Content uploaded by Andrea De Laurentiis
Author content
All content in this area was uploaded by Andrea De Laurentiis
Content may be subject to copyright.
64
Ann. N.Y. Acad. Sci. 1057: 64–84 (2005). © 2005 New York Academy of Sciences.
doi: 10.1196/annals.1356.004
The Nitric Oxide Theory of Aging Revisited
S. M. MCCANN,a C. MASTRONARDI,b A. DE LAURENTIIS,a AND V. RETTORIa
aCentro de Estudios Farmacológicos,
Consejo Nacional de Investigaciones Científicas y Técnicas (CEFYBO-CONICET),
School of Medicine, UBA, Paraguay 2155 piso 16, 1121, Buenos Aires, Argentina
bDivision of Neuroscience, Oregon National Primate Research Center,
Beaverton, Oregon 97006, USA
ABSTRACT: Bacterial and viral products, such as bacterial lipopolysaccharide
(LPS), cause inducible (i) NO synthase (NOS) synthesis, which in turn produc-
es massive amounts of nitric oxide (NO). NO, by inactivating enzymes and lead-
ing to cell death, is toxic not only to invading viruses and bacteria, but also to
host cells. Injection of LPS induces interleukin (IL)-1, IL-1␣, and iNOS syn-
thesis in the anterior pituitary and pineal glands, meninges, and choroid plex-
us, regions outside the blood–brain barrier. Thereafter, this induction occurs
in the hypothalamic regions (such as the temperature-regulating centers),
paraventricular nucleus (releasing and inhibiting hormone neurons), and the
arcuate nucleus (a region containing these neurons and axons bound for the
median eminence). Aging of the anterior pituitary and pineal with resultant de-
creased secretion of pituitary hormones and the pineal hormone melatonin, re-
spectively, may be caused by NO. The induction of iNOS in the temperature-
regulating centers by infections may cause the decreased febrile response in the
aged by loss of thermosensitive neurons. NO may play a role in the progression
of Alzheimer’s disease and parkinsonism. LPS similarly activates cytokine and
iNOS production in the cardiovascular system leading to coronary heart dis-
ease. Fat is a major source of NO stimulated by leptin. As fat stores increase,
leptin and NO release increases in parallel in a circadian rhythm with maxima
at night. NO could be responsible for increased coronary heart disease as obe-
sity supervenes. Antioxidants, such as melatonin, vitamin C, and vitamin E,
probably play important roles in reducing or eliminating the oxidant damage
produced by NO.
KEYWORDS: nitric oxide synthase; cyclic GMP; cyclooxygenase; bacterial
lipopolysaccharide; cytokines; hypothalamus; brain; pituitary gland; pineal
gland; degenerative diseases; inflammation; infection; stress; coronary heart
disease
INTRODUCTION
The causes of aging are undoubtedly multifactorial. One of the most prominent
current theories of aging is the free radical theory. According to this theory, free rad-
Address for correspondence: Samuel M. McCann M.D., Av. Cordoba 2465, piso 13 “A,” 1120
Buenos Aires, Argentina. Voice/fax: 54-11-49634473.
smmccann2003@yahoo.com
65MCCANN et al.: NITRIC OXIDE AND AGING
icals generated through mitochondrial metabolism can act to cause abnormal func-
tion and cell death. Various toxins in the environment can injure mitochondrial
enzymes, leading to increased generation of free radicals that, over the life span,
eventually play a major part in aging.1
At the Third International Symposium on the Neurobiology and Neuroendocri-
nology of Aging, we2 presented evidence to suggest that excessive production of the
free radical nitric oxide (NO) in the central nervous system (CNS) and its related
glands, such as the pineal and anterior pituitary, may be the most important factor in
the aging of these structures. Evidence for this hypothesis has been accruing rapidly.
Because of the fact that the synthesis of inducible NO synthase (iNOS) following in-
jection of bacterial lipopolysaccharide (LPS) in the rat was much greater outside the
blood–brain barrier,3 for example, in the anterior pituitary and pineal gland, than in-
side this barrier, it occurred to us that NO might play a role in aging of every organ
system of the body. The evidence for this concept is particularly well developed to
explain the pathogenesis of coronary arteriosclerosis.
NO has been found to play an ubiquitous role in the control of physiological func-
tions throughout the body. The toxic effects of the soluble gas occur following infec-
tion or inflammations that cause the production of iNOS, which floods the tissue
with toxic concentrations of NO. Here, we briefly review the methods of formation
of NO in the body and its physiological role in the organs where we believe it also
plays an important role in aging, and then present the evidence that NO is largely
responsible for the aging process.
PRODUCTION OF NO IN THE BODY
Three isoforms of NO synthase (NOS) occur in the body.4,5 The first of these is
iNOS. LPS or cytokines [interleukins (IL)-1, IL-2, IL-6 and tumor necrosis factor
alpha (TNF-α)] act via receptors on the cell surface of immune cells, particularly
macrophages and other cells, such as endothelial cells, to activate DNA-directed
mRNA synthesis which induces synthesis of iNOS. These LPS receptors are toll-like
receptors (TLRs).6 They activate nuclear factor κ B (NFκB), which activates DNA-
directed synthesis of not only pro-inflammatory cytokines (such as IL-1, IL-6, and
TNF-α), but also iNOS. A single injection of LPS leads to the release of massive
amounts of NO, beginning within a few hours, peaking at 18 h, and then declining
by 24 h. This enzyme is active as soon as it is induced, because it contains within
itself calcium and calmodulin, which are required for activation of all isoforms of
the enzyme. Like all forms of NOS, iNOS converts arginine and equimolar mole-
cules of oxygen in the presence of various cofactors, such as NADPH and tetrahy-
drobiopterin, into NO and equimolar amounts of citrulline. The NO decays in
solution with a half time of 5 to 10 sec, whereas citrulline remains in the cell and can
even be recycled into arginine to provide further substrate for conversion by NOS
into NO and citrulline. There is also a transport mechanism that carries arginine into
the cell to provide substrate.
NO is a soluble gas and diffuses to neighboring cells, bacteria, and viruses. The
high concentration formed after the induction of iNOS interferes with metabolism,
leading to death of viral and bacterial invaders and also host cells. NO blocks cellular
66 ANNALS NEW YORK ACADEMY OF SCIENCES
enzymes required in metabolism and also activates soluble guanylate cyclase (sGC),
a soluble enzyme present in the cytoplasm of cells. The activation occurs via inter-
action of NO with the Fe2+ in the heme portion of the molecule, thereby altering its
conformation and activating it. This causes conversion of guanosine triphosphate
(GTP) to cyclic guanosine monophosphate (cGMP), which mediates many of the
physiological actions of NO in mammalian cells. NO also activates cyclooxygenase,
generating prostaglandins and lipoxygenase, and forming leukotrienes, which are
toxic in high concentrations.
Endothelial NOS (eNOS) is a constitutive enzyme present in vascular endo-
thelium.7 Cholinergic stimulation by parasympathetic innervation of the vessels ac-
tivates the enzyme by increasing the intracellular free calcium concentration [Ca2+],
which combines with calmodulin. This interacts with eNOS, activating it. The NO
produced diffuses to overlying smooth muscle and activates sGC. The cGMP re-
leased reduces intracellular [Ca2+], thereby relaxing the vascular smooth muscle.
NO exercises tonic vasodilator tone in the vascular system since blockade of NO
synthase leads to rapid development of hypertension.
A third isoform of the enzyme, neural NOS (nNOS) is a constitutive enzyme, like
eNOS, present in many neurons in the central and autonomic nervous systems.7 It is
present in high concentrations in the cerebellum, cerebral cortex, and hippocampus,
and in very high concentrations in the hypothalamus, particularly in the paraventric-
ular and supraoptic neurons. The axons of these neurons project to the median emi-
nence and neural lobe of the pituitary gland, where very large amounts of enzyme
are present. This form of the enzyme, like eNOS, requires activation by synaptic in-
put, causing elevation of intracellular Ca2+ concentrations, which combines with
calmodulin, which activates the enzyme leading to the production of NO.
Recently, a mitochondrial form of the enzyme has been discovered localized in
the inner mitochondrial membrane. The NO produced by this enzyme has the poten-
tial to produce cellular death by oxidizing intramitochondrial structures.8,9
ROLE OF NO IN THE CNS
The first evidence that NO played a role in the CNS was derived from experi-
ments using cerebellar or hippocampal explants.7,10 These showed that NO plays an
important role in cerebellar function via activation of sGC and induction of cGMP
formation. However, because of the complexity of function in the cerebellum, it is
difficult to say what particular physiologic functions of the cerebellum are altered by
NO. Garthwaite and colleagues11 showed that NO may be important in induction of
long-term potentiation in the hippocampus, which is thought to play a role in mem-
ory.4–7 Palmer and colleagues10 showed by mass spectroscopy that the substance
was indeed NO. Earlier work had previously pointed clearly to NO; however, it is
possible that the major product is not NO itself, but nitroso compounds formed by
combination of NO with various other compounds. Indeed, slow release of NO from
nitroso compounds may account for a longer duration of action of NO than can be
explained on the basis of the free radical itself, which, as indicated above, decays
within seconds in solution.
