Skip to main content
NCBI home page
The Journal of Infectious Diseases logoLink to The Journal of Infectious Diseases
. 2026 Apr 18;234(3):e634–e642. doi: 10.1093/infdis/jiag218

Human Genetic Variation Associates With Infection by Derived Ugandan M. tuberculosis Lineage

Catherine M Stein 1,2,✉,2, Penelope Benchek 3, Lentlamatse Mantshoyane 4, Timothy Ciesielski 5, Michael L McHenry 6, Himiede Wilson-Sesay 7, Moses Joloba 8, Eddie Wampande 9, Kimberly A Dill Mc-Farland 10, Allison W Roberts 11, Ben Polacco 12, Max Bennett 13, Nevan Krogan 14,15, W Henry Boom 16, Jeffery S Cox 17, Harriet Mayanja-Kizza 18, Thomas R Hawn 19, Scott M Williams 20
PMCID: PMC13600372  PMID: 41999581

Abstract

Background

Several studies have examined host and pathogen genetic influences on tuberculosis (TB) susceptibility separately, but relatively few studied their combined effects. However, host-pathogen interactions or co-evolution may explain the inability to replicate many reported human genetic effects across global populations and provide additional insight into TB risk. In this study, we address such possible interactions by focusing on the outcome of infection with the L4-Uganda M. tuberculosis sub-lineage and human genetic variants as independent variables. This is possible because the L4-Uganda sub-lineage is both restricted to Uganda and nearby locations and is recent there, compared with other more ancestral L4 lineages.

Methods

Our study consisted of 276 culture-confirmed adult TB cases from a long-standing household contact study. We conducted a genome-wide association study, with infection with L4-Uganda versus L4-NonUganda as the outcome.

Results

Multiple loci with results suggestive of association (P < 10−5) also demonstrated convergent relevant evidence for strain specific infection via: evidence of gene expression in relevant cells and lung tissue, signatures of natural selection, eQTL expression, and CRISPR screens for immunity-related genes. We also replicated previously published host-pathogen interaction effects, demonstrating that effects seen for other sub-lineages were also present for L4-Uganda.

Conclusions

These results provide evidence for host-pathogen co-evolution in TB, consistent with our previous work, and indicate these interactions involve genes highly relevant to the host immune response to Mycobacterium infection.

Keywords: host-pathogen interaction, immunogenetics, GWAS, co-evolution, tuberculosis susceptibility

This study examined host genetic determinants of infection with the L4-Uganda tuberculosis lineage using a genome-wide association study. Supportive analyses for most significant effects included bioinformatics and eQTL data. Findings implicate immune response genes in host-pathogen interactions in TB.

Tuberculosis (TB) remains a major public health problem globally and is the leading infectious disease killer globally [1]. Caused by Mycobacterium tuberculosis (Mtb), TB disease ultimately results from a host-pathogen interaction, where susceptibility to disease is influenced by both host and microbial genetic factors [2–5]. The human genetic variants associated with TB often vary by global population. Not only do human genetic associations vary by geography, but Mtb genotypes do as well, and can be organized into 8 major lineages that have distinct geographical distributions and timelines of exposure to human populations [6, 7]. Within these 8 major lineages, there are also recently diverged sub-lineages. As ancient lineages are less likely to be virulent than the modern ones, this diversity of lineages and their historical coexistence with humans has led to the hypothesis that disrupted co-evolution between the host and Mtb complex genes increases virulence [8–13]. Consistent with this hypothesis, several candidate gene and genome-wide studies have been conducted [4, 14–23], demonstrating host-pathogen interaction effects at the human DNA—Mtb lineage level. However, most of these studies have been done in East Asia and the presence of such interaction effects in Africa is not as thoroughly explored. The role of host-pathogen interactions in Africa is critical as it is thought TB originated on the continent [24], thereby providing a unique opportunity to assess how interactions between the two species can affect disease risk in light of ancient versus recently diverged, sub-lineages.

In our previous work [25, 26], we posited that the relatively recent introduction into the Ugandan population of the L4-Uganda sub-lineage could result in co-evolution with host genotypes that affected disease severity. This is a different question than asking whether human genetic variation affects susceptibility to disease caused by a specific Mtb lineage, which the approach is taken by the other above cited studies. Given that the L4-Uganda sub-lineage is recently diverged from the ancestral L4 lineage and is unique to Uganda and surrounding countries, it is possible to identify novel human genetic variants that convey TB risk caused by this newer L4-Uganda lineage compared with older lineages that co-existed with humans in Uganda for much longer. Thus in this work, in contrast to our prior work on severity, we sought to identify human genetic factors significantly associated with pulmonary TB due to infection by the L4-Uganda sub-lineage. In Kampala, Uganda, both L4-Uganda and L4-NonUganda are common, providing a unique natural experiment in which to test the hypothesis that human genetic variants associate with strain specific disease.

