Skip to main content
Download PDF

Peripheral blood microbiome signature and Mycobacterium tuberculosis-derived rsRNA as diagnostic biomarkers for tuberculosis in human

Journal of Translational Medicine volume 23, Article number: 204 (2025) Cite this article

Abstract

Background

Tuberculosis (TB) is a major global health issue. Early diagnosis of TB is still a challenge. Studies are seeking non-sputum biomarker-based TB test. Emerging evidence indicates potential significance of blood microbiome signatures for diseases. However, blood microbiome RNA profiles are unknown in TB. We aimed to characterize the blood microbiome of TB patients and identify Mycobacterium tuberculosis (Mtb) genome-derived small RNA molecules to serve as diagnostic biomarkers for TB.

Methods

RNA sequencing data of the blood from TB patients and healthy controls were retrieved from the NCBI-SRA database for analyzing the blood microbiome and identifying rRNA-derived small RNA (rsRNA) of Mtb. Small RNA-seq was performed on plasma exosomes from TB patients and healthy controls. The levels of the candidate Mtb rsRNAs were determined by real-time quantitative reverse transcription PCR (RT-qPCR) on plasma from a separate cohort of 73 TB patients and 62 healthy controls.

Results

The blood microbiome of TB patients consisted of RNA signals from bacteria, fungi, archaea, and viruses, with bacteria accounting for more than 97% of the total. Reduced blood microbial diversity and abundance of 6 Mycobacterium-associated bacterial genera, including Mycobacterium, Priestia, Nocardioides, Agrobacterium, Bradyrhizobium, and Escherichia, were significantly altered in the blood of TB patients. A diagnostic model for TB based on the 6 genera achieved an area under the curve (AUC) of 0.8945. rsRNAs mapped to the Mtb genome were identified from blood and plasma exosomes of TB patients. RT-qPCR results showed that 2 Mtb-derived rsRNAs, 16 S-L1 and 16 S-L2, could be used as diagnostic biomarkers to differentiate TB patients from healthy controls, with a high co-diagnostic efficacy (AUC = 0.7197).

Conclusions

A panel of blood microbiome signatures and Mtb-derived rsRNAs can serve as blood biomarkers for TB diagnosis.

Introduction

Tuberculosis (TB), primarily caused by Mycobacterium tuberculosis (Mtb), remains the leading infectious cause of death worldwide. According to the Global Tuberculosis Report 2024 published by the World Health Organization (WHO), approximately 10.8 million people developed TB in 2023, leading to 1.25 million deaths [1].

To successfully eliminate TB, novel TB diagnostic methods have shifted from culture-based techniques to assays that are more rapid, less laborious, and eliminate the need for costly high-level biosafety laboratories. Xpert MTB/RIF, a rapid molecular diagnostic method currently recommended by WHO, is a polymerase chain reaction (PCR) test for simultaneous detection of Mtb DNA and the genetic mutations associated with rifampin resistance [2]. A recent large-scale meta-analysis indicated that the sensitivity of Xpert MTB/RIF for TB diagnosis in adults from high TB burden and high TB/HIV regions ranged from 62 to 69%, with a specificity of 99% [3]. Another study conducted in children found that the sensitivity of Xpert MTB/RIF ranged from 55.3 to 72.9%, with a specificity of 99% [4].

Sputum has long been the most commonly used sample in both the traditional culturing and emerging molecular diagnostic methods for TB [5]. Nevertheless, sputum remains a barrier for TB testing, since the success of these assays depends on the reliable collection and decontamination of expectorated sputum. Additionally, the complex and viscous matrix of sputum often limits the accuracy of the PCR analysis, as the chemically complex agents present in the sputum can inhibit nucleic acid extraction and amplification [6]. Although sensitivity of Xpert MTB/RIF is optimal for sputum, it decreases for non-sputum samples, ranged between 45.7% and 73.0% [4]. To enhance TB diagnosis, there is a need for non-sputum, culture-free testing methods that utilize biological samples that are easier to collect. In this context, peripheral blood biomarkers have gained considerable clinical interest due to the ease of sample collection, standardization, and lower invasiveness.

Recent studies suggest that the microbiome may play a role in TB infection [7]. The microbiome could influence both the risk of and progression of TB, primarily through its impact on immunity signaling. On the other side, Mtb infection and the prolonged anti-TB treatment would change the structure of microbiome [8, 9]. Although human blood is generally considered sterile, studies have suggested the presence of microbial biomolecules, including nucleic acids, (poly) peptides, and metabolites, in blood over the past decade [10,11,12]. Changes in blood microbiome have been observed in various diseases and clinical conditions [13,14,15]. Researchers have proposed that a dysregulated blood microbiome could serve as a potential biomarker for disease [11]. Although human studies have documented the changes and fluctuations in microbiome in TB using different specimens, including bronchoalveolar lavage fluid (BALF) [16], stool [17], and sputum [8, 18], the circulating microbiome profile in TB remains poorly understood.

We previously reported that non-coding sRNAs from Mtb can enter the bloodstream and may serve as potential diagnostic biomarkers for TB [19, 20]. rRNA-derived small RNA (rsRNA) is a novel class of precisely processed sRNAs in both eukaryotes and prokaryotes, which modulate biological processes by regulating gene expression [21, 22]. rsRNA can be packaged into or onto vesicles and delivered to host cells, thereby exerting its modulatory effects in host-microbe interactions [23]. However, the potential value of circulating bacterial rsRNA for TB prediction has not been reported.

Here, to develop a blood-based laboratory diagnostic method, we characterized the blood microbiome of TB patients by analyzing blood RNA sequencing data, verified the Mtb-derived rsRNA using plasma exosome RNA-seq, and evaluated their diagnostic potential for TB.

Materials and methods

Blood RNA sequencing data of TB patients

Blood RNA sequencing data were retrieved and chosen by strategy shown in Figure S1. SRA through NCBI was searched for datasets released up to Dec 31, 2021, using the term: (((blood) AND RNA) AND tuberculosis) AND homo sapiens Filters: Transcriptome; SRA. The inclusion criteria included the following: (1) the projects which contained subjects of both TB patients and healthy controls; (2) at least 30 samples in each group. Exclusion criteria were: (1) patients with concomitant disease; (2) patients underwent anti-TB therapy.

Following the exclusion criteria, 12 datasets were returned (Table 1). Among them, the datasets SRP092402 and SRP126691, which met all the inclusion and exclusion criteria, were selected as discovery and validation datasets, respectively (Figure S1). All the subjects in SRP092402 were enrolled in South Africa, a high-burden TB country; while SRP126691 includes subjects from the UK, a low-burden TB country.

Table 1 The details of the retrieved RNA sequencing data
Full size table

South Africa is among WHO’s list of 30 high-burden TB countries, which has one of the highest incidence rates of notified TB globally [1], approximately 3 times higher than the global average (134 vs. 427, per 100,000 of the population) [1]. In 2023, people fell ill with TB in the African region account for a quarter of new TB cases worldwide, second only to those from the Southeast Asia region (45%) [1]. China is also among the high-burden TB countries [1]. Therefore, we selected SRP092402 as the discovery dataset.

