combating neuropathic pain.236–238 EMA401 is acting via
peripheral mechanisms. Thus, both of the Ang II receptors, AT1R and AT2R are today serving
as targets for drugs.
2. Metabolism of angiotensins
The aspartyl protease renin liberates the essentially inactive angiotensin I (Ang I) from circu-
lating and tissue angiotensinogen, Fig. 16. Notably, a receptor for renin exits.239 This receptor
named the (pro)renin receptor binds both renin and prorenin, the inactive form of renin. Re-
ceptor binding triggers intracellular signaling.240 The next step, the proteolytic cleavage of
Figure 16. Angiotensin II is degraded to several fragments that have either similar biological effects or very
different ones. The degradation products Ang(1–7), which activates the Mas/Ang(1–7) receptor, and Ang(3–8)
(Ang IV), which inhibits the insulin-regulated aminopeptidase (IRAP) are examples of peptides with very different
profiles from that of the parent compound.
angiotensin I to produce Ang II, is mediated mainly by the drug target, the metalloproteinase
ACE. However, cleavage by chymase, carboxypeptidase, cathepsin G, or tonin provide alterna-
tive routes by which Ang II can be generated.241 Furthermore, carboxypeptidase A6 produces
Ang II.242 Ang II can also be produced from the dodecapeptide angiotensin (1–12) Ang (1–12)
by for example, ACE or chymase and accumulating evidence suggests that Ang (1–12) is a
functionally important substrate for the actions of AngII in the brain.243–245
Proteolytic cleavage of the major effector peptide Ang II by glutamyl aminopeptidase A
(AP-A) and membrane alanyl aminopeptidase N (AP-N), result in the sequential removal of sin-
gle amino acid residues from the N-terminal end to form Ang III (Ang II(2–8)) and Ang IV (Ang
II(3–8)), respectively,246 both of which are essential neuropeptide fragments in the CNS.247–250
Ang IV, in particular, plays an important role.251–254 It is noteworthy that Ang IV can be formed
also by the action of aminopeptidases on Ang I before it is converted to Ang II.255 Further-
more, a human Ang II–related peptide, Ang A, has recently been discovered.256 This peptide,
(Ala1 )-Ang II, is formed when the aspartic acid residue of Ang II is decarboxylated256 and it
acts as a full agonist with properties that are similar to those of Ang II.257 The heptapeptide
alamandine (Ala1 -Arg2 -Val3 -Tyr4 -Ile5 -His6 -Pro7 ) has recently been detected in human blood,
and characterized. Alamandine, can be formed by hydrolysis of (Ala1 )-Ang II by ACE2.258
Ang IV as well as the fragment Ang (3–7) is further processed by chymotrypsin and
dipeptidyl carboxypeptidase into inactive fragments and amino acid residues.259–264 Ang
Figure 17. The Mas (Ang (1–7)) receptor agonist AVE-0991 and compound XNT that activates ACE2 and
promotes degradation of Ang II to Ang(1–7).
of Mas, which is a GPCR.301, 303 Ang(1–7) has been demonstrated to antagonize cardiac hyper-
trophy and fibrosis, outcomes that are mediated by the Mas receptor.283, 304 The heptapeptide
ameliorates cardiac remodeling by decreasing hypertrophy and fibrosis,283, 305, 306 and genetic
depletion of the Mas receptor causes dyslipidaemia, insulin resistance, and marked fibrotic
and hypertrophic changes in rat myocardium.307, 308 It has been reported that chronic adminis-
tration of angiotensin(1–7) also prevented diabetes-induced cardiovascular dysfunction.309, 310
Sanofi–Aventis undertook a medicinal chemistry program aimed at discover drug-like ago-
nists to the Mas receptor. This program resulted in AVE-0991 (37),308 which has considerable
structural similarities to the AT2R agonist M024/C21 (35), Figs. 15 and 17. AVE-0991, how-
ever, does not bind to AT2R. One characteristic feature of AVE-0991 is an aldehyde group.
