![]() |
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
PERSPECTIVES IN PHARMACOLOGY
6* Nicotinic Acetylcholine Receptors: Potential Targets for Parkinson's Disease Therapy
The Parkinson's Institute, Sunnyvale, California (M.Q.); and Departments of Biology and Psychiatry, University of Utah, Salt Lake City, Utah (J.M.M.)
Received August 17, 2005; accepted October 4, 2005.
| Abstract |
|---|
6-containing nAChRs (designated
6*1 nAChRs) have a relatively selective localization to the nigrostriatal pathway and a limited number of other CNS regions. In addition to a unique distribution, this subtype has a distinct pharmacology and specifically interacts with
-conotoxinMII, a toxin key in its identification and characterization.
6* nAChRs are also regulated in a novel manner, with a decrease in their number after nicotine treatment rather than the increase observed for
4* nAChRs. Striatal
6* receptors were functional and mediate dopamine release, suggesting that they have a presynaptic localization. This is further supported by lesion studies showing that both
6* nAChR sites and their functions are dramatically decreased with dopaminergic nerve terminal loss, in contrast to only small declines in
4* and no change in
7* receptors. Although the role of nigrostriatal
6* nAChRs is only beginning to be understood, an involvement in motor behavior is emerging. This latter observation coupled with the finding that nicotine protects against nigrostriatal damage suggest that
6* nAChRs may represent unique targets for neurodegenerative disorders linked to the nigrostriatal system such as Parkinson's disease.
Acetylcholine can also regulate striatal dopamine levels by stimulating nicotinic acetylcholine receptors (nAChRs). These receptors are localized on nigrostriatal dopaminergic nerve terminals where they control dopamine release (Wonnacott, 1997
; MacDermott et al., 1999
; Gotti and Clementi, 2004
). Since striatal nAChRs seem to be involved in both motor control and neuroprotection against nigrostriatal damage (O'Neill et al., 2002
; Quik, 2004
), identification of the receptor subtypes is important for understanding basal ganglia function under physiological conditions and pathological states such as Parkinson's disease.
However, this has turned out to be a rather daunting task since multiple nAChRs are present in the CNS (Lukas et al., 1999
; Gotti and Clementi, 2004
). To date, six different
(
2-
7) and three
(
2-
4) subunits have been identified in mammalian brain that form pentameric ligand-gated ion channels.
7 receptors are generally homomeric, whereas different combinations of
2-
6 subunits coupled with
2-
4 subunits form heteromeric receptors. The acetylcholine binding site is present at an
-
interface consisting of an
2,
3,
4, or
6 subunit, coupled together with a
2 or
4 subunit. The
5 and
3 subunits do not seem to interact with cholinergic ligands but modulate receptor function (Lukas et al., 1999
; Gotti and Clementi, 2004
). Because receptors composed of different subunits vary in drug sensitivities and functional characteristics, it is important to identify the subtypes present in different brain regions. Interestingly, receptors containing the
6 subunit (designated
6* nAChRs) are concentrated in the visual and catecholaminergic pathways, including the dopaminergic nigrostriatal pathway (Whiteaker et al., 2000
; Gotti and Clementi, 2004
; Quik, 2004
), suggesting that they play a unique role in these systems.
Herein we review current findings on
6M* nAChRs in the brain, with a focus on the dopaminergic nigrostriatal pathway. We first discuss the unique characteristics of the
6* nAChR subtype, exemplified by its selective interaction with the marine snail peptide
-conotoxinMII. The use of this toxin has allowed for
6* nAChR localization in the brain, which is distinct compared with other nAChR populations and seems to be primarily presynaptic in the striatum, although it may be localized to other neuronal elements in different brain regions. The combination of a unique pharmacology and distribution may allow for selective therapeutic targeting of
6* nAChRs for CNS disorders involving this subtype.
