Tuberculosis (TB), caused by the Mycobacterium tuberculosis complex (MTBC) is an ancient disease and the 13^th cause of death globally (WHO, 2021). Only about 5 -10% of infected individuals progress to active TB disease in their lifetime; identified risk factors that may hasten progression to diseases (Flynn and Chan, 2001) include malnutrition, excessive alcohol consumption, smoking, HIV- coinfection, and diabetes mellitus (DM) (Narasimhan et al., 2013). According to the 2021 global TB report, among the 10 million new infections recorded, 8.2% were coinfected with HIV whereas 16% were comorbid with DM (WHO, 2021).

The gut microbiome, described as the last organ of the human body due to specific biochemical interaction and systemic integration with the host, comprises all microbial community and their genes inhabitant in the gut (Dumas et al., 2018; Kim and Benayoun, 2020). Gut microbes perform immunomodulatory role by sending out signals that promote immune cell maturation, influencing infection risks and disease progression (Maji et al., 2018; Hu et al., 2019). Short chain fatty acids (SCFA) produced by some of the normal flora of the gut (Makki et al., 2018) supports maintenance of intestinal barrier integrity and guard the host against infection and colorectal cancer (Peng et al., 2009; Lewis et al., 2010; Besten et al., 2013). The interaction between the gut microbes and the immune system is extremely homeostatic and carefully regulated such that in a disturbed and dysbiotic state the host immunity is negatively impacted (Honda and Littman, 2016).

The gut microbiota also influences distal organs such as the lungs through the gut-lung axis; early life gut depletion in Faecalibacterium, Lachnospira, and Rothia has been linked to asthma in humans (Barcik et al., 2020). Also, the gut microbiota has been shown to boost respiratory defenses through signaling granulocyte-macrophage colony-stimulating factor (GM-CSF), which accelerates pathogen clearance by alveolar macrophages via extracellular signal-regulated kinase signaling (Brown et al., 2017). Gut dysbiosis due to infections and/or use of antimicrobials damages this gut related host response predisposing the host to illness (Sood et al., 2018). Drug susceptible TB treatment includes narrow acting Isoniazid (H), Pyrazinamide (Z), Ethambutol (E) and broad-spectrum Rifampicin (R), (collectively termed (HZRE)), acts against a wide range of gram positive and negative organisms (Diallo et al., 2021). TB treatment is considered one of the longest antibiotic regimen globally taking a minimum of six months comprising an intensive 2-months (HRZE) phase and a continuous 4-months phase of H and R (CDC, 2016). The interplay between TB and other comorbidities such as DM and HIV infection and their impact of the gut microbiome remains unclear, also the specific effects of multiple antimicrobials in both comorbid and non-comorbid.

In this study, we analyzed the variation in gut microbiome of TB patients with and without comorbidities of either HIV or DM before the start of anti-TB therapy and after the month-2 intensive phase of TB therapy. The taxonomic and functional diversity of gut microbiome of the analyzed groups were compared in addition to changes that could occur after 2 months antibiotic use in TB patients.

The human gut microbiome is critical for the development of the host immune system and tissue homeostasis (Chunxi et al., 2020). Aside from the numerous beneficial roles of the microbiome within an individual, its role in the development of metabolic and infectious diseases in an altered (dysbiotic) state has been documented (Guinane and Cotter, 2013). In this study we explored the gut microbiome dynamics in TB patients with or without HIV or diabetes comorbidities before and after intensive phase anti-TB therapy. We found that the microbiome of the TB patients was enriched with inflammatory related microbial organisms with a general increase in these organisms after at 2 months HRZE intervention.

Several studies have reported a decreased gut microbiome diversity in TB patients (Luo et al., 2017; Hu et al., 2019; Cao et al., 2021; Wang et al., 2022), similar to findings in our study which shows a significantly lower gut microbial diversity in TB patients before HRZE intervention compared to healthy controls (Figures 3A–C). This reduced alpha diversity persisted after 2 months of HRZE intervention especially in the comorbid cases (Figures 4A, B). This could be indicative that in addition to TB, compositional shifts could be occurring in a TB-comorbidity specific manner.

Firmicutes are known to be a predominant phylum in the human gut (Qin et al., 2010; Arumugam et al., 2011; Morton et al., 2015). Likewise in this study, the most abundant phylum in both healthy controls (76.3%) and all TB patients (> 65.8%) was the phylum Firmicutes (Supplementary Figure 1). Several bacteria groups belonging to Firmicutes could utilize carbohydrates through anerobic fiber fermentation to produce short-chain fatty acids (SCFAs), such as acetates, propionate, and butyrate. These exert immunomodulatory effects including anti-inflammatory and regulation of gut’s pH to improve the availability of calcium (TeitelbaumAllan Walker, 2002; Broekaert et al., 2011; Venegas et al., 2019). Consistent with their beneficial role in the human gut, we detected an abundance of Firmicutes driven mostly by the genera Clostridium, Faecalibacterium, Blautia and Agathobacter in the gut of the healthy individuals (Figure 2A and Supplementary Table 2). Additionally, other genera such as Romboutsia, CAG-352, Subdoligranulum, and Dialister (Figure 2A) were also among the top 20 abundant genera in the healthy individuals. Several bacteria groups belonging to the Ruminococcaceae and Lachnospiraceae families (Figure 2A) which includes major human butyrate producing gut bacterial species (Venegas et al., 2019) were conspicuously low in TB patients. However, there was an abundance in other Firmicutes of the genera Streptococcus, Enterococcus, Bacteriodes and Erysipelatoclostridium in all the TB patients (Figures 1A, B).

