doi: 10.1111/iji.

12170

TLR8 gene polymorphism and association in bacterial load in southern

Punjab of Pakistan: an association study with pulmonary tuberculosis
M. Bukhari*,1, M. A. Aslam*,1, A. Khan, Q. Iram*, A. Akbar, A. G. Naz*, S. Ahmad,
M. M. Ahmad, U. A. Ashfaq**, H. Aziz & M. Ali*

Summary
Only 510% of people infected with Mycobacterium
tuberculosis develop active tuberculosis which suggests
a role of genetic variation in host immunity. Genetic
variants in TLRs are potential indicator for host
susceptibility and outcome of several diseases. We
explored the association of nonsynonymous genetic
variants (Met1Val) with Toll-like receptor 8 in Pakistani population. Genotypic and allelic distribution of
TLR8 polymorphism (rs3764880) in patients with TB
and healthy donors from different areas of southern
Punjab, Pakistan, was determined. Results provide that
our population is highly influenced by TLR8 Met1Val
SNP for TB, and G allele appeared to increase TB susceptibility. Mutant genotype GG or G/ and G allele
was significantly higher among all the categories of
cases than in controls. Among different levels of bacillary load and genotypes, GG or G/ and G allele significantly supports the incidence of 2 + class for
bacterial load.

Introduction
Pulmonary tuberculosis is an infectious and transmittable disease caused by Mycobacterium tuberculosis
(M. tuberculosis) in which lung cells are damaged.
*Institute of Molecular Biology and Biotechnology, Bahauddin Zakariya University, Multan, Pakistan, Institute of Microbiology, University of Agriculture Faisalabad, Faisalabad, Pakistan, Department of
Statistics, Bahauddin Zakariya University, Multan, Pakistan, Provincial TB Reference Laboratory, Nishtar Medical College and Hospital,
Multan, Pakistan, Social Security Hospital, Faisalabad, Pakistan,
**Department of Bioinformatics and Biotechnology, Government
College University (GCU), Faisalabad, Pakistan and Diagnostics Laboratories, Nuclear Medicine Oncology and Radiotherapy Institute, Islamabad, Pakistan
Received 16 January 2014; revised 13 October 2014; accepted
1 December 2014
Correspondence: Abrar ul Haq Khan, Institute of Microbiology,
University of Agriculture Faisalabad, Faisalabad, Pakistan.
Tel: +92321-7684893; Fax: +92(0)61 9210068;
E-mail: abrarulhaqkhan@gmail.com
1

These authors contributed equally.

46

This results in incompetent processing of inhaled air

leading to oxygen depletion in the brain (Smith,
2003). M. tuberculosis infects nearly one-third of the
worlds population (WHO, 2012). Approximately,
420 000 new TB cases emerge annually, and half of
these are sputum smear positive (WHO, 2012). Pakistan ranks fifth among the highest 22 TB burden countries in the world accounting 81% of cases. In
Pakistan, annually 231 new cases are reported of
100 000 people (Dogar et al., 2012).
In spite of a huge population infected with TB, it is
estimated that clinical disease is developed only in
10% of the infected individuals (Bloom & Small,
1998). The phenomena that only minority develop
progressive mycobacterial disease depict that host
genetics and its interaction with the environment has a
major contribution for its pathogenesis (Frieden et al.,
2003). Recently, a rapid method for the detection of
certain single nucleotides polymorphisms (SNPs) that
can help diagnosis and treatment of TB is developed;
showing the great potential of SNPs diagnosis in biological system (Chen et al., 2013).
Human Toll-like receptors (TLRs) have driven great
attention as immune pattern sensors of non-selfnucleic acids following encountered by phagocytes.
Their stimulation leads to complex changes in cellular
microenvironment including release of various cytokines, induction of inflammatory and adaptive immune
response. Recent studies have confirmed the role of
single nucleotide polymorphisms (SNPs) within TLR
genes impairing an individuals ability to respond
appropriately to TLR ligands (Dalgic et al., 2011).
These findings encouraged to inspect genetic variants
in TLRs (Medzhitov, 2001; Akira & Takeda, 2004;
Ben-Ali et al., 2004; Bulut et al., 2005; Takeda &
Akira, 2005) in relevance to human susceptibility to
pulmonary TB.
TLR8 has shown its role both in combating viral
infection by recognizing single stranded RNA (ssRNA)
and also in different carcinomas. But recently, its
potential to determine susceptibility against bacterial
infections including pulmonary tuberculosis across various populations has been confirmed. TLR8 gene is