67MCCANN et al.: NITRIC OXIDE AND AGING
ROLE OF NO IN HYPOTHALAMIC FUNCTION
It has been clearly established that NO plays a key role in controlling the physi-
ological release of a number of hypothalamic peptides and classical neurotransmit-
ters. It has been shown to control the release of corticotropin-releasing hormone
(CRH) by the CRH neurons in the paraventricular nucleus (PVN).12 The stimulatory
effect is mediated by cholinergic neurons that act on muscarinic receptors to stimu-
late the release of NO from nNOS-containing neurons, termed NOergic neurons, in
the PVN. This NO diffuses to the CRH-containing neurons, and activates the release
of CRH. The mechanism involves stimulation of release of prostaglandin E2 (PGE2)
and leukotrienes in the CRH neurons via activation of cyclooxygenase and lipoxy-
genase. In the case of the cyclooxygenase (COX), it is clear that NO activates con-
stitutive COX (COX-1) by interaction of the free radical with the heme group of the
enzyme, altering its conformation. Lipoxygenase also contains Fe2+, but a heme
group has yet to be demonstrated. Nonetheless, this enzyme is definitely activated
because sodium nitroprusside (NP), which releases NO, increases conversion of 14C
arachidonic acid to leukotrienes.13
NO is thought to act physiologically, mainly by activation of sGC.5 Indeed, NP
increases the release of cGMP from hypothalamic explants.14 We hypothesize that
NO releases CRH and other releasing hormones via the activation of GC, which re-
leases cGMP, which in turn increases intracellular [Ca2+], leading to activation of
phospholipase-A2. This converts membrane phospholipids into arachidonate, which
is then converted to PGE2 and leukotrienes by the activated cyclooxygenase and li-
poxygenase, respectively.7 The result is activation of adenylate cyclase, conversion
of ATP to cAMP, which activates protein kinase A, causing extrusion of CRH gran-
ules. The CRH then enters the hypophyseal portal vessels, reaches the anterior lobe
of the pituitary, and causes release of adrenocorticotropic hormone (ACTH).12
ROLE OF NO IN LUTEINIZING HORMONE-RELEASING
HORMONE RELEASE
Release of LH is pulsatile and controlled by pulsatile release of luteinizing hor-
mone-releasing hormone (LHRH) into hypophyseal portal vessels that deliver it to
the sinusoids of the anterior pituitary, where it acts on the gonadotropes to promote
the release of LH. LHRH also increases follicle-stimulating hormone (FSH) release
to a lesser extent.
Release of FSH can occur without LH release and vice versa, but both hormones
are released concomitantly in about 50% of the pulses. Immunocytochemistry re-
vealed a different localization in the hypothalamus of the two peptides. Most neu-
rons that contain LHRH or FSH-releasing hormone (FSHRH) contain only one of
the peptides, but a few contain both peptides.15 Specific receptors for each peptide
activate NOS, which activates release of each peptide by cGMP.16
On the basis of in vitro studies, we postulate that the pathway of activation of
LHRH release is as follows: either direct activation of noradrenergic terminals in the
region of the arcuate-median eminence region or their activation by glutamergic neu-
rons, via N-methyl-D-aspartic (NMDA) receptors, causes the release of norepineph-
rine, which acts by α1 adrenergic receptors on the NOergic neurons to cause the
68 ANNALS NEW YORK ACADEMY OF SCIENCES
release of NO.17 This diffuses to the LHRH terminals intermingled with NOergic
neurons nearby in the median eminence–arcuate region, stimulating the release of
LHRH by the identical mechanism described above for CRH—namely, by activation
of sGC, COX-1 and lipoxygenase—leading to the release of LHRH secretory gran-
ules into the hypophyseal portal vessels.18 The intermingling of NOergic and LHRH
terminals in this region has been demonstrated.18
NO controls LHRH release, which mediates mating behavior in both male and fe-
male rats by activating brain stem neurons controlling lordosis in the female and pe-
nile erection in the male rat. The role of NO in mediating LHRH release controlling
both of these activities has been demonstrated in vivo.19 The penile erection is con-
trolled by pelvic nerve cholinergic activation of nNOS. This NO activates guanylyl
cyclase in the corpora cavernosa penis, causing relaxation of the corpora and erec-
tion. When stimulation stops, the cGMP is degraded, causing loss of erection.
Sildenafil citrate (Viagra) inhibits the phosphodiesterase responsible for degrading
cGMP, thereby prolonging erection.20
We have demonstrated that oxytocin stimulates the release of LHRH via NO. The
low concentrations required (10–8–10–10 M) probably fall within the physiological
range in view of the high concentrations of oxytocin in the median eminence in juxta-
position to LHRH terminals.21 There is an ultrashort-loop negative feedback by which
NO released from NOergic neurons feeds back to inhibit the release of oxytocin.
Not only does NO stimulate the release of many hypothalamic peptides, but it also
stimulates the release of the inhibitory neurotransmitter, gamma amino-butyric acid
(GABA).22 This blocks the response of the LHRH terminal to NO. Therefore, GABA
mediates a feed-forward negative feedback to inhibit pulsatile LHRH release. Further-
more, NO also inhibits the release of both norepinephrine and dopamine from the me-
dial basal hypothalamus, constituting another negative feedback of pulsatile LHRH
release by feeding back on the terminals of the noradrenergic and dopaminergic neu-
rons to inhibit the release of both of these transmitters, one of which, and probably
both of which, stimulate the release of NO that drives LHRH release.23
Interestingly, 30 min after addition of norepinephrine or NMDA, both stimulators
of LHRH release, there is an increase in content of NOS in the hypothalamus mea-
sured by the citrulline method.14 This method is an index of the quantity of NOS in
the tissue because the labeled arginine is added to the homogenate and the labeled
citrulline formed is measured. That this represents de novo synthesis of NOS is in-
dicated by the fact that this effect is blocked by the inhibitor of DNA-directed RNA
synthesis, actinomycin D (Rettori and colleagues, unpublished observations).
Thus, pulsatile LHRH release is mediated by noradrenergic neuronal terminals
that activate NOergic neurons. The NO synthesized diffuses to the LHRH neurons
and activates LHRH release. The pulses of LHRH are terminated by NO-induced re-
lease of GABA that blocks the response to the LHRH neuron to NO and NO-induced
inhibition of further norepinephrine and dopamine release.
ROLE OF NO IN CONTROL OF OTHER HYPOTHALAMIC PEPTIDES
AND NEUROTRANSMITTERS
Growth hormone (GH) release is also pulsatile, but the greatest release occurs
during early sleep at night in humans. In vivo studies have shown that the inhibitor
69MCCANN et al.: NITRIC OXIDE AND AGING
of NO synthase, Nγ-monomethyl-L-arginine (NMMA), can block pulsatile release of
GH in the rat.24 Furthermore, NO can also stimulate somatostatin release and its
messenger RNA levels in the PVN by activation of sGC as determined in in vitro
studies utilizing explants of the paraventricular region.25 We hypothesize that the
pulsatile release of GH that occurs under normal conditions is brought about princi-
pally by NO stimulation of GH-releasing hormone (GHRH) release. At the same
time, somatostatin release is probably inhibited. In the interpulse interval when
GHRH release is absent, somatostatin release mediated by NO increases. IL-1 not
only inhibits GHRH release, but also stimulates somatostatin release, thereby inhib-
iting GH release during infections. The IL-induced prolactin release is also mediated
by NO26 probably by NO stimulation of prolactin-releasing peptides, such as oxyto-
cin,21 and by inhibition of the release of dopamine, a potent prolactin release-inhib-
iting hormone, into the hypophyseal portal vessels.23,27
All of these results indicate that under physiological conditions, NO plays a fun-
damental part in the control of neurotransmitter and neuropeptide release in the hy-
pothalamus. Therefore, these areas where NO is released physiologically will be
subjected to low levels of pulsatile NO throughout the life of the individual. It is not
clear whether such levels can be toxic, but they may be. No studies have been done
to determine whether long-term exposure to low levels of NO is damaging to neu-
rons and/or glia. Interestingly, Schultz and colleagues28 reported neurofibrillary de-
generation in nerve fibers in the arcuate nucleus of aged men but not women.
Possibly, pulsatile release of NO driving LHRH release over the life span may have
caused this degeneration. Because aged women failed to show the degeneration,
there appears to be a sex difference, probably mediated by sex hormones such as es-
trogen, which has been shown to alter NOS concentrations.
