Identification of inflammatory protein biomarkers for predicting the different subtype of adult with tuberculosis: an Olink proteomic study

doi:10.21203/rs.3.rs-4876065/v1

Objective

The objective of this study was to identify the potential inflammatory molecular biomarkers that could be utilized for early prediction of different subtype of tuberculosis (TB) in adults.

Methods

Plasma samples were obtained from a cohort of adults diagnosed with 48 cases of active tuberculosis (TB), including drug susceptible TB (DS-TB, n=28), multidrug resistant TB (MDR-TB, n=20), latent TB infection (LTBI, n=20), as well as a control group of healthy individuals without any infection (HC, n=20). The expression level of 92 inflammatory-related proteins was detected by using the high-throughput OLINK proteomics platform.

Results

There were 47 inflammatory proteins showing significant difference (p<0.05) among TB, LTBI and control healthy group, and 7 of them differed significantly between HC and LTBI groups, 46 proteins differed significantly between HC and TB groups, 43 proteins differed significantly between LTBI and TB groups, and overall CXCL10 and TGF-alpha proteins differed significantly among the three groups which could be used as potential diagnostic biomarkers. Furthermore, SCF demonstrates remarkable discriminatory power in distinguishing TB from LTBI, with an area under the curve (AUC) score of 0.920. It was revealed that IL-2RB possesses significant predictive value for MDR-TB, achieving an AUC of 0.709, while CXCL9 (AUC = 0.843) and IFN-alpha (AUC=0.843) show promising diagnostic value in discriminating between active TB and healthy controls. Particularly noteworthy is the emergence of SLAMF1 as the most effective predictor for differentiating between negative and positive tuberculosis cases, with an AUC of 0.779. Additionally, IL6 exhibits a high predictive value for distinguishing between non-severe and severe pulmonary TB, achieving an AUC of 0.92. Correlation analyses revealed both positive and negative relationships among co differentiated proteins, such as a strong positive correlation between TGF-alpha and CXCL10 in LTBI versus HC. Additionally, a strong positive correlation was observed for CXCL10 and CXCL9, as well as TNF and CCL3 in non-severe versus severe pulmonary TB, alongside a negative correlation for IL-6 and SCF. These co-differentiated proteins were found to be enriched in various biological processes and molecular functions related to immune regulation and signaling pathways, such as the p53 signaling pathway, the TNF signaling pathway, and NF-kappa B signaling pathway, highlighting the complex interplay of these proteins in the immune response to TB infection.

Conclusion

Inflammation-related proteins were differentially expressed in adults with TB compared with controls or LTBI. The co-differentiated proteins are intercorrelated, which is involve the pathogenesis of TB via regulation of immune response and immune cell proliferation and apoptosis and phosphorylation. The integration of these proteins offers enhanced diagnostic capabilities for various subtypes of TB in adults.

Biomaker

Inflammation

Tuberculosis

OLINK Proteomics

Tuberculosis, a contagious disease caused by Mycobacterium tuberculosis (M. tuberculosis), continues to present a substantial global health burden, with an estimated annual incidence of 10 million cases, rendering it a prominent contributor to morbidity and mortality on a global scale [1]. The disease manifests in two primary forms: active tuberculosis (ATB), which presents clinical symptoms, and latent tuberculosis infection (LTBI), where individuals have an immune response to the bacteria but show no symptoms. The risk of progressing from LTBI to ATB is notably higher in the first five years post-infection, particularly for children and those with compromised immune systems [2]. Early detection of LTBI is essential for effective prevention and management strategies [3].

Traditional diagnostic methods for tuberculosis, such as sputum smear microscopy, suffer from limited sensitivity and specificity, especially in cases of extrapulmonary TB or low bacillary loads. The gold standard, mycobacterial culture, is time-consuming and requires specialized facilities, making it less accessible in resource-limited settings. Molecular diagnostic tests, while faster, are often expensive and do not always differentiate between latent and active infections, nor do they detect antimicrobial resistance effectively [4]. To address these challenges, there is an urgent need for the development of rapid, cost-effective, and user-friendly diagnostic methods that can be effectively deployed in the field, ensuring quick isolation of infected individuals and monitoring the efficacy of treatments.

Next-Generation Sequencing (NGS) technology offers a swift and cost-effective approach to DNA sequencing, facilitating the identification of drug-resistance mutations in the M. tuberculosis genome and enhancing patient care outcomes [5]. However, the high costs of implementing and maintaining NGS, along with the requirements for robust internet infrastructure and cloud computing, may limit its application in low- and middle-income countries, necessitating the development of more low-cost diagnosis method and affordable diagnostic solutions [6]. Understanding the pathophysiology of TB, especially in its early stages, it is challenging to diagnose or differentiate from TB and LTBI. Standardized TB management is key to understanding its mechanisms, identifying biomarkers, and optimizing prevention, diagnosis, and treatment. This is vital given the growing epidemic of drug-resistant TB [7].

Our hypothesis suggests that immune and inflammation-related biomarkers may play a pivotal role in the progression of TB. To test this, we used either healthy individuals or LTBI patients as control groups to compare with TB cases. Utilizing the OLINK technology, an antibody-based proximity extension assay platform, we aimed to screen for circulating inflammatory biomarkers associated with TB in adults. This approach is intended to provide a more effective means of diagnosing TB at its earliest stages, thereby improving patient outcomes and the strategic handling of this significant global health challenge.

Patient recruitment

Between January 2022 and May 2023, adult individuals suspected of having TB were recruited in a prospective manner at Beijing Chest Hospital (Beijing, China). Peripheral blood specimens were collected from these individuals. The inclusion criteria for the health control (HC) group were as follows: (1) no clinical signs of TB, (2) no abnormalities on X-ray or plain chest radiographs, (3) IGRA test was negative. The inclusion criteria for latent tuberculosis infection (LTBI) were: (1) no clinical signs of TB, (2) no abnormalities on X-ray or plain chest radiographs, (3) IGRA test was positive. The inclusion criteria for active tuberculosis (TB) were: (1) With bacteriological evidence by any of culture or Xpert from sputum; (2) Administered anti-TB drug for ≤ 3 days in the past 6 months.