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

TLR8 polymorphism in Pakistan

predominantly expressed in lung and peripheral blood

leucocytes. Although the molecular basis for ligand
recognition and activation of signalling by TLR8 is
unknown, but its crystal structure of both unliganded
and ligand induced activated human TLR8 dimer is
solved (Tanji et al., 2013). It has been hypothesized
that after bacterial ingestion by host cell certain bacterial secreted product might become involved in triggering TLR8 activation. TLR8 gene is located on X
chromosome (Xp22.3-p22.2), and thus, males carrying
a single copy of defective allele make them highly vulnerable for tuberculosis. Among others a missense variant, rs3764880 (Met1Val; A>G) (Davila et al., 2008)
in TLR8 has shown a protective role of variant allele
G against TB along with the fact that its presence
strikingly reduces the expression of the most abundant
isoform TLR8v2. Literature reveals only a couple of
studies on the association of SNPs in TLR8 with the
incidence of tuberculosis. But both of them were carried on populations different from Pakistan (Russia,
Indonesia and Turkey). Due to a great need to explore
the genetic reasons for the larger incidence of TB, this
study was designed to find whether methionine to
valine shift in TLR8 is associated with pulmonary
tuberculosis in Pakistani population.

Materials and methods

Collection of samples

A total of 190 individuals with 103 clinically diagnosed patients with TB and 87 unrelated healthy controls were recruited in the study. Absence of active or
latent TB among healthy group was confirmed by
interferon-gamma release assay (QuantiFERON-TB 76
Gold In-Tube). The healthy controls were sputum negative for acid fast bacilli test as per WHO guidelines.
All subjects were HIV-negative and were clear with
any type of serious pathologies such as diabetes,
asthma, cardiovascular, cancer and atopy or autoimmune diseases like systemic lupus erythematosus. All
the patients had positive history of BCG vaccination.
Among patients, 45 were males and 58 were females
while in control group, 22 were males and 65 were
females. TB diagnosis for patients by chest radiograph
and their categorization according to sputum positivity
level based on culture using LowensteinJensen medium, that is sputum 1 + (1099/100 oil immersion
field), 2 + (110/oil immersion field), 3 + (More than
10/oil immersion field) was carried out by an expert
clinician at social security hospital Faisalabad, according to the described method (Rajpal et al., 2002).
Category four patients were those who were either
taking antituberculosis therapy or have completed
DOTs course. Table 1 shows the summary for disease
categories along with the fields examined and number
of patients under each group.
A standard interview-administered questionnaire
was designed to collect all the characteristic informa-

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

47

Table 1. Sputum positivity test-based categorization of patients

Sputum grading levels

Fields examined

Number of patients

1+
2+
3+
Other (Under treatment)
Total

100
50
20

16
13
9
65
103

tion of the participants in the study. After taking a

written informed consent, about 3 mL of peripheral
blood was collected from each participant for further
analysis. All the protocols were approved by the ethical review committee of the Institute of Molecular
Biology and Biotechnology, Bahauddin Zakariya University, Multan, and were adhering to the tenets of the
Declaration of Helsinki.
Genotyping

Genomic DNA was extracted from blood lymphocytes

using the standard phenol chloroform extraction
method (Sambrook and Russell, 2001). PCR-RFLP
assay was used to determine TLR8 genotype (Engin
et al., 2010). PCR was performed in a total volume of
25 lL having approximately 100 ng genomic DNA,
0.2 mM each of primers, 200 mM of dNTPs, 1.5 mM
MgCl2, 19 reaction buffer (20 mM (NH4)2SO4,
0.01% Tween 20, 75 mM Tris HCI pH 8.8 at 25C,
MBI Fermentas, Lahore, Pakistan) and 2 units Taq
polymerase (MBI Fermentas). Thermal profile was as
follows: initial denaturation was carried out at 94C
for 5 min, followed by 35 cycles of denaturation at
94C for 30 s, annealing at a 57C for 30 s and chain
elongation at 72C for 1 min. A final extension was
carried out at 72C for 5 min. PCR product of 393 bp
was visualized under UV illumination by Gel Doc XR
(Bio-Rad, Philadelphia, PA, USA) in a 2% agarose gel
containing ethidium bromide (0.5 mg mL 1; Sigma,
Northbrook, IL, USA). PCR product was digested with
NlaIII (Fermentas) at 37C overnight. Presence of
three fragments of 97, 137 and 156 bp indicated
homozygous wild genotype (A/A), while two fragments of 137 and 256 bp indicated homozygous
mutant genotype (G/G) of TLR8 Met1Val.
Statistical analysis

version 13.0 (SPSS Inc., Chicago, IL, USA) was

used for carrying statistical analysis. Genotypic and
allelic frequencies were calculated by direct counting
method followed by HardyWeinberg equilibrium
(HWE) test for genotypes using chi-square analysis.
Cross-tabulation in complex samples analysis was used
to measure the association between the disease incidence and various demographic and genetic parameters. Statistical significance of the association was