METHODS

Subject Ascertainment and Characterization

As described previously [26], this study includes TB cases ascertained through a household contact study in Kampala, Uganda from 2002 to 2012. Culture-confirmed adult TB cases with human genome-wide association (GWAS) data and Mtb lineage data, as described previously [27], were included in these analyses. Because Mtb isolates can only be obtained from active TB cases, this study employed a case-only design, which is a classic design for studying gene-environment interactions [28]. Subjects were grouped into two cohorts, based on which human genotyping chips were accessible at the time of genotyping. We refer to them as Cohort 1 and Cohort 2 (N = 149 and N = 127, respectively). The study protocol was approved by the National HIV/AIDS Research Committee of Makerere University and the institutional review board at University Hospitals Cleveland Medical Center. Final clearance was given by the Uganda National Council for Science and Technology. All participants provided written informed consent. Additional details about the original study protocol are described elsewhere [29]. The two cohorts differed in percentage of HIV-1 positive individuals (Table 1); therefore, HIV status was used as a covariate in all analyses.

Table 1.

Sample Characteristics

Cohort 1 Cohort 2 Total P
N = 149 N = 127
Mean Age (SD) 28.7 (8.1) 28.7(9.8) 28.7 (9.0) .99
# Male (%) 81 (54.4) 73 (57.5) 154 (55.8) .69
# HIV+ (%) 31 (24.4) 15 (10.1) 46 (16.7) .0025
Mean TBscore (SD) 6.2 (2.1) 5.4 (2.2) 5.8 (2.2) .0032
# L4.6/Ugandan (%) 93 (62.4) 75 (59.1) 168 (60.9) .49
# L4 (%) 17 (11.4) 21 (16.5) 38 (13.7)
# L3 (%) 37 (24.8) 31 (24.4) 68 (24.6)
Open in a new tab

Human Genotyping and Lineage Typing

Genotyping methods have been described in detail elsewhere [26, 30]. Briefly, Cohort 1 was genotyped on the Illumina Infinium MegaEX chip, and Cohort 2 used the Illumina HumanOmni5 microarray. Both cohorts were separately imputed to the TOPMED-r2 reference panel. Prior to imputation, only SNPs that had a call rate >0.98, minor allele frequency (MAF) > 0.05, and did not show deviation from Hardy-Weinberg equilibrium (P < 10−6) were retained. The total number of well-imputed (R2 > 0.8) SNPs that overlapped between the two cohorts was 18 468 458 (7 925 870 at MAF > 5%). Principal components were computed using PCAiR [31].

Mtb was isolated from sputum of each of the subjects, and lineages were classified according to lineage-identifying SNPs using real-time PCR and validated with long sequence polymorphism (LSP) PCR as described previously [27]. Lineage was determined using three SNPs that accurately distinguish all lineages found in Uganda, L4.6 Uganda (aka L4-Uganda), L3, and L4-NonUganda lineages [6, 7]. L4-Uganda is a sub-lineage that recently diverged from the parent L4 lineage that is only found in Uganda and countries immediately surrounding it [6, 7, 9, 32].

GWAS Analysis

In our primary analysis, we assessed as outcomes disease due to L4-Uganda versus all other L4 lineages; this analysis contrasts the recently evolved sub-lineage with the “older” sub-lineages of the same phylogenetic family [6, 9]. GWAS analyses were first performed in the two separate cohorts on well-imputed (R2 > 0.8) SNPs with a within cohort minor allele frequency (MAF) > 5% and a within cohort minor allele count (MAC) ≥ 15 using the penalized quasi-likelihood approximation to the generalized linear mixed model in the R, Genesis package. Covariate adjustment included age, sex, HIV status and the first two principal components. A fixed effects meta-analysis was run over the two GWAS utilizing the metafor package in R. Only SNPs that were marginally (P < .05) significant in each individual cohort were considered in the meta-analysis. We considered meta-analysis results attaining P < 10−5 as “suggestive”.

As a sensitivity analysis, we contrasted L4-Uganda with other L4 lineages and the L3 lineage combined. Unlike our primary analysis, this analysis does not strictly capture recently evolved versus older lineages, as this also includes L3, which is present in other global populations and diverged from L4 much longer ago [6, 33]. Additional sensitivity analyses included restricting the GWAS to one subject per household since four households included two TB cases, as well as restricting the analysis to HIV-negative subjects.

Identification of Regions With Signatures of Selection

We examined SNPs with association P < 10−5 for signatures of natural selection using the extended haplotype homozygosity (EHH) approach [34]. Under this model, variants under recent positive selection will demonstrate a large region of surrounding haplotype homozygosity. We leveraged a newer measure based on EHH that accounts for local recombination hotspots: nSL scores (number of segregating sites by length) [35]. We examined whole genome sequence data available in the 1000 Genomes database (phased 30× coverage build 38 data) [36], using Selscan 2.0 to obtain 26 nSL scores for each SNP (one nSL score per 1000 Genome population). Then for each population, we normalized the nSL scores in 100 bins of similar frequency to adjust for the impact of allele frequency on mean EHH. Creating these bins is essential as nSL scores are sensitive to MAF; hence our analyses, as performed, explicitly adjusted for MAF. Finally, we evaluated if any of these normalized |nSL| scores were in the top 1% of scores for the population they were measured in. This represents an empirically derived P-value for selection scores of 0.01, a commonly used threshold [37–39]. SNPs with a top1% |nSL| score were considered to be under recent positive selection in that population.

