consangunity

1. doi: 10.1111/j.1365-2796.2006.01741.x
Is susceptibility to tuberculosis acquired or inherited?
E. Schurr
McGill Centre for the Study of Host Resistance, McGill University and The Research Institute of the McGill University Health
Centre, Montreal, QC, Canada
Abstract. Schurr E (McGill University and The
Research Institute of the McGill University Health
Centre, Montreal, QC, Canada). Is susceptibility
to tuberculosis acquired or inherited? (Review).
J Intern Med 2007; 261: 106–111.
Tuberculosis is an ongoing major public health prob-
lem on a global scale. One of the striking features of
the disease is that only an estimated 10% of immuno-
competent persons infected by the causative pathogen
Mycobacterium tuberculosis will develop clinical
signs of disease. This well-established epidemiological
observation has prompted an intense search for the
factors that trigger advancement of infection to dis-
ease in the small proportion of susceptible individuals.
Central to this search is the questions if tuberculosis
patients are inherently susceptible to the disease or if
disease development is promoted by specific environ-
mental factors. It is known that genetic and non-
genetic factors of both the bacterium and the host
have impact on the host response to M. tuberculosis.
Yet, little is known about the interaction of these dif-
ferent factors and the resulting impact on disease
development. Recent work suggests that in addition to
common host susceptibility genes a second group of
susceptibility loci exists the action of which strongly
depends on the individual’s clinical and exposure his-
tory. The latter genes may have a very strong effect
on promoting advancement from infection to disease
only in specific epidemiological settings. These find-
ings suggest that a more detailed knowledge of gene–
environment interactions in tuberculosis is necessary
to understand why a small proportion of individuals
are susceptible to the disease whilst the majority of
humans are naturally resistant to tuberculosis.
Keywords: complex traits, gene–environment interac-
tions, host genetics, tuberculosis.
Tuberculosis, primarily caused by the pathogenic bac-
terium Mycobacterium tuberculosis, continues to be
an enormous public health problem, despite the wide-
spread use of BCG vaccine, multidrug therapy and
the existence of national and international tuberculosis
control programmes. Statistics show that annually
there are 8.8 million new cases and 2 million deaths,
including 250 000 children [1]. It is presently estima-
ted that 2.4 billion people carry M. tuberculosis and
modelling studies suggest that 70 million people will
die of tuberculosis from 1998 to 2018 [2]. These
numbers are likely to continue upwards as HIV infec-
tions increase and new multidrug resistant strains of
M. tuberculosis arise [3].
Assessment of the contribution of genetics of host
resistance to human tuberculosis is a long-standing
challenge of human genetics research, and tuberculo-
sis has long been considered as a complex disease
with a strong genetic component [4]. Several lines of
epidemiological and genetic investigations support
this conclusion. For example, in the Northern German
city of Luebeck a vaccination cohort of babies was
accidentally given three successive per os administra-
tions of a virulent strain of M. tuberculosis within
their first 10 days of life. Of the 251 immunologically
naı̈ve infants, 44 become tuberculin positive but
showed no clinical signs of disease, 72 died of tuber-
culosis within 1 year of infection, and 135 developed
clinical tuberculosis but overcame the disease [5].
Several large twin studies revealed a consistent
significant excess of monozygous twins where both
siblings had developed tuberculosis (disease concor-
dance) when compared with dizygous twins or other
106 ª 2007 Blackwell Publishing Ltd
Symposium |

2. pairs of siblings [6–8]. As monozygous twins are gen-
etically identical these studies provide strong data in
support of genetic factors being involved in tuberculo-
sis susceptibility.
More recent investigations have focused on identifica-
tion of the genetic factors underlying tuberculosis sus-
ceptibility in otherwise healthy persons. In two
genome-wide scans of tuberculosis susceptibility in
South Africa [9] and Brazil [10] a series of linked
regions were discovered leading to the identification
of UBE3A on chromosome 15 as candidate tuberculo-
sis susceptibility gene [11]. A recent genome scan in
Moroccan multiplex tuberculosis families identified a
major dominant acting tuberculosis susceptibility gene
located on chromosome region 8q12-q13 [12]. In
addition, numerous candidate gene studies identified
genetic tuberculosis susceptibility markers in the
VDR, IL12, IL12RB1, INFG, MBL, DRB1, SFTPA1/2
and NRAMP1 genes [13–16] but also provided evi-
dence for genetic heterogeneity and ethnicity-specific
association of risk variants with tuberculosis [17, 18].