THE ROLE OF NO IN CONTROL OF POSTERIOR
PITUITARY FUNCTION
The neural lobe of the pituitary is one of the regions richest in NOS in the rat,
with a predominance of the nNOS isoform, suggesting that NO may play a role in
controlling the release of neuropeptides and neurotransmitters from the posterior pi-
tuitary.7,29 NADPH-diaphorase, used as a marker of NOS, was co-localized with va-
sopressin (VP) and oxytocin (OT) in the hypothalamic-neurohypophyseal
system.30,31 The synthesis of NO by oxytocinergic or vasopressinergic neurons sug-
gests that it may participate in auto- and/or cross-regulation of OT and VP secretion.
Our studies indicate that NO donors reduce OT secretion from the neural pituitary
lobe and we postulated that released NO may suppress OT secretion through an ul-
trashort-loop negative feedback mechanism.21 It has been reported that intracere-
broventricular administration of L-NAME enhanced plasma levels of both OT and
VP peptides.32 Since NOS activity increases following salt loading and dehydration,
it has been suggested that this increase may provide a negative feedback to prevent
overstimulation of OT and VP release.33
Tachykinins belong to a family of peptides that includes substance P and neuro-
kinin A (NKA). They are contained in hypothalamic neurons and nerve fibers and
secretory cells of the posterior and anterior pituitary lobes, suggesting that these pep-
tides may have a physiological role in the control of pituitary function.34 Some ac-
70 ANNALS NEW YORK ACADEMY OF SCIENCES
tions of tachykinins are known to be exerted through NO release,35,36 and NOS
immunoreactivity was detected in some nerve terminals with NK-2 receptors,37 the
receptor subtype to which NKA binds with preferential affinity. Although we ob-
served an inhibitory effect of NKA on OT release by activation of NOergic neurons,
when NOS activity was blocked, NKA stimulated OT release, thus suggesting that
NKA may have a dual effect on OT release, decreasing it through NO and increasing
OT release by an NO-independent mechanism.38 Such opposite responses could re-
sult from different signaling events linked to NK-2 receptors. Alternatively, these
NKA effects could occur following paracrine activation of local modulators that
could play a crucial role in the biological actions of NKA.39 Nevertheless, the net
effect of NKA on OT release from posterior pituitary seems to be inhibitory and may
intervene in the control of OT response to osmotic stimuli.
Our study also shows that NO can inhibit the release of GABA from the posterior
pituitary. Furthermore, the inhibition of NOS activity by L-NMMA and L-NAME in-
creased GABA release, indicating that endogenous NO has an inhibitory effect on
GABAergic activity in posterior pituitary. However, we used 8-Br-cGMP, which did
not affect GABA release, suggesting that the inhibitory effect of NO on GABA re-
lease from posterior pituitary may be exerted through a cGMP-independent mecha-
nism.38 Support for this is in the report that NO activates K+ channels in posterior
pituitary nerve terminals by a cGMP-independent signaling pathway.40 GABA re-
lease could result from vesicular exocytosis by terminals of tuberohypophyseal
GABA system. Also, a carrier-mediated release from nerve terminals or pituicytes
could be another source of GABA in the posterior pituitary.41,42
GABAergic terminals in the neural and intermediate lobes participate in the con-
trol of OT, VP, and α-MSH release43–45 raising the possibility that NO modulation
of GABAergic activity in the posterior pituitary may be involved in the regulation of
the secretory function of this pituitary lobe. However, since NO decreased both OT
and GABA release from the posterior pituitary and GABA was reported to inhibit
OT release from nerve terminals of the neural lobe, it is unlikely that GABA medi-
ates the inhibitory effect of NO on OT release from posterior pituitary. It is possible
that NO might influence the release of OT at the level of the cell bodies and on nerve
terminals of the neural lobe by different mechanisms.
NKA inhibits GABA release from posterior pituitary without affecting hypotha-
lamic GABA release. We showed that the inhibition of NOS activity by L-NAME
completely blocked the inhibitory effect of NKA on GABA release. The stimulatory
effect of NKA on NO synthesis, together with the blockade of the inhibitory effect
of NKA on GABA release by L-NAME, indicates that NO is involved, at least par-
tially, in the reduction of the GABAergic activity induced by NKA. GABA arising
from the posterior pituitary may arrive at the anterior pituitary and directly decrease
prolactin secretion. The reduction in GABA release may contribute to the stimula-
tory effect of NKA on lactotroph function.46
THE ROLE OF NO IN CONTROL OF ANTERIOR
PITUITARY FUNCTION
Neural NOS has been localized in anterior pituitary cells. At least two cell types
contain the enzyme; one of these, the folliculostellate cells, are modified macro-
71MCCANN et al.: NITRIC OXIDE AND AGING
phages and known to secrete IL-6 and other cytokines. The other type is the gonado-
tropes that secrete LH. The in vitro secretion rates of most pituitary hormones are
low because of withdrawal of hypothalamic stimulation, but the secretion of prolac-
tin is greatly enhanced because of withdrawal of hypothalamic inhibition by dopam-
ine. In the case of prolactin, its secretion can be increased by inhibiting NOS with
NMMA, a competitive inhibitor of NOS or NAME, another inhibitor of the enzyme.
On the other hand, NP spontaneously releases NO and lowers prolactin release. The
prolactin-inhibiting action of dopamine, the principal prolactin-inhibiting hormone,
appears to be mediated via NO, because the action of dopamine to lower prolactin
release was blocked by inhibition of NOS. We hypothesize that DA released into
portal vessels reaches the anterior lobe, where it acts on D2 receptors in folliculo-
stellate cells and or gonadotropes to activate NOS, which activates NO release in
turn activating sGC, increasing cGMP, which then suppresses secretion of prolactin
from the lactotrophs. Indeed, cGMP decreased prolactin secretion.27
Adenosine is secreted by the folliculostellate cells and is the most powerful stim-
ulant of prolactin secretion from anterior pituitaries in vitro yet identified, increasing
release at concentrations of 10–10 –10–5 M with maximal release of three times basal
levels at 10–8 M. The action appears to be mediated by an autocrine activation of ad-
enosine 1 receptors on the surface of the folliculostellate cells, which activates in-
hibitory G proteins (Gi) that lower intracellular [Ca2+], thereby inhibiting nNOS
within the folliculostellate cells that decreases NO production. The reduced para-
crine NO inhibition of the lactotropes increases prolactin release.47
In contrast to prolactin, the release of which is inhibited by the hypothalamus, the
release of the gonadotropins LH and FSH is stimulated by the hypothalamic peptide
LHRH, which stimulates LH and to a lesser extent FSH, and by FSHR factor
(FSHRF) (lamprey III LHRH or a closely related peptide), which preferentially
stimulates FSH release. The mechanism is by stimulation of LHRH and FSHRF re-
ceptors, respectively, leading to increased [Ca2+], and activation of nNOS in the go-
nadotropes with resultant generation of cGMP, which in turn, activates PKG, leading
to extrusion of gonadotropin secretory granules.48
Gonadotropin secretion is pulsatile. Pulses can consist of the simultaneous re-
lease of FSH and LH, brought about by prior simultaneous release of FSHRF and
LHRH, or selective pulses of FSH or LH, brought about by prior release of the re-
spective releasing hormone. The relative abundance of the pulses of each type is gov-
erned by sex hormones. Thus, on the basis of the research done so far, it appears that
the pituitary gland is exposed to NO throughout normal life. Again, whether or not
these concentrations could be toxic is not clear.
THE EFFECT OF INFECTION ON CYTOKINE AND NO FORMATION
IN BRAIN, PITUITARY, AND PINEAL GLANDS
CNS infection is a powerful inducer of cytokine production in glia and neurons
of the brain. It causes induction of iNOS and production of potentially toxic quanti-
ties of NO. Injection of bacterial LPS also induces the pattern of pituitary hormone
secretion that characterizes infection. Intravenous LPS induces a dose-related re-
lease of ACTH and prolactin, a transient release of GH followed by profound inhi-
bition, decreased secretion of thyrotropin-releasing hormone (TSH), and inhibition
72 ANNALS NEW YORK ACADEMY OF SCIENCES
of LH, and to a lesser extent FSH release, in rats. It is believed that this pattern is
caused by effects of LPS directly on the brain because, after intravenous injection of
an intermediate dose of LPS, this pattern of pituitary hormone response occurred.