Bioinformatic Annotation

We further investigated our most significant findings using a number of bioinformatic annotation tools. We prioritized SNPs that attained P < 10−6 and any other suggestive SNPs that showed signatures of selection as identified above (Table 2). First, we used HumanBase (formerly known as GIANT [40]) to identify tissues and cell types where proximal genes within 5000 kb of the most significant SNP were expressed and how these genes were connected in gene expression networks. Second, we used FAVOR [41] to identify whether associated SNPs had epigenetic features, transcription factor binding sites, showed conservation across species according to GERP scores, and potential gene function implications through CADD scores. Third, we queried the most significantly associated genes from each region in the GWAS catalog for evidence of pleiotropy [42].

Table 2.

Evidence for Loci With P < 10−5 and at Least One Other Point of Supportive Evidence

Locus Position
(build38)		 Top Gene(s) P < 10−6 Signature of Selection Expression in Relevant Tissues Pathway Association EQTL (P < 0.05) CRISPR Screen P-value GERP Score Relevant Associated Traits In GWAS Catalog (for Top Genes)
chr 1 156760371
156765602
156776463
156776816		 PRCC, HDGF Yes … yes In network with SLC8B1 RRNAD1 … 4.27 Neutrophil count, leukocyte quantity
chr2 148613271 (none) Yes … N/A … … … … N/A
chr 12 91916625
113783216
113783825		 SLC8B1 … yes … In network with chr 1 genes … … … …
chr 16 840155
840260
840308
840327		 LMF1 Yes … … … NARFL/
CIAO3 		0.026 (NARFL) 6.3 Diabetes
Open in a new tab

Query of eQTLs and CRISPR Screens

In a previous study [43] we identified SNPs that were associated with differential gene expression in response to macrophage stimulation with M. tuberculosis in vitro. We queried the regions associated with Mtb lineage at the P < 10−5 levels for evidence of such eQTLs (Table 2). In addition, a recent genome-wide CRISPR knockout screen in cells [44] was conducted independently, to identify macrophage genes that influence the induction of TNF and iNOS upon infection with Mtb. We also queried genes attaining P < 10−6, a gene identified with a selection signature, and the eQTLs (Table 2) in this dataset.

Examination of Loci Reported in Prior Studies

We first examined association with previously-identified TB candidate genes, as summarized in our previous review [2]. Second, we examined loci that have been implicated in host-pathogen interaction analyses in other studies [14–23]. Because these are hypothesis-driven queries, we did not apply any multiple testing corrections.

RESULTS

The sample of TB cases was split into two independent cohorts, as described previously [26]; the first cohort had 149 TB cases, and the second had 127 cases (Table 1). The cohorts differed in their prevalence of HIV-infection and TB severity based on the Bandim TBscore, so HIV status was retained as a covariate in the GWAS. We did not adjust for disease severity as our previous work in this same dataset has shown there is no association between lineage and TBscore [25, 26]. Four households included two TB cases each, though interestingly, one case pair was discordant for lineage, and the other three were concordant, each for different lineages (Supplementary Table 1).

GWAS Results

In the meta-analysis, ten SNPs associated with disease due to L4-Uganda infection at P < 10−6 and an additional 81 SNPs at P < 10−5 (Figure 1, Supplementary Table 2). Among the regions that attained significance at the P < 10−6 threshold, there were four SNPs on chromosome 1 (approximately 156 Mb) (Figure 2A) with an additional six SNPs in the region with P < 10−5. On chromosome 2 (Figure 2B), there was one SNP with P < 10−6 (approximately 149 Mb) with one flanking SNP with P < 10−5 in this gene-poor region. In addition, there were four SNPs on chromosome 16 (approximately 890 kb) with P < 10−6 with an additional 11 flanking SNPs with P < 10−5 (Figure 2C). Furthermore, one SNP on chromosome 9 had a SNP with P < 10−5 but there were no other associated SNPs in that region at even the suggestive threshold (P < 10−5), so this region was not considered further. SNPs on chromosome 12 barely missed the P < 10−6 threshold, with P = 1.1 × 10−6 and another with P = 1.2 × 10−6 (Figure 2D). To assess consistency across the two cohorts and cross sample replication, we also present results for each cohort separately (Supplementary Table 3). We examined the impact of restricting the analysis to one TB case per household (Supplementary Table 4) and only HIV-uninfected subjects (Supplementary Table 5), and the results are highly concordant.

Figure 1.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Manhattan plot of L4-Uganda versus L4-NonUganda GWAS for both cohorts combined.

Figure 2.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

LocusZoom plots for most significant loci: (A) Chromosome 1; (B) Chromosome 2; (C) Chromosome 16; (D) Chromosome 12.

We examined all of the SNPs yielding association P < 10−5 for selection signatures in each of the 1000 Genome populations independently. Four of the 81 SNPs demonstrated recent evidence of positive selection in at least one of the 26 populations. All these SNPS were located within the same region on chromosome 12 (within an intron of SLC8B1). This region was only under selection in African populations (6 of 7 African populations; Supplementary Table 6). The one African population that showed no evidence of selection in this region was the Mandinka of the Gambian Western Division.