SPSS

48

M. Bukhari et al.

figured by Pearson chi-square test. Comparison of allelic and genotypic frequencies between cases and control was done by 2 9 2 contingency tables, chi-square
or Yates corrected chi-square as appropriate. Allelic
and genotypic association with the occurrence of TB
was measured by calculating odd ratio (OR), and
95% confidence intervals (95% CI) as an estimate of
relative risk. The results were considered significant
with the P value lesser than 0.05.

Results
Impact of age and sex as risk factor for the incidence of
pulmonary tuberculosis

Table 2 showed the analysis of both the factors for the

occurrence of TB where sex appeared as a major deciding agent for determining the fate of infection. Males
appeared to be more susceptible for TB as males were
104.54% higher in cases than in control. On the other
side, females were 10.77% lesser in number in cases
than control group. But regarding different age groups,
the one ranging from 5 to 45 years failed to express
their effect on the incidence of tuberculosis. However,
within age category highest number was seen in the age
group more than 45 years irrespective of the sex. Combinational analysis of both the sexes within each age
group also did not show any statistical effect.
Association of toll-like receptor 8 gene polymorphism with
pulmonary tuberculosis in southern Punjab

Genotypic and allelic frequencies distribution among

cases and controls group are given in Table 3. Significant difference for genotypic polymorphism was
observed between cases and control group. Mutant
genotype for Met1Val polymorphism was significantly

Table 2. Demographic factors studied for their inuence on

phenotype

Parameters
Sex
Male
Female
Age (in years)
520
2145
>45
Sex 9 Age
Male 9 520
Male 9 2145
Male 9 >45
Female 9 520
Female 9 2145
Female 9 >45
Total

Control

Cases

Chi-square
value

d.f.

P value

0.006

22
65

45
58

6.996

25
29
33

25
38
40

0.537

0.765

8
4
10
17
25
23
87

8
20
17
17
18
23
103

5.199

0.074

0.744

0.690

P values differing signicantly appear as bold.

(P < 0.0001) higher among cases than in controls as

no GG candidate was seen in healthy individuals.
GG was also highest in number among all three
genotypes within cases and AA, wild-type being lowest in this group. Both wild and heterozygous genotypes have also shown highly significant difference
among cases and control as in either case their frequency was higher in control than cases.
Consequently, cases have shown significant higher
(P < 0.0001) allelic frequency of variant allele G of
Met1Val as compared to controls with OR of 3.758
(2.4525.759). This depicts that there is nearly fourfold increase in chance of having TB if G allele is present. For A allele, the situation was reversed as it was
significantly higher in control against cases.
Inuence of Met1Val mutation in TLR8 gene on bacillary
load

To check the role of the gene variants with the progression of disease, patients were categorized in four
subgroups (including the one who were taking ATT or
completed DOTs course) and compared with the
healthy controls (Table 4). When we distribute
patients in 2 + class, we found a significant association with GG or G/ genotype (v2 = 7.80, d.f. = 2, P
value = 0.02) as frequency of GG or G/ was higher
as compared to other genotypes. When we compare
the allele frequency in 2 + class, we found a significant difference between G and A allele (P
value < 0.0001). Frequency of G allele is much higher
than A. In therapy/DOTS group genotypic comparison
showed that frequency of AA or A/ is very less as
compared to AG and GG or G/
(v2 = 16.66,
d.f. = 2, P value = 0.0002) while in control group frequency of GG or G/ is zero, and frequency of AA or
A/ and AG was much higher. This difference showed
a significant association of AA or A/ and AG genotypes with control group (P value < 0.0001). Allelic
distribution in control group also showed much higher
frequency of allele A (118) as compared to G allele
(56) and showed a significant association (P
value < 0.0001).
Sex-wise comparison of TLR8 polymorphism