In the TB-DM cohort, the genus Bacteroides was significantly abundant (Figure 6), comparable to studies which reported a positive association of Bacteroides with diabetic individuals (Sircana et al., 2018; Sharma and Tripathi, 2019; Gurung et al., 2020). The Bacteroides genera are beneficial for glucose metabolism as well as regulating the expression of tight junction genes in the colon to reduce gut permeability (Yoshida et al., 2018; Gurung et al., 2020). In the TB-Only cohort, the Streptococcus and Erysipelatoclostridium (Figure 5) which have been associated with high inflammatory conditions (Zeng et al., 2017; Wang et al., 2020), was abundant while in the TB-HIV in addition to Erysipelatoclostridium, Escherchia-Shigella, Dialister and Corynebacterium were abundant (Figure 7). Even though most of these bacteria are beneficial in maintaining a healthy gut microbiome, their association with inflammatory conditions could pose some health risk when it is overly abundant in the gut (Wang et al., 2020).

Contradictory to other studies in Africa which reported reduced abundance in the Actinobacteriota phylum (Zoetendal et al., 2008; Filippo et al., 2017; Costea et al., 2018), interestingly, in this study, we observed an abundance in the Actinobacteriota phylum (9.5%) in the healthy individuals, making it the second most abundant phylum in healthy individuals in this study (Supplementary Figure 1). The abundance of Actinobacteriota was driven mainly by the Bifidobacterium genera (Figure 1A and Supplementary Table 2) which has been shown to be among the first microbes to colonize the human gut and very beneficial in gut homeostasis and digestion of fiber to produce SCFAs (O’Callaghan and Sinderen, 2016). Contrarily, in all the TB cohorts, Actinobacteria was significantly decreased compared to the controls (Supplementary Figure 1).

Similar to other studies which associated Prevotella with high fiber-carbohydrate diet mostly in rural areas in sub-Saharan (Filippo et al., 2010; Parbie et al., 2021a), we observed an abundance in the genera Prevotella as the dominant group of phylum Bacteriodota in healthy individuals. Among the TB cohorts, Prevotella was more abundant in the TB-DM cohorts (3.4%) (Supplementary Table 2) allowing for some association with studies that recorded enrichment of this genera in metabolic disorders such as type 2 diabetes (Precup and Vodnar, 2019; Ding et al., 2022). Also, in addition to Prevotella, in the TB-DM cohort, the Bacteriodes which has been reported to be positively associated with diabetes (Murphy et al., 2017; Wu et al., 2017), was significantly differentially abundant in the TB-DM cohort (Figure 6).

Irrespective of the overall decrease in microorganism of the Firmicutes, Bacteroidetes and Actinobacteria phyla in all TB cohorts compared to healthy individuals, an enrichment of diverse commensal bacteria of the phylum Proteobacteria containing potential pathogenic bacteria and inflammatory related organisms and other pathobionts was observed. Specifically, the genera Escherichia- Shigella, documented to be abundant in a Ghanaian cohort (Parbie et al., 2021a), as well as other African populations (de Filippo et al., 2017; Mancabelli et al., 2017). Interestingly, in this study, Escherichia- Shigella was not detected as a predominant bacterium in the gut of healthy individuals (Figures 1A, B and Supplementary Figure 1). This may be attributable to difference in population dynamics including socioeconomic status and dietary habits; the participants in the present study were from urban communities in Accra, the capital city of Ghana, compared to participants from “peri-urban” communities in the previous study (Parbie et al., 2021a). Also, a study in Korea reported an abundance of the Proteobacteria phylum in type 2 diabetes individuals (Shin et al., 2015) but no such association was found in this study. We however observed a high abundance of Proteobacteria in TB-HIV individuals (Supplementary Table 2) driven mostly by the Escherichia- Shigella family. This genera have been associated with high inflammatory activities as seen in tuberculosis allowing for proliferation of inflammatory-related microorganisms (Eribo et al., 2020) which may not cause disease itself but rather serve to increase susceptibility to intestinal inflammation (Zeng et al., 2017). Additionally, abundances of Protobacteria have been associated with translocation and inflammation of HIV infected individuals (Parbie et al., 2021b).