SRA data were decompressed into FASTQ data using SRA-Tools (https://github.com/ncbi/sra-tools). Quality control and pre-process were conducted by FastQC and Trim Galore [24] software to remove adaptor sequences and bases of lower quality than Q30 (Figure S2). Rigorous quality control and filtering of raw data during the data analysis stage help to improve RNA data processing and enhance the signal-to-noise ratio [25].

Microbiome characterization

Following read filtering, trimming and quality check, the RNA-seq data of TB patients and healthy individuals were analyzed using Kraken2 [26] and Bracken [27] to compute the species-level abundance. R packages were used to combine all samples into one matrix and calculate alpha and beta diversity. Linear discriminant analysis of effect Size (LEfSe) [28] was used to identify the characteristic microbiome from healthy control and TB patient samples. The details of method of microbiome characterization were presented in the Supplementary Materials.

Identification of bacterial rsRNAs

The reference genome of Mtb H37Rv (NC_000962.3) was indexed using Bowtie2 software. The clean reads were aligned with the H37Rv reference genome using Tophat software to obtain the sam file storing the sequence alignment information. The sam files were converted to bam files by SAMtools software and the bam files were indexed. Transcript abundance and depth of coverage of the whole genome were quantified using BEDtools software. Candidate rsRNAs were screened based on Fragments Per Kilobase per Million mapped fragments (FPKM), and P value (P < 0.05).

Based on the depth of site coverage in the ribosomal region, rsRNAs were defined with lengths of 50–600 nt. Nucleotide sequences of rsRNAs were obtained based on defined rsRNA locus. The genus specificity of the sequences was analyzed using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) and a Sankey plot was drawn. rsRNAs with mean reads > 50 were identified as high-abundance rsRNAs, while those with mean reads < 50 were classified as low-abundance rsRNAs.

Patients and healthy controls

TB patients (N = 73) were enrolled at the Infectious Disease Hospital of Heilongjiang Province from 2019 to 2020. Diagnosis of TB was based on clinical manifestations, chest radiography, microscopic detection, and/or Mtb culture. Peripheral blood samples were obtained before anti-TB treatment. Healthy controls (N = 62) without clinical symptoms, with normal results in routine blood tests and chest radiographs, were collected from the physical examination center of the Second Affiliated Hospital of Harbin Medical University from 2019 to 2020.

Basic demographic and clinical data for all subjects were collected during the initial hospital evaluation. All subjects were long-term residents of Heilongjiang Province, China. There were no significant differences in age or gender between the two groups of participants recruited (Table 2). Exclusion criteria for the participants including: (1) people infected with HIV, or had diabetes, lung diseases, or gastrointestinal disorders; (2) people used probiotics, prebiotics, or antibiotics one month before the enrollment.

Table 2 Demographic information of the RT-qPCR cohort participants
Full size table

Written informed consent was obtained from all participants. Ethical permission was approved by the Ethics Committee of Harbin Medical University, with the certificate number (HMUIRB20190014PR1). Venous blood samples collected with EDTA-tubes were placed in a cooler with ice packs until they were transported to the laboratory within 5 h. Samples were spun down by centrifugation for 10 min at 2500 RPM at 4 ℃ to get plasma. Aliquots of 500 µl of plasma in 2 ml microcentrifuge tubes were stored at -80 ℃ until use.

Plasma exosome small RNA-seq

The plasma exosomes of 3 TB patients and 3 healthy controls were obtained by ultracentrifugation. Total RNA of exosome was extracted and sequenced. Details of RNA-seq were provided in the Supplementary Materials.

Plasma RNA extraction and RT-qPCR

Total RNA was extracted from 500 µl of human plasma with TRIzol LS Reagent (Thermo Fisher Scientific, USA). Plasma level of the rsRNAs were determined by reverse transcription-quantitative PCR (RT-qPCR). The cDNA was synthesized from 1 µg of total RNA using 5x All-In-One RT MasterMix kit (abm, Canada). Real-time PCR was performed in the LightCycler® 96 (Roche, USA) with BlasTaq 2x qPCR MasterMix (abm, Canada). Specificity of amplification was assessed by melt curve analysis. All experiments were performed in triplicate. GAPDH served as the reference gene. The relative expression level of rsRNAs in plasma were assessed using the 2 − ΔCt method (ΔCt = target gene Ct - reference gene Ct). See the Supplementary Materials for the method details. Primers were listed in Table S1.

Statistical analysis

IBM SPSS Statistics 26 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis, Origin 2022 (OriginLab Co., Northampton, MA, USA) and GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA) software was used to generate graphs. The Student’s t test and Mann-Whitney U test were used to compare Mtb rsRNA FPKM and relative expression differences between TB patients and healthy controls. P < 0.05 was considered statistically significant.

Machine learning including SVM [29], Random Forest (RF) [30], XGBoost [31], LightGBM [32], and KNN [33] were conducted by R. SHapley Additive exPlanation (SHAP) is a method used to interpret machine learning model predictions. Based on Shapley value theory, it breaks down the prediction result into the impact of each feature, providing both global and local interpretability for the model [34]. The calculation of Shapley values and the creation of SHAP summary plots were also performed using R.

Results

Microbial composition of blood from TB patients and healthy controls

After read trimming, filtering and quality check, microbial abundance analysis revealed the distribution of microorganisms across 4 kingdoms, including bacteria, viruses, fungi, and archaea, in the blood from TB patients (n = 96) and healthy controls (n = 26). Bacteria were the most abundant, accounting for 96.95%, followed by fungi, archaea, and viruses (Figure S3A).

The predominant bacterial phyla Pseudomonadota (synonym Proteobacteria), Bacillota (formerly Firmicutes), Actinomycetota, and Bacteroidota (Figure S4A). These are the previously reported core phyla of blood microbiome [35,36,37,38]. Bacterial genera with higher abundance included Pseudomonas, Priestia, Bacillus, Rhizobium, Salmonella, Pararhizobium, Klebsiella, Xanthomonas, Escherichia, and Staphylococcus (Figure S3B). The microbial composition for fungi and viruses were shown in Figure S3 and Figure S4 and introduced in detail in the Supplementary Materials.

Reduced alpha diversity of bacteria, fungi and virus were observed in TB patients, as indicated by the Shannon, Simpson, Richness, obs indices (Figure S5). No significant changes in beta diversity of all the microbiome were observed between TB patients and healthy controls (Figure S5).

Differential microbiome in the blood of TB patients

Analysis of the differential microbiome in the blood of TB patients and healthy individuals using the LEfSe method revealed a number of differentially abundant microbial taxa.

As shown in the distribution histogram and evolutionary branching diagram, there were bacterial taxa differed between the blood of TB patients and healthy individuals (Fig. 1A, B). At the phylum level, the abundance of the phyla Bacillota, Actinomycetota, Bacteroidota, and Cyanobacteriota were notably increased while Pseudomonadota was notably decreased in TB patients. Alterations of these phyla were also reported in previous studies on the gut microbiome of TB patients [39, 40].