Bioinformatic Annotation

We focused our next analyses using bioinformatic annotation and eQTL queries on the three loci attaining P < 10−6 plus the one locus showing evidence with P < 10−5 and a signature of selection (Table 2). The most significant SNP associations on chromosome 1 mapped to two genes with overlapping coding regions, PRCC and HDGF (Figure 2A). Gene expression profiles in HumanBase revealed that these genes are highly expressed in lung, leukocytes, and lymphocytes, and are expressed at a lower level in monocytes, and neutrophils, (Figure 3A). A network analysis including these two genes (Figure 3B) revealed that these genes are co-expressed with several genes key to the TB immune response, including IL12B, IL12A, and TNF, along with SLC8B1, the latter of which was the gene shown to have signatures of selective pressure in African populations (Supplementary Table 6, Figure 2D). The GWAS catalog revealed that PRCC and HDGF are associated with neutrophil count and leukocyte quantity, and the GERP score indicates potential functional importance of this SNP.

Figure 3.

For image description, please refer to the figure legend and surrounding text.

Open in a new tab

Gene expression profiling from HumanBase for genes on chromosome 1: (A) tissue and cell type expression, and (B) network diagram depicting co-expression with other genes.

Another region, on chromosome 16, showed that the associated SNPs in closest proximity to LMF1 (Figure 2C); the most closely related phenotype in the GWAS catalog for this gene was diabetes (Table 2). The associated gene also had a high GERP score (6.3), indicating functional importance.

eQTL and CRISPR Query

On chromosome 1, there were significant eQTLs for RRNAD1 (aka METTL25B) with three SNPs P < .05 (Supplementary Table 7). This gene appears in our network diagram (Figure 3B). We further examined gene expression profiles in HumanBase, and found this gene was also highly expressed in lung and leukocytes (Figure 3). On chromosome 16, our eQTL query revealed an association with NARFL (aka CIAO3) with nine SNPs P < .05 (Supplementary Table 7). In addition, the CRISPR macrophage screen showed a tentative association with NARFL (P = .0266, data not shown) [44]; two of the four guides that targeted NARFL were enriched in the TNF+ iNOS− population. These results did not attain significance after multiple testing corrections.

Query of Previously Studied TB Candidate Genes and Host-pathogen Interaction Effects

First, we examined whether any candidate genes previously associated with TB susceptibility [2] were associated with infection by L4-Uganda lineage (Supplementary Table 8). While some of these results fell within type I error rate (proportion of SNPs with P < .05 was approximately 5%), there were some candidate genes with well more than 5% of SNPs that reached P < .05. SLC11A1, aka NRAMP1, was associated with infection by L4-Uganda (most significant SNP P = .0055; 19% of SNPs reached P < .05); interestingly, this gene was previously indicated in an interaction effect with L4-Uganda in its association with disease severity as measured by the Bandim TBscore [25] along with susceptibility to infection by Beijing lineage [21]. Similarly, SNPs in TLR2 were associated with infection with L4-Uganda (P = .0043); this gene has previously been associated with infection with Beijing lineage in Vietnam [16]. Another noteworthy association was the vitamin D receptor gene (most significant P = .007).

Second, we queried genes and regions that were previously implicated in similar host-Mtb lineage interaction studies (Supplementary Table 9). While many results approached significance at P < .10, a few loci were notably more significant. Many of these have been previously associated with infection with the Beijing lineage, including CD53 (P = .000003), HLA-B (P = .0006), as well as the L2.2 lineage that is likely related to Beijing, including FSTL5 (P = .0035), CSGAL/NCAT1 (P = .0061), along with one gene associated with an L1 sub-lineage (RIMS3, P = .0077).

Sensitivity Analysis Including L3 Lineage

Lastly, we conducted a sensitivity analysis to evaluate the impact of including L3 lineage within the referent group of the outcome variable. There were too few subjects with L3 lineage for this to be considered a separate contrast (Table 1). Thus, we examined the correlation of P-values from our primary analysis, with those found when L4-Uganda was contrasted with both L4-NonUganda and L3 (Supplementary Figure 1). As shown in the correlation plot, the P-values at the most significant loci detected by the L4-Uganda versus L4-NonUganda GWAS had P-values in the L4-Uganda versus L4-NonUganda plus L3 GWAS that were barely significant at P < .001, orders of magnitude less significant than the initial finding for L4-Uganda versus L4. In other words, these loci would not have been detected at the “suggestive” P < 10−5 thresholds if the L3 lineage was included in the reference category. This analysis reveals that inclusion of the historically-diverged primary lineage of L3 decreases power, and that any co-evolutionary effects may have been missed if the contrast was not between the recently evolved lineage and its parent phylogenetic branch.