Statistically significant different results regarding genotypic distribution were obvious between males and
females irrespective of their health status (Table 5).
TLR8 being located on X chromosome males and
females were separately analysed. Among females,
there were no significant differences between the pulmonary TB and control group for all the genotypes
(P > 0.05). But frequency of allele A was significantly
higher in female control group as compared to female
patients with TB (P > 0.05). In female group, the presence of G allele increases the incidence of disease by
more than two and half folds. On the other hand, a
strong association with genotype G/( ) at rs3764880

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

TLR8 polymorphism in Pakistan

49

Table 3. Frequency distribution of different genotypes and alleles among southern Punjab subjects

Genotype/allele

Cases

Control

AA or A/
AG
GG or G/
Total
A
G

16
42
45
103
74
132

31
56
0
87
118
56

(0.16)
(0.41)
(0.44)
(0.36)
(0.64)

(0.36)
(0.64)

(0.68)
(0.32)

Chi-square
value

P value
(Yates corrected)

10.233
10.510
49.806
50.800
38.384

0.0014
0.0012
<0.0001

0.3322 (0.16650.6626)
0.3811 (0.21150.6868)

<0.0001

0.2661 (0.17370.4078)
3.758 (2.4525.759)

OR (95% CI)

OR, Odds ratio; CI, condence interval.

All P values given in the table are Yates corrected. The values differing signicantly appear as bold. P values for chi-square test used to test the
signicance of differences among cases and controls of Sahariya.

Table 4. Comparative distribution of genotypic and allelic frequencies between controls and sub categorizes of patients
Genotypes

Alleles

AA or A/
1+
2+
3+
Therapy/DOTS
Control

AG

3
2
2
9
31

5
4
5
28
56

GG or G/

P value

6
9
2
28
0

v
v2
v2
v2
v2
2

=
=
=
=
=

1.50, d.f. = 2, P value = 0.4724

7.80, d.f. = 2, P value = 0.0202
3.00, d.f. = 2, P value = 0.2231
16.66, d.f. = 2, P value = 0.0002
81.41, d.f. = 2, P value = 0.0000

P value

11
8
9
46
118

17
22
9
84
56

P
P
P
P
P

=
=
=
=
=

0.101
0.000
1.000
0.000
0.000

1 + , 2 + and 3 + are the three categories of bacillary load during active disease development. The fourth category other refers to patients
who were on therapy or completed their DOTS course. The values differing signicantly appear as bold.

Table 5. Gender based comparison of genotypic and allelic frequencies between patients with TB and healthy individuals
Genotype
distribution

Female case

Female control

P value

OR (95% CI)

AA
AG
GG
Total
A
G

12
42
4
58
24
50

9
56
0
65
74
56

(13.85)
(86.15)

0.443
0.095
0.1003

1.623 (0.62894.1896)
0.421 (0.16991.0475)

(0.57)
(0.43)

0.0013

0.363 (0.19980.6604)
2.753 (1.51415.0055)

(20.69)
(72.41)
(6.89)
(0.57)
(0.43)

Genotype
distribution

Male TB case

Male control

P value

A/( )
G/( )

4 (8.89)
41 (91.11)

22 (100%)
0

<0.0001
<0.0001

The values differing signicantly appear as bold.

with susceptibility to pulmonary TB in males

(P < 0.0001) was found. In addition, A/( ) genotype
frequency was significantly lower (P < 0.0001) in the
male patients with pulmonary TB when compared
with the male healthy controls.

Discussion
Toll-like receptors are unique pathogen recognition
receptors (PRRs) that recognize various conserved
motifs in micro-organisms (Medzhitov, 2001). Several
TLR ligands expressed by MTB stimulate the transcription factor NF-QB, CD4+ T-cell proliferation and

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

several proinflammatory cytokines production. The

response to TLR ligands is influenced by certain SNPs
that cause defective intracellular signal transductions
and ultimately result in susceptibility to infectious
diseases (Oh et al., 2008). TLR8 is an intracellular
membrane bounded protein expressed on myeloid dendritic cells and monocytes/macrophages. TLR8 locus
seems to be highly synchronized with development of
active TB. Many studies have confirmed the association of SNPs within TLR8 with several diseases.
Davila et al. found that among four polymorphisms
(rs3764879, rs3788935, rs3761624, rs3764880) in
the TLR8 gene rs3764880 showed greater proof of