Fig. 1
figure 1

Differentially abundant taxa (bacteria) identified by linear discriminant analysis (LDA) effect size (LEfSe). (A) Significantly differential taxa in terms of relative abundance (LDA score of ≥ 2) between TB and healthy control groups. (B) Significantly different taxa in the cladogram according to a LDA score of ≥ 2 (each circle represents phylogenetic levels from phylum to genus [inside to outside], and each diameter is proportional to the taxon’s abundance). (C) Significantly differential genera in the heatmap according to a LDA score of ≥ 2 between TB and healthy control groups. Genera marked with green asterisks were correlated to Mycobacterium

Full size image

Twenty-five bacterial genera, including Priestia, Paenibacillus, Streptomyces, Mycobacterium, Corynebacterium, were enriched in the blood of TB patients. In contrast, 10 bacterial genera, such as Thermanaerosceptrum, Shigella, Halomonas, Photobacterium, Vibrio were enriched in the blood of healthy individuals (Fig. 1C). In total, 35 differential bacterial genera were found between TB patients and healthy controls.

Alteration of fungal and viral taxa (Figure S6 and S7) were introduced in the Supplementary Materials.

Differential microbiome correlated to Mycobacterium

Microorganisms correlated with the genus Mycobacterium included 69 bacterial genera, 6 fungal genera and 4 viral genera, with a minimum Pearson correlation coefficient of 0.55 (Figure S8A). The SVM algorithm identified 30 bacteria genera from the differential bacteria identified by LEfSe (LDA score ≥ 2) (Figure S8B). The Venn diagram revealed 11 differential bacterial genera correlating with Mycobacterium (Figure S8C), which have been marked on Fig. 1C. No correlated differential fungal genera or viral genera were found.

Diagnostic efficacy of microbiome signatures

The abundance of 11 differential bacterial genera were compared between TB patients and the controls with Student’s t test. Six genera, Priestia, Mycobacterium, Nocardioides, Agrobacterium, Bradyrhizobium, and Escherichia, showed significant differences between the two groups (all P < 0.01) (Fig. 2A).

Fig. 2
figure 2

The diagnostic efficacy of differential bacteria. (A) Comparisons of abundance of candidate bacteria in TB patients (N = 96) and healthy controls (N = 26). (B) Receiver operating characteristics (ROC) curve analysis of the predictive model using the 6 differentially abundant bacteria for discriminating between TB patients and healthy controls. **P < 0.01; ***P < 0.001; ****P < 0.0001

Full size image

In the validation dataset SRP126691 sourced from the UK (a low TB burden region), 4 genera, including Priestia, Mycobacterium, Agrobacterium, and Bradyrhizobium, exhibited significantly higher abundance in the TB patients (Figure S9), suggesting that some altered taxa in TB patients presented in both low-burden and high-burden regions.

The diagnostic efficacy of the microbiome signature comprising the 6 differential genera were evaluated using the receiver operating characteristic (ROC) curve. The area under the curve (AUC) for the combination of the 6 genera was 0.8954 (Fig. 2B). Under a cutoff value of 0.8908, sensitivity was 67.71%, specificity was 100%, positive predictive value (PPV) was 100%, and negative predictive value (NPV) was 39.76% (Table S2).

Identification of rsRNAs

Reads unmapped to the human genome in blood RNA sequencing data were aligned to the reference genome of Mtb H37Rv. The results showed fragments aligned with 16 S and 23 S rDNA regions. No fragments were found in the 5 S rDNA region.

Four peaks in the 16 S rDNA region and 2 peaks in the 23 S rDNA region exhibiting high abundance (mean reads > 50) and loosely distributed along the locus, were designated as high-abundance rsRNAs (Fig. 3A, C) (Table 3). In contrast, low-abundance sRNAs were numerous and closely spaced (Fig. 3B, D).

Fig. 3
figure 3

The depth of site coverage of rsRNAs. (A) rsRNAs in 16 S rDNA region with high abundance. (B) rsRNAs in 16 S rDNA region with low abundance. (C) rsRNAs in 23 S rDNA region with high abundance. (D) rsRNAs in 23 S rDNA region with low abundance. (E) IGV view for 16 S-H1

Full size image
Table 3 rsRNAs with high abundance
Full size table

We merged the peaks in low-abundance to ensure that (1) subsequent PCR amplification covers the low abundance peaks as much as possible, and (2) the nucleotide sequences of merged regions have Mtb species-specificity. As a result, 7 low-abundance rsRNAs (3 from16S rDNA region, 4 from 23 S rDNA region), ranging from 50 to 600 nt in length, were identified (Fig. 3B, D) (Table 4).

Table 4 rsRNAs with low abundance
Full size table

All the identified rsRNAs were visualized with Integrative Genomics Viewer (IGV) (Fig. 3E and Figure S10). The rsRNAs 16 S-H1 and 16 S-H2 were located in the conserved regions of rDNA, while 16 S-H3, 16 S-H4, 16 S-L1, 16 S-L2, and 16 S-L3 spanned both the conserved and variable regions (Fig. 3A, B).

Compared with healthy controls, the levels of high-abundance rsRNAs were lower, while the levels of low-abundance rsRNAs were higher in TB patients (Fig. 3A to D).

Analysis of the genus specificity of rsRNAs

These rsRNAs were analyzed for genus specificity and aligned with the genomes of 6 genera in the microbiome signature by using CLSI guidelines [41]. Briefly, rsRNA sequences with an identity 97% to < 99% were annotated at the genus level. The BLAST results revealed that all low-abundance rsRNAs were exclusively matched with Mycobacterium (100% identity with the Mtb reference genome). High-abundance rsRNAs were annotated to multiple bacterial genera, including Mycobacterium, Escherichia, Nocardioides, Priestia, Agrobacterium, and Bradyrhizobium (Fig. 4).

Fig. 4
figure 4

The bacterial origin of high-abundance rsRNAs and low-abundance rsRNAs. The left side of the Sankey diagram shows high-abundance rsRNAs derived from multiple bacterial genera, the right side shows low-abundance rsRNAs specific to Mycobacterium genus, and the middle represents 6 blood microbiome signatures

Full size image

Differential expression levels of rsRNAs in RNA sequencing data

Based on RNA sequencing analysis, we compared the expression levels of rsRNAs between TB patients and healthy controls. The results in the discovery dataset showed that among the high-abundance rsRNAs, all except for 23 S-H2 exhibited differential expression between the two groups (Figure S11A). Among the low-abundance rsRNAs, all except 23 S-L4 had higher abundance in TB patients (Figure S11B).

The results in the validation dataset (SRP126691) displayed statistically significant differences between the two groups for 16 S-H2, 16 S-H3, 23 S-H1, 16SL1, 16 S-L2, 16 S-L3, and 16 S-L3 (Figure S12).

Small RNA-seq of plasma exosome verified the presence of the rsRNAs

Plasma exosome small RNA-seq revealed the presence of small RNA fragments mainly enriched in the rDNA region of Mtb reference genome (Figure S13A). Comparison of site coverage showed that the abundance of rsRNAs were higher in TB patients than that in healthy controls (Figure S13B, C). Similar to the result in the blood RNA sequencing data analysis, rsRNAs derived from the 16 S rDNA region were also located in either the variable or conserved regions (Figure S13B).