Selective Interaction of 6* nAChRs with -ConotoxinMII
|
|---|
6 subunit was first reported in 1990 (Lamar et al., 1990
6* receptor characteristics were described primarily because of difficulties in their expression and identification (Gerzanich et al., 1997
6* nAChRs expression was first achieved in oocytes with the chick
6 and human
4 subunits (Gerzanich et al., 1997
6
4* nAChR (Fucile et al., 1998
6 subunit also formed functional heteromeric nAChR with chick
2, although at a much lower abundance (Fucile et al., 1998
6* nAChR composed of only mammalian subunits was subsequently achieved, although not without difficulty (Kuryatov et al., 2000
6 and
4 subunits expressed only poorly in oocytes, whereas injection of
6 with
2 resulted in the formation of nonfunctional epibatidine binding aggregates; however, functional human
6
4
3, as well as
6
2
5 nAChRs, were observed (Kuryatov et al., 2000
6 was joined to the remaining portions of either
3 or
4 and expressed with either the
2 or
4 subunits (Kuryatov et al., 2000
3 subunit significantly improved expression of these chimeric
6* receptors (Dowell et al., 2003
6 subunits can assemble in heterologous expression systems with the
2 or
4, as well as the
5 and
3, subunit.
A striking pharmacological feature of
6* nAChRs is their high-affinity interaction with the naturally occurring snail toxin
-conotoxinMII (Kuryatov et al., 2000
; McIntosh et al., 2004
). This toxin binds to
6* nAChRs in a slowly reversible manner, which makes it a useful ligand for receptor identification and characterization. Indeed, receptor studies with [125I]
-conotoxinMII demonstrated a high-affinity (Kd,
0.8 nM) nAChR in rodent, monkey, and human brain (Whiteaker et al., 2000
; Quik et al., 2001
; Quik et al., 2004
). In addition,
-conotoxinMII also potently interacts with
3* nAChRs because of the high sequence homology to the
6 subunit (
75%). This presents a problem for receptor identification in tissues that contain both of these subtypes and has led to a search for compounds that can distinguish between
3* and
6* nAChRs (Table 1). One such agent is
-conotoxin PIA, a toxin from Conus purpurascens, which is
75-fold more selective for heterologously expressed chimeric
6/
3
2* versus
3
2 nAChRs (Dowell et al., 2003
).
-ConotoxinMII analogs exhibited an even greater selectivity (up to 2000-fold) for
6* compared with
3* nAChRs (McIntosh et al., 2004
). The IC50 values of these peptide analogs for chimeric
6
2* nAChRs in functional assays are all in the picomolar to nanomolar range and correlate well with those obtained from [125I]
-conotoxinMII competition binding assays in mouse striatum (Fig. 1) that contains
6* and not
3* nAChRs (Whiteaker et al., 2002
; Champtiaux et al., 2003
; McIntosh et al., 2004
).
|
|
It is worth noting that
-conotoxinMII-sensitive
6* receptors also have high affinity for methyllycaconitine, a plant alkaloid historically considered selective for
7 nAChRs. Putative
6
4
5
2 nAChRs present on dopamine neurons in rat substantia nigra and ventral tegmental area are completely inhibited by 1 nM methyllycaconitine (Klink et al., 2001
). Moreover, data from knockout mice suggest that these
-conotoxinMII- and methyllycaconitine-sensitive nAChRs do not contain
7 subunits (Klink et al., 2001
). Methyllycaconitine is also a potent inhibitor of
-conotoxinMII-sensitive-mediated dopamine release in striatum (Mogg et al., 2002
; Karadsheh et al., 2004
). Thus, methyllycaconitine potently interacts with both
7 and
6* nAChRs. Altogether, these results show that
6* nAChR receptor sites represent a class of neuronal nAChRs with a unique pharmacological profile. Moreover,
-conotoxinMII and related peptides from Conus seem to be excellent tools to investigate their characteristics and function.
Selective CNS Distribution of 6* nAChRs
|
|---|
6 nAChR transcript exhibits a very restricted distribution in rodent and monkey brain (Table 2).
6 mRNA labeling is particularly strong in catecholaminergic nuclei, including the substantia nigra, ventral tegmental area, and locus coeruleus, as well as in the medial habenula and interpeduncular nucleus, with a less intense signal in other brain nuclei (Le Novere et al., 1996
6 mRNA co-incides with that of
3 mRNA in the substantia nigra and ventral tegmental area (Table 2), suggesting that these transcripts are expressed in parallel in dopaminergic neurons (Azam et al., 2002
|
Receptor Studies
[125I]
-ConotoxinMII selectively and saturably binds to nAChRs in neuronal pathways expressing
6 mRNA. This includes the substantia nigra and ventral tegmental area, as well as their projection areas, the striatum and nucleus accumbens.