Of note is the recent identification of MCP1 as a very
strong risk factor with a large population attributable
fraction in two independent samples from Mexico and
Korea [19, 20]. If similar results are found in addi-
tional samples, the MCP1 susceptibility variants are
expected to account for a large proportion of all tuber-
culosis cases.
Whilst the contribution of host genetic factors to tuber-
culosis susceptibility is now well established, it is not
surprising that tuberculosis has a strong environmental
component. Exposure intensities, poor health, malnutri-
tion or social stress are known to contribute to
increased risk of developing tuberculosis. The mecha-
nisms of how exposure to these environmental factors
is translated into tuberculosis disease are not known.
Another question is whether similar clinical disease,
e.g. smear-positive pulmonary tuberculosis, represents
the same pathogenesis of disease development. It
stands to reason that a complex disease like tuberculo-
sis is not simply dependent upon the clinical end-point,
one could say is not a ‘function-of-state’, but that the
mechanistic pathways that lead to the same clinical
end-points can be different in different epidemiological
settings. For example, in a high-income country like
Canada, it is likely that a new immigrant from a devel-
oping country who develops tuberculosis shortly after
arrival will display different triggers of progression
from infection to disease when compared with a Cana-
dian-born middle-class patient. Consequently, different
genetic susceptibility factors may affect such patients
with different life experience.
If the above proposition is correct, it is to be expected
that genetic variants will reveal themselves as disease
risk factors to highly variable degrees in different epi-
demiological settings. Assuming that the differences
in mechanisms of disease pathogenesis are largely the
result of environmental factors, we can expect exten-
sive gene–environment interactions that can masquer-
ade as genetic heterogeneity of disease susceptibility.
Moreover, it is to be expected that the sequence of
environmental exposures will be critical due to the
adaptive capability of the human immune response.
Another factor that can give rise to difficulties in rep-
licating genetic factors of disease susceptibility is that
the same diagnostic name is given to a disease syn-
drome. In tuberculosis, for example, extrapulmonary,
pulmonary or tuberculosis meningitis may or may not
share the same basic defects of host responsiveness.
Likewise, primary and reactivational tuberculosis, i.e.
tuberculosis before or after 2 years after initial infec-
tion with M. tuberculosis, almost certainly differ in
critical aspects of pathogenesis, yet in most studies of
adult tuberculosis it is not easily possible to make a
distinction between these two disease subforms. If dif-
ferent study samples are made up of different propor-
tions of disease subtypes, it is not surprising that
replication of genetic control elements is not possible.
Difficulties in replicating genetic findings across stud-
ies are often interpreted as evidence against an
important role of host genetics. There are many rea-
sons why even in the case of true susceptibility genes
genetic findings are difficult to replicate (rev [21]). Of
course, it is possible that genetic susceptibility factors
detected in one study but not in others are false posi-
tives; however, it is also possible that difficulties in
replicating genetic results are caused by our incom-
plete understanding of disease pathogenesis and the
role played by gene–environment interactions.
E. Schurr
| Symposium: Host genetics of tuberculosis
ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 106–111 107

3. On the other hand, it seems equally likely that certain
key host response pathways directly underlie the dis-
ease phenotype. Genetic defects in such pathways are
likely to be unaffected by environmental factors other
than exposure to M. tuberculosis. In this scenario,
environmental factors (such as pathogen exposure)
may select the trigger but effectiveness of this step of
pathogenesis will be relatively insensitive to how trig-
gering occurred. Genetic defects in such basic path-
ways will reveal themselves as genetic tuberculosis
risk factors across populations and epidemiological
settings. The above points will be illustrated by recent
work from our group.
In 1987 an outbreak of primary tuberculosis affected
an aboriginal community in Northern Alberta [22].
Within one extended family of 81 persons, a total of
24 tuberculosis patients were identified based on clin-
ical symptoms or microbiological detection of
M. tuberculosis in sputum samples. Except for one
patient who was diagnosed 18 months after the index
case and identified as having been infected by a sec-
ondary index case, all cases were diagnosed within
5–6 months of the index case. Based on routine tuber-
culosis surveys in the community prior to the out-
break, clinical data and vaccination records, it was
possible to group family members into three liability
classes [23]. The most common liability class was
made up of people who converted to tuberculin posi-
tivity during the outbreak indicating productive expo-
sure to M. tuberculosis. Those people had no record
of previous disease or vaccination and were given the
highest penetrance (85%) of a putative tuberculosis
susceptibility gene. A second liability class contained
persons with a documented record of BCG vaccin-
ation or previous disease or who had been tuberculin
positive prior to the outbreak for unknown reasons.