Diagnostic efficacy of plasma rsRNAs in an independent cohort

Three high-abundance rsRNAs and 3 low-abundance rsRNAs were selected to test their relative levels in human plasma from an independent cohort. As shown in the Figure S14, single peaks were observed for all the amplicons of the rsRNAs, representing a pure, single amplicon. The expression levels of 16 S-L1 and 16 S-L2 in the plasma of TB patients were significantly higher than those in the healthy control group (Fig. 5A). For their diagnostic performance, the AUC for 16 S-L1 and 16 S-L2 was 0.6535 and 0.7130, respectively. AUC of the combination of both rsRNAs was 0.7197 (Fig. 5B). Under the cutoff value of 0.4733, sensitivity was 71.01%, specificity was 64.15%, PPV was 70.79%, and NPV was 60.80% (Table S2).

Fig. 5
figure 5

The detection and diagnostic performance of rsRNAs in an independent cohort. (A) Comparisons of the relative expression of rsRNAs in plasma of TB patients (N = 73) and healthy controls (N = 62) determined by RT-qPCR. (B) ROC curve analysis of the predictive model using the 2 differentially plasma rsRNAs for discriminating between TB patients and healthy controls. **P < 0.01; ****P < 0.0001

Full size image

Discussion

Reduced blood microbiome diversity we found here is consistent with the altered microbiota diversity in the gut [39, 42] and lung [16, 43] in TB patients. Reduced microbial diversity has been found in various diseases, but the underlying mechanisms remain unclear. In TB, infection of Mtb triggers the activation of innate and adaptive immunity that helps to clear the pathogen, but also kill or eliminate the normal microbiota in the respiratory tract [44]. This process may lead to the reduction of microbial diversity. On the other side, reduced diversity in TB patients represents for a disrupted microbial ecosystem, which could contribute to worsened immune function, prolonged inflammation, or a reduced capacity to control secondary infections [45, 46]. It has been proposed that increased oxidative stress induce persistently reduced microbiome diversity in inflammatory bowel disease [47]. Thereby we postulated that the oxidative stress elicited by Mtb infection also involved in the alteration of the microbial diversity. In addition, the cytokines may also influence the microbial diversity, as exemplified by the correlation of the microbiome diversity with the levels of interleukins and the frequent exacerbator phenotype in chronic obstructive pulmonary disease (COPD) [48]. Changes in the gut microbiota of TB patients were showed to be associated with alterations in cytokine levels, such as IL-1β and IL-4 [49].

Several studies have confirmed that the blood microbiome can serve as diagnostic markers in various diseases [50]. However, no studies have characterized the blood microbiome in TB patients. In our study, a panel of 6 genera, including Mycobacterium, Escherichia, Nocardioides, Priestia, Agrobacterium, and Bradyrhizobium, could differentiate between TB patients and healthy controls with high diagnostic efficacy. These genera are all belong to the core phylum of human microbiome.

Mycobacterium and Nocardioides belong to the Actinomycetota. Organisms belonging to Mycobacterium are quite diverse with respect to their ability to cause human diseases; some are strict pathogens, such as Mtb and M. bovis, while others are opportunistic pathogens or nonpathogenic colonizers. Nocardioides was firstly known as Nocardia and classified as a new genus of Actinomycetes in 1976 [51]. In human Nocardioides has been identified as part of the conjunctival bacterial community [52]. This genus has gained more recognition in environmental remediation and ecological protection due to its strong capabilities in degrading refractory pollutants including ritalinic acid, atrazine, and polylactic acid [53].

Escherichia, the core genus of gut microbiome, was reported to be altered in the gut of TB patients previously [54]. The most common clinically significant species in the genus is Escherichia coli. Commensal E. coli combat against the colonization and invasion of pathogens through the competition for nutrients, the modification of metabolites, the depletion of oxygen, the induction of host antimicrobial peptides, the maintenance of mucous and cellular immunity, and direct killing of competitors by the production of bacteriocins or Type IV secretion [55]. High proportion of E. coli and increased levels of TNF-α in plasma were discovered in tuberculous meningitis (TBM) patients. E. coli was shown to increase the plasma level of TNF-α and downregulate brain tight junction protein claudin-5 in the murine TBM model [56]. The underlying molecular mechanism is worth to investigate.

Priestia belong to the core phylum of Bacillota, formerly known as Firmicutes. Priestia was proposed as a novel genus segregated from Bacillus in 2020 [57]. The bacteria are widely distributed in the environment, with all other characteristics of the genus Bacillus. Generally, the genera are known to be non-pathogenic and occasionally cause infection, most occurred in immunocompromised hosts [58, 59]. Recently, Priestia sp. was isolated from human faeces. It is capable of degrading mucin, the large glycoproteins that form the main component of mucus and protect epithelial surfaces [60]. Alteration of Priestia was also reported in gut of TB patients recently [54]. The functional significance and mechanism of this genus in TB merit further exploration.

Agrobacterium and Bradyrhizobium belong to the phylum of Pseudomonadota (the formerly Proteobacteria). Members of these genera are environmental bacteria mostly associated with plants. Compared with the foregoing genera, the clinical significance of the two genera remained poorly understood.

Agrobacterium tumefaciens is widely used as a tool for transient transformation in plant bioengineering [61]. Agrobacterium sp. frequently isolated from human beings with various underlying diseases and often enters the bloodstream, causing bacteremia [62, 63]. Bradyrhizobium are common soil-dwelling microorganisms that form symbiotic relationships with leguminous plants, where they fix nitrogen in exchange for carbohydrates [64]. Bradyrhizobium was identified as one of featured genera of circulating plasma microbiome profiles in patients with liver cirrhosis [65]. Bradyrhizobium enterica was recognized as an opportunistic pathogen capable of causing cord colitis syndrome [66].

It is well known that microbiome is influenced by environmental factors, diet, or comorbidities. There were no significant differences in age and gender between the two groups and the participants recruited in the present study did not suffer from common concomitant diseases and had no history of microbiome modulator administration. However, it is unclear whether their diets were similar. Future prospective study with defined standardized diet would help to clarify its influence to the microbiome.

In addition to characterizing the blood microbiome of TB patients and constructing a microbial diagnostic model, we also sought to identify a body fluid diagnostic marker using plasma samples. Previous research identified Mtb rRNA in sputum as an indicator of bacterial replication [67]. Based on the publicly available RNA sequencing data, we discovered differential rsRNAs mapped to the ribosomal region of Mtb in peripheral blood of TB patients. The high-abundance rsRNAs share 100% sequence identity with multiple members among the 6 featured genera. In contrast, the low-abundance rsRNAs were all specific to the genus Mycobacterium. Using the exosomes isolated from human blood plasma, we verified the presence of higher levels of rsRNAs from Mtb in the TB patients. In a separate cohort, we verified the differential plasma levels of the selected rsRNA candidates with high-abundance and low-abundance. We found that two low-abundance rsRNAs (16 S-L1 and 16 S-L2) demonstrated good diagnostic value for TB. When combined, their diagnostic efficacy yielded a sensitivity of 71.01%, specificity of 64.15%, and an AUC of 0.7197 in differentiating TB patients from healthy controls.