-ConotoxinMII-sensitive receptors are also present in the medial habenula, interpeduncular nucleus, and regions associated with the visual system in both rodent and monkey brain (Whiteaker et al., 2000
; Quik et al., 2001
; Champtiaux et al., 2002
).
Since
-conotoxinMII interacts with
6* and
3* nAChRs in heterologous expression systems, the question arises whether the toxin binds to both of these subtypes in mammalian brain. Studies with nAChR subunit knockout mice showed that [125I]
-conotoxinMII binding was not significantly decreased in striatum of
3 (-/-) mice but was virtually eliminated in
6 (-/-) mice, suggesting that the toxin binds to only
6* nAChRs in mouse striatum (Champtiaux et al., 2002
; Whiteaker et al., 2002
). [125I]
-ConotoxinMII binding was also abolished in other brain regions of
6(-/-) mice but only partially reduced in the medial habenula and interpeduncular nucleus. In addition, [125I]
-conotoxinMII was partially decreased in these latter regions in
3 (-/-) mice and unchanged in striatum and other regions. Altogether, these data indicate that [125I]
-conotoxinMII binds to both
6* and
3* receptors in the medial habenula-interpeduncular pathway but only to
6* sites in striatum and other CNS regions in mouse brain. A different situation seems to exist in the primate CNS. Immunoprecipitation studies with subunit-targeted antibodies showed that there was appreciable
6 subunit-like immunoreactivity in monkey striatum, as well as a smaller
3 signal (Quik et al., 2005b
). To conclude, only
6* nAChRs are detectable in rodent striatum, whereas both
6* and
3* nAChRs are localized in primate striatum.
Composition of 6* nAChRs in Striatum
|
|---|
6, antibodies targeted to specific nAChR subunits have been used (Zoli et al., 2002
6* but no detectable
3* nAChR expression in mouse striatum, consistent with results from nAChR subunit knockout mice (Zoli et al., 2002
4,
5,
7,
2,
3, and
4 but not
2 subunits in rodent striatum (Zoli et al., 2002
3 and
6 subunits were identified in monkey striatum, as well as
2,
4,
7,
2, and
3, but not the
5 and
4 subunits.
Dual-label immunoprecipitation shows that the subtypes common across species are
6
4
2
3 and
6
2
3, as well as
4
2 and
7 receptors (Fig. 2A; Table 3) (Zoli et al., 2002
; Champtiaux et al., 2003
; Gotti et al., 2005
; Quik et al., 2005b
). In addition, there also seems to be a population of less abundant striatal nAChR subtypes that are unique to different species (Table 3). To date, these include
4
5
2 receptors on dopamine terminals in mouse but not monkey (Zoli et al., 2002
; Champtiaux et al., 2003
) and
3
2* receptors present in monkey but not mouse (Quik et al., 2005b
). Identification of the precise mix of nAChR subtypes in striatum is important because it may allow for selective targeting with drugs that uniquely interact with these populations.
|
|
Striatal 6 nAChR Stimulation and Dopamine Release
|
|---|
6
2* nAChRs have focused on mammalian striatum because of the relatively high receptor density, the availability of assays to study their function, and putative links to addiction and neurodegenerative disorders. An approach that has proved particularly useful for studying function of striatal nAChRs is nicotine-evoked dopamine release. Stimulation of presynaptic striatal nAChRs results in dopamine release that is mediated by subtypes that are blocked by
-conotoxinMII and those that are not (Grady et al., 2002
-conotoxinMII-sensitive dopamine release is most likely mediated through
6
2* nAChRs and represents
40% of the total response (Kulak et al., 1997
-conotoxinMII-resistant release occurs through
4
2* nAChRs and represents
60% of the response (Fig. 3).
|
2(-/-) mice, indicating an absolute requirement for the
2 subunit (Champtiaux et al., 2003
4 (-/-)
6 (-/-) mice and significantly affected in
4 (-/-) or
6 (-/-) mice, suggesting a mandatory presence for either the
4 or
6 subunit (Champtiaux et al., 2003
5 (-/-) and
3 (-/-) mice suggest a modulatory role for these subunits (Salminen et al., 2004b
4 and
7 nAChRs had no effect on release. Since the
2 and
3 subunits are not present in mouse striatum, these combined results substantiate a role for receptors containing the
4,
5,
6,
2, and
3 nAChR subunits in nicotine-evoked dopamine release from mouse striatum (Cui et al., 2003
The use of
-conotoxinMII, coupled with nAChR knockout mice, has allowed for further identification of the receptors that mediate dopamine release.