Patients diagnosed with active tuberculosis are stratified into distinct subgroups according to the presence of Multidrug-resistant tuberculosis (MDR-TB) and the severity of the disease. These subgroups include individuals with MDR-TB and those with Drug-susceptible TB (DS-TB), as well as those with non-severe pulmonary TB and severe pulmonary TB. The multidrug-resistant tuberculosis (MDR-TB) was defined as TB resistant to at least both isoniazid (INH) and rifampicin (RIF). Drug-susceptible TB (DS-TB): A bacteriologically confirmed or clinically diagnosed case of TB, without evidence of infection with strains resistant to isoniazid and rifampicin. The Ethics Committee of Beijing Chest Hospital approved this study (Ethical approval No. YNLX-2022-006). Informed consent was obtained in writing from each participant. The Non-severe pulmonary TB: A form of TB defined as intrathoracic lymph node TB without airway obstruction; uncomplicated TB pleural effusion; or paucibacillary, non-cavitary disease confined to one lobe of the lungs and without a miliary pattern. Severe (or advanced) pulmonary TB disease: The presence of bilateral cavitary disease or extensive parenchymal damage on chest radiography (CXR).

Specimen collection

A blood sample of 200 µl, anticoagulated with heparin, was collected within 12h of the diagnosis of TB or LTBI, along with corresponding age-matched samples from the HC group. In cases where blood transfusion or surgical intervention was deemed necessary, the blood samples were procured prior to the administration of these treatment modalities. The plasma was separated by centrifugation (1000g, 5 min) at 4°C and stored at − 80°C freezer. This study was approved by the Ethics Committee of the Department of Clinical Laboratory, Beijing Chest Hospital, Capital Medical University, Beijing Tuberculosis and Thoracic Tumor Institute (2018-04). Moreover, written informed consent was obtained from the all the patients participating in this study.

OLINK assay

The protein levels were quantified utilizing the OLINK 92 panel (OLINK Proteomics AB, Uppsala, Sweden) according to the manufacturer's recommended instructions. The Proximity Extension Assay (PEA) technology employed in the OLINK protocol, which has been extensively documented [8], facilitates the concurrent analysis of 92 analytes utilizing just 1 µl of plasma. In brief, pairs of oligo-nucleotide-labeled antibody probes bind to targeted proteins, and if the two probes are brought in close proximity the oligonucleotides will hybridize in a pair-wise manner. The addition of a DNA polymerase leads to a proximity-dependent DNA polymerization event, generating a unique PCR target sequence. The resulting DNA sequence was subsequently detected and quantified using a microfluidic real-time PCR instrument (Signature Q100, OLINK). The resulting Ct-data is then quality controlled and normalized using a set of internal and external controls. The final assay results were presented in Normalized Protein eXpression (NPX) values, serving as an arbitrary unit on a log2-scale where increased values correlate positively with heightened protein expression levels. The flow chart of the study and the characteristics of proteins detected by OLINK in ATB, LTBI and HC groups (Fig. 1).

Statistics analysis

The mean ± standard deviation was used to present normally distributed data, while the median (interquartile range, IQR) was used for non-parametric data. Qualitative data were summarized using percentages. Demographic data, clinical characteristics, and inflammation protein concentrations were compared among groups using appropriate statistical tests such as Chi-square, Fisher exact, Mann–Whitney U, and Kruskal–Wallis tests. All inflammation protein concentrations were logarithm (2)-transformed for analysis. Univariate analysis was used to examine the correlation between levels of inflammation proteins and the risk of tuberculosis, comparing them with both healthy control and latent tuberculosis groups. ROC were used to determine the predictive accuracy of the nomogram model. Discrimination capacities were assessed by calculating the area under the receiver-operating characteristic curve (AUC), and the optimal cutoff value was determined based on the Youden index. Statistical analysis was conducted using R statistical software and SPSS 23.0 software (IBM, Armonk, USA). A significance level of p < 0.05 was considered statistically significant. In figures, ****P < 0.0001,***P < 0.001, **P < 0.01, and *P < 0.05.

Population characteristics

A total of 88 cases were included in this study, including 20 cases in the healthy group (HC), 20 cases with latent TB (LTBI), 28 cases with drug-sensitive TB (DS-TB), and 20 cases with drug-resistant TB (MDR-TB), 20 cases with severe pulmonary TB, 28 cases with non-severe pulmonary TB. The average age was 44.21±16.722, and there was no statistically significant difference among the groups (F=2.232, p=0.090). There were 49 males and 41 females, showing no statistically significant difference among the groups (H=5.680, p=0.123). There were no significant differences in the basic characteristics between the DS-TB and MDR-TB groups, except for elevated levels of D-Dimer (DD) and serum creatinine (CREA) in the DS-TB group relative to the MDR-TB group (Table 1).

