BradykMnandinflammatorypain
Andy Dray and Martin Perkins
There is compelling evidence linking bradykinin (BK)
with the pathophysiological processes that accompany
tissue damage and inflammation, especially the pro-
duction of pare and hyperalgesia. Several mechanisms
have been proposed to account for hyperalgesia in-
cluding the direct activation of nociceptors as well as
sensitization of nociceptors through the production of
prostanoids or the release of other mediators. In keeping
with this, antagonists of the BK B2 receptor are
efficacious analgesic and anti-inflammatory agents in
acute inflammatory pain. More recently it has been
suggested that when inflammation is prolonged, BK B1
receptors, which are not expressed in healthy tissues to a
significant degree, also play an important role in the
maintenance of hyperalgesia. This may be one of a
number of adaptive mechanisms that occur peripherally
and centrally following the prolonged activation of
nociceptors during inflammation or injury.
The major kinins, including the nonapeptide brady-
kinin (BK) and kallidin (Lys-BK), are made de novo
from high and low molecular weight kininogen pre-
cursors ~'2. This chain of events follows the activation
of tissue and plasma kallikreins by pathophysiological
stimuli such as tissue trauma, inflammation, anoxia
and low pH. Kallidin and BK appear to act on tissues
in a similar manner, although BK has been more
extensively studied. Newly formed BK acts locally
close to its site of production on a wide variety of
tissues, both neural and non-neural, with effects
ranging from smooth muscle contraction, glandular
secretion and immune cell stimulation, to sensitization
and activation of sensory and sympathetic neurones
(Fig. 1). The effects on neural tissues are of particular
significance as BK is increasingly implicated in the
aetiology of a number of pain conditions including
cardiac pain and pain following tissue anoxia, as well as
the pain associated with inflammation and rheumatoid
diseases 2-~.
The effects of BK are rapidly curtailed by the
actions of proteolytic enzymes, which are abundant in
blood, lung and liver. However a major metabolite
[des-Argg]BK, which is active at one subtype of BK
receptor, B1, is produced through kininase I activity
(Fig. 1), while other enzymes with tissue-specific lo-
cations (kininase II and endopeptidases) further de-
grade BK metabolites to produce inactive species 1.
Bradykinin receptors and acute pain
It has long been known that BK causes acute pain
following application to a blister base in man and after
an intra-arteriai injection in laboratory animals. Indeed
it has been described as the most potent endogenous
algogenic substance known ~-7. The effects produced
by BK, including those on nociceptive neurones, are
due to the activation of specific BK receptors. There
is good evidence for at least two types of bradykinin
receptor, B1 and B2, and both have been intensively
studied in recent years. The B1 receptor shows a
TINS, Vol. 16, No. 3, 1993 © 1993, ElsevierScience Publishers Ltd, (UK)
10-50-fold greater preference for the BK metabolite AndyDrayand
[des-Argg]BK than for BK itself s'9. However, the MartinPerkinsareat
B1 receptor is not expressed to any significant ex- theSandozInstitute
tent in normal tissues, and the acute administration for MedicalResearch,
of [des-Argg]BK is usually ineffective and does not 5 GowerPlace,
London,
elicit pain. Thus the physiological role for this recep- UKWCI E6BN,
tor is unclear. It is possible that B1 receptors may
be of greater significance in pathophysiological con-
ditions, particularly inflammation (see below). B1 re-
ceptors have been characterized best on vascular
smooth muscle where their expression increases
over a period of hours (in vitro or in vivo) and
can be enhanced by a number of agents produced
during infection and inflammation 9.
Most actions of BK, including the acute activation
of nociceptors and the production of pain, are
mediated through the B2 receptor 6'7'1°. A variety of
peptide analogues have been produced that act as
selective antagonists at the B2 receptor (Fig. 2),
so enabling this receptor to be comprehensively
\ /
Kallikreins
Inactive metabolites Kininogens
iininaseII
opeptidases
inaseI / /
[Des-Argg] Bradykin~//~2 Prostanoid
production
Immunecells Sympathetic
and mast cells neurones
NocicepUve neurone
Excitation and sensitization
Fig. 1. The major kinins, kallidin and bradykinin (BK), are made from high and
low molecular weight kininogen precursors following activation of tissue and
plasma kallikreins by pathophysiological stimuli such as tissue trauma,
inflammation, anoxia and low pH. Newly formed BK acts acutely on a variety
of tissues, as shown here, mainly through B2 receptors. Thus BK stimulates
vascular endothelial cells to produce nitric oxide and constricts vascular
endothelial cells to produce plasma extravasation. It stimulates epithelial cells
to induce secretion from the gastrointestinal tract, has chemotactic and
stimulatory effects on immune cells (macrophages and neutrophils) and
degranulates mast cells. BK stimulates sympathetic nerve fibres to induce
vascular reflexes, contracts a variety of smooth muscle tissues and stimulates
prostanoid production from many tissues. Pain and hyperalgesia result from
the excitation and sensitization of nociceptive neurones. BK is rapidly degraded
by several enzymes. A major metabolite [des-Argg]BK is produced through
kininase I activity, while other enzymes (kininase II and endopeptidases) with
tissue-specific locations produce further degradation to form inactive BK
metabolites.
