Genome-wide DNA methylation profiling after Ayurveda intervention to bronchial asthmatics identifies differential methylation in several transcription factors with immune process related function

Abstract

Background

The Indian traditional medicinal system, Ayurveda, describes several lifestyle practices, processes and medicines as an intervention to treat asthma. Rasayana therapy is one of them and although these treatment modules show improvement in bronchial asthma, their mechanism of action, particularly the effect on DNA methylation, is largely understudied.

Objectives

Our study aimed at identifying the contribution of DNA methylation changes in modulating bronchial asthma phenotype upon Ayurveda intervention.

Materials and methods

In this study, genome-wide methylation profiling in peripheral blood DNA of healthy controls and bronchial asthmatics before (BT) and after (AT) Ayurveda treatment was performed using array-based profiling of reference-independent methylation status (aPRIMES) coupled to microarray technique.

Results

We identified 4820 treatment-associated DNA methylation signatures (TADS) and 11,643 asthma-associated DNA methylation signatures (AADS), differentially methylated [FDR (≤0.1) adjusted p-values] in AT and HC groups respectively, compared to BT group. Neurotrophin TRK receptor signaling pathway was significantly enriched for differentially methylated genes in bronchial asthmatics, compared to AT and HC subjects. Additionally, we identified over 100 differentially methylated immune-related genes located in the promoter/5′-UTR regions of TADS and AADS. Various immediate-early response and immune regulatory genes with functions such as transcription factor activity (FOXD1, FOXD2, GATA6, HOXA3, HOXA5, MZF1, NFATC1, NKX2-2, NKX2-3, RUNX1, KLF11), G-protein coupled receptor activity (CXCR4, PTGER4), G-protein coupled receptor binding (UCN), DNA binding (JARID2, EBF2, SOX9), SNARE binding (CAPN10), transmembrane signaling receptor activity (GP1BB), integrin binding (ITGA6), calcium ion binding (PCDHGA12), actin binding (TRPM7, PANX1, TPM1), receptor tyrosine kinase binding (PIK3R2), receptor activity (GDNF), histone methyltransferase activity (MLL5), and catalytic activity (TSTA3) were found to show consistent methylation status between AT and HC group in microarray data.

Conclusions

Our study reports the DNA methylation-regulated genes in bronchial asthmatics showing improvement in symptoms after Ayurveda intervention. DNA methylation regulation in the identified genes and pathways represents the Ayurveda intervention responsive genes and may be further explored as diagnostic, prognostic, and therapeutic biomarkers for bronchial asthma in peripheral blood.

Graphical abstract

Abbreviations:

