Eight MCP1 variants were genotyped in 2010 Ghanaian pulmonary TB cases and 2346 healthy control individuals (Table 1). P values, including those of the previous study, were corrected according to the Bonferroni-Holm procedure 17 for the eight comparisons made.

In Table 2, allelic associations of the eight newly genotyped MCP1 variants and of the MCP1 -2581 and -362 variants that were previously typed are given. The deletion of 14 bases length located in the first MCP1 intron, int1del554-567, was associated with protection against pulmonary TB to a similar extent as were the promoter alleles -2581G and -362C (OR = 0.85, confidence interval [CI] 0.78-0.92, P_corr = 0.00098, OR = 0.81, CI 0.73-0.91, P_corr = 0.0012 and OR = 0.83, CI 0.76-0.90, P_corr = 0.00015, respectively).

The genotype frequencies did not deviate from Hardy-Weinberg equilibrium (HWE) among cases and controls. Trend tests were performed to compare the frequencies of genotypes of cases and controls in an additive model and results were adjusted for gender, age and ethnicity. The results are given in Table 3. As also observed in the computation of allelic associations, int1del554-567 was in the trend test significantly associated with protection against TB (OR_trend = 0.84, CI 0.77-0.92, P_corr = 0.00098). In a genotype test where heterozygous and homozygous genotypes were individually compared to the wildtype, a strong association was seen for both heterozygous and homozygous carriers of the int1del554-567 deletion (OR = 0.80, CI 0.70-0.91, P_corr = 0.0063 and OR = 0.73, CI 0.61-0.87, P_corr = 0.0042, respectively), indicating a dominant genetic effect. The association of int1del554-567 was of similar strength as that in heterozygous and homozygous MCP1 -362C carriers in the previous study (Ref. 14 ; OR_trend = 0.83, CI 0.76-0.91, P_corr = 0.00017). Both variants were in strong LD (r² = 0.82). int1del554-567 was also in weak LD (r² = 0.27) with the MCP1 promoter variant at position -2581. Figure 1 shows the r² values of pairwise LDs of all variants examined in the present and in the previous study 14 .

Stratification for mycobacterial species (M. tuberculosis vs. M. africanum) and phylogenetic lineages did not reveal any differences in the associations. Thus, possible confounding exerted by mycobacterial species or distinct genotypes was excluded.

We focused on haplotypic combinations comprising the polymorphisms genotyped in our previous study, MCP1 -2581A/G and -362C/G, and the deletion or wildtype (W) at intron 1 positions 554-567 (int1del554-567/W), because the variant alleles at these positions are in LD and associations of these variants are presented here and have been described previously 4 6 7 14 . As the combination -2581A/-362G/W occurred as the most frequent haplotype in our study population (frequency [f] = 0.55) it was referred to as wildtype reference in further comparisons (Table 4).

The haplotype combinations -2581G/-362C/int1del554-567 and -2581A/-362C/int1del554-567 were significantly associated with resistance to TB compared to the reference haplotype -2581A/-362G/W (OR = 0.78, CI 0.69-0.87, P = 0.00002 and OR = 0.87, CI 0.78-0.96, P = 0.008, respectively; Table 4, Figure 1). The global P value, adjusted through 10 000 permutations, was P_{global/adjusted} = 0.0028.

In order to test variant MCP1 haplotypes with regard to their impact on gene expression, a luciferase reporter gene assay was performed. Figure 2 shows the plots of the Firefly Luciferase/Renilla Luciferase ratios (FL/RL ratios) that were obtained for the constructs subjected to the assay. An overall ANOVA statistics revealed a significant difference between the FL/RL ratios (P = 0.0019). The calculated studentized range critical value in the post hoc pairwise comparisons for variable groups (Tukey-HSD test) was 4.03, and comparisons of the construct carrying the -2581A/-362G/W alleles with the -2581G/-362C/int1del554-567 and -2581A/-362C/int1del554-567 constructs yielded significant results that were above the studentized range critical value (Tukey-HSD test 4.53 and 5.44, respectively).

The P values of a t-test that was calculated with the haplotype -2581A/-362G/W set as reference are given in Table 5. The constructs -2581G/-362C/int1del554-567 and -2581A/-362C/int1del554-567 expressed the luciferase gene to a significantly lower degree than did the wildtype construct -2581A/-362G/W (P = 0.02 and P = 0.006, respectively).

The study design and the enrollment procedure have been described in detail previously 14 26 . In brief, participants were recruited at the two major Ghanaian teaching hospitals in Accra and Kumasi and at additional hospitals and polyclinics in these metropolitan areas and at regional district hospitals. 2010 HIV-negative individuals with smear- and/or culture-positive pulmonary TB were recruited as cases. The control group consisted of 2346 individuals, from whom 1211 were unrelated personal contacts of cases and 1135 were community members from the adjacent neighbourhood or working contacts. The proportion of ethnic groups did not differ significantly between cases and controls. Participants belonged to the ethnic groups of Akan, Ga-Adangbe, Ewe and groups from northern Ghana, including Dagomba, Sissala, Gonja and Kusasi. The male-to-female ratio in the total study group was 1:0.58, and the mean age of participants was 33 years without gender differences. The characterization of phenotypes included documentation of the medical history of cases on structured questionnaires, two independent examinations of non-induced sputum specimens, serological determination of the HIV status, culturing and molecular differentiation of phylogenetic lineages of mycobacterial clades and posterior-anterior chest X-rays. Positive HIV test results were verified in an alternate test system. Fine-typing of genotypes by spoligotyping, IS6110 fingerprinting and determination of drug resistance was performed as previously described 14 . TB-patients were included for specific treatment in the DOTS programme (Directly Observed Treatment Short-Course strategy) organized by the Ghanaian National Tuberculosis Programme.