99
Bradykinin
1 H.Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg.oH
PA2
Bradykinin B= receptor antagonists (Guinea-pig ileum)
2 H.Arg-Pro-Pro-Gly-Phe-SerlPhr.~e-Arg.oH 5.9
3 I~Arg-Pro-Hyp-Gly-~Ser'~_~hi Arg.oH 6.9
4 IoArg~Arg-Pro-Hyp-GlyT-LT__~ser-~L~_cj-Arg.oH 8.4
ICso(taM)
Bradykinin B~ receptor antagonists (Rabbit aorta)
5 H.Arg-Pro-Pro-Gly-Phe-Ser-Pro 0.1
6 r~g~Arg - Pro-Hyp-Gly-E~Se r ~ ] O H 0.07
Fig. 2. Bradykinin (BK) and selective bradykinm antagonists. This figure shows
the amino acid sequence of BK (1) and examples of a number of peptide
analogues. In compounds (2) and (3), replacement of the profine residue at
position 7 by D-phenylalanine (tinted rectangles) is critical in conferring
antagonist properties against the B2 receptor. There are many examples of this
first generation of B2 receptor antagonists where other substitutions have
produced an increase in potency and selectivity. Substitutions indicated in the
boxed, open regions either confer or improve antagonistic potency, selectivity,
or both. In addition, some of these changes improve metabolic stabifity.
Compound (4) exempfifies a second generation of antagonists in which a
number of synthetic amino acids have been substituted (Tic represents 1,2,3,4-
tetrahydroisoquinofin-2-yl-carbonyl; Oic represents (3as,7as)-octahydroindol-
2-ylcarbonyl). This compound, HOE 140, made by Hoechst Pharmaceuticals, is
the most potent BK B2 antagonist yet described and appears to be
metabofically stable. The potency of each compound to inhibit bradykinin-
induced contractions in the guinea pig ileum is expressed as the pA2.
Compounds (5) and (6) are the only reported B1 receptor antagonists. The
potency of [des-ArggLeuS]BK and the [des-Arg 1°] derivative of HOE 140
0C5o) is expressed as the concentration, in W~, required to reduce the effect of
the B~ agonist [des-Arg~]BK by 50% on the rabbit aorta preparation.
characterized. These substances have also been criti-
cal for establishing a physiological role for BK as well
as identifying some of its pathophysiological effects.
The B2 receptor has recently been cloned and its se-
quence of 366 amino acids shows a high degree of
homology between a number of species, including hu-
mans n'~2. This receptor belongs to a supeffamily of
receptors whose members all have seven membrane-
spanning domains and are coupled with a G protein.
A number of sites have been identified which are
important for BK binding and for receptor regulation.
M e c h a n i s m s of s e n s o r y n e u r o n e e x c i t a t i o n and
sensitization
BK receptors have been localized to nociceptive
pathways (sensory nerve terminals and small cells in
dorsal root ganglia) by autoradiography of [3H]-BK
(Ref. 6). In keeping with this, excitation of nocicep-
tors by BK has been demonstrated in a number of
circumstances. These range from direct recording of
C-fibre nociceptor activity in vitro and bradykinin-
elicited nociceptive reflexes 13-15, to recording brady-
kinin-mediated excitation of nociceptors in vivo in
skin, skeletal muscle, joints and visceral organs 16-19.
In most cases, the B1 agonist [des-Arg9]BK was
ineffective, and antagonists to the B2 but not B1
receptor blocked the acute activation of sensory
neurones by BK (Refs 10, 20).