Table 1 Population Demographic data and baseline characteristics

Characteristics	HC (n=20)	LTBI (n=20)	TB		p value
Characteristics	HC (n=20)	LTBI (n=20)	DS-TB group （n=28）	MDR-TB group（n=20）	p value
Gender，n（%）					0.428
male	11(55.0)	10(45.5)	15（53.6）	13（65.0）
female	9(45.0)	12(54.5)	13（46.4）	7（35.0）
Age（years）	32.00 （26.25，46.75）	44.50 （39.75，52.75）	44.50（29.25，60.75）	38.00（27.75，58.00）	0.530
Classification of TB，n（%）	-	-			0.692
Non-severe pulmonary TB	-	-	17（60.7）	11（55.0）
severe pulmonary TB			11（39.3）	9（45.0）
Smear，n（%）					0.843
negative			12（42.9）	8（40.0）
positive			16（57.1）	12（60.0）
WBC	-	-	7.81±4.27	7.76±3.55	0.968
Hemoglobin	-	-	107.86±24.00	113.45±24.67	0.435
PLT	-	-	298.50±99.95	265.75±119.73	0.308
Lymphocyte	-	-	1.13±0.57	3.37±7.48	0.195
Monocyte	-	-	0.48±0.24	0.70±0.61	0.093
Neutrophil	-	-	6.10±4.32	10.21±18.54	0.262
Albumin	-	-	12.86±8.64	21.15±32.12	0.273
Total bilirubin	-	-	11.19±5.70	10.98±3.98	0.886
CREA	-	-	49.31±14.96	64.08±25.74	0.016
LDH	-	-	159.00 （137.25，219.25）	144.00 （131.50，172.00）	0.143
CRP	-	-	44.18（8.03，73.04）	15.72（0.76，55.07）	0.094
PCT	-	-	0.04（0.02，0.14）	0.03（0.02，0.08）	0.718
NT-proBNP	-	-	619.70 （421.25，1744.75）	339.05 （66.90，1439.20）	0.102
PH	-	-	7.43（7.41，7.46）	7.45（7.44，7.46）	0.242
PaCO₂	-	-	37.00（34.50，41.50）	42.00（35.00，54.25）	0.138
PaO₂	-	-	100.62±24.87	123.86±52.28	0.140
DD	-	-	1.70±1.46	0.95±0.90	0.040
NLR	-	-	4.02（1.84，11.65）	3.78（2.43，8.62）	0.707

WBC white blood cell, N neutrophil, Hb hemoglobin, PLT platelet, CRP C-reactive protein, CREA serum creatinine (CREA).

Note: a represents using independent samples t-test, β represents using Mann-Whitney U test, c represents using chi-square test, d represents using Fisher's exact test.

Plasma Protein Profiling and PCA Analysis: Identifying Differential Expression and Potential Biomarkers in various subtypes of tuberculosis

All the 92 plasma proteins were identified in the three groups, and 7 of them differed significantly between HC and LTBI groups, 43 proteins differed significantly between LTBI and TB groups, 46 proteins differed significantly between TB and HC groups (Figure 2A). Principal component analysis (PCA) analysis revealed clear distinctions between LTBI and TB (Figure2B), as well as MDR-TB and DS-TB (Figure 2C). CXCL10, SCF, and TRANCE were the main contributors to variability in Dim1 (36.0%), while PDL-1 and CXCL11 were most important in Dim2 (29.1%). TRANCE and TWEAK were the key factors in the variability among MDR-TB in Dim1 (45.5%). Overall CXCL10 and TGF-alpha proteins differed significantly among the three groups which could be used as potential diagnostic biomarkers (Figure 2D).

Comparative Proteomic Analysis: Differential Expression and Diagnostic Markers in various subtypes of tuberculosis

In the comparison between the active TB group and the HC group, 46 differentially expressed proteins were identified. Compared to the HC group, IL-6, CXCL9, CXCL10, IFN-gamma, EN-RAGE, and MCP-3 were significantly elevated in active tuberculosis, while TRANCE, TWEAK, SCF, and TRAIL were notably decreased (Figure3).

We found that 7 inflammatory proteins including CCL23, CCL28, CXCL10, NT-3, IL-12B, IL-17A, TGF-alpha can potentially distinguish LTBI from HC groups. All the 7 proteins remained significantly different after statistical adjustments. Nevertheless, a combination of these 7 protein biomarkers showed significantly improved diagnostic value than the individuals.

Similarly, 43 inflammatory proteins can effectively distinguished TB from LTBI groups. Among them, the expression of IL-6, IFN-γ, EN-RAGE were higher in TB, while the expression of SCF (Stem cell factor ), DNER(Delta and Notch-like epidermal growth factor-related receptor (DNER), TRAIL (TNF-related apoptosis-inducing ligand) proteins were lower in LTBI (Figure 3).

The detection of differentially expressed inflammatory proteins among the three groups highlights the complexity of the immune response in TB and the importance of these proteins in disease diagnosis, progression, and treatment. This founding could be instrumental in advancing our understanding of TB and in the development of more effective diagnostic and therapeutic strategies. Further research is needed to fully understand the roles of these proteins in TB pathogenesis and to explore their potential clinical applications.

Identification and Diagnostic Potential of Inflammatory Proteins in MDR-TB, DS-TB, LTBI

In the comparative analysis between the Multi-Drug Resistant Tuberculosis (MDR-TB) and Drug-Sensitive Tuberculosis (DS-TB) groups, six proteins were found to be differentially expressed. Specifically, the levels of IL-2RB, GDNF, CST5, TRANCE, TWEAK, and IL10RA were significantly reduced in the MDR-TB group compared to the DS-TB group (Figure4). Among these, IL-2RB and TRANCE, which are inflammatory proteins, demonstrated moderate predictive power in distinguishing MDR-TB from DS-TB, with their respective area under the curve (AUC) values being 0.709.

In contrast, when assessing the diagnostic value for differentiating between Latent Tuberculosis Infection (LTBI) and Healthy Controls (HC), CCL28 (AUC = 0.677), TGF-alpha (AUC = 0.682), and NT-3 (AUC = 0.665) exhibited low predictive capabilities. On the other hand, CXCL9 (AUC = 0.843), IFN-alpha (AUC = 0.843), and EN-RAGE (AUC = 0.837) showed good diagnostic value when differentiating between Active Tuberculosis (ATB) and HC. These proteins indicate a stronger inflammatory response and immune activation in ATB patients compared to healthy individuals.

Furthermore, SCF (AUC = 0.921), IFN-alpha (AUC = 0.902), and EN-RAGE (AUC = 0.882) displayed superior diagnostic value for distinguishing between LTBI and ATB. This highlights the potential of these proteins as biomarkers to identify the transition from a latent to an active state of the disease (Figure5).

However, the differential expression of specific inflammatory-related proteins holds promise as diagnostic biomarkers for various stages and severities of tuberculosis. The study underscores the significance of cytokine and chemokine dysregulation in the progression of the disease, as indicated by the AUC values in Figure 4. These findings could contribute to a better understanding of TB pathogenesis and aid in the development of more accurate diagnostic tools.