DISCUSSION

Our objective in this analysis was to further examine the paradigm of host-pathogen interaction in TB susceptibility, by examining the human genome for variants associated with disease due to the recently evolved L4-Uganda lineage. Under the co-evolutionary model that we have previously proposed [25, 26], SNPs associated with L4-Uganda versus the older L4-NonUganda would provide evidence of disruptive co-evolution. Our genome-wide analysis revealed five loci attaining suggestive signals. Two of these are in loci that are eQTLs for genes exhibiting differential expression in response to M. tuberculosis stimulation in vitro, and a third shows signatures of recent natural selection. In addition, we replicated previously reported genes involved in host-pathogen interaction effects as well as TB susceptibility genes. Together, these results demonstrate that the effect of immune response genes on TB susceptibility depends on the context of the infecting Mtb lineage. A unique strength of this work is the availability of data on eQTLs that are specific to the in vitro response to M. tuberculosis stimulation, as well as the availability of results from a genome-wide CRISPR screen of macrophage response to Mtb, to evaluate the biologic plausibility of our findings. Another unique strength of this work is the ability to contrast a recently diverged lineage from its ancestral lineage, pointing to potential co-evolutionary effects, consistent with our previous work [25, 26]. Overall, our results provide further support for the concept that the effect of host genes on TB susceptibility may depend on the genetics of the pathogen and the historical concordance between human alleles and lineage. This work demonstrates that these interaction effects are most prominent in immune response genes.

Results for chromosome 1 showed several associated SNPs—two of these map to genes PRCC and HDGF, and these SNPs are eQTLs for the nearby gene RRNAD1. Interestingly, these genes are all expressed in lung and immune cells and are also connected in a co-expression network in dendritic cells along with other genes that are part of the immune response to TB. Both PRCC and HDGF have been associated with neutrophil count in prior GWAS studies; neutrophils and dendritic cells play key roles in the immune response to TB [45, 46]. Thus, our results support the hypothesis that genes in this region have a role in disease susceptibility most likely acting through host-pathogen interactions. There is extensive linkage disequilibrium in this region, and network analysis indicates these genes are co-expressed, so it is difficult to discern which gene(s) are causal, as most significant the SNP is an eQTL for RRNAD1, it is a strong candidate.

Chromosome 16 also includes some promising candidate genes. NARFL was weakly associated in the CRISPR screen, indicating a role in regulating early innate immune responses during the initial interaction of macrophages and Mtb. These results, though tentative, indicate that decreased NARFL results in more TNF production, so NARFL may inhibit TNF. NARFL (aka CIAO3) codes for cytosolic iron-sulfur assembly component 3, and recent work showed that iron-sulfur cofactors are essential for intracellular adaption of Mtb [47]. The role of bioelements such as iron and sulfur is currently an active area of TB research, thus making this gene another promising candidate for further investigation. On the other hand, the connection between LMF1 and TB susceptibility is less clear.

We note that the selection signal in SLC8B1 meets the commonly used criteria for interestingness in EHH-based measures of recent positive selection (the top 1%, equivalent to P < .01) [37–39], but the evidence supporting this finding is far more diverse and convergent than this one statistical criterion can convey. All 19 selection signals that coincide on four SNPs are within a single locus, SLC8B1, and are only in one superpopulation, which also happens to be the only superpopulation exposed to L4-Uganda. Finally, we note the alleles being selected are the same in every population. This convergent evidence strongly supports the conclusion that this locus is under selection in African populations and the results are not due to chance alone.

We also examined genes that have been previously associated with TB susceptibility as well as those previously implicated in host-pathogen interaction effects on disease risk and severity. Several of these genes appeared to be overrepresented in terms of the number of SNPs that associated with TB in our data, providing evidence for replication. Most notably, we found several immune genes that were associated with L4-Uganda infection, including NRAMP1 (SLC11A1), TLR2, HLA-B, and CD53. As we noted in our prior work [25], NRAMP1 is the most studied TB candidate gene, but association results have been inconsistent across world populations. Our current findings, along with others [21], indicate that the effect of NRAMP1 is dependent on pathogen lineage, which varies globally. While there has been a great emphasis on NRAMP1 in prior literature, this current work reveals that other immune genes are also subject to host-pathogen effects. The idea that immunological responses to TB may be pathogen-dependent [48, 49] is critical for understanding how to best design future vaccine and other therapeutic targets.

A limitation of this study is the sample size. As these results did not attain genome-wide significance, without additional supportive evidence, they may be false positives. This restricted our ability to examine effects of the entire L4 lineage versus L3 and may have resulted in reduced power to replicate other previously reported genetic effects. While our work focused on one recently diverged L4 sub-lineage, it is likely that similar effects might be seen for other Mtb sub-lineages elsewhere. In addition, our bacterial genetic data are limited to the lineage level, so these results cannot be translated to effects due to specific genetic variants in Mtb; this limitation constrains our ability to infer potential bacterial host effects mechanistically. Lastly, while we conducted sensitivity analyses for HIV-infection, we could not account for variation in CD4 counts and/or antiretroviral adherence, since neither of these datapoints were collected in the early days of the HIV epidemic in Uganda.