Intensive phase anti-TB therapy comprises drugs with specific TB targets (HZE) and a broad-spectrum antibiotics (R) (Jnawali and Ryoo, 2013). Broad spectrum antibiotics such as RIF, which are active against most bacterial pathogens, could be responsible for the antibiotic associated perturbations (Dethlefsen and Relman, 2011).

After HRZE treatment, there was a higher abundance of bacteria groups in all the follow-up groups, especially in the TB-HIV cohort (Figure 4B). There was an overall enrichment in the genera Veillonella, Erysipelatoclostridium, and Fusobacterium at follow-up compared to the baseline (Figure 8), like findings from a study reporting abundance of Erysipelatoclostridium in a United States population after HRZE treatment (Wipperman et al., 2017). The genera Erysipelatoclostridium and Fusobacterium are known to be abundant during inflammatory and immune reducing conditions as in TB patients as well as in HIV infection and diabetes (Neuhaus et al., 2010; Etna et al., 2014). While Erysipelatoclostridium possess the ability to flourish during broad-spectrum antibiotic treatment (Dinh et al., 2014), Fusobacterium produces chemical signals that can induce apoptosis in monocytes and macrophages through activation of free fatty acid receptors (Lee et al., 2018; Gu et al., 2020), which is also observed in gut mucosa of colorectal cancer patients (Kelly et al., 2018). Even though the significance of HRZE induced taxonomic perturbations is not well-understood, the majority of these bacteria have been associated with immune-inducing phenotypes relevant for TB immunity (Wipperman et al., 2017).

Contrary, to previous reports which identified vitamin B12 deficiency as a serious problem in diabetics (Liu et al., 2006; Pflipsen et al., 2009; Kibirige and Mwebaze, 2013) due to a negative relationship with metformin, this was not observed in our study population. We found a significant abundance of methylmalonyl-CoA mutase in the gut of TB-DM patients compared to controls. This enzyme is involved in Vitamin B12 synthesis, and degradation of amino acids, short-chain fatty acids and cholesterol (Blum et al., 2021). Thus, the abundance of this enzyme and MTB’s capacity for de novo biosynthesis of vitamin B12 (Gopinath et al., 2013), resulting in reports of elevated B12 vitamin in TB patients (Gebremicael et al., 2019), could account for our observation. The abundance of methylmalonyl-CoA mutase in only TB-DM group is intriguing and could be the body’s homeostasis activity (Takahashi et al., 1994). Also, a significant abundance of organisms involved in isopropanol biosynthesis in the TB-DM cohort was intriguing, since the metabolite produced is known to inhibit growth of Mycobacterium tuberculosis. However, this abundance was only detected in TB-DM cohort and not in the other TB groups. Additionally, an increase of L-rhamnose mutarotase which is involved in the L-rhamnose metabolism pathway for carbohydrate metabolism (Ryu et al., 2005), and very crucial in the diabetes control (Solis-Herrera et al., 2021), was abundant in the TB-DM cohorts.

We acknowledge some limitations especially in the sample size of our control and TB-HIV cohort. However due to the nature of our study, we believe the best representation of healthy controls would be those individuals that live in the same household as the test subjects to rule out any bias. But this number was fewer than expected and we were unable to use previously published gut microbiome data from Ghana as reference gut microbiome due to difference in population dynamics including socioeconomic status, dietary habits, and geographical location. Nevertheless our study reveals a profound impact of TB on the gut microbiome.

This study describes a profound gut dysbiosis in TB patients with or without diabetes or HIV comorbidities compared to the healthy controls. Several opportunistic pathogens of the bacteria groups including Escherichia-shigella, Streptococcus and Enterococcus were enriched in all TB cohorts, whereas beneficial bacteria groups such as SCFA-producing bacteria were depleted. Notably, there was significant alteration in microbial population in TB-DM individuals driven mainly by Bacteriodes, Blautia and Escherichia-shigella. Further studies focused at understanding the long-term changes in the gut microbiome of comorbid groups during the entire TB treatment, and its potential impact on the treatment outcome will provide further insight on the role of gut microbiome on progression of TB.

The data presented in the study are deposited in the NCBI repository, accession number PRJNA876282.

The studies involving human participants were reviewed and approved by Institutional Review Board of the Noguchi Memorial Institute for Medical Research with Federal wide Assurance number FWA00001824. The patients/participants provided their written informed consent to participate in this study.

PM, AB, IL, and EA contributed to sample collection and possessing. PM, AA-P, IO, SA, and DY-M contributed to the methodology. PM, DN, IL, CG, EA, SA contributed to the investigation. PM, PP, and IO did formal analysis. PM, DN, and CG performed all laboratory experiments. PM prepared the original draft of the manuscript. PM, PP, AB, PA, AA-P, IO, SA, and DN contributed to reviewing and editing manuscript. IO, AA-P, and DY-M supervised the work. DY-M provided recourses and conceptualized the study. All authors contributed to the article and approved the submitted version.