Identification and Diagnostic Potential of Inflammatory Proteins in negative tuberculosis and positive pulmonary TB

In this study, we conducted an analysis of inflammatory protein expression in individuals with negative and positive tuberculosis, identifying 6 inflammatory proteins that may serve as potential markers for distinguishing between the two groups. Of these proteins, SLAMF1, SCF, LIF, NRTN, EN-RAGE, and FGF-19 exhibited statistically significant differences (Figure 6). Furthermore, SLAMF1(AUC=0.779) and MMP (0.712) showed a significant diagnostic value for negative TB. Nevertheless, a combination of these 2 protein biomarkers showed significantly improved diagnostic value than the individuals (Figure 7).

Identification and Diagnostic Potential of Inflammatory Proteins in severe pulmonary tuberculosis

In our study, we have identified 43 inflammatory proteins that exhibit significant potential for distinguishing between different severities of pulmonary tuberculosis. Specifically, we observed a decrease in the expression of SCF (Stem Cell Factor), MCP-4 (Monocyte Chemoattractant Protein-4), and TRANCE (Tumor Necrosis Factor-Related Activation-Induced Cytokine) compared to non-severe cases. Conversely, the expression levels of the remaining 40 inflammatory proteins were found to be increased in severe cases (Figure 8). This observation is indicative of a possible role these proteins play in the progression of the disease.

Furthermore, IL6, EN-RAGE, CXCL10, and PD-L1 demonstrated a notable diagnostic efficacy, the efficacy was determined by their high area under the curve (AUC) values, which exceeded 0.800, a threshold often used to indicate strong predictive performance in diagnostic tests. This suggests that the differential expression of specific inflammatory proteins can be a significant indicator of disease severity in pulmonary tuberculosis. The identification of these proteins, particularly those with high diagnostic efficacy, may contribute to the development of more precise diagnostic tools and personalized treatment strategies for patients with severe pulmonary tuberculosis (Table 2) .

Table 2 The diagnostic value of potential protein biomarkers between Non-severe pulmonary TB and severe pulmonary TB

Name	Cutoff1	Sensitivity	Specificity	AUC	95%CI	p_values
IL6	0.521	0.850	0.926	0.933	0.864-1	8.85E-09
EN-RAGE	0.382	0.950	0.778	0.915	0.822-1	7.99E-08
CXCL10	0.506	0.850	0.889	0.907	0.817-0.998	3.37E-08
PD-L1	0.552	0.800	0.926	0.904	0.818-0.989	2.77E-07
OSM	0.474	0.800	0.852	0.891	0.798-0.983	4.31E-07
IL-18R1	0.361	0.950	0.742	0.889	0.797-0.981	6.87E-07
CXCL11	0.381	0.950	0.778	0.881	0.782-0.981	1.15E-06
CXCL9	0.328	0.950	0.741	0.881	0.782-0.981	1.16E-06
CCL20	0.493	0.750	0.926	0.880	0.779-0.98	3.18E-06
IL18	0.468	0.900	0.815	0.865	0.76-0.97	8.51E-06
CSF-1	0.462	0.800	0.852	0.857	0.748-0.967	1.92E-05
SCF	0.314	0.900	0.741	0.852	0.74-0.964	2.70E-06

Correlation Analysis of Differentially Expressed Inflammatory Proteins in various subtypes of tuberculosis.

To further investigation the correlation of inflammatory proteins within each group, a clustered heat map was generated by computing the Pearson correlation coefficients among pairwise differentially expressed proteins. The X and Y axes of the graph denote the names of these differentially expressed proteins, with red denoting positive correlation and blue denoting negative correlation. The intensity of the color reflects the strength of the correlation.

The co-differential proteins in individuals with TB and HC exhibit a robust positive correlation between SIRT2 and STAMBP, and a negative correlation between SCF and IL-6. Conversely, the co-differential proteins in individuals with LTBI and HC demonstrate a strong positive correlation between TGF-alpha and CXCL10, with a correlation coefficient of 0.638. Furthermore, in individuals with LTBI and TB, a positive correlation is observed between CXCL9 and CXCL10, as well as STAMBP and CASP. The co-differential proteins in non-severe and severe tuberculosis exhibit a strong positive correlation for CXCL10 and CXCL9 (correlation coefficient = 0.920) and TNF and CCL3 (correlation coefficient = 0.924), as well as a negative correlation for IL-6 and SCF (correlation coefficient = -0.735). Conversely, the co-differential proteins in negative and positive tuberculosis demonstrate a significant negative correlation for SCF and FGF 19 (correlation coefficient = -0.6184). Additionally, the co-differential proteins in DS-TB and MDR-TB show a significant positive correlation for TRANCE and TWEAK (correlation coefficient = 0.6273). These findings indicate that inflammation-related proteins are intercorrelated (Figure 9).

GO and KEGG Analysis of Co-Differentially Expressed Proteins in various subtypes of tuberculosis

To confirm the role of co-differentially expressed proteins within three distinct groups, we conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis on these proteins. In the biological process (BP), co-differentially expressed proteins between HC and TB are primarily enriched in processes related to immune regulation, including positive regulation of interleukin-1 beta production, inflammatory response, lymphocyte chemotaxis, T cell proliferation, humoral immune response, macrophage differentiation, peptidyl-tyrosine and peptidyl-serine phosphorylation, MAP kinase activity, chemokine-mediated signaling pathway, and cellular response to lipopolysaccharide. In the molecular function (MF), co-differential expressed proteins are mainly enriched in chemokine activity and CXCR chemokine receptor binding (Figure 10A). Furthermore, analysis of KEGG pathways indicates that these proteins are enriched in pathways such as the p53 signaling pathway, apoptosis Toll-like receptor signaling pathway, and TNF signaling pathway (Figure 10B).