50

M. Bukhari et al.

relationship with tuberculosis inclination in adult

males among diverse populations.
These conclusions propose crucial role of TLR8 in
sensing phagocytosed bacteria and also suggest logical
consequences that TLR8 function is modulated by the
(rs3764880: p.Met1Val; A>G) single nucleotide polymorphism (Davila et al., 2008). In an over expression
assay, it has been revealed that mutant allele G
affects NF kappa B activation (Oh et al., 2008) and
thus results a varied response to different TLR8
ligands. Furthermore, it has been proposed that the
loss of initial amino acids could influence protein folding, intracellular trafficking or stability of the mature
protein.
Our study is the first casecontrol report confirming
the relationship of TLR8 SNP with the susceptibility
of Pulmonary TB in southern Punjab, Pakistan. We
found a strong association of Met1Val polymorphism
of TLR8 gene with the susceptibility of TB in cases as
compared to controls. Furthermore, this effect is
clearer among males rather than females. Location of
TLR8 gene on X chromosome helps to propose that
being hemizygous; presence of any disease susceptible
allele have a higher impact among males. Our results
agree with the findings by Davila et al. (2008), who
reported the same association with the susceptibility of
pulmonary TB in Russian and Indonesian population.
This study reports that nearly 30% of Indonesian male
subjects had methionine allele coupled with threat for
TB whereas allelic frequency of the same allele was
78% among Russian patients. Our findings are also in
accordance with another study regarding paediatric
tuberculosis carried on Turkish population showing
that 34.3% of TB infected male children had the Met
allele (Dalgic et al., 2011).
But in contrast to both the studies, we found G
allele responsible for causing susceptibility for tuberculosis in our population. This is in agreement with
the International HapMap Project report 2003 showing that Asian population has an unusually elevated
variant allele frequency (Gibbs, 2003). This could be
due to selective advantage or disadvantage for this
missense mutation in Asian environment. It is notable
to mention that in spite of huge variation in allelic
frequency among all studied populations, rs3764880
has shown the genetic association with the occurrence
of TB. Diversity of host immune response against
pathogens could be due to the ethnic differences in
TLR 8 polymorphism. That is why most of the studies on TLR polymorphisms have been performed on
ethnically matched populations to avoid genetic diversity.
Our data report that among TB cases, 64% carried
variant allele putting nearly fourfold increases in susceptibility for tuberculosis. Gender wise both males
and females have shown significant effect of this SNP
with 2.7 times increase of TB incidence among
females. All the female patients were homozygous for
G allele showing that the presence of A allele either in

homozygous or heterozygous confers resistance against

TB. The frequency of polymorphic genotype G/
was significantly more common among males cases
than in controls with P value <0.05. Hence, our findings draw possible evidence about gender-specific
effects where an increased susceptibility to pulmonary
TB has been shown in male with G/
allele of
rs3764880 being hemizygous for this condition.
To the best of our knowledge, this is the first report
establishing an association with TLR8 SNP severity of
pathogenic load. A significant link was observed
between the Met1Val SNP and disease severity.
Among different levels of bacillary load and genotypes, GG or G/ and G allele significantly support
the incidence of 2 + class for bacterial load. For the
last category of patients with TB, AG and GG genotypes appeared in equal frequency again concluding G
allele as a major factor for disease susceptibility. On
the other hand, it cleared that A allele has a protective
role against TB.
In conclusion, it is the first evidence that like Indonesia, Russia and Turkish population the data from
southern Punjab population of Pakistan, put very
strong influence of TLR8 SNP on TB susceptibility
especially among males. As in southern Punjab, G
allele is reported to have protective role against TB
appeared to be the causative agent in contrast to other
studies so the molecular basis of this incidence upon
the fate of infection must be studied. Deviation in
results of TLR8 polymorphism in southern Punjab
population from other studied populations are may be
due to racial differences. Hence, these SNPs by interacting with other genetic and nongenetic risk factors
may contribute in differential immune response in different ethnic populations. It is worth mentioning that
other populations may present a nonsignificant role of
this SNP for TB susceptibility due to certain other protective genetic factors pointing the need of haplotype
analysis or complete absence of the inappropriate environmental condition in that geographic region could
be another factor.

Conclusion
We conclude that Met1Val of TLR8 is significantly
associated with risk of PTB susceptibility in southern
Punjab population. This effect was highly pronounced
among males. Difference from other populations was
that we found G allele decreases protection against
mycobacterial infection, which could be due to ethnic
difference. Furthermore, for the first time an association of TLR8 SNP with bacterial load was established
describing that double dose of G allele is potentially
deleterious. Further investigations on interaction of
Met1Val polymorphisms with other genetic polymorphisms for genes related to immune system can show
deep insight for understanding its effects on TB resistance, pathogenesis, severity and prognosis especially
in cases of drug resistance.