5AC

5-Azacytidine

aPRIMES

Array-Based Profiling of Reference-Independent Methylation Status

AADS

Asthma-Associated DNA methylation Signatures

AHR

Airway Hyper Responsiveness

AT

After Treatment

BEC

Bronchial epithelial cells

BH

Benjamini Hochberg

BT

Before Treatment

DAC

5-Aza-2′-Deoxycytidine

DAVID

Database for Annotation, Visualization and Integrated Discovery

DNMT

DNA Methyl Transferase

HC

Healthy Controls

HDAC2

Histone Deacetylase 2

Hg18

Human genome build 18

EWAS

Epigenome wide association study

FC

Fold Change

FDR

False Discovery Rate

FEV1

Forced Expiratory Volume in 1 s

FVC

Forced Vital Capacity

GO

Gene Ontology

ICD

International Classification of Disease

IEC

Institutional Ethical Committee

IER

Immediate-Early Response

IL-4

Interleukin-4

KEGG

Kyoto Encyclopedia of Genes and Genomes

LTC4

Leukotriene C4

LTD4

Leukotriene D4

LTE4

Leukotriene E4

NEB

New England Biolabs

PCA

Principal Component Analysis

TADS

Treatment-Associated DNA Methylation Signatures

TAMV

Tilak Ayurveda Mahavidhyalaya

TCM

Traditional and Complementary Medicine

TF

Transcription Factors

Th2

Helper T cell 2

Treg

Regulatory T cells

TSS

Transcription Start Site

1. Introduction

Asthma is one of the major respiratory disorders, affecting 334 million people worldwide, with a global prevalence of 4.3% [1]. In India, the estimated burden of asthma is 17–30 million [2] with a prevalence varying from 2.3% to 11.9% in children and 0.96%–11.03% among adults [3]. It is a heterogeneous disease with genetic and environmental factors influencing disease susceptibility, inception, progression and therapeutic outcome [4]. The inhaled corticosteroids, by far the most effective anti-inflammatory treatment, have become first-line therapy, and bronchodilators and oral cysteinyl leukotriene receptor antagonist (montelukast) which inhibit inflammatory actions of leukotrienes (peptides) such as LTC4, LTD4 and LTE4 are administered depending upon patient's clinical presentation [5,6]. Bronchial asthma condition is described as ‘Tamaka Śwāsa’, a type of Swasa Roga (breathing disorder) and manageable with medication in Ayurveda, a traditional and complementary medicine (TCM) [7]. Ayurveda based treatments involve panchakarma (purification therapies) and rasayana medicines (curative & rejuvenating) along with diet plan and lifestyle modification for the treatment of asthma in India [8]. Ayurveda multiherbal formulations such as Shirishavaleha [7] Bharangyadi Avaleha [9] and Vasa Avaleha [10] displayed symptom relief, improved the quality of life and treatment with decreased asthma attack frequency in bronchial asthma patients.

Epigenetic mechanisms are recognized to play an important role in the establishment and maintenance of the Helper T cell 2 (Th2) bias in asthma, the release of Interleukin-4 (IL-4), Interleukin-5 (IL-5) and Interleukin-13 (IL-13) inflammatory cytokines [[11], [12], [13]]. Up-regulation of such key genes in asthma is preceded by activation of transcription factors (TF) such as NF-kB, STATs, NF-ATs, and AP-1 in response to external stimuli [14,15]. Cell-specific TFs such as T-bet, GATA3, FOXP3, RORC2 that play important roles in various T-subtype differentiations into Th1 cell, Th2 cells, Regulatory T cells (Treg cells) and Th17 cells respectively, are known to be regulated by DNA methylation and histone modifications [14,16]. Other inflammation-related genes such as IFN-γ, FOXP3, STAT5A, CRIP1, IL-4R, ARG2, ASCL3, ALOX12, ADRB2, IL-6, iNOS, IL-1R2, IL-2 have been reported to be associated with altered DNA methylation in asthma, with a few showing inverse correlation with gene expression in asthmatics [[17], [18], [19], [20]]. Moreover, various environmental factors are identified to mediate their influence through DNA methylation changes that contribute to disease development or protection [21]. Such strong influence of environmental factors mediating reversible DNA methylation regulation has paved the way for the potential use of DNA methyltransferase (DNMT) inhibitors, azacytidine (5-azacytidine or 5AC) and decitabine (5-aza-2′-deoxycytidine or DAC) as therapeutic agents in asthma [22,23].

Consumption of a healthy diet, following a good lifestyle regime and stress management, demonstrated positive effects in therapy for prostate cancer in men [24,25]. Such influence of environmental factors can be explained on the basis of epigenetic changes. A similar study on bronchial asthmatics, which involved Ayurveda intervention with controlled diet and lifestyle changes, showed reduced asthma symptoms and significantly varied plasma cytokine profiles in treated individuals [8]. We hypothesized that asthma pathophysiology and treatment response might be influenced by the DNA methylation of specific genes and could be altered by treatment with Ayurvedic interventions. Therefore, a prospective comparative study was undertaken to evaluate the effect of Ayurveda treatment mediated by DNA methylation in the reversal of asthma phenotype and to examine its molecular and functional basis in the therapeutic outcome. We observed DNA methylation changes in various regulatory regions in the genome showing consistent methylation status in Ayurveda treated and healthy controls in comparison to bronchial asthmatics and suggests their important role in the clinical management of asthma.