In the context of the biological process, the co-differentially expressed proteins between LTBI and TB are predominantly enriched in functions related to tumor necrosis factor receptor binding, positive regulation of JNK cascade, tumor necrosis factor production, peptidyl-tyrosine and peptidyl-serine phosphorylation, interleukin-17 production, cell proliferation, inflammatory response, interleukin-1 beta production, T cell proliferation, and chemokine production. In the context of molecular function (MF), co-differentially expressed proteins demonstrate significant enrichment in cytokine activity, chemokine activity, and CXCR chemokine receptor binding (Figure 10C). Furthermore, KEGG pathway analysis reveals enrichment of co-differentially expressed proteins in pathways such as the p53 signaling pathway, Tuberculosis, apoptosis, necroptosis, neurodegeneration pathway, TNF signaling pathway, and NF-kappa B signaling pathway (Figure 10D).

In the biological process (BP), co-differential expressed proteins among Non-severe and severe TB are mainly enriched in protein kinase B signaling, positive regulation of interleukin-17 production, positive regulation of interleukin-12 production, cellular response to interferon-gamma, positive regulation of interleukin-1 beta production, positive regulation of chemokine production, positive regulation of NIK/NF-kappaB signaling, positive regulation of peptidyl-tyrosine and peptidyl-serine phosphorylation(figure 10E). The KEGG pathway analysis revealed that co-differentially expressed proteins exhibit enrichment in several pathways, including the PPAR signaling pathway, IL-17 signaling pathway, primary immunodeficiency, endocrine resistance, and NF-kappa B signaling pathway (Figure 10F).

The outcomes of the GO and KEGG analyses of co-differentiated proteins across different groups highlight the significant involvement of immune regulation and cell signaling in the development of tuberculosis. Specifically, the dysregulation of cytokine and chemokine production and signaling emerges as a crucial factor in distinguishing between various stages and severities of the disease. These findings contribute to the advancement of knowledge regarding the pathophysiology of tuberculosis and have the potential to facilitate the development of targeted therapeutic interventions and diagnostic strategies.

In this study, we utilized high-throughput OLINK proteomics analysis technology to screen and identify for highly specific and sensitive biomarkers for different subtypes of TB. Utilizing minimal sample requirements, we conducted a concurrent examination of inflammatory proteins, resulting in the identification of 47 such proteins that displayed significant expression among cases of ATB, LTBI, and control healthy cases.

Among these proteins, CXCL10 and TGF-alpha show potential as biomarkers for distinguishing between TB, LTBI and control healthy group. In addition, the co-differentiated proteins were found to be enriched in various biological processes and molecular functions related to immune regulation and signaling pathways, such as the p53 signaling pathway, the TNF signaling pathway, and the NF-kappa B signaling pathway. The above results suggests that these proteins may play a critical role in the development and progression of TB. Furthermore, previous studies have highlighted the essential function of innate immune cells, especially dendritic cells and macrophages, in the recognition of Mycobacterium tuberculosis and the initiation of adaptive immune responses. A balanced inflammatory response is crucial for effective control of the infection. An imbalance, marked by either excessive or insufficient cytokine production, can lead to uncontrolled bacterial growth or cause harmful immunopathology [9].

The regulatory protein neurotrophin-3 (NT-3), which is co-differentially expressed in HC and LTBI, plays a pivotal role in the development and survival of nervous system neurons, encompassing critical neurological functions like neuronal survival, differentiation, and growth [9–10]. TB predominantly impacts the lungs but can also infiltrate the brain and nervous system, causing neurotuberculosis. Although the direct role of NT-3 in TB has not been extensively studied, several studies have shown that neurotrophic factors, including NT-3, can modulate the immune response during infection [11]. They can regulate the activation and function of vital immune cells, like macrophages and T-cells which are pivotal in TB infection control. This implies that NT-3 may contribute to the immune response against TB, either by promoting or inhibiting immune cell function. Additionally, neurotrophic factors like NT-3 have been implicated in the process of nerve regeneration and repair [12]. Neurotuberculosis can cause damage to the nervous system, leading to neurological deficits [13]. NT-3, along with other neurotrophins, may have a role in promoting nerve regeneration and recovery in TB-related neurological complications.

CXCL10, a protein that is differentially expressed among individuals with ATB, LTBI, and HC, shows a notable increase in patients with severe pulmonary tuberculosis. This chemokine, induced by interferon-gamma (IFN-γ), is integral to the immune response against tuberculosis, as it aids in the recruitment of immune cells to the infection site and contributes to granuloma formation—a hallmark of TB pathology [14]. The association of high serum and bronchoalveolar lavage fluid CXCL10 levels with active TB underscores its potential as a diagnostic biomarker [15]. Furthermore, CXCL10's expression levels offer prognostic value, indicating the effectiveness of treatment and monitoring disease progression. The chemokines CXCL10 and CXCL9 also demonstrate promise as markers for discriminating between drug-resistant and drug-sensitive TB, suggesting their enhanced ability to differentiate disease stages [16].

The co-differentiated expression protein IFN-gamma in LTBI and ATB, significant increase in active TB, Prior research has indicated that The CD5 + and CD10 + B cell subpopulations exhibit potential as biomarkers for distinguishing between latent tuberculosis (LTB) and active tuberculosis (ATB). Specifically, LTB is characterized by elevated levels of CD5 + B cells, which contribute to a cytokine-rich microenvironment containing IFN-gamma, IL-10, and IL-4. Conversely, ATB demonstrates an anti-inflammatory response only in the presence of stimulation with mycobacterial proteins or lipids [17]. IFN-gamma might play important roles in the immune response to TB.

Matrix metalloproteinase 1 (MMP1), a significantly co-differentiated protein present in both negative and positive pulmonary TB, serves as an important inflammatory factor. Mycobacterium tuberculosis (Mtb) triggers the secretion of Oncostatin M (OSM) by monocytes through a prostaglandin (PG)-dependent mechanism. OSM is found in human tuberculous granulomas and works in conjunction with TNF-α to induce the secretion of fibroblast MMP-1 and MMP-3. The release of these matrix-degrading and cytokine-processing enzymes by peripheral stromal cells within the granuloma may contribute to continuous tissue destruction and a persistent inflammatory response [18].