Various approaches have been used to elucidate the
receptor-coupled mechanisms of nociceptor acti-
100
vation. Because of the difficulties of studying fine un-
myelinated nociceptive fibres in vivo these studies
have relied heavily on the use of tissues and cells
maintained in vitro. For example clonal cell lines such
as the neuroblastoma-glioma (NG 108-15) hybrid re-
spond to BK and have been used to correlate the
changes in membrane ion conductance with intra-
cellular biochemical events. In these cells BK pro-
duces a biphasic response. Initially there is a transient
hyperpolarization due to the activation of a Ca 2+-
sensitive K + conductance initiated by the release of
Ca2+ from intracellular stores following the formation
of inositol 1,4, 5-trisphosphate (IP3). This hyperpolar-
ization is followed by a prolonged depolarization due
to inhibition of the voltage-dependent K + conduc-
tance 21,22.
Sensory neurones respond somewhat differently to
BK, but some similarities are evident. Thus, BK
produces an immediate depolarization of sensory
neurones and nociceptive fibres that is directly related
to an increase in membrane permeability, mainly to
Na + (Refs 10, 23; Fig. 3). Membrane excitability is
also altered due to effects on a number of cellular
messenger systems. For example, BK stimulates
membrane phospholipase C to generate diacylglycerol
(DAG) and IP3, which activate intracellular protein
kinase C (PKC) and elevate intracellular Ca 2+, re-
spectively 1°'23'24. These events may also be associ-
ated with cell activation. However IP3-mediated re-
lease of intracellular Ca 2+ appears less important in
sensory neurones than in other cell types. Stimulation
of PKC has also been shown to inhibit membrane
Ca2+ conductance 25 but the relationship between
this and increased cellular excitability is less clear,
although it is possible that a Ca2+-dependent K +
conductance may be altered (see below). The BK-
induced changes in sensory neuronal excitability are
coupled to a G protein since the effects were reduced
by GDP-[~-S and prolonged by GTP-y-S, even though
neural activation by BK was insensitive to pertussis
toxin24. BK may also stimulate the production of
arachidonic acid in sensory neurones, although this
appears to be derived from the breakdown of DAG
(Refs 26, 27). It is unclear whether arachidonic acid
acts by itself upon release or whether it is metabolized
further to form prostanoids. However, it is well
known that BK can directly activate membrane
phospholipase A2 (Refs 1, 3) in many cell types,
leading to the production of prostanoids via sub-
sequent cyclooxygenase and lipoxygenase activity.
In visceral sensory neurones, BK indirectly in-
creases excitability via the activation of cAMP and the
subsequent inhibition of Ca2+-dependent K + per-
meability which underlies a post-spike after-hyper-
polarization. This effect of BK allows the cell to fire
repetitively following a stimulus and therefore may be
one of the mechanisms underlying nociceptor sensi-
tization. This action of BK is abolished by indo-
methacin and mimicked by prostaglandins, suggesting
that it is mediated indirectly via prostanoid pro-
duction28,29.
BK-induced sensitization of nociceptors may well
be more important than the immediate short-lasting
nociceptor activation in the overall production of
pain. Nociceptor sensitization probably underlies the
phenomenon of peripheral hyperalgesia that results in
an increase in the perception of and response to
TINS, Vol. 16, No. 3, 1993
pain3°. A contribution to this phenomenon is also
made by a population of peripheral fibres that are
normally insensitive to noxious stimuli (silent noci-
ceptors) but behave as nociceptors after becoming
sensitized by BK or other mediators during inflam-
mation31. These fibres have been studied in a number
of situations including inflammation of the joint and
inflammation of the urinary bladder. The normal
function of silent nociceptors is unclear, and more
information on this type of nerve fibre is required.
A variety of other factors contribute to nociceptor
sensitization. As mentioned above, BK stimulates the
production and release of a number of endogenous
chemicals including prostanoids (PGE2, PGI2) from
many cell types and sympathetic neurones, and
cytokines [interleukin-1 (IL-1) and tumour necrosis
factor (TNF-00] from macrophages. Prostaglandins
have long been known to induce sensitization of
nociceptors and this is believed to involve activation
of cellular cAMP-dependent mechanisms. Postgangli-
onic sympathetic fibres have been considered to be
an important source of prostanoids in inflammation
and are involved in a number of pathological hyper-
algesic states. They have been directly implicated in
the mechanism of BK-induced plasma extravasation32
and BK-induced hyperaigesia in response to stimu-
lation by pressure 33, but not heat34.
The activation of sensory fibres by BK also causes
the release of neuropeptides such as substance P,
neurokinin A and calcitonin gene-related peptide
(CGRP) 35'36. These peptides can also mediate a
number of local pro-inflammatory effects and con-
tribute to nociceptor sensitization and hyperalgesia.