With the progress and application of RNA sequencing technology, numerous rsRNAs have been identified in both eukaryotic and prokaryotic cells. Similar to tRNA-derived sRNA (tsRNA), eukaryotic rsRNAs have been found to contribute to biological processes such as cell proliferation, tissue development, metabolism, and inflammation [21, 22]. In bacteria, rsRNAs can function as bacterial growth regulators during the transition from logarithmic to stationary phase [68], or in the expression of drug resistance [69]. Moreover, bacteria-produced rsRNAs can also participate in the host-pathogen interaction by acting as pathogen-associated molecular patterns, mediating host inflammatory responses, and inducing apoptosis. For instance, Staphylococcus aureus 23 S rDNA-derived rsRNA Sa19, recognized by host TLR13, mediates the production of interleukins and interferons, activates the immune response and plays an important role in bacterial infections [70]. Mtb-derived rsRNAs have been shown to activate apoptosis and disrupt the ability of human monocytes to control the growth of Mtb through caspase-8-dependent mechanism [71].

Both symbiotic and pathobiont bacteria produce membrane vesicles (MVs) containing nucleic acids, which can be distributed to distal organs and tissues through the bloodstream [72, 73]. Circulating nucleic acids derived from bacteria have been confirmed as useful diagnostic markers for diseases. For TB, several studies have identified the diagnostic potential of Mtb non-coding RNA in body fluids [19, 74]. Our present work reveals, for the first time, the presence of circulating microbial rsRNAs from altered bacterial taxa and verified their potential as plasma diagnostic markers for TB. rsRNAs have been recognized as pathogen pattern molecules that induce host defense mediated by host-encoded microbial sensors [75, 76] and immune cell subsets and cytokines coordinate immune responses against invaders [77]. Future work on their molecular action to the host pattern recognition receptors (PRRs) and other potential target would help to elucidate their roles in TB susceptibility and disease progress.

The AUC is widely used to measure the accuracy of diagnostic tests, which could evaluate the performance of a binary diagnostic classification method [78]. It is crucial to set a cutoff value with an appropriate sensitivity and specificity, and generate PPV and NPV to describe more clearly the strengths and weaknesses of the diagnostic model. Additional machine learning methods [79] can aid to enhance the selection of diagnostic biomarkers and the generation of AUC. With multiple algorithms we obtained an optimized AUC of 0.90 for microbiome signatures and 0.82 for rsRNAs (Figure S15 A, C). However, their diagnostic effectiveness still needs to be widely validated across different populations.

The sensitivity and specificity of the rsRNA diagnostic model we constructed here were 71.01% and 64.15%, respectively, which are comparable with other non-sputum TB diagnostic methods, primarily including lipoarabinomannan (LAM) detection and interferon-gamma release assay (IGRA). LAM is a component of the mycobacterial cell wall lipopolysaccharide and has high immunogenicity. The sensitivity of LAM detection is 59% and its performance was limited in TB patients with low CD4 cell counts [80]. IGRA detects interferon-γ production by T cells in response to stimulation with Mtb-specific antigens. The latest meta-analysis showed the sensitivity of IGRA at around 90% and specificity at 99.15% [81]. It requires more complex technical and costly experimental conditions. These requirements limit its widespread application in low- and middle-income facilities [82].

The rapid and accurate diagnosis for diseases depends on the development of clinical laboratory science and scientific collaboration [83, 84]. Non-sputum samples significantly reduce the risk of infection for laboratory personnel. Blood sample collection, high-quality RNA extraction, as well as cost-effective microbiome sequencing and RT-qPCR that can be routinely performed in standard diagnostic laboratories provide the basis for rapid integration of findings into clinical workflows. More interpretable diagnostic models for clinicians could be made with machine learning tools to enhance the significance for the scalability and real-world implementation of the results [34].

Microbiome dysbiosis caused by Mtb infection can lead to alterations in metabolism, immunity, and inflammation, which, in turn, can affect the treatment and prognosis of TB [8, 9]. Future work involving prospective studies would enhance the robustness of the association between plasma bacterial RNA and the progression of TB. By identifying specific bacterial signatures associated with the therapy response of TB, clinicians could tailor personalized treatment strategies to target the microbiome and complement existing therapies, potentially improving patient outcomes and reducing side effects.

Conclusions

In summary, we discovered distinct circulating microbial RNA signatures in TB patients and developed a blood microbiome diagnostic model for TB. Based on the homology and abundance of rsRNAs derived from the altered bacterial taxa, 13 candidate Mtb rsRNAs were identified. Using an independent cohort, we further identified 2 plasma rsRNA markers that can be used for TB diagnosis. Our work provides new insights into the blood microbiome and non-sputum, biomarker-based diagnostic approach for TB.

Data availability

The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Abbreviations

TB:

Tuberculosis

Mtb:

Mycobacterium tuberculosis

rsRNA:

rRNA-derived small RNA

AUC:

Area under the curve

ROC:

Receiver operating characteristic

BALF:

Bronchoalveolar lavage fluid

SHAP:

SHapley Additive explanation

LEfSe:

Linear discriminant analysis of effect size

TBM:

Tuberculous meningitis

MVs:

Membrane vesicles

References

  1. World Health Organization-. Global tuberculosis report 2024. 2024.

  2. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al. Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J Med. 2010;363:1005–15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Shapiro AE, Ross JM, Yao M, Schiller I, Kohli M, Dendukuri N, et al. Xpert MTB/RIF and Xpert Ultra assays for screening for pulmonary tuberculosis and rifampicin resistance in adults, irrespective of signs or symptoms. Cochrane Database Syst Rev. 2021;3:Cd013694.

    PubMed  Google Scholar 

  4. Kay AW, González Fernández L, Takwoingi Y, Eisenhut M, Detjen AK, Steingart KR, et al. Xpert MTB/RIF and Xpert MTB/RIF Ultra assays for active tuberculosis and rifampicin resistance in children. Cochrane Database Syst Rev. 2020;8:Cd013359.

    PubMed  Google Scholar 

  5. Kontsevaya I, Cabibbe AM, Cirillo DM, DiNardo AR, Frahm N, Gillespie SH, et al. Update on the diagnosis of tuberculosis. Clin Microbiol Infect. 2024;30:1115–22.

    Article  CAS  PubMed  Google Scholar 

  6. Beavis KG, Lichty MB, Jungkind DL, Giger O. Evaluation of Amplicor PCR for direct detection of Mycobacterium tuberculosis from sputum specimens. J Clin Microbiol. 1995;33:2582–6.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Mori G, Morrison M, Blumenthal A. Microbiome-immune interactions in tuberculosis. PLoS Pathog. 2021;17:e1009377.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Musisi E, Wyness A, Eldirdiri S, Dombay E, Mtafya B, Ntinginya NE, et al. Effect of seven anti-tuberculosis treatment regimens on sputum microbiome: a retrospective analysis of the HIGHRIF study 2 and PanACEA MAMS-TB clinical trials. Lancet Microbe. 2023;4:e913–22.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Naidoo CC, Nyawo GR, Wu BG, Walzl G, Warren RM, Segal LN, et al. The microbiome and tuberculosis: state of the art, potential applications, and defining the clinical research agenda. Lancet Respir Med. 2019;7:892–906.