TGF-alpha might significant biomarker in LTBI, TB and HC. The previous study discovered that recombinant human Hepatocyte Growth Factor (HGF) and Transforming Growth Factor-alpha (TGF-α) significantly boosted the secretion of IL-8 and GM-CSF from primary human Alveolar Epithelial Cells (AECs), with moderate effects on IL-6 and MCP-1. Influenza infection, through the activation of HGF/c-Met and TGF-α/EGFR pathways, likely augments the innate immune response in AECs, facilitated by prostaglandin E2 produced during the infection and the crosstalk between AECs and lung fibroblasts [19]. TGF-alpha may also play a role in the immunopathogenesis of tuberculosis, modulating both protective and pathological immune responses.

The co-differential expressed protein SCF are capable of distinguishing TB and LTBI. It has been reported that, various other immune cell populations also exhibit expression of the stem cell factor receptor, c-kit, and have the potential to be stimulated by SCF, thereby playing a role in the advancement of fibrotic disorders. Eosinophils, for instance, have been identified as expressing c-kit and, upon activation by SCF, are capable of generating pro-fibrotic cytokines such as TGFβ and FGF, in addition to numerous lipid mediators, proteases, and chemokines [20]. Integration of SCF and Leukotriene D4 (LTD4) signals may contribute to Mast cells (MCs) hyperplasia and hyper-reactivity during airway hyper-response and inflammation [21]. Its differential expression may indicate differences in inflammatory processes or the host’s attempt to control M.tuberculosis. Overall, SCF’s differential expression suggests its role in TB and LTBI pathogenesis and immune response.

TRANCE (Tumor Necrosis Factor-Related Activation-Induced Cytokine) and IL-2RB (Interleukin-2 Receptor Beta) immune-related molecules [22, 23], exhibit significant predictive value for multiple drug-resistant tuberculosis (MDR-TB). Previous research has highlighted their crucial roles in the immune response against pathogens, particularly M.tuberculosis. These molecules are integral to the activation and regulation of immune cells, including T cells and natural killer (NK) cells, which are vital for controlling and eliminating mycobacterial infections.

TRANCE and IL-2RB regulate cytokine production, crucial for immune response, inflammation, and mycobacterial killing [24–25]. Altered expression of these molecules in drug-resistant TB patients predicts MDR-TB risk. They serve as biomarkers for identifying risks and monitoring treatment. Targeting TRANCE and IL-2RB pathways could enhance the immune response, potentially overcoming drug resistance and improving outcomes. In summary, TRANCE and IL-2RB play a significant role in predicting MDR-TB due to their immunological functions and therapeutic potential.

SLAMF1 (Signaling Lymphocytic Activation Molecule Family Member 1), has been identified as a potential biomarker for Negative TB prediction, due to its association with immune system functions and inflammatory response [26]. previous study showed that SLAMF1 expression was reduced in active TB patients [27], suggesting impaired immune response. A high SLAMF1 level may indicate a normal immune response and lower TB risk. Conversely, higher SLAMF1 levels in active TB patients reflect an increased immune response and likelihood of infection. Thus, SLAMF1 can predict both negative and positive TB status.

IL-6, a pro-inflammatory cytokine, is crucial in the immune response to infections [28]. Several studies had revealed elevated IL-6 levels in pulmonary TB patients [29], especially those with severe disease [30]. In severe TB, the infection spreads wider, causing a higher bacterial load and stronger immune response, resulting in significantly elevated IL-6 levels. However, IL-6 is not the sole predictor; clinical symptoms, radiological findings, and other immune markers are also vital for accurate diagnosis and prognosis.

We found that PD-L1(Programmed Death Ligand-1), EN-RAGE (Extracellular Newly Identified RAGE-Binding Protein), and CXCL10 (Chemokine C-X-C Ligand 10) have significant diagnostic value for severe pulmonary TB. PD-L1 is a protein present on the surface of many immune cells [31]. It interacts with its receptor, PD-1, to regulate the immune system and prevent excessive inflammation [32]. In severe pulmonary TB, the bacterial load is high, triggering the expression of PD-L1 on immune cells as an immune evasion mechanism [33]. Elevated levels of PD-L1 have been detected in the sputum and blood samples of patients with severe pulmonary TB, making it a potential biomarker for diagnosing the severity of the disease.

EN-RAGE, a protein binding to RAGE (Receptor for Advanced Glycation End Products), augments inflammation [34–35]. In severe pulmonary TB, EN-RAGE production surges, significantly elevated in severe TB patients’ serum compared to non-severe cases. Its measurement may assist in diagnosing severe pulmonary TB. CXCL10, a chemokine attracting immune cells to infection sites, is elevated in severe pulmonary disease patients’ serum and bronchoalveolar lavage fluid, positively correlating with disease severity [36–37]. Measuring CXCL10 levels aids in identifying severe pulmonary TB patients.

The findings of this study have significant implications, especially in the context of potential advancements in TB diagnosis and treatment. The discovery of IL-2RB, TGF-α, Neurotrophin-3 (NT-3), and other proteins as potential biomarkers has the potential to transform the approach to TB diagnosis. These biomarkers offer the possibility of developing more accurate diagnostic tools capable of distinguishing between active TB, latent TB infection, and control groups with heightened sensitivity and specificity. Furthermore, the predictive value of immune-related molecules like TRANCE and IL-2RB for multi-drug-resistant tuberculosis (MDR-TB) offers a promising avenue for early identification of patients at risk and for the development of strategies to combat drug resistance. This could be crucial in the fight against the rising global threat of MDR-TB. The potential of proteins like MMP1 and SLAMF1 to distinguish between positive and negative could lead to earlier detection of the disease, allowing for interventions before the onset of symptoms. This could significantly impact public health strategies by contributing to the early containment of TB outbreaks. Understanding the role of these proteins in the immune response to TB could open new avenues for targeted therapies that enhance or modulate the immune system's reaction to the disease. Moreover, the study's findings could be instrumental in the development of prognostic models that assess treatment outcomes by monitoring changes in biomarker levels. This could provide clinicians with valuable insights into the efficacy of treatments and the progression of the disease.