Substance P and CGRP stimulate vascular endothelial
cells to release the smooth muscle relaxant nitric
oxide and thereby increase tissue blood flow and
induce an inflammatory flare due to dilation of
arterioles. In addition, substance P contracts the
endothelium and allows the leakage of plasma protein
that accounts for neurogenic oedema. Substance P
and neurokinin A may also induce degranulation of
mast cells and release of histamipe, another pro-
inflammatory mediator. In addition to this, a number
of other blood-borne factors such as 5-HT and ATP
(from platelets) may gain access to nociceptors and
contribute to sensitization 37"3s. Finally it is well known
that BK itself is a powerful inflammatory mediator:
inducing vasodilation through stimulation of nitric
oxide release from vascular endothelial cells; produc-
ing plasma extravasation; and inducing the release of
other mediators following mast cell degranulation
(histamine, 5-HT) and tissue stimulation (prosta-
noids) 1'3'4. With such a diversity of direct and indirect
actions it is quite likely that BK has major algesic and
pro-inflammatory influences.
Regulation of bradykinin activity
While all these BK-induced processes are going on,
there are powerful counteractive mechanisms that
limit the actions of this factor in vivo, including rapid
degradation by proteolytic enzymes and rapid desen-
sitization of the BK receptor. Cyclic GMP-dependent
kinase, which phosphorylates the BK receptor, is an
important factor in desensitization. In this case,
desensitization of the response of sensory neurones
to BK results from the production of nitric oxide,
which in turn activates guanylyl cyclase to increase
TINS, Vol. 16, No. 3, 1993
Fig. 3. The mechanisms of bradykinm (BK)-induced activation and sensitization
of nociceptors. BK B2 receptors are coupled with a G protein. Receptor
activation produces a depolarization mainly by increasing the permeability of
the membrane to Na +. This depolarization may activate voltage-dependent
Ca2+ influx. Activation of the B2 receptor also stimulates phospholipase C to
generate diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3) from
membrane phospholipids. Though IP3 may stimulate a rise in intracellular
Ca2+, it is not clear whether this is of significance in sensory neurones.
However, DAG activates protein kinase C (PKC), which can regulate ion
channel proteins and thereby alter membrane excitability as well as regulating
other cellular processes. In addition, BK stimulates the production of
arachidonic acid (AA). This may result from the metabolism of DAG (via the
activity of DAG lipase) rather than via activation of phospholipase A2.
Prostanoids (PGs; especially PGE2 and PGI2), produced mainly from AA
metabolism in other cells, act on nociceptors to induce sensitization of the
membrane. This is achieved via PG receptor-induced activation of adenylyl
cyclase (AC) to generate cAMP and to activate kinases that change the activity
of membrane ion channel proteins (e.g. Ca2+-dependent K + channels) and
other cellular processes. Not shown here, BK may also stimulate the production
of nitric oxide (NO) from L-arginine. NO stimulates guanylyl cyclase,
generating cGMP and thereby activating cGMP-dependent kinases. This may
be involved in regulating BK receptors or the activity of receptor-coupled ion
channels.
cellular cGMP (Refs 39, 40). In addition, desensitiz-
ation may involve membrane glycoproteins and down-
regulation of receptors following aggregation and
internalization of receptor proteins41,42. However, it
is unclear whether receptor desensitization occurs
during tissue injury as endogenous BK may be
metabolized too rapidly for this process to be of any
significance. Also, it is possible that local levels of BK
may never be high enough to induce desensitization.
It is clear though that the stimulatory effects of
endogenous BK can become more intense and per-
sistent due to the sensitization of sensory neurones
by prior exposure to BK as well as of a number of
other agents that are produced during tissue injury
and inflammation.
Bradykinin receptor antagonists
It seems certain, therefore, that the production and
subsequent actions of BK can initiate a powerful ~
positive feedback loop leading to a sustained increase
in the excitability of nociceptors. Because of this we
believe there is a compelling rationale for the develop-
ment of BK antagonists that are likely to have
analgesic activity and therapeutic potential.