-ConotoxinMII completely blocks nicotine-stimulated dopamine release in
4 (-/-) mice, showing that a component of release is mediated by
6* nAChRs (Salminen et al., 2004b
). Conversely,
-conotoxinMII does not block residual dopamine release in
6 (-/-) mice, demonstrating an
4* nAChR-sensitive component (Champtiaux et al., 2003
). Altogether, the above studies, coupled with immunoprecipitation data (Zoli et al., 2002
; Champtiaux et al., 2003
) and studies with
6*-selective conotoxins (McIntosh et al., 2004
; Azam and McIntosh, 2005
), suggest that
-conotoxinMII-sensitive sites in mice represent
6
2
3 and
6
4
2
3 subtypes, whereas
-conotoxinMII-resistant receptors are
4
2 and
4
5
2 nAChRs.
The situation in primate striatum bears resemblance and some differences compared with rodents. In monkey striatum, the greater portion (
70%) of nicotine-evoked dopamine release is mediated through
-conotoxinMII-sensitive (
6
2* and/or
3
2*) subtypes, whereas
-conotoxinMII-resistant or
4
2* receptors mediate only
30% release; these proportions were reversed in rodents (Fig. 3) (Kulak et al., 1997
; Grady et al., 2002
; McCallum et al., 2005a
). The
-conotoxinMII-sensitive release most likely occurs in response to stimulation of
6
2
3 and
6
4
2
3 subtypes (Quik et al., 2005b
) as in rodents, but in monkeys it may also involve
3
2* receptors.
Overall, there is a consensus that, in striatum, presynaptic
6*, as well as
4*, nAChRs play a critical role in receptor-evoked dopamine release with a greater contribution from
6* versus
4* nAChRs in primates than rodents. The localization and function of
6* nAChRs in the visual, habenular-interpeduncular, and other pathways remain to be investigated.
Down-Regulation of 6* nAChRs with Long-Term Nicotine Treatment
|
|---|
4
2* nAChRs. This is one of the most prevalent CNS subtypes, and effects on these sites may have masked changes in other populations. Indeed, chronic nicotine administration does not alter
3
4* nAChRs in the central or peripheral nervous system (Flores et al., 1997
7 nAChRs are generally unaffected and sometimes modestly increased across brain regions (Pauly et al., 1991
6* nAChRs are decreased in mouse striatum after several weeks of nicotine treatment administered via drinking water or by chronic jugular infusion (Salminen et al., 2004a
6* nAChR-evoked [3H]dopamine release, indicating that the receptor loss is of functional significance (LSalminen et al., 2004a; ai et al., 2005). Although some studies have not reported a decline in
6* sites after nicotine treatment, this may relate to the route of administration, species, age, and/or method of
6* receptor determination (Nguyen et al., 2003
Altogether, these findings show that nicotine treatment differentially influences nAChR subtypes, with increases, decreases, or no change. This disparate regulation suggests that distinct mechanisms control receptor expression. The increase in
4
2* sites may be due to nicotine-induced receptor desensitization that resembles an apparent receptor blockade, with a compensatory increase to ameliorate the functional loss. Nicotine may also decrease the turnover rate of already assembled nAChRs and/or act intracellularly on receptor precursors to enhance their maturation (Sallette et al., 2005
). The down-regulation of
6* nAChRs with chronic nicotine treatment suggests that this subtype is controlled in a fashion analogous to that for neurotransmitter receptors that decrease with persistent agonist exposure (Creese and Sibley, 1981
; Wonnacott, 1990
). This differential control by nicotine may occur through an interaction with specific residues on the
subunits, comparable to the regulatory microdomains identified on the
2 versus
4 subunits (Sallette et al., 2004
).
Overall, the presence of multiple nAChRs populations, including
6
4
2
3,
6
2
3,
4
2, as well as
7 and possibly others (Fig. 1; Table 2), provides the potential for a complex regulation by nicotine in striatum. There may thus be widely divergent nAChR-mediated functional changes in striatum after nicotine exposure in smokers or individuals on chronic nicotine therapy.