Those people were considered to have acquired partial
protection from tuberculosis because of previous
exposure to mycobacterial antigen. Based on results
from a BCG vaccination trial in Canadian aboriginal
groups, the penetrance of their susceptibility gene was
set to 57% of class I. A third group of persons had
not converted to positive tuberculin status during the
outbreak. Based on the known false negative rate of
tuberculin tests the penetrance of their susceptibility
gene was set to 10%. Finally, very young and very
old individuals were considered at high risk of tuber-
culosis because of age-specific constitutive factors
[23] (Fig. 1).
Fig. 1 Partial pedigree of extended family suffering from an outbreak of primary tuberculosis. Below each circle in the pedi-
gree the following information is given: result of Mycobacterium tuberculosis sputum culture (c+: culture positive, c): culture
negative); liability class for genetic analysis; NRAMP1 haplotypes; D2S426 alleles, TNF haplotype. Reproduced with permis-
sion from Greenwood et al. [23].
108 ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 106–111
E. Schurr
| Symposium: Host genetics of tuberculosis

4. What has been accomplished by the introduction of
such liability classes? This is simply a means to state
that the same genetic factors can have very different
effects on disease susceptibility dependent upon the
exposure history of each person in the pedigree. In
other words, liability classes are modelling gene–envi-
ronment interactions. This leaves the question, which
gene are we modelling? Of course, any gene can be
selected. In the above study, for a variety of reasons
including results from the mouse model [24], and
genetic studies in tuberculosis [25] and leprosy [26] it
had been decided to test the human NRAMP1 gene.
The function of the human NRAMP1 protein is
unknown. However, the mouse protein has been
shown to be a mediator of divalent cation fluxes
across the phagosomal membrane, and it appears
likely that human NRAMP1 will have a similar func-
tion. Intra-phagosomal cation concentrations are medi-
ators of phagosome maturation and pathogen survival
[27]. Based on results from the mouse model it was
decided to test a strong effect (relative risk = 10) of a
dominant acting susceptibility variant. When the cor-
responding linkage analysis was conducted, the results
very strongly supported linkage of the NRAMP1 gene
to tuberculosis susceptibility (P 10)5
). The truly
impressive result was that when the same analysis
was repeated without taking into account the above-
described gene–environment interactions, the evidence
for linkage was close to nil [23]. Hence, our genetic
study of the tuberculosis outbreak had not only detec-
ted a very strong genetic effect but had also shown
that even very strong genetic effects can go undetec-
ted if gene–environment interactions are not taken
into consideration.
The above-described study differed in two important
features from other studies of tuberculosis susceptibil-
ity. First, due to the geographical location and expo-
sure history the outbreak affected a group of people
largely immunologically naı̈ve to mycobacterial anti-
gens. Secondly, the time of infection was known for
all cases, and diagnosis occurred short periods after
exposure. Hence, all patients were categorized as pri-
mary tuberculosis cases [22]. It is possible that the
strong effect of the NRAMP1 gene on susceptibility
might not have been detected if the cases had largely
been suffering from reactivational tuberculosis or if
immunological sensitization had occurred prior to the
outbreak. To test if the NRAMP1 effect was more pro-
nounced amongst tuberculosis cases suffering from
primary disease, we enrolled a familial sample of pae-
diatric tuberculosis cases from Houston, TX [28]. In
the overall sample, we found a modest effect of
NRAMP1 alleles on susceptibility (approximate relat-
ive risk = 1.7). However, when stratifying the sample
on exposure intensity as determined by the proportion
of healthy co-sibs who were tuberculin positive, we
observed that in families with low exposure pressure
the strength of the NRAMP1 effect was significantly
increased. For example, in families with more than
one tuberculosis case the proportion of tuberculin-
positive co-sibs was approximately two-thirds indica-
ting high exposure to M. tuberculosis and the
NRAMP1 effect was not detectable. In contrast,
amongst simplex families (i.e. families with only a
single case) the proportion of tuberculin-positive
co-sibs was on average less than one-third and the
NRAMP1 effect was significantly stronger (approxi-
mate relative risk = 3.3). Moreover, amongst simplex
families the effect was mainly contained in those fami-
lies with no additional tuberculin-positive sibs. These
results demonstrate again that the effect of the
NRAMP1 gene on susceptibility strongly depends on
environmental parameters [28]. Considering the rela-
tively crude methods employed, it seems likely that
we have only seen the tip of an iceberg and that
gene–environment interactions in tuberculosis suscep-
tibility are playing a pivotal role in controlling the
flow of M. tuberculosis through exposed populations.