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

TLR8 polymorphism in Pakistan

Acknowledgements
We are thankful to the study participants of southern
Punjab for their active cooperation and support in this
work. This work was supported by the Institute of
Molecular Biology and Biotechnology, Bahauddin
Zakariya University Multan, Pakistan, and was conducted using facilities available at the institute developed through the research grants from institution.

References
Akira, S. & Takeda, K. (2004) Toll-like receptor signalling.
Nature Reviews Immunology, 4, 499.
Ben-Ali, M., Barbouche, M.R., Bousnina, S., Chabbou, A. &
Dellagi, K. (2004) Toll-like receptor 2 Arg677Trp
polymorphism is associated with susceptibility to tuberculosis
in Tunisian patients. Clinical and Diagnostic Laboratory
Immunology, 11, 625.
Bloom, B.R. & Small, P.M. (1998) The evolving relation between
humans and Mycobacterium tuberculosis. New England
Journal of Medicine, 338, 677.
Bulut, Y., Michelsen, K.S., Hayrapetian, L., Naiki, Y., Spallek,
R., Singh, M. & Arditi, M. (2005) Mycobacterium
tuberculosis heat shock proteins use diverse toll-like receptor
pathways to activate pro-inflammatory signals. Journal of
Biological Chemistry, 280, 20961.
Chen, S.X., Zhang, D.Y. & Seelig, G. (2013) Conditionally
fluorescent molecular probes for detecting single base changes
in double-stranded DNA. Nature Chemistry, 5, 782.
Dalgic, N., Tekin, D., Kayaalti, Z., Cakir, E., Soylemezoglu, T.
& Sancar, M. (2011) Relationship between toll-like receptor 8
gene polymorphisms and pediatric pulmonary tuberculosis.
Disease Markers, 31, 33.
Davila, S., Hibberd, M.L., Hari Dass, R., Wong, H.E.,
Sahiratmadja, E., Bonnard, C. et al. (2008) Genetic
association and expression studies indicate a role of toll-like

2015 John Wiley & Sons Ltd

International Journal of Immunogenetics, 2015, 42, 4651

51

receptor 8 in pulmonary tuberculosis. PLoS Genetics, 4,

e1000218.
Dogar, O.F., Shah, S.K., Chughtai, A.A. & Qadeer, E. (2012)
Gender disparity in tuberculosis cases in eastern and western
provinces of Pakistan. BMC Infectious Diseases, 12, 244.
Engin, A., Arslan, S., Kizildag, S., Ozturk, H., Elaldi, N.,
Dokmetas, I. & Bakir, M. (2010) Toll-like receptor 8 and 9
polymorphisms in crimean-congo hemorrhagic fever. Microbes
and Infection, 12, 1071.
Frieden, T.R., Sterling, T.R., Munsiff, S.S., Watt, C.J. & Dye, C.
(2003) Tuberculosis. Lancet, 362, 887.
Gibbs, R.A. (2003) The International HapMap Project. Nature,
426, 789.
Medzhitov, R. (2001) Toll-like receptors and innate immunity.
Nature Reviews Immunology, 1, 135.
Oh, D.Y., Taube, S., Hamouda, O., Kucherer, C., Poggensee, G.,
Jessen, H. et al. (2008) A functional toll-like receptor 8
variant is associated with HIV disease restriction. Journal of
Infectious Diseases, 198, 701.
Rajpal, S., Dhingara, V.K. & Agarwal, J.K. (2002) Sputum
grading as predictor of treatment outcome in pulmonary
tuberculosis. Indian Journal of Tuberculosis, 49, 139.
Sambrook, J. & Russell, D.W. (2001) Preparation and analysis of
eukaryotic genomic DNA. In: Molecular Cloning: A
Laboratorymanual, 3rd edn, vol 1 (eds J. Sambrook & D.W.
Russell), pp. 6.1 6.12. Cold Spring Harbor Laboratory Press,
New York.
Smith, I. (2003) Mycobacterium tuberculosis pathogenesis and
molecular determinants of virulence. Clinical Microbiology
Reviews, 16, 463.
Takeda, K. & Akira, S. (2005) Toll-like receptors in innate
immunity. International Immunology, 17, 1.
Tanji, H., Ohto, U., Shibata, T., Miyake, K. & Shimizu, T.
(2013) Structural reorganization of the toll-like receptor 8
dimer induced by agonistic ligands. Science, 339, 1426.
World Health Organization. (2012) Global Tuberculosis Report
2012. http://www.who.int/tb/publications/global_report/en/
index.html [Accessed on 17 July 2013].