AVE-0991 mimics the action of angiotensin (1–7) in many tissues by its vasodilating and na-
triuretic properties.301, 308 The magnitude of the infarcted area following an infarction induced
by left coronary artery ligation in rats is much smaller when the compound is administrated.311
Pronounced cardioprotective effects of AVE-0991 were also demonstrated in a diabetes rats
model,309, 312 and it mediates antifibrotic actions in a rat model of cirrhosis.313 It has recently
been reported that this Mas receptor agonist facilitates penile erection314 and that the compound
attenuates pulmonary remodeling in a model of asthma.315
Another heptapeptide, alamandine has actions that resemble those produced by angiotensin
(1–7); vasodilation, antifibrosis, anti-hypertensive, and central effects. Alamandine acts through
the Mas-related GPCR, MrgD.258 Interestingly, small molecules have been discovered that
are able to activate ACE2 and in this way promote the degradation of Ang II, and catalyze
the generation of Ang(1–7). Such compounds, such as the lead XNT (38) need to be further
optimized, but have demonstrated beneficial outcomes in animal models.316–319 Thus, this ACE2
activator (38) elicited reductions in blood pressure, improved cardiac function, and reversed the
myocardial and perivascular fibrosis observed in spontaneously hypertensive rats.316 However,
recently it was postulated that the biological effects of 38 is ACE2-independent and should not
be attributed to the activation of this enzyme.320 In summary, the preclinical data, in particular
with the Ang(1–7) peptidemimetic AVE-0991, are promising and the ACE2/Ang(1–7)/Mas
system is an interesting target for cardiovascular and pulmonary disorders, among others.280
Figure 18. Examples of small compounds that interac with the IRAP/Ang IV receptor. Compounds 39 and 40
are derived from the hexapeptide Ang IV and compound 41 from virtual screening. IRAP inhibitors enhance
cognition.
motif of Ang IV is critical to its binding affinity. Kobori et al. at Taisho Pharmaceuticals filed
patent applications in the late 1990s that covered compounds, such as compound 39, that bind
strongly to Ang IV binding sites in guinea pig hippocampus membranes,327, 328 Fig. 18. Albiston
et al. suggested in 2001 that insulin-regulated aminopeptidase, a zinc-metallopeptidase of the
M1 family is the main target for Ang IV.329 IRAP was identified as cystinyl aminopeptidase
(CAP, EC 3.4.11.3), placental leucine aminopeptidase (P-LAP, soluble human homologue),
oxytocinase, gp160 or vp165.330–333
Considerable efforts were subsequently devoted to identify drug-like compounds that
mimic Ang IV and inhibit IRAP. It was believed that such compounds could provide new
classes of cognitive enhancers that could be used to treat Alzheimer’s disease and other related
disorders.297, 334–337 The proven effects of Ang IV and related peptide analogues on memory
and learning were promising.297–299 Improved cognitive enhancers are needed since the clinical
studies of the cholinesterase inhibitors and N-methyl-D-aspartate antagonists utilized in ther-
apy today have been mostly disappointing.338–341 The major problem encountered in the design
process of IRAP inhibitors was to make compounds able to penetrate the BBB. One approach
to drug-like Ang IV mimics relied on stepwise alterations of Ang IV by various cyclizations
to determine the bioactive conformation of the hexapeptide when it binds to IRAP.342, 343 This
approach took advantage of the knowledge that oxytocin and vasopressin, both of which have
positive effects on memory, are substrates of IRAP.344 These large peptides encompass a macro-
cyclic disulphide at the cleavage site. Potent macrocyclic inhibitors of IRAP have subsequently
been developed, such as 40,345, 346 but these macrocycles are still too peptidic to cross BBB.347
A very promising series of IRAP inhibitors was identified in 2008, after the in silico screening
of 1.5 million commercially available compounds against a model structure homologous to
IRAP.348, 349 These low-molecular-weight benzopyran-based inhibitors, exemplified by 41, act
as cognitive enhancers in rats. The benzopyran class of IRAP inhibitors provides promising
leads for further development.350 It is not known whether the ligands synthesized by Kobori
et al. inhibit IRAP.