An example of a tuberculosis susceptibility locus that
has a strong effect independent of additional covariates
is given by the results of a recent genome scan in
Moroccan tuberculosis families [12]. For this study, 96
families comprising 227 sputum culture-positive pul-
monary tuberculosis cases were studied for genome-
wide evidence for the presence of a major tuberculosis
locus. A major locus was found on chromosome
region 8q12-q13 close to the anonymous marker
D8S1723 with high confidence (P = 3 · 10)5
). Closer
analysis revealed that 26 known genes underlie the
linkage peak and qualify as strong candidates for
E. Schurr
| Symposium: Host genetics of tuberculosis
ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 106–111 109

5. tuberculosis susceptibility genes [12]. The true suscep-
tibility variant can now be identified by applying the
same strategy that we have used successfully for the
cloning of leprosy susceptibility variants on chromo-
some region 6q25 in the PARK2/PACRG shared pro-
moter region [29, 30], and the corresponding
experiments are ongoing. However, a closer look at
the inheritance pattern of the susceptibility locus
revealed an unexpected finding. Of the 96 families
enrolled in the study, 39 had at least one parent affec-
ted with tuberculosis whilst in 57 families only chil-
dren are patients. When conducting linkage analysis
independently in both family samples it turned out that
there was very strong evidence for linkage only in the
affected parent families employing two different types
of linkage analysis [12]. These results are consistent
with a dominant mode of inheritance of the major
tuberculosis susceptibility locus detected. The mode of
inheritance of strong tuberculosis susceptibility loci
has some interest for the understanding of the natural
history of tuberculosis. For example, if in previously
exposed populations, such as the Europeans over the
last 400 years, a substantial proportion of cases was
caused by rare dominant acting susceptibility loci, then
selection against such loci is expected to have been
severe. In other words, such populations may have
developed increased resistance to tuberculosis due to
genetic selection. This is a tantalizing possibility that
clearly needs additional study, especially in the light
of its potential importance for tuberculosis vaccine
development.
To return to the initial question of whether tuberculo-
sis susceptibility is acquired or inherited, the answer
is probably both. Tuberculosis susceptibility is likely
a fleeting target that depends on the specific exposure
history of a population or person. As shown, there are
strong genetic risk factors that act independently of
environmental modulators, there are strong risk fac-
tors that can only be revealed in the context of gene–
environment interactions, and possibly there will be
instances when the environment overrides any exist-
ing genetic boundaries. Of particular interest is new
evidence suggesting that M. tuberculosis strains are
far from homogeneous in their ability to interact with
a human host [31]. Yet, no genetic study of tuberculosis
susceptibility has taken into account such pathogen
diversity. This suggests that we have merely begun
our understanding of the joint and independent contri-
butions of host genome and environment to tuberculo-
sis susceptibility.
Conflict of interest statement
No conflict of interest was declared.
References
1 Raviglione MC. The TB epidemic from 1992 to 2002. Tubercu-
losis 2003; 83: 4–14.
2 Murray CJ, Salomon JA. Modeling the impact of global tuber-
culosis control strategies. Proc Natl Acad Sci U S A 1998; 95:
13881–6.
3 WHO. Global Tuberculosis Control. Surveillance, Planning,
Financing. WHO Report. Geneva: WHO, 2004.
4 Neel JV, Schull WJ. Human Heredity. Chicago, IL: University
Chicago Press, 1954.
5 Schurmann A. Beobachtungen bei den Luebecker Sae-
uglingstuberkulosen. Beitr Klin Erforsch Tuberk Lungenkrankr
1929; 81: 694–703.
6 Diehl K, von Verscheur O. Der Erbeinfluss bei der Tuberkulose.
Jena: Gustav Fischer Verlag, 1936.
7 Kallmann FJ, Reisner D. Twin studies on the significance of
genetic factors in tuberculosis. Am Rev Tuberc 1943; 47: 549–
74.
8 Comstock GW. Tuberculosis in twins: a re-analysis of the Pro-
phit survey. Am Rev Respir Dis 1978; 117: 621–4.
9 Bellamy R, Beyers N, McAdam KP et al. Genetic susceptibility
to tuberculosis in Africans: a genome-wide scan. Proc Natl
Acad Sci U S A 2000; 97: 8005–9.
10 Miller EN, Jamieson SE, Joberty C et al. Genome-wide scans
for leprosy and tuberculosis susceptibility genes in Brazilians.
Genes Immun 2004; 5: 63–7.