In conclusion, this work provides further support for host genetic effects that are significant primarily in the context of specific Mtb genetic backgrounds. Both our novel findings, and replication of previously published findings, demonstrate that these host genes tend to have roles in immunity or other relevance to Mtb infection and progression to disease. Furthermore, the results indicate that replication of host genetic associations across global populations may not be expected when such co-evolutionary effects exist. Ancestry explains genetic differences in immunity-related genes [50], and genetic variation in immunity-related genes has been shaped by natural selection. Finally, genetic epidemiological studies need to incorporate cross species genetic effects to inform functional investigation in the pursuit of better treatment outcomes for TB.

Supplementary Material

jiag218_Supplementary_Data
jiag218_supplementary_data.pdf (388.2KB, pdf)

Notes

 Acknowledgments. We want to acknowledge the contributions made by senior physicians, medical officers, health visitors, laboratory and data personnel: Dr. Lorna Nshuti, Dr. Roy Mugerwa, Dr. Alphonse Okwera, Dr. Deo Mulindwa, Dr. Mary Nsereko, Denise Johnson, Dr. Allan Chiunda, Hussein Kisingo, Mark Breda, Dennis Dobbs, Mary Rutaro, Albert Muganda, Richard Bamuhimbisa, Yusuf Mulumba, Deborah Nsamba, Barbara Kyeyune, Faith Kintu, Gladys Mpalanyi, Janet Mukose, Grace Tumusiime, Pierre Peters, Annet Kawuma, Saidah Menya, Joan Nassuna, Keith Chervenak, Karen Morgan, Alfred Etwom, Micheal Angel Mugerwa, Emily Hellwig, and Lisa Kucharski. We would like to acknowledge the medical officers, nurses and counselors at the National Tuberculosis Treatment Center, Mulago Hospital, the Ugandan National Tuberculosis and Leprosy Program and the Uganda Tuberculosis Investigation Bacteriological Unit, Wandegeya, for their contributions to this study.

Author Contributions . Conceptualization: C. M. S., S. M. W.; Data curation: L. M., M. L. M.; Formal analysis: P. B., T. C., M. L. M., H. W. -S., E. W., A. W. R., K. A. D. -M.; Funding acquisition: W. H. B., J. S. C., T. R. H.; Project administration: W. H. B., H. M. -K.; Resources: B. P., M. B., N. K.; Writing—original draft: C. M. S.; Writing—review and editing—all authors.

Data availability statement. Due to Ugandan IRB restrictions, these data cannot be deposited publicly. Requests for data can be made to the chair of the data access committee, Assoc Prof Erisa Sabakaki Mwaka, erisamwaka@gmail.com. Summary statistics will be posted online.

Author approval: All authors have seen and approve of the manuscript.

Financial support. This work was supported by National Institutes of Health grants, including U19AI162583, N01 AI95383, R01HL096811, T32HL007567, and R01AI124348.

Contributor Information

Catherine M Stein, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA; Division of Infectious Diseases, Department of Medicine, Case Western Reserve University, Cleveland, Ohio, USA.

Penelope Benchek, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Lentlamatse Mantshoyane, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Timothy Ciesielski, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Michael L McHenry, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Himiede Wilson-Sesay, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Moses Joloba, School for Biomedical Sciences, College of Health Sciences, Makerere University, Kampala, Uganda.

Eddie Wampande, School for Biomedical Sciences, College of Health Sciences, Makerere University, Kampala, Uganda.

Kimberly A Dill Mc-Farland, Department of Allergy and Infectious Diseases, School of Medicine, University of Washington, Seattle, Washington, USA.

Allison W Roberts, Department of Molecular and Cell Biology, University of California, Berkeley, California, USA.

Ben Polacco, Quantitative Biosciences Institute and Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, California, USA.

Max Bennett, Quantitative Biosciences Institute and Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, California, USA.

Nevan Krogan, Quantitative Biosciences Institute and Department of Bioengineering and Therapeutic Sciences, University of California, San Francisco, California, USA; Gladstone Institute of Data Science and Biotechnology, J. David Gladstone Institutes, San Francisco, California, USA.

W Henry Boom, Division of Infectious Diseases, Department of Medicine, Case Western Reserve University, Cleveland, Ohio, USA.

Jeffery S Cox, Department of Molecular and Cell Biology, University of California, Berkeley, California, USA.

Harriet Mayanja-Kizza, Department of Medicine, School of Medicine, Makerere University and Mulago Hospital, Kampala, Uganda.

Thomas R Hawn, Department of Allergy and Infectious Diseases, School of Medicine, University of Washington, Seattle, Washington, USA.

Scott M Williams, Department of Population and Quantitative Health Sciences, Case Western Reserve University, Cleveland, Ohio, USA.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online (http://jid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.