Alternative macromolecular targets for Ang IV have been proposed, such as c-Met, a
tyrosine kinase receptor that binds hepatocyte growth factor and that is associated with memory
and learning consolidation.351, 352 No inhibitors of IRAP have still reached Phase I clinical trials.
5. In summary
Ang I and Ang II are degraded to bioactive fragments with pharmacological profiles that
differ from those of the parent peptides. At least two receptors of peptide fragments derived
from Ang II, the Mas receptor, and IRAP are being studied as targets for the systematic
development of drug-like new chemical entities. Furthermore, the concept of interfering with
the ACE2/Ang(1–7)/Mas receptor axis by applying ACE2 activators and manipulating in this
Figure 19. Kallidin is degraded to bradykinin, and both kallidin and bradykinin are agonists that bind prefer-
entially to the B2 receptor. Kallidin is degraded also to des-Arg10-kallidin, which has very high affinity for the
B1 receptor. Bradykinin is degradaded to des-Arg9-bradykinin with high affinity to the B1 receptor and to the
inactive bradykinin(1–7). The fragments have biological profiles that differ from those of the parent peptides.
way the protoelytic processing is an interesting approach. Angiotensin II AT1 and AT2 receptor
peptidemimetics, for example, 34 and 35, respectively, and drug-like molecules mimicking the
effects of the fragment Ang(1–7), such as 37 and the fragment Ang IV, such as 40 and 41 at
their receptors, have been developed.
D. Bradykinins
The nonapeptide bradykinin (Arg1 -Pro2 -Pro3 -Gly4 -Phe5 -Ser6 -Pro7 -Phe8 -Arg9 ) is a powerful
vasodilator that increases the permeability of capillaries. It constricts nonvascular smooth
muscle and is an active component in several physiological and pathological processes related to
inflammation and pain. The decapeptide kallidin (Lys1 -Arg2 -Pro3 -Pro4 -Gly5 -Phe6 -Ser7 -Pro8 -
Phe9 -Arg10 ), also known as “lysyl-bradykinin”, has similar effects to those of bradykinin. Both
bradykinin and kallidin play central roles in the CNS after injury and during infection and
inflammation.353 The two main kinin receptors, B1 and B2 354–356 have been cloned and sequenced
in several species.357–360 They belong to the GPCR family and both have been addressed as drug
targets. The B1 receptors play roles in chronic pain and inflammation361, 362 and are inducible
and poorly expressed under normal physiological conditions. B2 receptors, in contrast, are
constitutively expressed in several cell types and mediate vasodilation.363
1. Metabolism of bradykinins
Bradykinin and kallidin are produced by proteolytic processing of the low-molecular-weight
and high-molecular-weight glycoprotein kininogen by tissue or plasma kallikreins.364, 365
Trypsin and plasmin are also able to affect this conversion. Kallikreins are endopeptidases
of trypsin type, and are activated by autoproteolysis or alternatively by other proteases, for
example, after injury. Thus, the kallikreins play roles that are similar to that of renin in
the renin-angiotensin system, but no kallikrein inhibitors have reached the clinical practice.
Kallikrein, bradykinin, and kallidin have all been identified in the CNS, where bradykinin is
most abundant in the hypothalamus and the pituitary.366–368 Bradykinin can be formed from
kallidin by cleavage of the Lys1 -Arg2 bond by aminopeptidases, such as aminopeptidase N,369
Fig. 19 and Table I. It is worth noting that high-molecular-weight kininogen is the precursor
of bradykinin in plasma.
Figure 20. Three B1 receptor antagonists. The B1 receptor is the major target for des-Arg10-kallidin and des-
Arg9-bradykinin. MK-0686 has been evaluated in clinic for postherpetic neuralgia, osteoarthritis, and postsurgery
dental pain.