This study is subject to several limitations. Firstly, additional validation with a larger sample size is necessary to elucidate the role of differential proteins in predicting tuberculosis across various disease states. Secondly, ongoing monitoring of specific proteins in patients with similar conditions is essential to understand the potential mechanisms of action of these proteins in the disease. Future research endeavors will focus on addressing these limitations and advancing the identification of early diagnostic biomarkers.

In summary, the importance of co-differentially expressed proteins in diagnosing the severity or subtype of TB lies in their correlation with immune responses and inflammatory processes during TB infection. Nevertheless, it is imperative to underscore the need for further research to validate their diagnostic value and establish standardized cut-off thresholds for their clinical application.

C-C motif chemokine 3 (CCL3); Interleukin-2 receptor subunit beta (IL-2RB); Macrophage colony-stimulating factor1(CSF-1); Neurotrophin-3(NT-3); Programmed cell death 1 ligand 1 (PD-L1); Protein S100-A12 (EN-RAGE); Signaling lymphocytic activation molecule (SLAMF1); STAM-binding protein (STAMPB); Stem cell factor(SCF); TNF-beta (TNFB);TNF-related activation-induced cytokine (TRANCE); TNF-related apoptosis-inducing ligand (TRAIL); Transforming growth factor alpha(TGF-alpha); Tumor necrosis factor (Ligand) superfamily, member 12 (TWEAK); Tumor necrosis factor (TNF);

Acknowledgments

Authors acknowledges funding from the State Key Laboratory of Pathogenesis, Prevention, Treatment of Central Asian High Incidence Diseases Fund. We are grateful to Guirong Wang, who contributed to editing of the manuscript.

Author Contribution

Yunlin Song and Xiaobo Lu were responsible for design, Guirong Wang and Xinting Yang collected samples and clinical data. Jiyuan Zhang and Buzukela Abuduaini conducted data analysis and statistics. Buzukela Abuduaini contributed to writing of the original draft, Yunlin Song guided and financial support the study, Yunlin Song, Xiaobo Lu and Guirong Wang edited the manuscript. All the authors reviewed and approved the final version of the manuscript. Buzukela Abuduaini and Xinting Yang contributed equally.

Declaration of interests

The authors have no conflict of interest to declare.

Funding

This study was supported by the State Key Laboratory of Pathogenesis, Prevention, Treatment of Central Asian High Incidence Diseases Fund of China (SKL-HIDCA-2021-JH13) and Tianshan Outstanding Medical and Health High-level Talent Training Project (TSYC202301B013), and the National Natural Science Foundation of China (Grant No:82360381).

Data availability

All data pertaining to this study are available upon justified request through the author in correspondence.