11 Cervino AC, Lakiss S, Sow O et al. Fine mapping of a putative
tuberculosis-susceptibility locus on chromosome 15q11-13 in
African families. Hum Mol Genet 2002; 11: 1599–603.
12 Baghdadi JE, Orlova M, Alter A et al. An autosomal dominant
major gene confers predisposition to pulmonary tuberculosis in
adults. J Exp Med 2006; 203: 1679–84.
13 Casanova JL, Abel L. Genetic dissection of immunity to myco-
bacteria: the human model. Annu Rev Immunol 2002; 20: 581–
620.
14 Malik S, Schurr E. Genetic susceptibility to tuberculosis. Clin
Chem Lab Med 2002; 40: 863–8.
15 Casanova JL, Abel L. The human model: a genetic dissection
of immunity to infection in natural conditions. Nat Rev Immunol
2004; 4: 55–66.
16 Cooke GS, Hill AV. Genetics of susceptibility to human infec-
tious disease. Nat Rev Genet 2001; 2: 967–77.
E. Schurr
| Symposium: Host genetics of tuberculosis
110 ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 106–111

6. 17 Delgado JC, Baena A, Thim S, Goldfeld AE. Ethnic-specific
genetic associations with pulmonary tuberculosis. J Infect Dis
2002; 186: 1463–8.
18 El Baghdadi J, Remus N, Benslimane A et al. Variants of the
human NRAMP1 gene and susceptibility to tuberculosis in
Morocco. Int J Tuberc Lung Dis 2003; 7: 599–602.
19 Flores-Villanueva PO, Ruiz-Morales JA, Song C-H et al. A
functional promoter polymorphism in monocyte chemoattractant
protein-1 is associated with increased susceptibility to pulmon-
ary tuberculosis. J Exp Med 2005; 202: 1649–58.
20 Alcais A, Fieschi C, Abel L, Casanova JL. Tuberculosis in chil-
dren and adults: two distinct genetic diseases. J Exp Med 2005;
202: 1617–21.
21 Zondervan KT, Cardon LR. The complex interplay among fac-
tors that influence allelic association. Nat Rev Genet 2004; 5:
89–100.
22 Mah AW, Fanning EA. An epidemic of primary tuberculosis in
a Canadian aboriginal community. Can J Infect Dis 1991; 2:
133–41.
23 Greenwood CM, Fujiwara TM, Boothroyd LJ et al. Linkage of
tuberculosis to chromosome 2q35 loci, including NRAMP1, in
a large aboriginal Canadian family. Am J Hum Genet 2000; 67:
405–16.
24 Vidal SM, Malo D, Vogan K, Skamene E, Gros P. Natural
resistance to infection with intracellular parasites: isolation of a
candidate for Bcg. Cell 1993; 73: 469–85.
25 Bellamy R, Ruwende C, Corrah T, McAdam KP, Whittle HC,
Hill AV. Variations in the NRAMP1 gene and susceptibility to
tuberculosis in West Africans. N Engl J Med 1998; 338: 640–4.
26 Abel L, Sanchez FO, Oberti J et al. Susceptibility to leprosy is
linked to the human NRAMP1 gene. J Infect Dis 1998; 177:
133–45.
27 Gros P, Schurr E. NRAMP1 and resistance to intracellular path-
ogens. In: Bellamy R, ed. Susceptibility to Infectious Diseases.
Cambridge: Cambridge University Press, 2004; 221–58.
28 Malik S, Abel L, Tooker H et al. Alleles of the NRAMP1 gene
are risk factors for pediatric tuberculosis disease. Proc Natl
Acad Sci U S A 2005; 102: 12183–88.
29 Mira MT, Alcais A, Van Thuc N et al. Chromosome 6q25 is
linked to susceptibility to leprosy in a Vietnamese population.
Nat Genet 2003; 33: 412–5.
30 Mira MT, Alcais A, Van Thuc N et al. Susceptibility to leprosy is
associated with PARK2 and PACRG. Nature 2004; 427: 636–40.
31 Reed MB, Domenech P, Manca C et al. A glycolipid of hyper-
virulent tuberculosis strains that inhibits the innate immune
response. Nature 2004; 431: 84–7.
Correspondence: Erwin Schurr, Montreal General Hospital Research
Institute, 1650 Cedar Avenue, Room L11-521, Montreal, QC, H3G
1A4, Canada.
(fax: 1 514 934 8238; e-mail: erwin.schurr@mcgill.ca).
E. Schurr
| Symposium: Host genetics of tuberculosis
ª 2007 Blackwell Publishing Ltd Journal of Internal Medicine 261; 106–111 111