    Article  PubMed  Google Scholar 

  10. Castillo DJ, Rifkin RF, Cowan DA, Potgieter M. The healthy human blood microbiome: fact or fiction? Front Cell Infect Microbiol. 2019;9:148.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. Giacconi R, Donghia R, Arborea G, Savino MT, Provinciali M, Lattanzio F et al. Plasma bacterial DNA load as a potential biomarker for the early detection of Colorectal Cancer: a case-control study. Microorganisms. 2023; 11.

  12. Allen RM, Zhao S, Ramirez Solano MA, Zhu W, Michell DL, Wang Y, et al. Bioinformatic analysis of endogenous and exogenous small RNAs on lipoproteins. J Extracell Vesicles. 2018;7:1506198.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. Gedgaudas R, Bajaj JS, Skieceviciene J, Varkalaite G, Jurkeviciute G, Gelman S, et al. Circulating microbiome in patients with portal hypertension. Gut Microbes. 2022;14:2029674.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Ormseth MJ, Wu Q, Zhao S, Allen RM, Solus J, Sheng Q, et al. Circulating microbial small RNAs are altered in patients with rheumatoid arthritis. Ann Rheum Dis. 2020;79:1557–64.

    Article  CAS  PubMed  Google Scholar 

  15. Nganou-Makamdop K, Talla A, Sharma AA, Darko S, Ransier A, Laboune F, et al. Translocated microbiome composition determines immunological outcome in treated HIV infection. Cell. 2021;184:3899–e39143816.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Vázquez-Pérez JA, Carrillo CO, Iñiguez-García MA, Romero-Espinoza I, Márquez-García JE, Falcón LI, et al. Alveolar microbiota profile in patients with human pulmonary tuberculosis and interstitial pneumonia. Microb Pathog. 2020;139:103851.

    Article  PubMed  Google Scholar 

  17. Maji A, Misra R, Dhakan DB, Gupta V, Mahato NK, Saxena R, et al. Gut microbiome contributes to impairment of immunity in pulmonary tuberculosis patients by alteration of butyrate and propionate producers. Environ Microbiol. 2018;20:402–19.

    Article  CAS  PubMed  Google Scholar 

  18. Cheung MK, Lam WY, Fung WY, Law PT, Au CH, Nong W, et al. Sputum Microbiota in Tuberculosis as revealed by 16S rRNA pyrosequencing. PLoS ONE. 2013;8:e54574.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Han X, Li T, Fan Y, Wang X, Gu W, Lu W, et al. Screening of 20 Mycobacterium tuberculosis sRNAs in plasma for detection of active pulmonary tuberculosis. Tuberculosis (Edinb). 2021;129:102086.

    Article  CAS  PubMed  Google Scholar 

  20. Gu W, Tu X, Lu W, Yin Y, Meng Q, Wang X, et al. Identification of RNU44 as an endogenous reference gene for Normalizing Cell-Free RNA in tuberculosis. Open Forum Infect Dis. 2022;9:ofac540.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Chu C, Yu L, Wu B, Ma L, Gou LT, He M, et al. A sequence of 28S rRNA-derived small RNAs is enriched in mature sperm and various somatic tissues and possibly associates with inflammation. J Mol Cell Biol. 2017;9:256–9.

    Article  CAS  PubMed  Google Scholar 

  22. Lyons SM, Gudanis D, Coyne SM, Gdaniec Z, Ivanov P. Identification of functional tetramolecular RNA G-quadruplexes derived from transfer RNAs. Nat Commun. 2017;8:1127.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Lambert M, Benmoussa A, Provost P. Small non-coding RNAs derived from eukaryotic ribosomal RNA. Noncoding RNA. 2019; 5.

  24. Bush SJ. Read trimming has minimal effect on bacterial SNP-calling accuracy. Microb Genom. 2020; 6.

  25. Abualigah L, Al-Okbi NK, Awwad EM, Sharaf M, Daoud MS. Boosted Aquila arithmetic optimization algorithm for multi-level thresholding image segmentation. Evol Syst. 2024:1–28.

  26. Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome Biol. 2019;20:257.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Lu J, Breitwieser FP, Thielen P, Salzberg SL. Bracken: estimating species abundance in metagenomics data. PeerJ Comput Sci. 2017;3:e104.

    Article  Google Scholar 

  28. Segata N, Izard J, Waldron L, Gevers D, Miropolsky L, Garrett WS, et al. Metagenomic biomarker discovery and explanation. Genome Biol. 2011;12:R60.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Got A, Zouache D, Moussaoui A, Abualigah L, Alsayat A. Improved Manta Ray foraging optimizer-based SVM for feature selection problems: A Medical Case Study. J Bionic Eng. 2024;21:409–25.

    Article  Google Scholar 

  30. Hu J, Szymczak S. A review on longitudinal data analysis with random forest. Brief Bioinform. 2023; 24.

  31. Li J, Liu S, Hu Y, Zhu L, Mao Y, Liu J. Predicting Mortality in Intensive Care Unit patients with heart failure using an interpretable machine learning model: Retrospective Cohort Study. J Med Internet Res. 2022;24:e38082.

    Article  PubMed  PubMed Central  Google Scholar 

  32. Yan J, Xu Y, Cheng Q, Jiang S, Wang Q, Xiao Y, et al. LightGBM: accelerated genomically designed crop breeding through ensemble learning. Genome Biol. 2021;22:271.

    Article  PubMed  PubMed Central  Google Scholar 

  33. Uddin S, Haque I, Lu H, Moni MA, Gide E. Comparative performance analysis of K-nearest neighbour (KNN) algorithm and its different variants for disease prediction. Sci Rep. 2022;12:6256.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Cakiroglu C, Demir S, Ozdemir MH, Aylak BL, Sariisik G, Abualigah L. Data-driven interpretable ensemble learning methods for the prediction of wind turbine power incorporating SHAP analysis. Expert Syst Appl. 2024;237:121464.

    Article  Google Scholar 

  35. Hidi L, Kovács GI, Szabó D, Makra N, Pénzes K, Juhász J, et al. Human blood vessel microbiota in healthy adults based on common femoral arteries of brain-dead multi-organ donors. Front Cell Infect Microbiol. 2022;12:1056319.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Khan I, Khan I, Jianye Z, Xiaohua Z, Khan M, Hilal MG, et al. Exploring blood microbial communities and their influence on human cardiovascular disease. J Clin Lab Anal. 2022;36:e24354.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Wang Q, Wang Q, Zhao L, Bin Y, Wang L, Wang L, et al. Blood bacterial 16S rRNA gene alterations in women with polycystic ovary syndrome. Front Endocrinol (Lausanne). 2022;13:814520.

    Article  PubMed  Google Scholar 

  38. Qiu J, Zhou H, Jing Y, Dong C. Association between blood microbiome and type 2 diabetes mellitus: a nested case-control study. J Clin Lab Anal. 2019;33:e22842.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Hu Y, Yang Q, Liu B, Dong J, Sun L, Zhu Y, et al. Gut microbiota associated with pulmonary tuberculosis and dysbiosis caused by anti-tuberculosis drugs. J Infect. 2019;78:317–22.