Also there was an induction of IL-1α immunoreactive neurons in the preoptic-
hypothalamic region. These cells were shown to be neurons by the fact that double
staining revealed the presence of neuron-specific enolase. The neurons are present
in saline-injected control animals, but increased in number by a factor of two within
two hours after LPS injection. They were located in a region that also contains the
thermosensitive neurons. They may be the cells that are stimulated to induce fever
after LPS injection. They have short axons that did not clearly project to the areas
containing the various hypothalamic releasing and inhibiting hormones, but they
could also be involved in the stimulation or inhibition of their release, which occurs
following infection.49
This study led to further research in which we determined the effect of intraperi-
toneal injection of a moderate dose of LPS on the induction of Il-1β and iNOS
mRNA in the brain, pituitary, and pineal gland. The results were very exciting, be-
cause an induction of IL-1β and iNOS mRNA occurred with the same time course
as found in the periphery following injection of LPS, namely, clear induction of
iNOS mRNA within two hours, reaching a peak in two or six hours, followed by a
decline to near basal levels at the next measurement by 24 hours after the single in-
jection of LPS. The induction of both mRNAs occurred in the meninges, the choroid
plexus, the circumventricular organs (such as the subfornical organ and median em-
inence) in the ependymal cells lining the ventricular system, and, very surprisingly,
in parvocellular neurons of the PVN and arcuate nucleus (AN), areas of particular
interest because they contain, not only the hypothalamic releasing and inhibiting
hormones, but also other neurotransmitters controlled by NO.3
The greatest induction occurred in the anterior lobe of the pituitary, where the
iNOS mRNA was increased at two hours by a factor of 45, and the pineal, where the
activity was increased by a factor of 7 at six hours, whereas the increase in the PVN
was five-fold. At six hours, the medial basal hypothalamus was found to have an in-
creased content of NOS measured in vitro, and the collected cerebrospinal fluid
(CSF) had increased concentrations of the NO metabolite, nitrate. These results indi-
cate that the increase in iNOS mRNA was followed by de novo synthesis of iNOS that
liberated NO into the tissue and also into the CSF. Presumably, LPS was bound to its
receptors in the circumventricular organs and in the choroid plexus. These receptors,
as in macrophages, activated DNA-directed IL-1β mRNA synthesis, which in turn
caused the synthesis of IL-1β. IL-1β then activated iNOS mRNA and synthesis.
How can neurons in the AN and PVN—neurons inside the blood–brain barrier—
be activated? In the case of the AN, the neurons may have axons that project to the
median eminence. These neurons may have LPS receptors on their cell surface,
which then induce IL-β mRNA and IL-β synthesis. This may then induce iNOS
mRNA. Alternatively, LPS acting on its receptors may simultaneously induce IL-β
mRNA and iNOS mRNA.
Active transport mechanisms for IL-1 and other cytokines,50 and perhaps LPS,
are present in the choroid plexus. The cells of the choroid plexus on the basis of our
results must have LPS receptors on them. LPS must stimulate IL-1β and iNOS
mRNA followed by synthesis of IL-1β and iNOS in the choroid plexus. LPS and
IL-1β are then transported into the CSF. LPS is carried by CSF flow to the third ven-
73MCCANN et al.: NITRIC OXIDE AND AGING
tricle, where it either crosses the ependyma or acts on terminals of PVN neurons in
the ependyma to induce IL-1β and iNOS mRNA.
These results raise the possibility that even moderate infection, without direct
CNS involvement, can increase iNOS levels and lead to production of toxic levels of
NO. Therefore, it is possible that repeated infections over the life span could lead to
brain damage in areas where there is large induction of iNOS in neurons, such as the
PVN—the site of the cell bodies of most of the releasing and inhibiting hormone
neurons—and the AN–median eminence region, which is also the site of production
of GHRH, many neurotransmitters, and the site of passage of axons of many of the
releasing hormone neurons, such as LHRH neurons, which project to the median em-
inence. There may also be damage to glial elements, meninges, and to the choroid
plexus over the life span. The induction of IL-1α neurons in the temperature-
regulating regions of the preoptic area should also be followed by induction of iNOS.
Exposure to high levels of NO in this region may kill thermosensitive neurons and
thus be responsible for the decreased febrile response to infection in the elderly.
Measurement of iNOS activity in the hypothalamus of aged male rats (greater than
two years of age) revealed a significant increase in NOS activity in comparison with
that in young adults, which provided the first experimental support for this concept
(Rettori, unpublished data).
The greatest increase in iNOS mRNA after LPS injection occurred in the anterior
pituitary gland. Therefore, the likelihood of damage to the cells in this gland during
infection is great. This, coupled with the damage to the releasing hormone neurons,
could account for aging changes in secretion of pituitary hormones. For example,
GH and prolactin are released largely at night, and nocturnal GH release is known to
be impaired with age.
The massive induction of iNOS in the pineal could very well contribute to the
gradual reduction in function of this gland associated with decreased nocturnal me-
latonin levels and finally even calcification of the gland, which occurs with aging.51
Melatonin is an antioxidant, and has been shown to reduce oxidative damage pro-
duced by brain ischemia and reperfusion. There is evidence that exogenous melato-
nin increases the life span of mice.51 Therefore, NO-induced pineal “aging” may
play a role in aging. The pineal hormone melatonin releases LHRH from MBH of
male rats that was prevented by blockade of NOS or GC activation.52 Not only are
proinflammatory cytokines found in tissue following injection of LPS but they are
also present in blood partly by activation of NFκB.6
EFFECT OF LPS ON TNF-␣ LEVELS IN PLASMA
OF ADULT MALE RATS
On the morning following insertion of a jugular catheter the night before, there
was no detectable plasma TNF-α. TNF-α concentrations rose within 30 min after
withdrawal of the first blood sample (0.6 mL). By 2 h, the concentration had risen
40-fold to 1,000 pg/mL. At 3 h, it had dropped precipitously to 125 pg/mL. It was
zero the following morning but rose to 125 pg/mL at 1 h after the initial blood sample
only to fall rapidly back to zero at 4 h. These results indicate that the rat, under rest-
ing conditions, has little or no circulating TNF-α, but can respond with a rapid syn-
74 ANNALS NEW YORK ACADEMY OF SCIENCES
thesis and release of this cytokine into the circulation under bleeding stress.
Previously, the rat had only been thought to respond to bacterial products.53
The release of TNF-α is mediated by the brain by inhibition of dopamine release
leading to release of prolactin, which acts on its receptors on the immune cells to
stimulate synthesis and release of TNF-α. There is inhibitory β-adrenergic control
and stimulatory α-adrenergic control of TNF-α release, as indicated by studies with
α- and β-adrenergic drugs.54 TNF-α may well induce iNOS. If so we would have
potentially damaging effects of stress. Indeed, Kishimoto and colleagues55 reported
that immobilization stress increased iNOS, another indication that stress may pro-
duce damaging levels of NO.
ROLE OF NO IN OTHER NEURODEGENERATIVE DISEASES
There is already considerable evidence that NO plays a role in neuronal cell
death, which brings on Parkinson’s disease by loss of the neurons of the nigrostriatal
dopaminergic system. Indeed, the toxin 1-methyl-4-phenyl-1,2,3,4-tetrahydropyri-
dine (NPTP) induces a parkinsonism-like syndrome. It is transported into the sub-
stantia nigra dopaminergic neurons by the dopamine transporter. It then interferes
with mitochondrial metabolism, leading to the production of oxygen free radicals.
Apparently, the basal production of NO can then be sufficient to cause toxicity via
its diffusion into the dopaminergic neurons and combination with superoxide to gen-
erate peroxynitrite, a much more potent free radical than either superoxide or NO it-
self.56 Another probable mechanism for toxicity of NO in all sites is in combination
with the heme groups, with various enzymes, thereby inactivating them and blocking
cellular respiration leading to cell death.5,56
These findings provide an explanation for the high incidence of early-onset Par-
kinson’s disease in many people who served in World War I and developed influenza.
At that time there was a major epidemic of influenza with encephalitis, which pre-
sumably led to generation of large amounts of NO in the region of the substantia ni-
gra that then caused loss of dopaminergic neurons and eventual development of
Parkinson’s symptoms many years before it would have appeared as a result of nor-
mal aging. The appearance of parkinsonism with age is probably related to the quite
rapid decline, beginning at age 45, in dopaminergic neurons in this region, even in
normal individuals,57 which may also be caused by enhanced NO generation during
infection.
In Alzheimer’s disease, an important neurodegenerative disease, plaques form
consisting of amyloid. Surrounding these plaques are many abnormal astrocytes,
which are seen on immunocytochemical study to contain IL-1.58 The IL-1 should
cause the induction of iNOS and production of NO, which may be a large factor in
neuronal cell death in the vicinity of the plaques in this condition. Indeed, pros-
tanoids, presumably formed by action of NO, accumulate adjacent to these plaques.
Even in normally aging brain, there is an increased incidence of these abnormal
astrocytes.58 Their production of IL-1 and NO could be partly responsible for the
general neuronal cell loss that occurs with aging. NO is probably also involved in
producing cell death around any area of inflammation in the CNS, for example, in
multiple sclerosis, or after brain trauma.58
75MCCANN et al.: NITRIC OXIDE AND AGING
In Huntington’s chorea, there is a mutation of the huntingtin protein that is asso-
ciated with the selective loss of basal ganglion neurons that characterize this disease.