References

  • 1. World Health Organization . Global tuberculosis report 2024. 2004. Available at: https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2024. Accessed 1 July 2025.
  • 2. Stein  CM, Sausville  L, Wejse  C, et al.  Genomics of human pulmonary tuberculosis: from genes to pathways. Curr Genet Med Rep  2017; 5:149–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Schurz  H, Naranbhai  V, Yates  TA, et al.  Multi-ancestry meta-analysis of host genetic susceptibility to tuberculosis identifies shared genetic architecture. Elife  2024; 13:e84394. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Ndong Sima  CAA, Smith  D, Petersen  DC, Schurz  H, Uren  C, Möller  M. The immunogenetics of tuberculosis (TB) susceptibility. Immunogenetics  2022; 75:215–30. [DOI] [PubMed] [Google Scholar]
  • 5. McHenry  ML, Williams  SM, Stein  CM. Genetics and evolution of tuberculosis pathogenesis: new perspectives and approaches. Infect Genet Evol  2020; 81:104204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Stucki  D, Brites  D, Jeljeli  L, et al.  Mycobacterium tuberculosis lineage 4 comprises globally distributed and geographically restricted sublineages. Nat Genet  2016; 48:1535–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Ngabonziza  JCS, Loiseau  C, Marceau  M, et al.  A sister lineage of the Mycobacterium tuberculosis complex discovered in the African Great Lakes region. Nat Commun  2020; 11:2917. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Gagneux  S, Small  PM. Global phylogeography of Mycobacterium tuberculosis and implications for tuberculosis product development. Lancet Infect Dis  2007; 7:328–37. [DOI] [PubMed] [Google Scholar]
  • 9. Gagneux  S. Ecology and evolution of Mycobacterium tuberculosis. Nat Rev Microbiol  2018; 16:202–13. [DOI] [PubMed] [Google Scholar]
  • 10. Coscolla  M, Gagneux  S. Consequences of genomic diversity in Mycobacterium tuberculosis. Seminars Immunol  2014; 26:431–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Comas  I, Coscolla  M, Luo  T, et al.  Out-of-Africa migration and neolithic coexpansion of Mycobacterium tuberculosis with modern humans. Nat Genet  2013; 45:1176–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Gagneux  S. Host-pathogen coevolution in human tuberculosis. Philosophical transactions of the Royal Society of London. Ser B, Biol Sci  2012; 367:850–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Kodaman  N, Sobota  RS, Mera  R, Schneider  BG, Williams  SM. Disrupted human-pathogen co-evolution: a model for disease. Front Genet  2014; 5:290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Luo  Y, Huang  C-, Howard  NC, et al.  Paired analysis of host and pathogen genomes identifies determinants of human tuberculosis. Nat Commun  2024; 15:10393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Salie  M, van der Merwe  L, Möller  M, et al.  Associations between human leukocyte antigen class I variants and the Mycobacterium tuberculosis subtypes causing disease. J Infect Dis  2014; 209:216–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Caws  M, Thwaites  G, Dunstan  S, et al.  The influence of host and bacterial genotype on the development of disseminated disease with Mycobacterium tuberculosis. PLoS Pathog  2008; 4:e1000034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Thuong  NTT, Tram  TTB, Dinh  TD, et al.  MARCO variants are associated with phagocytosis, pulmonary tuberculosis susceptibility and Beijing lineage. Genes Immun  2016; 17:419–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Intemann  CD, Thye  T, Niemann  S, et al.  Autophagy gene variant IRGM -261T contributes to protection from tuberculosis caused by Mycobacterium tuberculosis but not by M. africanum strains. PLoS Pathog  2009; 5:e1000577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Thye  T, Niemann  S, Walter  K, et al.  Variant G57E of mannose binding lectin associated with protection against tuberculosis caused by Mycobacterium africanum but not by M. tuberculosis. PLoS One  2011; 6:e20908. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Songane  M, Kleinnijenhuis  J, Alisjahbana  B, et al.  Polymorphisms in autophagy genes and susceptibility to tuberculosis. PLoS One  2012; 7:e41618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. van Crevel  R, Parwati  I, Sahiratmadja  E, et al.  Infection with Mycobacterium tuberculosis Beijing genotype strains is associated with polymorphisms in SLC11A1/NRAMP1 in Indonesian patients with tuberculosis. J Infect Dis  2009; 200:1671–4. [DOI] [PubMed] [Google Scholar]
  • 22. Omae  Y, Toyo-oka  L, Yanai  H, et al.  Pathogen lineage-based genome-wide association study identified CD53 as susceptible locus in tuberculosis. J Hum Genet  2017; 62:1015–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Toyo-oka  L, Mahasirimongkol  S, Yanai  H, et al.  Strain-based HLA association analysis identified HLA-DRB1*09:01 associated with modern strain tuberculosis. HLA  2017; 90:149–56. [DOI] [PubMed] [Google Scholar]
  • 24. Barberis  I, Bragazzi  NL, Galluzzo  L, Martini  M. The history of tuberculosis: from the first historical records to the isolation of Koch's bacillus. J Prev Med Hyg  2017; 58:E9–E12. [PMC free article] [PubMed] [Google Scholar]