Dheda K, Perumal T, Moultrie H, et al. The intersecting pandemics of tuberculosis and COVID-19: population-level and patient-level impact, clinical presentation, and corrective interventions. Lancet Respir Med. 2022;10(6):603–22.
Moreno Guillén S, Rodríguez-Artalejo FJ, Ruiz-Galiana J, et al. Tuberculosis in Spain: An opinion paper. Rev Esp Quimioter. 2023;36(6):562–83.
Ketchanji Mougang YC, Endale Mangamba LM, Capuano R, et al. On-Field Test of Tuberculosis Diagnosis through Exhaled Breath Analysis with a Gas Sensor Array. Biosens (Basel). 2023;13(5):570.
The Lancet Microbe. The unsustainable anachronism of tuberculosis diagnosis. Lancet Microbe. 2023;4(6):e379.
The CRyPTIC Consortium and the 100,000 Genomes Project, Allix-Beguec C, Arandjelovic I et al. Prediction of susceptibility to first-line tuberculosis drugs by DNA sequencing. N Engl J Med.. 2018; 379: 1403–1415.
Meehan CJ, Goig GA, Kohl TA, et al. Whole genome sequencing of Mycobacterium tuberculosis: current standards and open issues. Nat Rev Microbiol. 2019;17:533–45.
Kontsevaya I, Lange C, Comella-Del-Barrio P, et al. Perspectives for systems biology in the management of tuberculosis. Eur Respir Rev. 2021;30(160):200377.
Blokzijl A, Nong R, Darmanis S, et al. Protein biomarker validation via proximity ligation assays. Biochim Biophys Acta. 2014;1844(5):933–9.
Ravesloot-Chavez MM, Van Dis E, Stanley SA. The Innate Immune Response to Mycobacterium tuberculosis Infection. Annu Rev Immunol. 2021;39:611–37.
Manti S, Brown P, Perez MK, Piedimonte G. The Role of Neurotrophins in Inflammation and Allergy. Vitam Horm. 2017;104:313–41.
Scuri M, Samsell L, Piedimonte G. The role of neurotrophins in inflammation and allergy. Inflamm Allergy Drug Targets. 2010;9(3):173–80.
Zheng Y, Huang C, Yang X, Zhang Z. Altered expression of glycoprotein non–metastatic melanoma protein B in the distal sciatic nerve following injury. Int J Mol Med. 2020;45(6):1909–17.
Shirazinia M, Sheybani F, Naderi H, Haddad M, Hajipour P, Khoroushi F. Chronic meningitis in adults: a comparison between neurotuberculosis and neurobrucellosis. BMC Infect Dis. 2024;24(1):441.
Liu M, Guo S, Hibbert JM, Jain V, Singh N, Wilson NO, et al. CXCL10/IP-10 in infectious diseases pathogenesis and potential therapeutic implications. Cytokine Growth Factor Rev. 2011;22(3):121–30.
Delemarre EM, van Hoorn L, Bossink AWJ, Drylewicz J, Joosten SA, Ottenhoff THM, et al. Serum Biomarker Profile Including CCL1, CXCL10, VEGF, and Adenosine Deaminase Activity Distinguishes Active From Remotely Acquired Latent Tuberculosis. Front Immunol. 2021;12:725447.
Sampath P, Rajamanickam A, Thiruvengadam K, Natarajan AP, Hissar S, Dhanapal M, et al. Plasma chemokines CXCL10 and CXCL9 as potential diagnostic markers of drug-sensitive and drug-resistant tuberculosis. Sci Rep. 2023;13(1):7404.
Taneja V, Kalra P, Goel M, Khilnani GC, Saini V, Prasad G, et al. Impact and prognosis of the expression of IFN-alpha among tuberculosis patients. PLoS ONE. 2020;15(7):e0235488.
O'Kane CM, Elkington PT, Friedland JS. Monocyte-dependent oncostatin M and TNF-alpha synergize to stimulate unopposed matrix metalloproteinase-1/3 secretion from human lung fibroblasts in tuberculosis. Eur J Immunol. 2008;38(5):1321–30.
Ito Y, Correll K, Zemans RL, Leslie CC, Murphy RC, Mason RJ. Influenza induces IL-8 and GM-CSF secretion by human alveolar epithelial cells through HGF/c-Met and TGF-alpha/EGFR signaling. Am J Physiol Lung Cell Mol Physiol. 2015;308(11):L1178–88.
Oliveira SH, Taub DD, Nagel J, Smith R, Hogaboam CM, Berlin A, Lukacs NW. Stem cell factor induces eosinophil activation and degranulation: mediator release and gene array analysis. Blood. 2002;100:4291–7.
Al-Azzam N, Kondeti V, Duah E, Gombedza F, Thodeti CK, Paruchuri S. Modulation of mast cell proliferative and inflammatory responses by leukotriene d4 and stem cell factor signaling interactions. J Cell Physiol. 2015;230(3):595–602.
Sordillo EM, Pearse RN. RANK-Fc: a therapeutic antagonist for RANK-L in myeloma. Cancer. 2003;97(3 Suppl):802–12.
Fernandez IZ, Baxter RM, Garcia-Perez JE, et al. Correction: A novel human IL2RB mutation results in T and NK cell-driven immune dysregulation. J Exp Med. 2019;216(6):1465.
Josien R, Wong BR, Li HL, Steinman RM, Choi Y, TRANCE. a TNF family member, is differentially expressed on T cell subsets and induces cytokine production in dendritic cells. J Immunol. 1999;162(5):2562–8.
Fernandez IZ, Baxter RM, Garcia-Perez JE, et al. A novel human IL2RB mutation results in T and NK cell-driven immune dysregulation [published correction appears in J Exp Med. J Exp Med. 2019;216(6):1255–67.
Karampetsou MP, Comte D, Suárez-Fueyo A, et al. Signaling Lymphocytic Activation Molecule Family Member 1 Engagement Inhibits T Cell-B Cell Interaction and Diminishes Interleukin-6 Production and Plasmablast Differentiation in Systemic Lupus Erythematosus. Arthritis Rheumatol. 2019;71(1):99–108.
Barbero AM, Trotta A, Genoula M, et al. SLAMF1 signaling induces Mycobacterium tuberculosis uptake leading to endolysosomal maturation in human macrophages. J Leukoc Biol. 2021;109(1):257–73.
Rose-John S, Winthrop K, Calabrese L. The role of IL-6 in host defence against infections: immunobiology and clinical implications. Nat Rev Rheumatol. 2017;13(7):399–409.
Correia JW, Freitas MV, Queiroz JA, PereiraPerrin M, Cavadas B. Interleukin-6 blood levels in sensitive and multiresistant tuberculosis. Infection. 2009;37(2):138–41.
Boni FG, Hamdi I, Koundi LM, Shrestha K, Xie J. Cytokine storm in tuberculosis and IL-6 involvement. Infect Genet Evol. 2022;97:105166.
Wei R, Guo L, Wang Q, Miao J, Kwok HF, Lin Y. Targeting PD-L1 Protein: Translation, Modification and Transport. Curr Protein Pept Sci. 2019;20(1):82–91.
Chen R-Y, Zhu Y, Shen Y-Y, Xu Q-Y, Tang H-Y, Cui N-X, Jiang L, Dai X-M, Chen W-Q, Lin Q, Li X-Z. The role of PD-1 signaling in health and immune-related diseases. Front Immunol. 2023;14:1163633.
Pan SW, Shu CC, Huang JR, et al. PD-L1 Expression in Monocytes Correlates with Bacterial Burden and Treatment Outcomes in Active Pulmonary Tuberculosis. Int J Mol Sci. 2022;23(3):1619.
Yue Q, Song Y, Liu Z, Zhang L, Yang L, Li J. Receptor for Advanced Glycation End Products (RAGE): A Pivotal Hub in Immune Diseases. Molecules. 2022;27:4922.
Kalousová M, Kuběna AA, Benáková H, Dusilová-Sulková S, Tesař V, Zima T. EN-RAGE (extracellular newly identified receptor for advanced glycation end-products binding protein) and mortality of long-term hemodialysis patients: A prospective observational cohort study. Clin Biochem. 2012;45(7–8):556–60.
Liu M, Guo S, Hibbert JM, et al. CXCL10/IP-10 in infectious diseases pathogenesis and potential therapeutic implications. Cytokine Growth Factor Rev. 2011;22(3):121–30.
Al-Adwi Y, Atzeni IM, Doornbos-van der Meer B, et al. High serum C-X-C motif chemokine ligand 10 (CXCL10) levels may be associated with new onset interstitial lung disease in patients with systemic sclerosis: evidence from observational, clinical, transcriptomic and in vitro studies. EBioMedicine. 2023;98:104883.

No competing interests reported.

Identification of inflammatory protein biomarkers for predicting the different subtype of adult with tuberculosis: an Olink proteomic study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Patient recruitment

Specimen collection

OLINK assay

Statistics analysis

Results

Discussion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1