Whole blood RNA signatures in leprosy patients identify reversal reactions before clinical onset: a prospective, multicenter study - PubMed (original) (raw)

Multicenter Study

. 2019 Nov 29;9(1):17931.

doi: 10.1038/s41598-019-54213-y.

Anouk van Hooij 1, Kidist Bobosha 1 2, Jolien J van der Ploeg-van Schip 1, Sayera Banu 3, Saraswoti Khadge 4, Pratibha Thapa 4, Chhatra B Kunwar 4, Isabela M Goulart 5, Yonas Bekele 2, Deanna A Hagge 4, Milton O Moraes 6, Rosane M B Teles 7, Roberta Olmo Pinheiro 6, Erik W van Zwet 8, Jelle J Goeman 8, Abraham Aseffa 2, Mariëlle C Haks 1, Tom H M Ottenhoff 1, Robert L Modlin 7 9, Annemieke Geluk 10

Affiliations

Multicenter Study

Whole blood RNA signatures in leprosy patients identify reversal reactions before clinical onset: a prospective, multicenter study

Maria Tió-Coma et al. Sci Rep. 2019.

Abstract

Early diagnosis of leprosy is challenging, particularly its inflammatory reactions, the major cause of irreversible neuropathy in leprosy. Current diagnostics cannot identify which patients are at risk of developing reactions. This study assessed blood RNA expression levels as potential biomarkers for leprosy. Prospective cohorts of newly diagnosed leprosy patients, including reactions, and healthy controls were recruited in Bangladesh, Brazil, Ethiopia and Nepal. RNA expression in 1,090 whole blood samples was determined for 103 target genes for innate and adaptive immune profiling by dual color Reverse-Transcription Multiplex Ligation-dependent Probe Amplification (dcRT-MLPA) followed by cluster analysis. We identified transcriptomic biomarkers associated with leprosy disease, different leprosy phenotypes as well as high exposure to Mycobacterium leprae which respectively allow improved diagnosis and classification of leprosy patients and detection of infection. Importantly, a transcriptomic signature of risk for reversal reactions consisting of five genes (CCL2, CD8A, IL2, IL15 and MARCO) was identified based on cross-sectional comparison of RNA expression. In addition, intra-individual longitudinal analyses of leprosy patients before, during and after treatment of reversal reactions, indicated that several IFN-induced genes increased significantly at onset of reaction whereas IL15 decreased. This multi-site study, situated in four leprosy endemic areas, demonstrates the potential of host transcriptomic biomarkers as correlates of risk for leprosy. Importantly, a prospective five-gene signature for reversal reactions could predict reversal reactions at least 2 weeks before onset. Thus, transcriptomic biomarkers provide promise for early detection of these acute inflammatory episodes and thereby help prevent permanent neuropathy and disability in leprosy patients.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1

Figure 1

Transcriptional profiles of BL/LL and TT/BT patients without reactions. Gene expression levels of 103 target genes were assessed by dual-color RT-MLPA performed on ex vivo RNA isolated from whole blood of newly diagnosed, untreated BL/LL (n = 228; black circles) and TT/BT (n = 131; grey circles) leprosy patients without reactions from Bangladesh, Brazil, Ethiopia and Nepal. Log2-transformations of peak areas (normalized to the housekeeping gene GAPDH) of genes that were significantly differentially expressed between BL/LL and TT/BT are shown on the _y_-axis. Raw p-values were calculated using the Mann–Whitney test and adjusted for multiple comparisons using the Benjamini-Hochberg correction. *adjusted p-values < 0.05; **adjusted p-values < 0.01; (see Table S5).