(43), which has been tested in Phase II clinical trials for postherpetic neuralgia, osteoarthritis,
and postsurgery dental pain, is one example from this group of substances. Human PK data
from MK-0686 are available.391 Interestingly, it has recently been reported that B1 receptor
antagonists, such as LF22–0542 (44), a water- soluble arylsulfonamide are promising candidates
for treating diabetic retinopathy after ocular application.392 Common binding motifs of aryl
sulfonamides and biarylmethyl amines have been proposed393 and all data from structure-
activity relationship and homology modeling394, 395 should hopefully promote the discovery of
investigational drugs antagonizing the B1 receptor and that are successful in clinical use.396, 397
The B2 receptor antagonists may be useful after traumatic brain injury.398, 399 The first B2
antagonist to be used in the clinic icatibant (45), has the length of ten amino acid residues and
is injected in connection with acute attacks of the rare hereditary angioedema that may be life
threatening and that is caused by a deficiency of the C1 esterase inhibitor in the complement
system,400–403 Fig. 21. Large programs have been devoted during the past two decades to
developing efficient, orally bioavailable drug-like B2 receptor antagonists.404, 405 The quinoline
derivative FR165649 (46), which has a glycine linker, is one of a large number of selective
B2 antagonists. Remarkably, this molecule can be transformed into the proinflammatory B2
receptor agonist FR190997 (47) by attaching a pyridine unit to the quinoline nucleus via a
methyleneoxy spacer.406–408 The agonist FR190997 induces a hypotensive response in rats that
is blocked by icatibant.409
3. In summary
Kallidin and bradykinin, which bind preferentially to the B2 receptor, are degraded to des-
Arg10-kallidin and des-Arg9-bradykinin, respectively, that both exhibit high affinity to the
Figure 21. The B2 receptor antagonists icatibant and FR165649, and the B2 receptor agonist FR190997. Icat-
ibant is used in the clinic to treat acute attacks of the rare hereditary angioedema (HAE). The proinflammatory
agonist FR190997 induces a hypotensive response in rats that is blocked by icatibant.
inducible B1 receptor. Thus, the proteolytic processing can result in fragments with biological
profiles that differ significantly from those of the parent peptides. Drug-like agonists and
antagonists have been discovered and most of the nonpeptide drug-like B1 and B2 antagonists
under development have been identified after screening and a subsequent systematic lead-
optimization process.
E. Neuropeptide Y
Neuropeptide Y (NPY, neuropeptide tyrosine) comprises 36 amino acid residues and is a mem-
ber of the pancreatic polypeptide family. This family of neuropeptides includes the structurally
related gastrointestinal hormone peptide YY (PYY, peptide tyrosine tyrosine) and the pancre-
atic polypeptide (PP).410, 411 NPY, is widely distributed in both the central and peripheral ner-
vous system and is involved in several physiological systems.412–415 NPY acts in humans through
four GPCRs; Y1, Y2, Y4, Y5,416–418 which have a relatively low sequence homology.415, 419 PP
is more likely to serve as the endogenous ligand to the Y4 receptor than NPY or PPY, and
this receptor also shows a greater variability than Y1, Y2, and Y5.420 In fact, the Y1, Y2, and
Y5 receptors are the most different GPCRs (with 27–32% overall identity) known that bind to
the same peptide.421 The Y1 and Y2 receptors are mainly responsible for the periphery effects
mediated by NPY.422
The impact of the neuropeptide on body weight regulation has attracted particular
interest.423 Hence, after chronic central administration of NPY to normal rats, a pathophys-
iological profile of similar nature to that observed in connection with human obesity was
developed.424, 425 Fasting leads to higher levels of NPY than normal in brain regions associated
with energy homeostasis, such as the paraventricular nucleus and the arcuate nucleus.426 It is
Figure 22. NPY is nonselective. The peptide is degraded to NPY(3–36), which is selective for the Y2 receptor
and has a different biological profile from that of NPY. NPY(3–36) is further degraded to the inactive NPY(3–35)
by elimination of the important tyrosine amide in the C-terminal.