Recently, a brain-specific protein that is associated with huntingtin has been identi-
fied, and has been termed huntingtin-associated protein (HAP-1). The location of
this protein with neurons containing nNOS mRNA, with dramatic enrichment in
both the pseudopedunculopontine nuclei, the accessory olfactory bulb, and the su-
praoptic nucleus of the hypothalamus, with co-localization of HAP-1 and nNOS in
some of these neurons, suggests that, here again, NOS could generate sufficient NO
to produce the neuronal cell loss responsible for Huntington’s disease.59
THE ROLE OF NO IN CORONARY ATHEROSCLEROSIS
Because we had found such profound induction of iNOS in the anterior pituitary
and pineal glands, areas outside the blood–brain barrier, it occurred to us that LPS
would probably induce similar changes in all organs outside the blood–brain barrier,
a prime example being the coronary arteries. Indeed, we have determined that there
is induction of IL-1β and iNOS mRNA in the endothelium of the vascular system in
the same rats given LPS. Induction of IL-1β and iNOS was also dramatic in renal
vessels.60
A great deal of evidence has accrued, suggesting the possibility that chronic in-
fections may have a relationship with coronary heart disease (CHD).61 In the 1970s,
experimental infection of germ-free chickens with avian herpes virus induced patho-
logic changes resembling those in human CHD.62 There have been many studies
showing the presence of high titers of antibodies against various organisms in pa-
tients with CHD. Although there is always some question about such studies, the in-
cidence is such as to make it appear very likely that antibodies against Helicobacter
pylori, Chlamydia pneumonia, cytomegalovirus, or other herpes viruses are very
common in these patients. There is even an association with severe dental caries.61
Stimulated by these reports, there have now been two reports of treatment of pa-
tients with CHD with tetracycline derivatives.63,64 In both studies, further complica-
tions of CHD were significantly reduced in the treated groups. In one study,
treatment reduced the complications ten-fold.63
Tetracyclines have now been studied in chondral cell cultures from patients with
osteoarthritis and in cell cultures from animals with experimentally produced arthri-
tis. They have been shown to have chondro-protective effects.65 NO is spontaneously
released from human cartilage affected by osteo- or rheumatoid arthritis in quantities
sufficient to cause cartilage damage. In a recent report, tetracyclines have been
shown to reduce the expression and function of human osteoarthritis-effected NOS
(iNOS).65 It appears that in addition to the antibacterial action of these drugs, tetra-
cyclines inhibit the expression of NOS, leading to reduction in the toxic consequenc-
es of production of NO. It is likely that these compounds will be beneficial in the
treatment of osteoarthritis, as well as CHD. They will also probably be of therapeutic
value in rheumatoid arthritis and cardiomyopathy, both thought to be autoimmune
diseases caused largely by excess NO.
The current theory of CHD is that it is induced by an elevation of plasma choles-
terol above the normal limit of 200 mg%. However, if one looks at the incidence of
CHD versus the concentration of plasma cholesterol, one finds that as cholesterol
76 ANNALS NEW YORK ACADEMY OF SCIENCES
passes the 200 mg% concentration, there is only a very slight increase in the inci-
dence of the disease. The slope of incidence begins to rise between 250 and 300 and
then rises quite rapidly as it approaches 400 mg%. There are many cases of CHD in
patients with perfectly normal cholesterol. Indeed, increased LDL cholesterol has
been considered particularly ominous, whereas HDL cholesterol has been thought to
be protective. However, in many cases, CHD develops and has its downward pro-
gression in the presence of normal cholesterol and other lipids.
In a reported case study, a 72-year-old male had gradually rising cholesterol val-
ues from 200 mg% at the age of 40 to 220 at the age of 60, but had no symptoms of
CHD. In late November, 1996, 72 h after return from a trip to England, he contracted
influenza, followed by pneumonia, with a fever of 103.5°F and a pulse of 120. The
pneumonia was probably bacterial, because it was responsive to ampicillin. This se-
vere infection was followed by the development of angina pectoris within five
weeks, which progressed to the point that he finally sought medical attention five
months later. At this time, he had advanced CHD. His weight was normal. His cho-
lesterol was 240 mg%, with a slightly elevated LDL and low HDL, but normal trig-
lycerides. Angiography revealed extensive disease that was judged unsuitable for
either balloon angioplasty or coronary bypass surgery, and he was placed on a cho-
lesterol synthesis inhibitor that normalized his total cholesterol, LDL, and HDL
within one month. The angina gradually improved over the next five months.
At that point in mid-November 1997, the subject took a trip to Europe and the
Middle East. Beginning in Frankfurt, Germany, then to Cairo, Egypt, Luxor, and by
Nile steamer to Aswan, back to Cairo, then to Israel and back to Germany. He devel-
oped a bad cold, necessitating antibiotic treatment just before departure. He reacti-
vated his osteoarthritis while tramping through the ruins of Egypt. The cold recurred
one week after it had ceased, and there was additional activation of the arthritis by
walking the streets of Jerusalem. Additional stress was occasioned by the marked
time changes, alternating cold and hot temperatures, dry weather, and the stress of
the political tensions in the Middle East at that time. However, his angina was still
ameliorating. In Germany, it was cold and damp, and his cold and arthritis were ac-
tive. On return to this country, there was an extraordinary exacerbation of the os-
teoarthritis followed by a rapid downhill progression of angina pectoris. Because of
the success with zithromycin in reducing the complications of CHD, he was treated
with this drug. The treatment was followed by marked amelioration of arthritis, but
the angina continued to worsen, and he developed angina during sleep, unrelieved
by nitroglycerin.
Finally, he was admitted to the hospital. Angiography demonstrated that his cor-
onary arteries had deteriorated greatly since first examined, even in the presence of
perfectly normal lipids (cholesterol of 170 mg%, normal HDL and LDL, and low
triglycerides). Fortunately, there was no evidence of myocardial infarction, and he
survived and recovered from quadruple bypass surgery.2
ROLE OF LEPTIN IN THE RESPONSE TO LPS
AND ITS ROLE IN NO PRODUCTION
Leptin release from the adipocytes is also under neural control. Its release is stim-
ulated by prolactin and inhibited by bromocriptine, a stimulator of dopamine release
77MCCANN et al.: NITRIC OXIDE AND AGING
that inhibits prolactin release. It is decreased by ketamine anesthesia and inhibited
by activation of β- and α-adrenergic agonist drugs with the antagonists having the
opposite effects. Leptin release is also stimulated within 10–30 min by LPS. It is
stored in microcytic vesicles lying just beneath the plasma membrane and is rapidly
released from them. The relative increase of leptin release in response to LPS is less
than that of TNF-α, but the absolute release is greater. Not only is the release of lep-
tin stimulated by LPS, but also its synthesis is stimulated, as indicated by a dramatic
increase in leptin mRNA in epididymal fat pads at 6 h.66
In both male rats and humans, plasma leptin levels are directly related to body
weight and fat mass. There is a diurnal release of leptin, most apparent in humans,
but also present in the rat. Amazingly, there is a highly significant direct correlation
of NO, as evidenced by its metabolites NO2-NO3, with plasma leptin throughout the
24-h period in male rats. The quantity of NO metabolites (NO2-NO3) in plasma is
roughly 1,000 times greater than that of leptin and reached high millimolar concen-
trations. Since micromolar concentrations of NO have been considered to be cyto-
toxic, one wonders if these concentrations of NO in fact could be detrimental and of
course they would be increased further after LPS or during infection.
To determine whether leptin could stimulate NO production in hemisected epid-
idymal fat pads, we incubated them in vitro for 1 h and found a high production of
NO2-NO3 (380 µM NO2-NO3 in the medium). Incubation with 10–7 or 10–6 M leptin
did not alter this production, but 10–5 M leptin highly significantly increased it to
750 µM. Surprisingly, leptin evoked a significantly dose-related in vitro stimulation
of TNF-α. Stimulation was small but significant at the lowest dose of leptin (10–7
M) and was increased 25-fold at the 10–5 M dose.