To this end, studies in vivo using the first
generation of B2 receptor antagonists, which are
101
 100 -- 100

HOE 140 [des-Arg9Leu 8] Bradykinin

80 I

60
i

e~
40-
l-

20-
iD
>
t9
rr

°t
--20
O-

-20-
J
0.0 1.0 2.0 3.0 4.0 5.0 6.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0

Time (h) Time (h)

Fig. 4. B1 receptors play a major role in inflammatory hyperalgesia. Injection of Freund's complete adjuvant into the
knee joint of an anaesthetized rat induces a mild inflammation lasting several days. When animals were placed on a
transducer 64-70 hours later they easily tolerated a lOO g load on the joint of the untreated leg, but would only tolerate
40 g or so on the inflamed joint, indicating discomfort and hyperalgesia. The hyperalgesia could be reversed by the
administration of analgesic drugs. In the two graphs shown, the degree of analgesia (% reversal of hyperalgesia),
indicating the efficacies of the analgesic drugs, is plotted against time after drug administration (indicated by the
downward arrow at time 1 hour). On the left, the potent B2 antagonist HOE 140 was administered intravenously but
was almost completely ineffective at 51~mol/kg (rifled circles) and lO t~mol/kg (filled squares). Filled triangles represent
the saline control. However, on the right, the B1 antagonist [des-ArggLeuB]BK at I nmol/kg intravenously (filled
squares) and 10 nmol/kg intravenously (filled circles) produced a prolonged and dose-related reversal of hyperalgesia.
Filled triangles represent the saline control These experiments indicated that the activation of B1 receptors is much more
important than activation of B2 receptors in maintaining the hyperalgesia during prolonged inflammation of the joint.

peptide analogues, showed that these compounds
have analgesic and anti-inflammatory activities 2'6' 15,43.
Thus, hyperalgesia induced by injecting urate crystals
into the rat paw or abdominal contractions produced
by intraperitoneal injections of acetic acid in mice
were blocked by Be receptor antagonists 7. In addition,
the acute oedema induced by both BK and car-
rageenan was also shown to be due to B2 receptor
stimulation 44. Studies of BK antagonists in humans
have also shown that the kinin receptors exposed in
the blister base are of the B2 subtype 45. More
recently, a second generation of BK antagonists,
including HOE 140 (Fig. 2), have been developed that
have improved antagonist potency at the Be receptor
and increased resistance to proteolytic enzyme
degradation 46. HOE 140 shows analgesic and anti-
inflammatory activity in a number of acute tests 47 but
seems to us to be less efficacious in vivo than
expected in light of its potency at B2 receptors in
vitro. This was particularly noticeable when its activity
was tested in conditions of persistent inflammatory
hyperalgesia (Fig. 4)48.
Bradykinin receptors and persistent
inflammatory hyperalgesia
The involvement of BK receptors in nociception
has been studied mainly in animal models in which
there was little underlying pathology. More recently, a
number of reports have suggested that the expression
of B~ receptors 49 may be induced by inflammatory
processes or by bacterial infection. Indeed, earlier
observations showed that endotoxin from Escherichia
coli could induce B1 responses in vascular smooth
muscle in rabbits 9'5° and that this action might explain
the profound hypotension seen in animals that had
been pretreated with bacterial lipopolysacchafide
(LPS), a model for septiceamic shock 51.
Many factors may be involved in regulating the
expression of B1 receptors. Recently, a role for
cytokines has been highlighted. Thus IL-1 and IL-2
can increase responses to [des-Arg9]BK in vitro,
and agents that stimulate macrophages, a major
source of interleukins, also increase responses to
[des-Argg]BK in vitro 51'52. More directly, it has been
shown that activation of the B1 receptor stimulates
the release of TNF-o~ and IL-1 from macrophages and
prostanoid release from synovial cells 52'53. These
observations raise the possibility that macrophage
activity can be regulated by both cytokines and kinins
(Fig. 5). In this regard it is worth noting that
cytokines also activate macrophages to express cyto-
static/cytotoxic activity due to the production of nitric
oxide following the induction of a CaZ÷-independent
form of nitric oxide synthase 54.
Recently, cytokines, particularly interleukins (IL-I[3,
IL-6 and IL-8) and TNF-o~, have been strongly im-
plicated in mechanisms of inflammatory hyperal-
gesia 55-57. Their effects are mediated in part via
prostanoid production, being reduced by indomethacin
pretreatment, and also indirectly through the activation
of peripheral sympathetic neurones (Fig. 5). Pres-
ently, it is unclear whether BK activates the release

102 TINS, Vol. 16, No. 3, 1993


of cytokines or vice versa, or whether both processes
can occur together. It is, however, probable that a
specific role for BI receptors in hyperalgesia may only
become apparent in conditions that involve a per-
sistent inflammation. Support for this has been
obtained in two models: persistent inflammatory
hyperalgesia produced by exposing the rat hind paw
to ultraviolet irradiation, and hyperalgesia that ac-
companies the prolonged inflammation induced by
Freund's adjuvant being injected into the rat knee
joint. In these studies the B~ receptor antagonist
[des-Arg9LeuS]BK reversed the hyperalgesia in
both models, but the B2 receptor antagonist HOE 140
was relatively inactive. In addition, the B~ agonist
[des-Arg9]BK increased the degree of hyperalgesia
(Fig. 4) 48. Finally, it is significant that the concen-
tration (normally 0.1-1.0ng/ml) of both BK and
[des-Argg]BK rise (approximately 5-10-fold) during
inflammation 8,58.