believed that the feeding response is mediated by Y1 and/or the Y5 receptors expressed in
the hypothalamus.427 Peptide YY, occurs in two forms, PYY(1–36) and PYY(3–36), and these
are released when food is consumed. PYY(1–36) is an unselective Y1/Y2 agonist, while the
PYY(3–36) fragment is a highly selective Y2 agonist. Hence, proteolytic processing determines
the level of receptor selectivity. Peripheral administration of PYY(3–36) to humans significantly
reduces appetite and food intake,428, 429 which suggests that the Y2 receptor is also involved
in energy homeostasis.430 Thus, the experimental data suggest that stimulation of the Y1 and
Y5 receptors plays an important role in increased feeding while stimulation of the Y2 and Y4
receptors plays an important role in appetite inhibition. Furthermore, the activation of NPY re-
ceptors is associated with anxiety and depression,415, 431, 432 the experience of pain,433, 434 alcohol
dependence,435, 436 bone formation437, 438 as well as angina pectoris439 and cardiac stability.440
1. Metabolism of neuropeptide Y
Cleavage of the prepro-NPY sequence441 produces pro-NPY, which has a length of 69 amino
acid residues. This sequence contains (i) the 36 residues that will subsequently form NPY, (ii) the
three-peptide unit Gly-Lys-Arg, required for the cleavage and amidation, and (iii) a sequence
of length 30 amino acids that forms the C-flanking peptide of NPY. This latter sequence, often
named CPON, is liberated from the prohormone by a convertase. NPY is degraded by, for ex-
ample, NEP leading to cleavage of the Tyr20 -Tyr21 and the Leu30 -Ile31 bonds, which is consistent
with the specificity of this enzyme.442 Peptide bond cleavages in hippocampal synaptosomes
deliver NPY(1–30) and NPY(31–36). These reactions can be suppressed by pepstatin, which
suggests that an aspartic protease is involved.443 Removal of the N-terminal dipeptide moiety
Tyr1 -Pro2 by dipeptidyl-aminopeptidase IV yields NPY(3–36),444 a fragment that is inactive at
the Y1 receptor subtype but active at the Y2 receptor,445, 446 Fig. 22 and Table I. It is worth
noting that NPY (13–36) is a selective Y2 receptor agonist.447 The NPY/PYY receptor Y2 is
pharmacologically characterized by its ability to retain high affinity for N-terminally truncated
peptide fragments, not only NPY(3–36) and NPY(13–36) but also PYY(3–36) and PYY(13–
36). Thus, the enzymatic processing of the parent NPY and of PYY, can lead to fragments
with receptor subtype selectivity. Not only enzymes that are similar to dipeptidyl peptidase
IV, but also specific endopeptidases (such as meprin and neprilysin-like enzymes), postargi-
nine hydrolysing enzymes, aminopeptidase P, etc., are engaged in the processing. Deletion of
the C-terminal amino acid residue(s) by specific proteases yields products such as NPY3–35,
PYY3–35, NPY3—34, or PYY3–34, that are completely inactive.448
Medicinal Research Reviews DOI 10.1002/med
METABOLISM TO BIOACTIVE FRAGMENTS r 29
Figure 23. Examples of NPY Y5 receptor antagonists. The NPY Y5 receptor antagonists may be useful agents
to control appetite.
3. In summary
The nonselective neuropeptide Y that is involved in many physiological systems can be converted
not only to biologically inert metabolites but also to degradation products that are bioactive
and more receptor selective than the parent peptide. NPY(3–36) is an example of a fragment
that exhibits selectivity for the Y2 receptor. Selective drug-like receptor antagonists are reported
Medicinal Research Reviews DOI 10.1002/med
30 r HALLBERG
but no drug-like Y2 or Y4 receptor agonists with potential use as antiobesity drugs are yet
disclosed.