  • 25. McHenry  ML, Bartlett  J, Igo  RP, et al.  Interaction between host genes and Mycobacterium tuberculosis lineage can affect tuberculosis severity: evidence for coevolution?  PLoS Genet  2020; 16:e1008728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. McHenry  ML, Wampande  EM, Joloba  ML, et al.  Interaction between M. tuberculosis lineage and human genetic variants reveals novel pathway associations with severity of TB. Pathogens  2021; 10:1487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Wampande  EM, Hatzios  SK, Achan  B, et al.  A single-nucleotide-polymorphism real-time PCR assay for genotyping of Mycobacterium tuberculosis complex in peri-urban Kampala. BMC Infect Dis  2015; 15:396. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Gauderman  WJ. Sample size requirements for association studies of gene-gene interaction. Am. J. Epidemiol  2002; 155:478–84. [DOI] [PubMed] [Google Scholar]
  • 29. Stein  CM, Zalwango  S, Malone  LL, et al.  Resistance and susceptibility to Mycobacterium tuberculosis infection and disease in tuberculosis households in Kampala, Uganda. Am J Epidemiol  2018; 187:1477–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. McHenry  ML, Simmons  J, Hong  H, et al.  Tuberculosis severity associates with variants and eQTLs related to vascular biology and infection-induced inflammation. PLoS Genet  2023; 19:e1010387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Conomos  MP, Miller  MB, Thornton  TA. Robust inference of population structure for ancestry prediction and correction of stratification in the presence of relatedness. Genet Epidemiol  2015; 39:276–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Gagneux  S. Host-pathogen coevolution in human tuberculosis. Philos Trans R Soc Lond B Biol Sci  2012; 367:850–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Brites  D, Gagneux  S. The nature and evolution of genomic diversity in the Mycobacterium tuberculosis Complex. Adv Exp Med Biol  2017; 1019:1–26. [DOI] [PubMed] [Google Scholar]
  • 34. Sabeti  PC, Reich  DE, Higgins  JM, et al.  Detecting recent positive selection in the human genome from haplotype structure. Nature  2002; 419:832–7. [DOI] [PubMed] [Google Scholar]
  • 35. Ferrer-Admetlla  A, Liang  M, Korneliussen  T, Nielsen  R. On detecting incomplete soft or hard selective sweeps using haplotype structure. Mol Biol Evol  2014; 31:1275–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Byrska-Bishop  M, Evani  US, Zhao  X, et al.  High-coverage whole-genome sequencing of the expanded 1000 genomes project cohort including 602 trios. Cell  2022; 185:3426–40.e19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Rahman  A, Smith  TQ, Szpiech  ZA. Selection scans and downstream analysis with selscan. BioRxiv 681670. [Preprint]. October 12, 2025. Available at: 10.1101/2025.10.10.681670 [DOI] [PMC free article] [PubMed]
  • 38. Zhang  Y, Xue  X, Liu  Y, et al.  Genome-wide comparative analyses reveal selection signatures underlying adaptation and production in Tibetan and Poll Dorset sheep. Sci Rep  2021; 11:2466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Voight  BF, Kudaravalli  S, Wen  X, Pritchard  JK. A map of recent positive selection in the human genome. PLoS Biol  2006; 4:e72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Greene  CS, Krishnan  A, Wong  AK, et al.  Understanding multicellular function and disease with human tissue-specific networks. Nat Genet  2015; 47:569–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Zhou  H, Arapoglou  T, Li  X, et al.  FAVOR: functional annotation of variants online resource and annotator for variation across the human genome. Nucleic Acids Res  2023; 51:D1300–d1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Cerezo  M, Sollis  E, Ji  Y, et al.  The NHGRI-EBI GWAS catalog: standards for reusability, sustainability and diversity. Nucleic Acids Res  2025; 53:D998–d1005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Hong  H, Dill-McFarland  KA, Simmons  JD, et al.  Mycobacterium tuberculosis-dependent monocyte expression quantitative trait loci, cytokine production, and TB pathogenesis. Front Immunol  2024; 15:1359178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Roberts  AW, Del Cid  LN, Garelis  NE, Cox  JS. Genome-wide screen in Mycobacterium tuberculosis infected macrophages reveals innate regulation of antibacterial mediators by IRF2. BioRxiv 678671v1 [Preprint]. September 29, 2025. Available at: https://www.biorxiv.org/content/10.1101/2025.09.26.678671v1
  • 45. Rahman  F. Characterizing the immune response to Mycobacterium tuberculosis: a comprehensive narrative review and implications in disease relapse. Front Immunol  2024; 15:1437901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Kroon  EE, Coussens  AK, Kinnear  C, et al.  Neutrophils: innate effectors of TB resistance?  Front Immunol  2018; 9:2637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47. Pandey  M, Talwar  S, Bose  S, Pandey  AK. Iron homeostasis in Mycobacterium tuberculosis is essential for persistence. Sci Rep  2018; 8:17359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Wang  C, Peyron  P, Mestre  O, et al.  Innate immune response to Mycobacterium tuberculosis Beijing and other genotypes. PLoS One  2010; 5:e13594. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Tientcheu  LD, Koch  A, Ndengane  M, Andoseh  G, Kampmann  B, Wilkinson  RJ. Immunological consequences of strain variation within the Mycobacterium tuberculosis complex. Eur J Immunol  2017; 47:432–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50. Sanz  J, Randolph  HE, Barreiro  LB. Genetic and evolutionary determinants of human population variation in immune responses. Curr Opin Genet Dev  2018; 53:28–35. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

jiag218_Supplementary_Data
jiag218_supplementary_data.pdf (388.2KB, pdf)

Articles from The Journal of Infectious Diseases are provided here courtesy of Oxford University Press

RESOURCES