Genetics of tobacco use and dependence

As early as 1958, Fisher hypothesized that the link
between smoking and lung cancer could be ex-
plained at least in part by shared genes that predis-
pose individuals to begin smoking as young adults
and to develop lung cancer later in adulthood [88].
More recently, tobacco researchers have begun to
explorewhether genetic factors do in fact contribute
toward tobacco use and dependence.
Tobacco use and dependence are hypothesized to
result from an interplay of many factors (includ-
ing pharmacologic, environmental and physiologic)
[77]. Some of these factors are shared within fam-
ilies, either environmentally or genetically. Studies
of families consistently demonstrate that, compared
to family members of nonsmokers, family members
of smokers aremore likely to be smokers also. How-
ever, in addition to shared genetic predispositions,
it is important to consider environmental factors
that promote tobacco use—siblings within the same
family share many of the same environmental in-
fluences as well as the same genes. To differentiate
the genetic fromthe environmental influences, epi-
demiologists use adoption, twin, twins reared apart,
and linkage study designs [89].
Key to the adoption studies is the assumption that
if a genetic link for tobacco use exists, then tobacco
use behaviors (e.g., smoking status, number of years
smoked, number of cigarettes smoked per day) will
be more similar for persons who are related geneti-
cally (i.e., biologically) than for persons who are not
related genetically. Hence, one would expect to ob-
serve greater similarities between children and their
biological parents and siblings than would be ob-
served between children and their adoptive parents
or adopted siblings. Indeed, research has demon-
strated stronger associations (i.e., higher correlation
coefficients) between biologically-related individu-
als, compared to nonbiologically-related individu-
als, for the reported number of cigarettes consumed
[90]. In recent years, it has become more difficult
to conduct adoption studies, because of the reduced
number of intranational children available for adop-
tion [91]. Additionally, delayed adoption (i.e., time
elapsed between birth and entry into the new fam-
ily) is common with international adoptions and
might lead to an overestimation of genetic effects
if early environmental influences are attributed to
genetic influences [92].
In twin studies, identical (monozygotic) twins
and fraternal (dizygotic) twins are compared. Iden-
tical twins share the same genes; fraternal twins,
like ordinary siblings, share approximately 50% of
their genes. If a genetic link exists for the phe-
nomenon under study, then one would expect to
see a greater concordance in identical twins than
in fraternal twins. Thus, in the case of tobacco
use, one would expect to see a greater proportion
of identical twins with the same tobacco use be-
havior than would be seen with fraternal twins.
Statistically, twin studies aim to estimate the per-
centage of the variance in the behavior that is
due to (1) genes (referred to as the “heritability”),
(2) shared (within the family) environmental ex-
periences, and (3) nonshared (external from the
family) environmental experiences [91]. A num-
ber of twin studies of tobacco use have been con-
ducted in recent years. These studies have largely
supported a genetic role [91,93]; higher concor-
dance of tobacco use behavior is evident in identical
twins than in fraternal twins. The estimated aver-
age heritability for smoking is 0.53 (range, 0.28–
0.84) [93,94]; approximately half of the variance
in smoking appears to be attributable to genetic
factors.
Recent advances in the mapping of the human
genome have enabled researchers to search for
genes associated with specific disorders, including
tobacco use. Using a statistical technique called link-
age analysis, it is possible to identify genes that pre-
dict a trait or disorder. This process is not based on
prior knowledge of a gene’s function, but rather it
is determined by examining whether the trait or
disorder is coinherited with markers found in spec-
ified chromosomal regions. Typically, these types
of investigations involve collection of large family
pedigrees, which are studied to determine inheri-
tance of the trait or disorder. This method works
well when a single gene is responsible for the out-
come; however, it becomes more difficult when
multiple genes have an impact, such as with to-
bacco use. In linkage studies of smoking, it is com-
mon for investigators to identify families, ideally
with two or more biologically-related relatives that
have the trait or disorder under study (referred to
as affected individuals, in this case, smokers) and
other unaffected relatives. For example, data from
affected sibling pairs with parents is a common de-
sign in linkage analysis. A tissue sample (typically
blood) is taken fromeach individual, and the sample
undergoes genotyping to obtain information about
the study participant’s unique genetic code. If a
gene in a specific region of a chromosome is as-
sociated with smoking, and if a genetic marker is
linked (i.e., in proximity), then the affected pairs
(such as affected sibling pairs) will have increased
odds for sharing the same paternal/maternal gene
[91].
As genetic research moves forward, new clues
provide insight into which genes might be promis-
ing “candidates” as contributors to tobacco use and
dependence. Currently, there are two general lines
of research related to candidate genes for smoking.
One examines genes that affect nicotine pharmaco-
dynamics (the way that nicotine affects the body)
and the other examines genes that affect nicotine
pharmacokinetics (the way that the body affects
nicotine). A long list of candidate genes are being
examined—some of the most extensively explored
involve (a) the dopamine reward pathway (e.g.,
those related to dopamine synthesis, receptor acti-
vation, reuptake, and metabolism) and (b) nicotine
metabolism via the cytochrome P450 liver enzymes
(specifically, CYP2A6 and CYP2D6).
In summary, each of these types of study designs
supports the hypothesis that genetics influence the
risk for a wide range of tobacco-related phenotypes,
such as ever smoking, age at smoking onset, level
of smoking, ability to quit, and the metabolic path-
ways of nicotine (e.g., see [45,89,95–99]). But given
that there are many predictors of tobacco use and
dependence, of which genetic predisposition is just
one piece of a complex puzzle, it is unlikely that so-
ciety will move toward widespread genotyping for
early identification of individuals who are at risk
for tobacco use. Perhaps a more likely use of ge-
netics as related to tobacco use is its potential for
improving our treatment for dependence [91]. If
genetic research leads to new knowledge regarding
the mechanisms underlying the development and
maintenance of dependence, it is possible that new,
more effective medications might be created. Fur-
thermore, through pharmacogenomics research we
might gain improved knowledge as to which pa-
tients, based on their genetic profiles, would be best
treatedwithwhichmedications. Researchers are be-
ginning to examine how DNA variants affect health
outcome with pharmacologic treatments, with a
goal of determining which genetic profiles respond
most favorably to specific pharmaceutical aids for
cessation (e.g. [98,100–103]).