Thus, many of the components required for the
induction, expression and activation of B1 receptors
exist. The lack of B1 receptor involvement in studies
of acute nociception, described earlier, could have
been due to the absence of conditions necessary to
express the receptor. Indeed there are at present a
number of examples of dynamic and adaptive changes
in nociceptive pathways that only occur during in-
flammation and persistent activation of nociceptive
afferent neurones. Among these changes are the
production of endogenous opioids by immune cells and
the expression of opioid receptors in peripheral nerve
fibres 59. Changes in the expression, production and
release of sensory neuropeptides from sensory
neurones also occur, and NMDA-type excitatory
amino acid receptors play a more prominent role in
increasing the excitability of the spinal cord 6°.
Further studies will be required to determine
whether B~ receptors occur on sensory nerves, or
whether nociceptors are sensitized indirectly follow-
ing B~ receptor-mediated release of cytokines or
other inflammatory mediators from surrounding
tissues. The view invited by these recent data is that
TINS, Vol. 16, No. 3, 1993
Fig. 5, During inflammation, bradykinin (BK) and [des-
Argg]BK initiate a cascade of events which activate and
sensitize nociceptors. The various reactions produce a self-
reinforcing cycle of events. During acute inflammation,
bradykinin directly activates nociceptors (via B2 receptors),
and also stimulates the production of a variety of other
mediators, including prostanoids from a variety of cells
including sympathetic post-ganglionic fibres, cytokines
from macrophages or other cells, nitric oxide from
macrophages and vascular endothefium, other blood-
~etic borne factors (e.g. 5-HT) released into interstitial fluid
following plasma extravasation, histamine and 5-HT
following mast cell degranulation, and neuropeptides
released from peptidergic afferent fibres. These sub-
stances sensitize nociceptors and alter the sensitivity of the
BK receptors. During chronic inflammation, sensitivity to
the BK metabollte, [des-Argg]BK is increased, and the B1
Nociceptive receptor plays an increasingly dominant role m the
neurone accompanying hyperalgesia. However, it is not known if
BI receptors occur on nociceptors or on sympathetic
fibres. The production of cytokines (TNF-m IL- 1, IL-6 and
IL-8) causes hyperalgesia, but the mechanisms involved
are poorly understood. Hyperalgesia produced by these
agents is mediated partly via the production of prosta-
noids and in some cases via the activation of sympathetic
fibres.
BK B2 receptors may have a more significant role to
play in the earlier stages of inflammatory pain, and
during prolonged inflammation [des-Arg9]BK acting
on the B1 receptor becomes more important for the
maintenance of the hyperalgesic state. The desen-
sitization of B2 receptors or an enhancement of BK
degradation may have a role to play in shifting the
balance from an acute to a chronic inflammatory state.
The speculation about which BK receptor is of most
importance may be premature as it is possible that
further receptor subtypes may be found in patho-
logical conditions. Ultimately, however, the import-
ance of BK in inflammatory pain and hyperalgesia will
be resolved when a BK antagonist with specific
receptor selectivity is shown to be efficacious in the
clinic.