Of the control group, the medical history was obtained and a clinical examination was performed. Chest X-rays did not reveal any signs of actual or past pulmonary TB. In addition, a tuberculin skin test (TST, Tuberculin Test PPD Mérieux, bioMérieux, Nürtingen, Germany) was performed. The TST was positive in 2217 controls and 129 control individuals were TST-negative.

Ethical approval of the study design was obtained by the Committee on Human Research, Publications and Ethics, College of Health Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana, and the Ethics Committee of the Ghana Health Service, Accra, Ghana. Informed consent was given by study participants either by signature or, in case of illiteracy, by thumbprint in the presence of a witness. The aims of the study and the procedure of venous blood collection were explained before blood samples were taken.

According to the most recent data of the haplotype structure of MCP1 obtained from the innate immunity database https://pharmgat.org/IIPGA2/PGAs/InnateImmunity/CCL2/ we selected eight MCP1 polymorphisms that are in LD with the MCP1 -362 promoter variant which has previously shown the strongest association 14 .

Table 1 lists the variants that were selected, including their rs numbers and PCR amplification primers as well as sensor/anchor nucleotides for LightTyper-based genotyping. Three variants are located in the promoter region, two in the first intron and three in the 3'-UTR.

Standard methods were applied to extract DNA from full venous blood and genotypes of the MCP1 variants were determined by fluorescence resonance energy transfer (LightTyper^®; Roche Diagnostics, Mannheim, Germany) with dynamic allele specific hybridization.

Demographic and self reported data was double entered into a Fourth Dimension database (San José, CA, USA). Genotype frequencies and odds ratios as well as Hardy-Weinberg equilibria (HWE) were calculated with the Stata 10 software (Stata Corporation, College Station, TX, USA) and logistic regression was applied to adjust for age, gender and ethnicities. Allelic and haplotype frequencies and associations were used to reconstruct haplotypes, calculated with the Unphased software (version 3.1.4; http://www.mrc-bsu.cam.ac.uk/personal/frank/software/unphased). P values were adjusted through 10 000 permutations. Haploview version 4.1 http://www.broad.mit.edu/mpg/haploview/ was used to calculate linkage disequilibria (LD, given as r²) and to generate the graphical output. The Tukey Honestly Significant Difference test (Tukey-HSD test) was performed for post hoc comparisons of variable groups in the evaluation of the reporter gene assay.

The power to determine a genetic effect (CaTS software; http://www.sph.umich.edu/csg/abecasis/CaTS/) with a genotype relative risk of 1.4 was, with 2010 cases and 2346 controls and assuming a disease allele frequency of 0.2, a prevalence 0.003 and a significance level of 1 × 10^-7 was 89%.

The PGL2-Control Vector (Promega, Mannheim, Germany) was used for cloning of all constructs of interest. Four fragments of the MCP1 gene, each of 3569 bp length and containing the promoter, the first exon and the first intron, were PCR-amplified with primers 5'- caccaagaggagcttttcca-3' and 5'-gcgcacgcgtcctctgcactgagatcttcct-3'. The MCP1 -2581A/G and -362C/G variants as well as the deletion int1del554-567/W were examined. Only haplotype combinations occurring with frequencies >1% were subjected to the reporter gene assay. The following combinations were included: MCP1 -2581A/-362G/W; -2581G/-362C/int1del554-567; -2581A/-362C/int1del554-567; -2581A/-362C/W. The Expand Long Template PCR System (Roche, Mannheim, Germany) was used for PCR-amplification.

PCR conditions were: Initial denaturation (94°C, 2 min), 10 amplification cycles (98°C, 10'; 60°C, 30'; 68°C, 10''), 25 amplification cycles (98°C, 15'; 62°C, 30'; 68°C, 20'') and final elongation (78°C, 7''). SMA1 and MLU1 restriction sites at the 5' and 3'ends, respectively, were engineered on each PCR product. In an intermediate step, the fragments were gel-purified, ligated into a pCR-XL-TOPO plasmid (Invitrogen, Carlsbad, USA) and subsequently transfected into Top10 cells (Invitrogen, Carlsbad, USA) according to the manufacturer's instructions. After overnight incubation, cells were lysated and plasmids were digested with SMA1 and MLU1 restriction enzymes (New England Biolabs, Ipswich, USA). The resulting fragments were then ligated into the PGL2 Cloning Vector and transfected into Top10 cells. The final constructs were isolated using an EndoFree Plasmid Maxi Kit (Qiagen, Hilden, Germany).

The Bio-RAD Gene Pulser Xcell system (Bio-Rad Laboratories Ltd., Hertfordshire, UK) was used for co-transfection of 6 × 10⁶ THP1-cells (German Resource Centre for Biological Material, DSMZ [Deutsche Sammlung für Mikroorganismen und Zellkulturen], Braunschweig, Germany) with 0,5 μg of the phRL-CMV vector and either 0,5 μg of the pGL2-Control vector or 0,5 μg of one of the four plasmid constructs. Four hours after transfection, cells were harvested and luciferase activities were measured using a single tube Junior LB9509 luminometer (Berthold Technologies, Bad Wildbad, Germany) and the Dual-Luciferase Reporter Assay System (Promega, Mannheim, Germany). After a 10 second period of Firefly luminescence measurement, 100 ml 1× Stop & Glo Reagent that is supplied with the Dual-Luciferase Reporter Assay System kit were added and Renilla luminescence was detected in another 10 second measurement period. Ten independent transfections and measurements were performed for each construct.