Loss of Striatal 6* nAChRs with Nigrostriatal Damage
|
|---|
6* nAChRs have focused on the nigrostriatal system because of the availability of toxins that selectively destroy this pathway and the relevance to neurodegenerative disorders, such as Parkinson's disease. Initial studies using monkeys lesioned with the selective dopaminergic neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) showed that there was a dramatic decline in [125I]
-conotoxinMII binding or
6* nAChRs, with a smaller loss of
4* nAChRs (Quik et al., 2001
6* receptors are located presynaptically. Experiments to investigate the link between lesion-induced receptor declines and function showed there was a regionally selective compensation in both
6* and
4* receptor-mediated [3H]dopamine release with nigrostriatal damage (McCallum et al., 2005a
There was also a decline in nAChRs with nigrostriatal damage in rodents, but in this species there were similar losses in both
6* and
4* subtypes (Zoli et al., 2002
; Champtiaux et al., 2003
; Quik et al., 2003
). Interestingly, striatal nAChR function in mice was not decreased until the dopamine transporter was reduced by
20% (Quik et al., 2003
), again suggesting some form of functional adaptation.
The results of lesion studies suggest that
6* nAChRs are localized to dopaminergic terminals in the striatum, with a loss of
6
4
2
3 and
6
2
3 nAChRs after nigrostriatal damage in both primates and rodents (Fig. 2B). These studies also showed declines in
4
2* nAChRs (Zoli et al., 2002
; Champtiaux et al., 2003
; Salminen et al., 2004b
; Gotti et al., 2005
; Quik et al., 2005b
). These receptor losses seem to be biologically relevant, with a decline in dopamine release after nigrostriatal damage, although significant functional compensation occurred, particularly in primates. These adaptive mechanisms may be responsible, at least in part, for the observation that Parkinson's disease symptoms only develop after >80% declines in striatal dopamine.
6* nAChRs in Human Brain: Declines in Parkinson's Disease
|
|---|
4*,
7*, and more recently
6* subtypes (Court et al., 2000
-conotoxinMII binds with high affinity (
0.5 nM) to numerous regions in human brain that include, in order of decreasing intensity of labeling, optic tract, nucleus accumbens, caudate, and putamen, consistent with results in rodents and primates (Quik, 2004
-conotoxinMII binding sites were identified in hippocampus, globus pallidus, frontal cortex, thalamus, and cerebellum, distinct from results in rodent and nonhuman primates. This may reflect binding to
3* nAChRs in human brain, although it is also possible that
6* nAChRs are present in these latter regions in humans. Such an interpretation is consistent with in situ hybridization results with
3-subunit mRNA probes, which identified the
3 transcript in the cortex, hippocampus, and thalamus (Rubboli et al., 1994
3-subunit directed antibodies show that
3-like immunoreactivity is present in these same regions (Guan et al., 2002
Since the nigrostriatal pathway degenerates in Parkinson's disease, the question arose whether
6* nAChRs are decreased as in experimental models. Indeed, there were 50 to 90% reductions in [125I]
-conotoxinMII sites in Parkinson's disease striatum, the region containing dopaminergic nerve terminals (Quik et al., 2004
; Bohr et al., 2005
). In human brain, this decline in [125I]
-conotoxinMII binding did not parallel the dopamine nerve terminal loss as closely as anticipated, with greater losses of the transporter compared with nAChRs (Quik et al., 2004
). These data suggest that [125I]
-conotoxinMII sites may be located both pre- and postsynaptically in human striatum.
The composition of [125I]
-conotoxinMII receptors, that is, whether they contain
3 and/or
6 subunits in human brain, is currently not known. One study reported a decrease in
3-like immunoreactivity in Parkinson's disease striatum (Guan et al., 2002
), although others found no change in
2-
7,
2, and
3 nAChR subunit immunoreactivity (Martin-Ruiz et al., 2002
) despite undisputed declines in striatal nAChRs using radioligand binding studies. Thus, although there are clearly alterations in both
4* and
6*/
3 (
-conotoxinMII-sensitive) subtypes in Parkinson's disease striatum, their composition requires further study.