In vivo leptin (6 or 60 nM/kg) was tested and it evoked a significant, nearly dou-
bling of the area under the plasma curve NO3-NO2 (mM × min) with the 60 nM/kg
but not the ten-fold lower dose. There was a dose-related significant increase in plas-
ma TNF-α peaking at 90 min but still significant at 120 min. The ability of leptin to
increase NO release may be a means by which the adipocyte increases blood flow
when it is active metabolically.67
The large amounts of NO3-NO2 in vitro could be toxic since it is known that NO
concentrations greater than 1 µM are toxic, whereas nanomolar concentrations are
associated with the physiological functions of NO. The stimulation of TNF-α by lep-
tin both in vitro and in vivo was unexpected and raises the question of a possible con-
tribution of TNF-α to NO secretion in stress.67
NO ACTIVATES SECRETION OF CORTICOSTERONE
All sorts of stresses activate ACTH secretion, which then acts to evoke release of
pinocytotic vesicles containing corticosterone by activation of NO. NO activates cy-
clooxygenase-1, which produces PGE2, evoking exocytosis of corticosterone-
containing vesicles.68
Toll-like receptors (TLRs) are located in many organs including the adrenal cor-
tex, and TLR-2–deficient mice have a deficient corticosterone response to inflam-
matory stress, probably mediated by a decreased response to proinflammatory
cytokines, which act directly on cortical cells to induce corticosterone release.6
78 ANNALS NEW YORK ACADEMY OF SCIENCES
Overwhelming infection or stress may cause massive production of NO in the ad-
renal, leading to excessive dilation of adrenal vessels and excessive permeability,
leading to hemorrhage into the adrenal, which can lead to permanent adrenal insuf-
ficiency (Waterhouse Freidrichson’s syndrome). In overwhelming infection massive
production of NO and similar hemorrhage into the anterior pituitary gland may cause
anterior pituitary gland insufficiency, known as Sheehan’s syndrome.
Ascorbic Acid and Vitamin E
Ascorbic acid is synthesized by most mammals, the exceptions being man, the
great ape, and the guinea pig. It is present in highest amount in the adrenal cortex
followed by the corpus luteum and the brain. It and vitamin E are the principal anti-
oxidants in the body. Ascorbic acid is present in far greater quantities than vitamin
E and must be the major antioxidant in the body.
The pineal hormone melatonin releases hypothalamic ascorbic acid and LHRH
presumably stored in synaptic vesicles by activation of NOS. The released NO acti-
vates GC, which in turn activates LHRH and ascorbic acid release by cGMP. Inhib-
itors of NOS and GC can block the release of both.52
Similarly, corticosterone and ascorbic acid are released together from the adrenal
cortex following stimulation by ACTH released in response to stress. The concen-
tration of ascorbic acid in the adrenal cortex is by far the highest in the body. Further
study of its role in corticosterone release is urgently needed. Since NO and GC bring
about ascorbic acid release from the hypothalamus, it is likely that they are similarly
involved in ascorbic acid release from the adrenals.
Mitochondrial NOS
Recently a NOS has been shown to be associated with the inner mitochondrial
membranes. It appears to be identical to nNOS.8 If activated by infection or cell
damage, it could produce mitochondrial damage and damage to other cell compo-
nents, leading to cell death.9 More work is necessary to determine the significance
of mitochondrial NOS.
A significant 20% prolongation of life occurred in transgenic mice with overex-
pression of human catalase to the peroxisome, nucleus or mitochondria, supporting
a pathophysiological role of mitochondrial NO inactivated by catalase.69 Cardiac pa-
thology and cataract formations were delayed and oxidative damage reduced and mi-
tochondrial deletions reduced. This study provides direct support of the free radical
theory of aging.
DISCUSSION
The data presented above indicate that there are many areas in the brain where
there is regular periodic physiological release of NO throughout the life span. This
probably occurs in the hippocampus, cerebellum, and, in particular, in the hypothal-
amus, in which NO controls most of the hypothalamic peptidergic neurons (such as
CRH, LHRH, GHRH, somatostatin, oxytocin, and vasopressin) and also activates
the release of GABA and inhibits that of norepinephrine and dopamine. It is not def-
79MCCANN et al.: NITRIC OXIDE AND AGING
inite that this physiological release could ever reach levels that would produce neu-
ronal cell damage; however, Schultz and colleagues28 have reported that in men, but
not women, over 75 years of age, there were neurodegenerative changes in the arc-
uate-median eminence region associated with an increase in neurofibrillary protein.
This is the site of the interaction between NO and LHRH neurons.
The fact that injection of moderate amounts of LPS to mimic the effect of bacte-
rial infection induces increased numbers of IL-1α immunoreactive neurons in the re-
gion of the thermosensitive neurons in the preoptic hypothalamic region, plus
increased IL-β mRNA and iNOS mRNA in the PVN, AN, median eminence, choroid
plexus, meninges, and in massive amounts in the anterior pituitary and pineal with
consequent release of NO, suggests that toxic amounts of NO could exist in these
regions during moderate infections, even though there is no direct involvement of the
brain.
Destruction of neurons in the temperature-regulating centers and in the paraven-
tricular and arcuate median-eminence region after multiple infections during the life
span may be responsible for the reduced febrile response and pituitary hormone se-
cretion in response to infection, respectively. Presumably, LPS acts on receptors in
the choroid plexus and after its transport into the CSF on neuronal terminals in the
wall of the third ventricle, which induce IL-1 and iNOS mRNA with resultant pro-
duction of NO at their cell bodies in the PVN. Repeated bouts of infection over the
life span of the individual, even without direct CNS involvement, could lead to neu-
ronal cell loss in the hippocampus, because NO is important in memory formation.
Cerebellar and cortical dysfunction could also ensue.
The evidence that NO is involved in a number of neurodegenerative diseases,
among them Huntington’s and Parkinson’s disease, is already impressive. In chronic
CNS infections such as AIDS, there would certainly be even greater responses in the
aforementioned areas, plus increased iNOS in glia producing large amounts of NO.
Indeed, CNS AIDS has led to Alzheimer-like changes in the brain.58 Therefore, NO
may cause much of the neuropathologic changes in CNS AIDS.
Even more impressive is the induction of iNOS in the pituitary after LPS, which,
with age, could alter the responses of the pituitary to infection. The dramatic induc-
tion of iNOS mRNA in the pineal should lead to high concentrations of NO that
could result in the death of pineal cells and reduction of melatonin secretion, leading
to impaired resistance to free radicals normally scavenged by melatonin and accel-
eration of aging.
In humans, it is already well known, as indicated before, that infections with re-
lease of bacterial or viral products, such as LPS, causes the induction of cytokines,
which are released and travel through the bloodstream. LPS and the released cytok-
ines combine with their receptors on the coronary artery endothelial cells. They in-
duce iNOS in the endothelial cells and in macrophages that might be adherent to or
resident within the vessel. The result would be production of 1,000 times more NO
than would be released by eNOS. NO would oxidize LDL and cause the production
of prostaglandins, leukotrienes that are damaging to the vessel. NO itself would have
toxic effects to bring about cell injury and death. There would be generation and en-
largement of the atherosclerotic plaques, producing a rapid, downhill course of the
coronary disease. It is also known that inflammation, as for example in severe
osteoarthritis65 (as in the patient described earlier), causes the induction of cytokines
that would circulate to the coronary vessels and also induce iNOS.
80 ANNALS NEW YORK ACADEMY OF SCIENCES
Finally, in rats, it has been shown that stress itself, even without tissue damage,
can cause the induction of nNOS in the same areas that have been studied in the case
of iNOS induction by LPS,55 and therefore, presumably also in the vascular system,
although this has not yet been studied. Also, removal of blood from conscious rats
produced a dramatic elevation of TNF-α that might induce NOS.53 Relatively large
concentrations of NO2-NO3 with a diurnal rhythm that parallels that of leptin have
been described that increases with fat mass, suggesting that these could be toxic at
least in obese subjects. Whether or not psychological or physical stress can cause in-
duction of NOS in the coronary endothelium of humans has not been determined, but
it is a well-known fact that stress predisposes to CHD and myocardial infarction. In
fact, executives who fired their employees were twice as prone to have a heart attack
around that time as on ordinary days. Even if stress-induced heart attacks are not di-
rectly caused by NO, they may be caused by increased vasoconstriction associated
with the stress-induced withdrawal of NOergic vasodilator tone or augmented adr-
energic vasoconstrictor tone. Further studies are needed to determine which of these
possibilities is correct in this case; however, the evidence is rapidly mounting that
the final mediator of the effects of inflammation and infection on the coronaries is
the massive amounts of NO released by iNOS that cause a rapid progression of CHD,
leading to the development of angina pectoris and finally myocardial infarction. It
appears that the triad of stress, infection, and inflammation are the main factors that
precipitate rapid deterioration of the coronary vessels mediated in large part by NO.
NO can cause CHD even in the presence of a normal lipid profile.
OTHER ORGAN SYSTEMS AND THERAPEUTIC IMPLICATIONS
Space does not allow development of this hypothesis in other organ systems, but
quite clearly it is probable that the aging effects of sunlight on the skin are also me-
diated by the inflammatory response and production of massive amounts of NO. Sim-
ilarly, infections could mediate aging changes in the gonads, digestive system, and
every other organ of the body. The reduction in incidence of infectious disease via
public health and sanitation measures, from immunization and from their successful
treatment with chemo- and antibiotic therapy, may account for the increased longev-
ity in developed countries by reducing exposure to toxic concentrations of NO.