3. CONCLUSION
The proteolytic processing of neuropeptides has an important regulatory function and the
peptide fragments resulting from the enzymatic degradation often exert essential physiological
roles. A large number of the peptide fragments display very different activities than their parent
peptides. This is well illustrated in the RAS system, where the short-lived metabolites of the
octapeptide Ang II, Ang IV (Ang(3–8)), and Ang(1–7) are generated. Both of these peptides
exert very different effects from Ang II. Drug-like substances that mimic the biological effects
of these fragments are now being examined as potential pharmaceutics. A metalloprotease is
the target for the Ang II degradation fragment Ang IV and the Mas receptor for the Ang II
degradation fragment Ang(1–7). Drug-like mimetics of the substance P fragment SP(1–7) have
been reported and provide another example from the tachykinin system, but significantly more
efforts have been devoted to finding selective and nonselective NK1, NK2, and NK3 antago-
nists for clinical use. NPY, although not receptor selective, is degraded to the agonistic selective
Y2 receptor NPY(3–36) fragment. No selective drug-like Y2 receptor agonists have been de-
veloped, even though selective drug-like Y2 receptor antagonists are known. The degradation
of dynorphin A (a κ opioid receptor agonist that produces dysphoria) to Leu-enkephalin (that
acts as a δ and μ opioid receptor agonist and induces euphoria) is one example from the opioid
Medicinal Research Reviews DOI 10.1002/med
METABOLISM TO BIOACTIVE FRAGMENTS r 31
system, where the protolytic processing leads to different biological effect profiles. It is worth
noting that with the exception of the opioid research area, most attention has been directed
toward drug-like receptor antagonists as new chemical entities.
Drug-like antagonists such as B2 receptor antagonists could be transformed into B2 recep-
tor agonists by small structural manipulations, and selective AT1 and AT2 receptor agonists
could be generated from nonselective AT1/AT2 receptor agonists, originally derived from the
sartans. Furthermore, opioid receptor agonists have been transformed to antagonists, and opi-
oid antagonists to agonists, and drug-like NK receptor agonists were obtained after minor
structural alterations of NK antagonists. Thus, herein, a representative series of important
drug-like molecules that act as agonists or antagonists at neuropeptide receptors has briefly
been discussed.
As exemplified in this review, many common neuropeptides are degraded into smaller
fragments that are often bioactive but relatively few of these fragments (e.g., substance P(1–7)
and angiotensin II(3–8)) have yet been examined in detail with respect to their biological
activities. Such short bioactive fragments should be less complicated to convert into drug-
like substances mimicking the desired effects of the degradation products (e.g., the substance
P(1–7) mimic 2 and the angiotensin II(3–8) mimic 39) and could provide a potential rich
source of new pharmaceuticals. However, very few examples heretofore have been reported of
successful design of drug-like agonists or antagonists using the endogenous peptide fragments
as starting points. The transformation of peptides to drug-like nonpeptides displaying favorable
pharmacokinetic profiles, through procedures involving rigidifications, displacement of peptide
bonds, replacement of amino acids with other moieties, truncations, etc. remains a challenge.
4. ABBREVIATIONS
ACE angiotensin-converting enzyme ARB angiotensin II receptor blocker AT1R AT1
receptor AT2R AT2 receptor BBB blood–brain barrier CGRP calcitonin gene-related peptide
CINV chemotherapy-induced nausea and vomiting CNS central nervous system CPE car-
boxypeptidase E GPCR G-protein-coupled receptor IRAP insulin-regulated aminopeptidase
NEP neutral endopeptidase NOP nociceptin opioid receptor NPY Neuropeptide Y RAS
renin-angiotensin system SPE substance P endopeptidase
ACKNOWLEDGMENTS
This study was supported by the Kjell and Märta Beijer Foundation and Swedish Research
Council. The author acknowledges Dr. Christer Westerlund, AstraZeneca, Dr. Rolf Johansson,
AstraZeneca, Dr. Ingrid Påhlman, AstraZeneca/Albireo Pharma, Professor Anders Hallberg,
Uppsala University, and Dr. Hanna Andersson, Gothenburg University for carefully scruti-
nizing the manuscript and for constructive criticism, not least concerning the drawing of the
chemical formula.
REFERENCES
1. Burger E. Peptide hormones and neuropeptides. Proteolytic processing of the precursor regulatory
peptides. Arzneimittelforschung 1988;38(5):754–761.
2. Hallberg M, Nyberg F. Neuropeptide conversion to bioactive fragments–an important pathway in
neuromodulation. Curr Protein Pept Sci 2003;4(1):31–44.
3. Nyberg F, Hallberg M. Peptide conversion–a potential pathway modulating G-protein signaling.
Current Drug Targets 2007;8(1):147–154.