Selected references
1 Bhoola, K. D., Figueroa, C. D. and Worthy, K. (1992)
Pharmacol. Rev. 44, 1-80
2 Bathon, J. M. and Proud, D. (1991) Annu. Rev. Pharmacol.
Toxicol. 31, 129-162
3 Farmer, S. G. and Burch, R. M. (1992)Annu. Rev. PharmacoL
ToxicoL 32, 511-536
4 Meller, S. T. and Gebhart, G. F. (1992) Neuroscience 48,
501-524
5 Keele, C. A. and Armstrong, D. (1964) Substances Producing
Pain and Itch, Arnold
6 Steranka, L. R. et al. (1988) Proc. Natl Acad. 5ci. USA 85,
3245-3249
7 Steranka, L. R. etaL (1987) Eur. J. PharmacoL 136, 261-262
8 Regoli, D. and Barabe, J. (1988) Pharmacol. Rev. 32, 1-46
9 Bouthillier, J., Deblois, D. and Marceau, F. (1987) Br. J.
Pharmacol. 92,257-264
10 Dray, A., Patel, I. A., Perkins, M. N. and Rueff, A. (1992) Br. J.
Pharmacol. 107, 1129-1134
11 McEachern, A. et al. (1991) Proc. Natl Acad. 5ci. USA 88,
7724-7728
12 Hess, J., Borkowski, J., Young, G., Strader, C. and Ransom, R.
(1992) Biochem. Biophys. Res. Commun. 184, 260-268
13 Lang, E., Novak, A., Reeh, P. W. and Handwerker, H. O.
(1990) J. NeurophysioL 63,887-901
14 Juan, H. and Seewann, S. (1980) Eur. J. Pharmacol. 65,
267-278
15 Griesbacher, T. and Lembeck, F. (1987) Br. J. Pharmacol. 92,
333-340
16 Foch, S. and Mense, S. (1976) Brain Res. 105, 459-467
17 Berkely, K. J., Hatta, H., Robbins, A. and Stao, Y. (1990)
103
 J. Neurophysiol. 63, 256-272
18 Mizumura, K., Minagawa, M., Tsujii, Y. and Kumazawa, T.
(1990) Pain 40, 221-227
19 Sato, J., Mizumura, K. and Kumazawa, T. (1989) J. Neuro-
physiol. 62, 119-125
20 Haley, J. E., Dickenson, A. H. and Schachter, M. (1989)
Neurosci. Lett. 97, 198-202
21 Brown, D. A. and Higashida, H. (1988) J. Physiol. 297,
167-184
22 Higashida, H., Streaty, R. A., Klee, W. and Nirenberg, M.
(1986) Proc. Natl Acad. 5ci. USA 83,942-946
23 Burgess, G. M., Mullaney, I., McNeill, M., Dunn, P. M. and
Rang, H. P. (1989) J. Neurosci. 9, 3314-3325
24 McGuirk, S. M. and Dolphin, A. C. (1992) Neuroscience 49,
117-128
25 Boland, L. M., Allen, A. C. and Dingledine, R. (1991)
J. Neurosci. 11, 1140-1149
26 Gammon, C. M., Allen, A. C. and Morell, P. (1989)
J. Neurochem. 53, 95-101
27 Allen, A. C., Gammon, C. M., Ousley, A. H., McCarthy, K. D.
and Morell, P. (1992) J. Neurochem. 56, 1130-1139
28 Weinreich, D. (1986) Eur. J. PharmacoL 132, 61-63
29 Weinreich, D. and Woderlin, W. F. (1987) J. Physiol. 394,
415-427
30 Raja, S. N., Meyer, R. A. and Campbell, J. N. (1988)
Anaesthesiology 68, 571-590
31 McMahon, S. B. and Kolzenburg, M. (1990) Trends Neurosci.
13, 199-201
32 Lee, A., Coderre, T. J., Basbaum, A. I. and Levine, J. D. (1991)
Brain Res. 557, 146-148
33 Levine, J. D., Taiwo, Y., Collins, S. D. and Tam, J. K. (1986)
Nature 323, 158-160
34 Koltzenburg, M., Kress, M. and Reeh, P. W. (1992) Neuro-
science 46, 465-473
35 Maggi, C. A., Patacchini, R., Rovero, P. and Giachetti, A.
J. Autonom. Pharmacol. (in press)
36 Geppetti, P. eta/. (1990) Neuroscience 38, 687-692
37 Rueff, A. and Dray, A. (1992) Neuroscience 50, 899-905
38 Taiwo, Y. O. and Levine, J. D. (1992) Neuroscience 48,
485-490
39 Burgess, G. M. etal. (1989)J. Neurochem. 53, 1212-1218
Neuronatrophyduringaging:programmedor sporadic?
C a l e b E. F i n c h
CalebE. Finch Atrophy of neurons is a common change during aging
is at thein laboratory rodents and humans. However, cholin-
Neurogerontology ergic neurons of the same type have been found to
Division, Andrus atrophy, hypertrophy or not change at all, according to
GerontologyCenter various reports on different species and genotypes.
and Dept of Biological
Sciences, University
Possible factors responsible for these diverse outcomes
of Southern include species- and genotype-specific aging changes
Cafifomia, and age-related diseases. An open question is whether
LosAngeles, slowly evolving changes in neuronal size share any
CA 90089-0191, mechanisms with the rapid programmed death of
USA. neurons that occurs during development. Progress in
the study of neuronal atrophy with aging may be
furthered by using fewer rodent genotypes.