In conclusion, although much work needs to be done, it is already known that
treatment of patients with antioxidants, vitamins C and E, which would reduce the
toxic effects of NO, is of value in patients with CHD. This is probably the mecha-
nism of their protective effects against CHD. Melatonin, as indicated above, is a nat-
urally occurring antioxidant that has been shown to increase the life span of mice.
Finally, compounds that inhibit the production of NO directly, such as inhibitors of
NOS or agents that inhibit the production of NOS, such as corticoids, the tetracy-
clines, and α-MSH, may prove useful in slowing the aging process. Aspirin blocks
cyclooxygenase 1, thereby reducing production and toxicity of prostanoids produced
by NO, accounting for its protective effect in CHD. It may even be beneficial to de-
crease the production of NO in infections by the use of inhibitors of NOS, such as
NAME, or if they can be developed, specifically of iNOS.
81MCCANN et al.: NITRIC OXIDE AND AGING
REFERENCES
1. KIRKWOOD, T.B.L. & A. KOWALD. 1997. Network theory of aging. Exp. Gerontol. 32:
395–399.
2. MCCANN, S.M., J. LICINIO, M.L. WONG, et al. 1997. The nitric oxide hypothesis of
brain aging. Exp. Gerontol. 32: 431–440.
3. WONG, M.L., V. RETTORI, A. AL-SHEKHELEE, et al. 1996. Inducible nitric oxide synthase
gene expression in the brain during systemic inflammation. Nat. Med. 2: 581–584.
4. MCCANN, S.M. & V. RETTORI. 1996. The role of nitric oxide in reproduction. Proc.
Soc. Exp. Biol. Med. 211: 7–15.
5. MCDONALD, L.J. & F. MURAD. 1996. Nitric oxide and cyclic GMP signaling. Proc.
Soc. Exp. Biol. Med. 211: 1–6.
6. BORNSTEIN, S.R., R.R. SCHUMANN, V. RETTORI, et al. 2004. Toll-like Receptor 2 and
toll-like receptor 4 expression in human adrenals. Horm. Metab. Res. 36: 470–473.
7. BREDT, D.S., P.M. HWANG, S.H. SNYDER. 1990. Localization of nitric oxide synthase I
indicating a neural role for nitric oxide. Nature 347: 768–770.
8. GIUVILI, C., J.J. PODEROSO & A. BOVERIS. 1998. Production of nitric oxide by mito-
chondria. J. Biol. Chem. 273: 11038–11043.
9. NAVARRO, A., M.J. SÁNCHEZ DEL PINO, C. GÓMEZ, et al. 2002. Behavioral dysfunction,
brain oxidative stress, and impaired mitochondrial electron transfer in aging mice.
Am. J. Physiol. Regul. Integr. Comp. Physiol. 282: 985–992.
10. PALMER, R.M.J., A.G. FERRIGE & S. MONCADA. 1987. Nitric oxide release accounts for
the biological activity of endothelium-derived relaxing factor. Nature 327: 524–526.
11. GARTHWAITE, J., S.L. CHARLES & R. CHESS-WILLIAMS. 1988. Endothelium-derived
relaxing factor release on activation of NMDA receptors suggests role as intercellu-
lar messenger in the brain. Nature 336: 385–388.
12. KARANTH, S., K. LYSON, & S.M. MCCANN. 1993. Role of nitric oxide in interleukin 2-
induced corticotropin releasing factor release from incubated hypothalami. Proc.
Natl. Acad. Sci. USA 90: 3383–3387.
13. RETTORI, V., M. GIMENO, K. LYSON & S.M. MCCANN. 1992. Nitric oxide mediates
norepinephrine-induced prostaglandin E2 release from the hypothalamus. Proc. Natl.
Acad. Sci. USA 89: 11543–11546.
14. CANTEROS, G., V. RETTORI, A. FRANCHI, et al. 1995. Ethanol inhibits luteinizing hor-
mone releasing hormone (LHRH) secretion by blocking the response of LHRH neu-
ronal terminals to nitric oxide. Proc. Natl. Acad. Sci. USA 92: 3416–3420.
15. HINEY, J.K., W.H. SOWER, W.H. YU, et al. 2002. Gonadotropin releasing hormone neu-
rons in the preoptic-hypothalamic region of the rat contain lamprey gonadotropin
releasing hormone III, mammalian luteinizing hormone-releasing hormone, or both
peptides. Proc. Natl. Acad. Sci. USA 99: 2386–2391.
16. MCCANN, S.M., S. KARANTH, C.A. MASTRONARDI, et al. 2002. Hypothalamic control of
gonadotrophin secretion. Prog. Brain Res. 141: 153–166.
17. KAMAT, A., W.H. YU, V. RETTORI & S.M. MCCANN. 1995. Glutamic acid stimulated
luteinizing-hormone releasing hormone release is mediated by alpha adrenergic stim-
ulation of nitric oxide release. Brain Res. Bull. 37: 233–235.
18. RETTORI, V., N. BELOVA, W.L. DEES, et al. 1993. Role of nitric oxide in the control of
luteinizing hormone-releasing hormone release in vivo and in vitro. Proc. Natl. Acad.
Sci. USA 90: 10130–10134.
19. MANI, S.K., J.M.C. ALLEN, V. RETTORI, et al. 1994. Nitric oxide mediates sexual behavior
in female rats by stimulating LHRH release. Proc. Natl. Acad. Sci. USA 91: 6468–6472.
20. MCCANN, S.M., C. MASTRONARDI, A. WALCZEWSKA, et al. 2003. The role of NO in the
control of LHRH release that mediates gonadotrophin release and sexual behavior.
Curr. Pharmacol. Des. 9: 381–390.
21. RETTORI, V., G. CANTEROS, R. REYNOSO, et al. 1997. Oxytocin stimulates the release of
luteinizing hormone–releasing hormone from medial basal hypothalamic explants by
releasing nitric oxide. Proc. Natl. Acad. Sci. USA 94: 2741–2744.
22. SEILICOVICH, A., B.H. DUVILANSKI, D. PISERA, et al. 1995. Nitric oxide inhibits hypo-
thalamic luteinizing hormone–releasing hormone release by releasing γ-aminobu-
tyric acid. Proc. Natl. Acad. Sci. USA 92: 3421–3424.
82 ANNALS NEW YORK ACADEMY OF SCIENCES
23. SEILICOVICH, A., M. LASAGA, M. BEFUMO, et al. 1995. Nitric oxide inhibits the release
of norepinephrine and dopamine from the medial basal hypothalamus of the rat.
Proc. Natl. Acad. Sci. USA 92: 11299–11302.
24. RETTORI, V., N. BELOVA, W.H. YU, et al. 1994. Role of nitric oxide in control of growth
hormone release in the rat. Neuroimmunomodulation 1: 195–200.
25. AGUILA, M.C. 1994. Growth hormone-releasing factor increases somatostatin release
and mRNA levels in the rat periventricular nucleus via nitric oxide by activation of
guanylate cyclase. Proc. Natl. Acad. Sci. USA 91: 782–786.
26. RETTORI, V., N. BELOVA, M. GIMENO & S.M. MCCANN. 1994. Inhibition of nitric oxide
synthase in the hypothalamus blocks the increase in plasma prolactin induced by
intraventricular injection of interleukin-1a in the rat. Neuroimmunomodulation 1:
116–120.
27. DUVILANSKI, B.H., C. ZAMBRUNO, A. SEILICOVICH, et al. 1995. Role of nitric oxide in
control of prolactin release by the adenohypophysis. Proc. Natl. Acad. Sci. USA 92:
170–174.
28. SCHULTZ, C., H. BRAAK & E. BRAAK. 1997. A sex difference in neurodegeneration of
the human hypothalamus. Abstract presented at the 3rd International Symposium on
Neurobiology and Neuroendocrinology of Aging. Bregenz, Austria.
29. SAGAR, S.M. & D.M. FERREIRO. 1987. NADPH diaphorase activity in the posterior
pituitary: relation to neuronal function. Brain Res. 400: 348–352.
30. CALKA, J. & C.H. BLOCK. 1993. Relationship of vasopressin with NADPH-diaphorase
in the hypothalamo-neurohypophyseal system. Brain Res. Bull. 32: 207–210.
31. MIYAGAWA, A., H. OKAMURA & Y. IBATA. 1994. Coexistence of oxytocin and NADPH-
diaphorase in magnocellular neurons of the paraventricular and supraoptic nuclei of
the rat hypothalamus. Neurosci. Lett. 171: 13–16.
32. KADEKARO, M., M.L. TERRELL, H. LIU, et al. 1998. Effects of L-NAME on cerebral,
vasopressin, oxytocin, and blood pressure responses in hemorrhaged rats. Am. J.
Physiol. 274: 1070–1077.
33. LIU, Q.S., Y. JIA & G. JU. 1997. Nitric oxide inhibits neuronal activity in the supraoptic
nucleus of the rat hypothalamic slices. Brain Res. Bull. 43: 121–125.