Functional Consequence of Striatal 6 nAChR Stimulation
|
|---|
4* nAChRs. In addition, the relatively dense distribution of
6* nAChRs in brain regions linked to these and other functions suggest this latter subtype may also play a role (Whiteaker et al., 2000
6* nAChRs in the optic tract and its target regions may be important in vision; the occurrence of
6* receptors in nucleus accumbens suggests a possible involvement in addiction, whereas their presence in striatum may imply a role in locomotor activity. | Putative Symptomatic Effect of Nicotinic Receptor Stimulation on Motor Function |
|---|
4* nAChRs, most likely through release of striatal dopamine (Grady et al., 1992
6* subtypes may also be involved since nicotine-induced locomotor activity is attenuated in mice treated with
6 antisense (Le Novere et al., 1999
6* nAChRs may represent a somewhat more select target for modulating motor behaviors since these receptors exhibit a more restricted localization than the
4* subtype, which is widespread throughout the brain (Gotti and Clementi, 2004
In neurodegenerative disorders such as Parkinson's disease, there is a loss of presynaptic dopamine terminals and an accompanying decline in nAChRs (Quik et al., 2004
; Bohr et al., 2005
). This raises the possibility that stimulation of residual
6* (as well as
4*) nAChRs could enhance release from remaining terminals to result in symptomatic improvement in motor symptoms. A question that arises is how effective
6* (and/or
4*) nAChR stimulation is when studies show there is
50 to 90% decline in the receptors in Parkinson's disease striatum (Quik et al., 2004
; Bohr et al., 2005
). Interestingly, our recent work in monkeys with nigrostriatal damage shows that striatal
6* (and also
4*) nAChR function is at normal levels despite 50% receptor declines (McCallum et al., 2005a
,b
). These compensatory changes in function in the presence of significant nAChR losses suggest that subtype-selective agonists would be beneficial despite nigrostriatal damage. In addition, the use of nicotine or subtype-selective nicotinic receptor agonists may offer the advantage that released dopamine from the nerve terminal represents a more physiologic mode of stimulation of postsynaptic dopamine function than that which occurs in response to administration of L-dopa or dopamine agonists. There may be advantages in synchronizing postsynaptic dopamine receptor stimulation with presynaptically evoked action potentials that are not conserved with directly acting dopamine agonists. Indeed, previous work has shown that administration of nAChR agonists to monkeys in combination with L-dopa allowed for a reduction in L-dopa dosage without a loss in the antiparkinsonian efficacy of L-dopa (Schneider et al., 1998
). The reduction in L-dopa dose may result in a decline in debilitating side effects, including dyskinesias and psychiatric disturbances, while maintaining the therapeutic response.
To date, studies investigating effects of nicotine administration for Parkinson's disease therapy have been very limited and have yielded mixed results. Reductions in tremor and/or bradykinesia, as well as other improvements in motor performance, have been observed in some studies but not others (Ishikawa and Miyatake, 1993
; Fagerstrom et al., 1994
; Ebersbach et al., 1999
; Kelton et al., 2000
; Vieregge et al., 2001
; Lemay et al., 2004
). This inconsistency may relate to the small number of patients in the different studies as well as the short duration of nicotine treatment (a few weeks). The mode of nicotine administration has also been quite variable and includes i.v. infusion and/or use of the nicotine patch, gum, or lozenge. These dosing regimens are quite distinct, and the particular one selected may impact the behavioral response. For instance, chronic delivery with the nicotine patch will lead to steady-state nicotine levels that may result in greater receptor desensitization than intermittent regimens such as the gum or lozenge (Giniatullin et al., 2005
; Wang and Sun, 2005
). Further study to evaluate the potential for nicotine in the symptomatic treatment of Parkinson's disease is critical.
| Neuroprotective Effect of Nicotine |
|---|
7 nAChRs in contrast to protection in most culture models in which
7 nAChRs seem to be involved (O'Neill et al., 2002
1 year) study in primates, we observed a clear-cut protective effect of nicotine against MPTP-induced striatal damage (Quik et al., 2005a
Although the neuroprotective potential against nigrostriatal damage in Parkinson's disease remains to be evaluated, epidemiological studies overwhelmingly demonstrate
50% reduced incidence of Parkinson's disease in smokers (Morens et al., 1995
; Checkoway and Nelson, 1999
; Allam et al., 2004
). This relationship is directly correlated to the duration of smoking and the number of cigarettes smoked, and the protective effect is reduced when smoking is discontinued. Admittedly, the mechanism(s) for this inverse correlation are not known; however, the work described above in culture and animal models supports a role for nicotine in tobacco products (Quik et al., 2005a