Neuron atrophy is described by a decrease in the size
of the perikaryon, nucleus, nucleolus or dendritic
arbor. This process is a feature of Alzheimer's disease
(AD) and Parkinson's disease (PD), but also occurs to
a modest degree during normal aging in many animals.
Neuron atrophy during aging, particularly of forebrain
cholinergic (FB-Ch) neurons, is being studied in
relation to cognitive functions. If a larger perikaryal
size is used to distinguish neurons from glia it is
possible that apparent neuron loss will be over-
estimated - neurons may still be present, but just
reduced in size. This should be borne in mind when
104 © 1993, ElsevierScience PublishersLtd, (UK)
40 McGehee, D. S., Goy, M. F. and Oxford, G. S. (1992) Neuron
9, 315-324
41 Dray, A., Patel, I. A., Perkins, M. N., Rueff, A. and Urban, L.
(1992) in Recent Progress in Kinins (Bonnet, G. et al., eds),
pp. 93-97, Birkhauser
42 Roberts, R. A. and Gullick, W. J. (1990) Biochemistry 29,
1975-1979
43 Steranka, L. R., Burch, R. M., Vavrek, R. J., Stewart, J. M. and
Enna, S. J. (1988)Adv. Exp. Med. Biol. 236, 111-127
44 Neppl, N. etal. (1991)Acta PhysioL Scand. 142, 141-147
45 Whalley, E. T., Clegg, S., Stewart, J. M. and Vavrek, R. J.
(1987) Naunyn-Schmiedeberg's Arch. Pharmakol. 336,
652-655
46 Hock, F. J. et al. (1991) Br. J. Pharmacol. 102, 769-774
47 Beresford, I. J. M. and Birch, P. J. (1992) Br. J. PharmacoL
105, 135P
48 Perkins, M. N., Campbell, E. and Dray, A. Pain (in press)
49 Farmer, S. G., McMillan, B. A., Meeker, S. N. and Burch,
R. M. (1991)AgentsActions 34, 191-193
50 Regoli, D., Marceau, F. and Lavigne, J. (1981) Eur. J.
Pharmacol. 71, 105-115
51 DeBIois, D., Bouthillier, J. and Marceau, F. (1989) Immuno-
pharmacology 17, 187-198
52 Bathon, J. M., Manning, D. C., Goldman, D. W., Towns,
M. C. and Proud, D. (1992) J. Pharmacol. Exp. Ther. 260,
384-392
53 Tiffany, C. W. and Burch, R. M. (1989) FEBS Lett. 247,
189-192
54 Moncada, S., Palmer, R. M. and Higgs, E. A. (1991)
Pharmacol. Rev. 43, 109-142
55 Ferreira, S. H., Lorenzetti, B. B., Bristow, A. F. and Poole, S.
(1988) Nature 334, 698-700
56 Cunha, F. Q., Lorenzetti, B. B., Poole, S. and Ferreira, S. H.
(1991) Br. J. Pharmacol. 104, 765-767
57 Cunha, F. Q., Poole, S., Lorenzetti, B. B. and Ferreira, S. H.
(1992) Br. J. Pharmacol. 107, 660-664
58 Damas, J., Bourdon, V., Remacle-Volon, G. and Adam, A.
(1990) Br. J. Pharmacol. 101,418-422
59 Stein, C. (1991)J. Pain Symptom Manage. 6, 119-124
60 Dubner, R. and Ruda, M. A. (1992) Trends Neurosci. 15,
96-103
considering the present controversies about neuron
loss during aging and AD; neuron atrophy may
produce an apparent cell loss in these situations 1"2.
This review considers size changes in the FB-Ch and
dopaminergic neurons of the substantia nigra pars
compacta (SNc); FB-Ch neurons show diverse
changes in size during aging. I will discuss whether
changes in cell size are programmed or sporadic, the
roles of genotype and species differences in aging
patterns, and mechanisms responsible for the changes
in neuronal size, including systemic influences.
Cell atrophy is not restricted to senescence and is
not necessarily detrimental to cell function. Neurons,
like many other cells, may atrophy reversibly during
adult life history. For example, song birds show
hormonaUy driven seasonal cycles in perikaryal size
and dendritic complexity3. In addition, atrophied cells
can remain metabolically active. For example, the
liver shrinks dramatically during starvation; yet, the
shrunken hepatocytes remain active in synthesizing
RNA and protein and rapidly recover their size and
nutrient depots after feeding recommences.
In order to anticipate the vexing variability dis-
played by aging neurons, it helps to know that some
aging changes are canonical, that is, they occur during
aging in all individuals of related species (Ref. 2,
TINS, Vol. 16, NO. 3, 1993