34. NUSSDORFER, G.G. & L.K. MALENDOWICZ. 1998. Role of tachykinins in the regulation
of the hypothalamo-pituitary-adrenal axis. Peptides 19: 949–968.
35. EUTAMENE, H., V. THEODOROU, J. FIORAMONTI & L. BUENO. 1995. Implication of NK1
and NK2 receptors in rat colonic hypersecretion induced by interleukin 1 beta: role
of nitric oxide. Gastroenterology 109: 483–489.
36. FIGGINI, M., C. EMANUELI, C. BERTRAND, et al. 1996. Evidence that tachykinins relax
the guinea pig trachea via nitric oxide release and by stimulation of a peptide-insen-
sitive NK1 receptor. Br. J. Pharmacol. 117: 1270–1276.
37. PORTBURY, A.L., J.B. FURNESS, B.R. SOUTHWELL, et al. 1996. Distribution of neurokinin-2
receptors in the guinea pig gastrointestinal tract. Cell Tissue Res. 286: 281–292.
38. DE LAURENTIIS, A., D. PISERA, B. DUVILANSKI, et al. 2000. Neurokinin A inhibits oxy-
tocin and GABA release from the posterior pituitary by stimulating nitric oxide syn-
thase. Brain Res. Bull. 53: 325–330.
39. EISTETTER, H.R., A. MILLS & S.J. ARKINSTALL. 1993. Signal transduction mechanisms
of recombinant bovine neurokinin-2 receptor stably expressed in baby hamster kid-
ney cells. J. Cell. Biochem. 52: 84–91.
40. AHERN, G.P., S.F. HSU & M.B. JACKSON. 1999. Direct actions of nitric oxide on rat
neurohypophysial K+ channels. J. Physiol. Lond. 520: 165–176.
41. THEODOSIS, D.T. & B. MACVICAR. 1996. Neurone-glia interactions in the hypothalamus
and pituitary. Trends Neurosci. 19: 363–367.
42. VIZI, E.S. & B. SPERLAGH. 1999. Separation of carrier mediated and vesicular release
of GABA from rat brain slices. Neurochem. Int. 34: 407–413.
43. CROWLEY W. & W. ARMSTRONG. 1992. Neurochemical regulation of oxytocin secretion
in lactation. Endocr. Rev. 13: 33–65.
44. SCHNEGGENBURGER, R. & A. KONNERTH. 1992. GABA-mediated synaptic transmission
in neuroendocrine cells: a patch-clamp study in a pituitary slice preparation. Pflugers
Arch. 421: 346–373.
83MCCANN et al.: NITRIC OXIDE AND AGING
45. SLADEK, C. & W. ARMSTRONG. 1987. Gamma-aminobutyric acid antagonists stimulate
vasopressin release from organ cultured hypothalamo-neurohypophyseal explants.
Endocrinology 120: 1576–1580.
46. PISERA, D., A. DE LAURENTIIS, B. DUVILANSKI, et al. 1996. Neurokinin A affects the
tubero-hypophyseal GABAergic system. Neuroreport 7: 2236–2240.
47. YU, W.H., M. KIMURA, A. WALCZEWSKA, et al. 1998. Adenosine acts by A1 receptors to
stimulate release of prolactin from anterior-pituitaries in vitro. Proc. Natl. Acad. Sci.
USA 95: 7795–7798.
48. YU, W.H., A. WALCZEWSKA, S. KARANTH & S.M. MCCANN. 1997. Nitric oxide medi-
ates leptin-induced luteinizing hormone releasing hormone (LHRH) and LHRH and
leptin-induced LH release from the pituitary gland. Endocrinology 138: 5055–5058.
49. RETTORI, V., W.L. DEES, J.K. HINEY, et al. 1994. An interleukin-1a-like neuronal sys-
tem in the preoptic-hypothalamic region and its induction by bacterial lipopolysac-
charide in concentrations which alter pituitary hormone release.
Neuroimmunomodulation 1: 251–258.
50. BANKS, W.A., A.J. KASTIN & R.D. BROADWELL. 1995. Passage of cytokines across the
blood–brain barrier. Neuroimmunomodulation 2: 241–248.
51. PIERPAOLI, W., D. BULIAN, A. DALL’ARA, et al. 1997. Circadian melatonin and young-
to-old pineal grafting postpone aging and maintain juvenile conditions of reproduc-
tive functions in mice and rats. Exp. Gerontol. 32: 587–602.
52. KARANTH, S., W.H. YU, C.A. MASTRONARDI & S.M. MCCANN. 2004. Inhibition of stim-
ulated ascorbic acid and luteinizing hormone-releasing hormone by nitric oxide syn-
thase or guanylyl cyclase inhibitors. Exp. Biol. Med. 229: 72–79.
53. MASTRONARDI, C.A., W.H. YU & S.M. MCCANN. 2001. Comparisons of the effects of
anesthesia and stress on release of tumor necrosis factor–alpha, leptin, and nitric
oxide in adult male rats. Exp. Biol. Med. 226: 296–300.
54. MASTRONARDI, C.A., W.H. YU, S.M. MCCANN. 2001. Lipopolysaccharide-induced
tumor necrosis factor-alpha release is controlled by the central nervous system. Neu-
roimmunomodulation 9: 148–156.
55. KISHIMOTO, J., T. TSUCHIYA, P.C. EMSON & Y. NAKAYAMA. 1996. Immobilization-
induced stress activates neuronal nitric oxide synthase (nNOS) mRNA and protein in
hypothalamic–pituitary–adrenal axis in rats. Brain Res. 720: 159–171.
56. PRZEDBORSKI S., V. JACKSON-LEW IS, R. YOKOYAMA, et al. 1996. Role of neuronal NO
in 1-methyl-4-phenyl-1,2,3,4-tetrahydropyridine (MPTP)-induced dopaminergic tox-
icity. Proc. Natl. Acad. Sci. USA 93: 4565–4571.
57. KNOLL, J. 1997. Sexual performance and longevity. Exp. Gerontol. 32: 539–552.
58. GRIFFIN, S. 1998. Neuroimmunomodulation Symposium. Am. Assoc. Immunol. Meet.
San Francisco. April, 1998.
59. XIAO-JIANG, L., A.H. SHARP, L. SHI-HUA, et al. 1996. Huntington-associated protein
(HAPI): discrete neuronal localizations in the brain resemble those of neuronal nitric
oxide synthase. Proc. Natl. Acad. Sci. USA 93: 4839–4844.
60. WONG, M.L., F. O’KIRWAN, N. KHAN, et al. 2003. Identification, characterization, and
gene expression profiling of endotoxin-induced myocarditis. Proc. Natl. Acad. Sci.
USA 100: 14241–14246.
61. DANESH, J., R. COLLINS & R. PETO. 1997. Chronic infections and coronary heart dis-
ease: is there a link? Lancet 350: 430–436.
62. FABRICANT, C.G., J. FABRICANT, M.M. LITRENTA & C.R. MINICK. 1978. Virus-induced
atherosclerosis. J. Exp. Med. 148: 335–340.
63. GURFINKEL, E., G. BOZOVICH, A. DAROCA, et al. 1997. For the ROXIS Study Group.
Randomised trial of roxithromycin in non-Q-wave coronary syndromes: ROXIS pilot
study. Lancet 350: 404–407.
64. GUPTA, S., E.W. LEATHAM, D. CARRINGTON, et al. 1997. Elevated Chlamydia pneumo-
niae antibodies, cardiovascular events, and azithromycin in the male survivors of
myocardial infarction. Circulation 96: 404–407.
65. AMIN, A.R., M.G. ATTUR, G.D. THAKKER, et al. 1996. A novel mechanism of action of
tetracyclines: effects on nitric oxide synthases. Proc. Natl. Acad. Sci. USA 93:
14014–14019.
84 ANNALS NEW YORK ACADEMY OF SCIENCES
66. MASTRONARDI, C.A., W.H. YU, V.K. SRIVASTAVA, et al. 2001. Lipopolysaccharide-induced
leptin release is neurally controlled. Proc. Natl. Acad. Sci. USA 98: 14720–14725.
67. MASTRONARDI, C.A., W.H. YU & S.M. MCCANN. 2002. Resting and circadian release of
nitric oxide is controlled by leptin in male rats. Proc. Natl. Acad. Sci. USA 99:
5721–5726.
68. MOHN, C., J. FERNANDEZ SOLARI, A. DE LAURENTIIS, et al. 2005. The rapid release of
corticosterone from the adrenal induced by ACTH is mediated by NO acting by
PGE2. Proc. Natl. Acad. Sci. USA 102: 6213–6218.
69. SCHRINER, S.E., N.J. LINFORD, G.M. MARTIN, et al. 2005. Extension of murine lifespan
by overexpression of catalase targeted to mitochondria. Science 308: 1909–1911.