    Article  PubMed  Google Scholar 

  40. Luo M, Liu Y, Wu P, Luo DX, Sun Q, Zheng H, et al. Alternation of gut microbiota in patients with pulmonary tuberculosis. Front Physiol. 2017;8:822.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Petti C, Bosshard P, Brandt M, Clarridge J, Feldblyum T, Foxall P et al. Interpretive criteria for identification of bacteria and fungi by DNA target sequencing; approved guideline. 2008.

  42. Li W, Zhu Y, Liao Q, Wang Z, Wan C. Characterization of gut microbiota in children with pulmonary tuberculosis. BMC Pediatr. 2019;19:445.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Hu Y, Kang Y, Liu X, Cheng M, Dong J, Sun L, et al. Distinct lung microbial community states in patients with pulmonary tuberculosis. Sci China Life Sci. 2020;63:1522–33.

    Article  CAS  PubMed  Google Scholar 

  44. Cai X, Luo Y, Zhang Y, Lin Y, Wu B, Cao Z, et al. Airway microecology in rifampicin-resistant and rifampicin-sensitive pulmonary tuberculosis patients. BMC Microbiol. 2022;22:286.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Han M, Wang X, Zhang J, Su L, Ishaq HM, Li D, et al. Gut bacterial and fungal dysbiosis in tuberculosis patients. BMC Microbiol. 2024;24:141.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Xiao G, Cai Z, Guo Q, Ye T, Tang Y, Guan P, et al. Insights into the Unique Lung Microbiota Profile of Pulmonary Tuberculosis patients using Metagenomic Next-Generation sequencing. Microbiol Spectr. 2022;10:e0190121.

    Article  PubMed  Google Scholar 

  47. Schirmer M, Garner A, Vlamakis H, Xavier RJ. Microbial genes and pathways in inflammatory bowel disease. Nat Rev Microbiol. 2019;17:497–511.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Pragman AA, Hodgson SW, Wu T, Zank A, Reilly CS, Wendt CH. Sputum microbiome α-diversity is a key feature of the COPD frequent exacerbator phenotype. ERJ Open Res. 2024; 10.

  49. Huang SF, Yang YY, Chou KT, Fung CP, Wang FD, Su WJ. Systemic proinflammation after Mycobacterium tuberculosis infection was correlated to the gut microbiome in HIV-uninfected humans. Eur J Clin Invest. 2019;49:e13068.

    Article  PubMed  Google Scholar 

  50. Sciarra F, Franceschini E, Campolo F, Venneri MA. The diagnostic potential of the human blood microbiome: are we dreaming or Awake? Int J Mol Sci. 2023; 24.

  51. Prauser H. Nocardioides, a new genus of the order Actinomycetales. Int J Syst Evol MicroBiol. 1976;26:58–65.

    Google Scholar 

  52. Mohamed YH, Uematsu M, Morinaga Y, Nguyen HT, Toizumi M, Sasaki D et al. Conjunctival Sac Microbiome in Infectious Conjunctivitis. Microorganisms. 2021; 9.

  53. Ma Y, Wang J, Liu Y, Wang X, Zhang B, Zhang W, et al. Nocardioides: specialists for hard-to-degrade pollutants in the Environment. Molecules. 2023;28:7433.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Shi W, Hu Y, Ning Z, Xia F, Wu M, Hu YOO, et al. Alterations of gut microbiota in patients with active pulmonary tuberculosis in China: a pilot study. Int J Infect Dis. 2021;111:313–21.

    Article  CAS  PubMed  Google Scholar 

  55. Foster-Nyarko E, Pallen MJ. The microbial ecology of Escherichia coli in the vertebrate gut. FEMS Microbiol Rev. 2022; 46.

  56. Li S, Guo J, Liu R, Zhang F, Wen S, Liu Y, et al. Predominance of Escherichia-Shigella in Gut Microbiome and its potential correlation with Elevated Level of Plasma Tumor Necrosis Factor Alpha in patients with tuberculous meningitis. Microbiol Spectr. 2022;10:e0192622.

    Article  PubMed  Google Scholar 

  57. Gupta RS, Patel S, Saini N, Chen S. Robust demarcation of 17 distinct Bacillus species clades, proposed as novel Bacillaceae genera, by phylogenomics and comparative genomic analyses: description of Robertmurraya kyonggiensis sp. nov. and proposal for an emended genus Bacillus limiting it only to the members of the subtilis and Cereus clades of species. Int J Syst Evol Microbiol. 2020;70:5753–98.

    Article  CAS  PubMed  Google Scholar 

  58. Guo J, Chen J, Sun X. Bacillus megaterium infection presenting as pulmonary alveolar proteinosis, a case report. BMC Infect Dis. 2024;24:868.

    Article  PubMed  PubMed Central  Google Scholar 

  59. Uçar AD, Ergin ÖY, Avcı M, Arı A, Yıldırım M, Erkan N. Bacillus flexus outbreak in a tertiary burn center. Burns. 2016;42:948–9.

    Article  PubMed  Google Scholar 

  60. Deswal G, Selwal MK, Nirvan H, Selwal KK. Priestia Flexa KS1: a new bacterial strain isolated from human faeces implicated in mucin degradation. Int Microbiol. 2023;26:475–86.

    Article  CAS  PubMed  Google Scholar 

  61. Krenek P, Samajova O, Luptovciak I, Doskocilova A, Komis G, Samaj J. Transient plant transformation mediated by Agrobacterium tumefaciens: principles, methods and applications. Biotechnol Adv. 2015;33:1024–42.

    Article  CAS  PubMed  Google Scholar 

  62. Aujoulat F, Marchandin H, Zorgniotti I, Masnou A, Jumas-Bilak E. Rhizobium pusense is the main human pathogen in the genus Agrobacterium/Rhizobium. Clin Microbiol Infect. 2015;21:e472471–475.

    Article  Google Scholar 

  63. Kaselitz T, Hariadi N, LiPuma J, Weinberg J. Rhizobium radiobacter bacteremia in a neonate. Infection. 2012;40:437–9.

    Article  CAS  PubMed  Google Scholar 

  64. Ormeño-Orrillo E, Martínez-Romero E. A genomotaxonomy view of the Bradyrhizobium Genus. Front Microbiol. 2019;10:1334.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Trebicka J, Macnaughtan J, Schnabl B, Shawcross DL, Bajaj JS. The microbiota in cirrhosis and its role in hepatic decompensation. J Hepatol. 2021;75(Suppl 1):S67–81.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Bhatt AS, Freeman SS, Herrera AF, Pedamallu CS, Gevers D, Duke F, et al. Sequence-based discovery of Bradyrhizobium Enterica in cord colitis syndrome. N Engl J Med. 2013;369:517–28.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  67. Walter ND, Born SEM, Robertson GT, Reichlen M, Dide-Agossou C, Ektnitphong VA, et al. Mycobacterium tuberculosis precursor rRNA as a measure of treatment-shortening activity of drugs and regimens. Nat Commun. 2021;12:2899.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  68. Luidalepp H, Berger S, Joss O, Tenson T, Polacek N. Ribosome shut-down by 16S rRNA fragmentation in stationary-phase Escherichia coli. J Mol Biol. 2016;428:2237–47.