Figure 2

Figure 2

Transcriptional profiles of healthy household contacts and endemic controls. Gene expression levels of 103 target genes were assessed by dual-color RT-MLPA performed on ex vivo RNA isolated from whole blood of healthy household contacts (HHC; n = 83; black circles) and endemic controls (EC; n = 200; grey circles) from Bangladesh, Brazil, Ethiopia and Nepal. Log2-transformations of peak areas (normalized to GAPDH) of genes significantly differentially expressed between HHC and EC are shown on the _y_-axis. Raw p-values were calculated using the Mann–Whitney test and adjusted for multiple comparisons using the Benjamini-Hochberg correction. *adjusted p-values < 0.05; **adjusted p-values < 0.01; (see Table S6).

Figure 3

Figure 3

Identification of biomarker risk signature for developing reversal reactions. Gene expression levels of 103 target genes were assessed by dual-color RT-MLPA performed on ex vivo RNA isolated from whole blood from BL/LL patients who developed reversal reactions (RR) at least two months later during the study (n = 30; grey circles) and BL/LL who did not develop reactions (n = 184; black circles) from Bangladesh, Brazil, Ethiopia and Nepal. Samples were analyzed at t = 0: in the absence of clinical signs of reactions. Log2-transformations of peak areas (normalized for GAPDH expression) of genes with significantly different expression between both groups (at t = 0) are shown on the _y_-axis. (A) Genes with a significant (p-value < 0.05) different expression only using the Mann-Whitney test are show. P-values were adjusted for multiple comparisons using the Benjamini-Hochberg correction (see Table S7). (B) Genes with a significant different expression in the global test and Mann-Whitney or the global test only are shown.

Figure 4

Figure 4

Identification of a minimal biomarker risk signature for developing reversal reactions. Biomarker signature to assess the risk of BL/LL patients to develop reversal reactions (RR). Gene expression data obtained by dual-color RT-MLPA of RNA isolated from whole blood of BL/LL patients from Bangladesh, Brazil, Ethiopia and Nepal at t = 0 were analyzed using the global test cluster analysis. The global test is a cluster analysis based on absolute correlation difference and average linkage developed for data sets in which many covariates (or features) have been measured for the same subjects, together with a response variable. Graphs (A,B) indicate genes that are higher expressed in future RR patients (red) or in non-reactional BL/LL patients (green). In (A) all genes analyzed are shown and in (B) only significant branches are shown. Table (C) shows values for the 5 genes that were statistically significant after correction for multiple testing (inheritance <0.05), representing the output signature of the global test shrinkage model.

Figure 5

Figure 5

Intra-individual longitudinal expression of patients who developed RR. Direct ex vivo RNA expression values were assessed by dual-color RT-MLPA on whole blood of 10 leprosy patients who developed RR during this study. Blood was analyzed at three time points: in the absence of any clinical signs of reactions and at least two months before RR (t = 0), at RR diagnosis before steroids (t = x) or after MDT and at least one month after end of steroids, in the absence of reactions (t = end). Log2-transformations of peak areas (normalized for GAPDH expression) are shown on the _y_-axis. Wilcoxon signed-rank test was performed. Genes with a significant difference (p-value < 0.05) in expression between before RR-at RR and at RR-after RR are shown.

References

    1. WHO. Global leprosy update, 2017: reducing the disease burden due to leprosy. Weekly epidemiological record. 2018;93:11.
    1. Barnes PF, et al. Cytokine production at the site of disease in human tuberculosis. Infect. Immun. 1993;61:3482–3489. -PMC -PubMed
    1. Saini C, Ramesh V, Nath I. CD4+ Th17 cells discriminate clinical types and constitute a third subset of non Th1, Non Th2 T cells in human leprosy. PLoS Negl Trop Dis. 2013;7:e2338. doi: 10.1371/journal.pntd.0002338. -DOI -PMC -PubMed
    1. Modlin RL, et al. Learning from lesions: patterns of tissue inflammation in leprosy. Proc. Natl. Acad. Sci. USA. 1988;85:1213–1217. doi: 10.1073/pnas.85.4.1213. -DOI -PMC -PubMed
    1. Scollard, D. M. In International textbook of leprosy (eds D. M. Scollard & T. P. Gillis) (2016).

Publication types

MeSH terms

Substances

LinkOut - more resources