    Article  CAS  PubMed  Google Scholar 

  69. Cafiso V, Stracquadanio S, Lo Verde F, Dovere V, Zega A, Pigola G, et al. COLRAcinetobacter baumannii sRNA signatures: computational comparative identification and biological targets. Front Microbiol. 2020;10:3075.

    Article  PubMed  Google Scholar 

  70. Hidmark A, von Saint Paul A, Dalpke AH. Cutting edge: TLR13 is a receptor for bacterial RNA. J Immunol. 2012;189:2717–21.

    Article  CAS  PubMed  Google Scholar 

  71. Obregón-Henao A, Duque-Correa MA, Rojas M, García LF, Brennan PJ, Ortiz BL, et al. Stable extracellular RNA fragments of Mycobacterium tuberculosis induce early apoptosis in human monocytes via a caspase-8 dependent mechanism. PLoS ONE. 2012;7:e29970.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Toyofuku M, Schild S, Kaparakis-Liaskos M, Eberl L. Composition and functions of bacterial membrane vesicles. Nat Rev Microbiol. 2023;21:415–30.

    Article  CAS  PubMed  Google Scholar 

  73. Díaz-Garrido N, Badia J, Baldomà L. Microbiota-derived extracellular vesicles in interkingdom communication in the gut. J Extracell Vesicles. 2021;10:e12161.

    Article  PubMed  PubMed Central  Google Scholar 

  74. Lu G, Jiang X, Wu A, Zhou J, Liu H, He F, et al. Two small Extracellular Vesicle sRNAs Derived from Mycobacterium tuberculosis serve as diagnostic biomarkers for active pulmonary tuberculosis. Front Microbiol. 2021;12:642559.

    Article  PubMed  PubMed Central  Google Scholar 

  75. Frantz R, Teubner L, Schultze T, La Pietra L, Müller C, Gwozdzinski K et al. The secRNome of Listeria monocytogenes harbors small noncoding RNAs that are potent inducers of Beta Interferon. mBio. 2019; 10.

  76. Aghaei M, Khademi R, Far MAJ, Bahreiny SS, Mahdizade AH, Amirrajab N. Genetic variants of dectin-1 and their antifungal immunity impact in hematologic malignancies: a comprehensive systematic review. Curr Res Transl Med. 2024;72:103460.

    CAS  PubMed  Google Scholar 

  77. Aghapour SA, Torabizadeh M, Bahreiny SS, Saki N, Jalali Far MA, Yousefi-Avarvand A, et al. Investigating the dynamic interplay between cellular immunity and tumor cells in the fight against cancer: an updated comprehensive review. Iran J Blood Cancer. 2024;16:84–101.

    Article  Google Scholar 

  78. Nahm FS. Receiver operating characteristic curve: overview and practical use for clinicians. Korean J Anesthesiol. 2022;75:25–36.

    Article  PubMed  PubMed Central  Google Scholar 

  79. Abualigah L. Metaheuristic Optimization Algorithms: optimizers, analysis, and applications. Elsevier; 2024.

  80. Li Z, Tong X, Liu S, Yue J, Fan H. The value of FujiLAM in the diagnosis of tuberculosis: a systematic review and Meta-analysis. Front Public Health. 2021;9:757133.

    Article  PubMed  PubMed Central  Google Scholar 

  81. Krutikov M, Faust L, Nikolayevskyy V, Hamada Y, Gupta RK, Cirillo D, et al. The diagnostic performance of novel skin-based in-vivo tests for tuberculosis infection compared with purified protein derivative tuberculin skin tests and blood-based in vitro interferon-γ release assays: a systematic review and meta-analysis. Lancet Infect Dis. 2022;22:250–64.

    Article  CAS  PubMed  Google Scholar 

  82. Goletti D, Delogu G, Matteelli A, Migliori GB. The role of IGRA in the diagnosis of tuberculosis infection, differentiating from active tuberculosis, and decision making for initiating treatment or preventive therapy of tuberculosis infection. Int J Infect Dis. 2022;124(Suppl 1):S12–9.

    Article  CAS  PubMed  Google Scholar 

  83. Aghaei M, Khademi R, Bahreiny SS, Saki N. The need to establish and recognize the field of clinical laboratory science (CLS) as an essential field in advancing clinical goals. Health Sci Rep. 2024;7:e70008.

    Article  PubMed  PubMed Central  Google Scholar 

  84. Saki N, Haybar H, Aghaei M, Subject. Motivation can be suppressed, but scientific ability cannot and should not be ignored. J Transl Med. 2023;21:520.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to thank Yuqin Liu, Zhiguo Yue, Na Li and Long Jin from Infectious Disease Hospital of Heilongjiang Province, for their assistance with sample collection. We would also like to thank Yanhong Liu (the second Affiliated Hospital of Harbin Medical University) for the significant support and effort in promoting the study and helping it run smoothly.

Funding

This study was supported by the National Science and Technology Major Project [2017ZX10201301–003–005] and Heilongjiang Provincial Natural Science Foundation [LH2024H014].

Author information

Author notes

  1. Yunfan Fan & Ting Li

    Present address: Department of Clinical Laboratory, Chongqing Public Health Medical Center, Chongqing, China

Authors and Affiliations

  1. Department of Microbiology, School of Basic Medical Sciences, WU Lien-Teh Institute, Harbin Medical University, Harbin, China

    Wei Gu, Zhigang Huang, Yunfan Fan, Ting Li, Xinyuan Yu, Zhiyuan Chen, Yan Hu, Aimei Li, Fengmin Zhang & Yingmei Fu

  2. Heilongjiang Provincial Key Laboratory of Infection and Immunity, Harbin Medical University, Harbin, China

    Aimei Li, Fengmin Zhang & Yingmei Fu

Authors
  1. Wei Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Zhigang Huang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Yunfan Fan
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Ting Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Xinyuan Yu
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Zhiyuan Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Yan Hu
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Aimei Li
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Fengmin Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  10. Yingmei Fu
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Wei Gu, conceptualization, investigation, methodology, formal analysis, data curation, writing original draft, validation, software; Zhigang Huang, investigation, methodology, formal analysis, data curation, validation, software; Yunfan Fan and Ting Li, investigation, validation, software. Xinyuan Yu, Zhiyuan Chen, and Yan Hu, original draft, validation. Aimei Li and Fengmin Zhang, validation, supervision. Yingmei Fu: conceptualization, validation, supervision, project administration, writing - review & editing.

Corresponding author

Correspondence to Yingmei Fu.

Ethics declarations

Ethics approval and consent to participate

Informed consent was obtained from the patients and healthy volunteers to participate in the study. The design of the work was approved by the Ethics Committee of Harbin Medical University, China (certificate number HMUIRB20190014PR1).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gu, W., Huang, Z., Fan, Y. et al. Peripheral blood microbiome signature and Mycobacterium tuberculosis-derived rsRNA as diagnostic biomarkers for tuberculosis in human. J Transl Med 23, 204 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-025-06190-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-025-06190-2

Keywords

Journal of Translational Medicine

ISSN: 1479-5876

Contact us