2. Materials and methods

2.1. Study design and clinical characteristics of bronchial asthmatics

This study comprised 40 subjects with mild-moderate asthma/bronchial asthma (Tamaka Śwāsa) and 40 healthy age and gender matched control. Participants visiting the outpatient department of Tilak Ayurveda Mahavidhyalaya and Sheth Tarachand Ramnath Charitable Ayurveda Hospital, Pune (TAMV); were enrolled in the study after obtaining written informed consent. The standard operating protocol for the Ayurveda treatment on bronchial asthmatics was followed as per guidelines of the Declaration of Helsinki and Tokyo for humans and approved by the TAMV Institutional Ethical Committee and informed consent was obtained as reported earlier (IEC Approval No. RSTH/RES/IEC/429/2011) [8]. The treatment regimen included panchakarma, shaman (disease alleviating treatment) and rasayana treatment (immunomodulator) with a controlled diet and lifestyle changes [8]. The Ayurvedic intervention used in this study is provided in Joshi et al. [8]. The intervention consisted of oil massage (Abhyanga) using Narayan taila, steam bath, and intake of the medicine combinations according to the asthma disease/stage in the patient with diet advice based on the clinical approach mentioned in Ayurveda classical texts [8].

Bronchial asthma was defined as per the international classification of disease (ICD)-9-CM guidelines (ICD, 2005) for diagnosis by chest physician along with the consultation of expert Ayurveda physicians [8]. Lung functional measurements such as forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC) was performed using spirometry according to 2005 ATS/ERS standards for recruitment of bronchial asthmatics [8]. The inclusion and exclusion criteria for recruitment and demographic data of the participants in the study including age, gender, absolute eosinophil count, and spirometry values are presented in Supplementary file 1, Table S1.

Peripheral blood samples were collected from healthy and bronchial asthmatics before and after Ayurveda intervention. The DNA was isolated from the peripheral blood of healthy control (HC), and bronchial asthmatics before treatment (BT) and after treatment (AT) using the Flexigene-DNA isolation kit (Qiagen, Germany) according to the manufacturer's instructions. The DNA was stored at −20 °C until used. The quality and quantity of the DNA were assessed by agarose gel electrophoresis and spectrophotometry. The overall study design has been summarized in Fig. 1.

Fig. 1.

Flowchart showing the stepwise analysis followed in this study.

2.2. Genome-wide methylation profiling using microarray technology

Genome-wide methylation was analyzed with aPRIMES [26] coupled to microarray using pooled DNA of ten individual samples from male and female belonging to BT, AT and HC group. In brief, 5 μg of genomic DNA was digested using 50 U MseI (New England Biolabs (NEB), UK) for 16 h as recommended by the manufacturers. The MseI-fragments were ligated to annealed double stranded linkers (H-12: 5′-TAA TCC CTC GGA-3′ and H-24: 5′-AGG CAACTG TGC TAT CCG AGG GAT-3′) at 16 °C and heat inactivated the T4 DNA ligase (2U) (NEB, UK) at 65 °C for 10 min. The resulting ligated MseI fragments were equally divided and were digested with the methylation-specific restriction enzyme McrBC (NEB, UK) for 8 h. The other half was subjected to digestion consecutively with HpaII (NEB, UK) and BstUI (NEB, UK) for 16 h each as recommended by manufacturers. The digested DNA was later purified using the NucleoSpin® Gel and PCR Clean-up kit (Macherey–Nagel, Germany) according to the manufacturer's instruction. Enriched DNA fragments in both the aliquots were amplified by linker-PCR using H-24 primers in Veriti® Thermal Cycler (Thermo Fisher Scientific, USA). Amplification conditions were programmed to 72 °C for 5 min, followed by 20 cycle loops at 97 °C (1 min) and 72 °C (3 min). Final elongation was carried out at 72 °C for 10 min. The PCR product was purified using the NucleoSpin® Gel and PCR Clean-up kit (Macherey–Nagel, Germany), subsequently labeled with cyanine3 and cyanine5 (Amersham Biosciences, USA) by using BioPrime array CGH genomic labeling system (Thermo Fisher Scientific, USA). Equal concentrations of labeled cyanine3 and cyanine5 enriched for unmethylated and methylated DNA fragments respectively were co-hybridized onto the Agilent 244 K CpG island microarray chip (Agilent Technologies, USA) at 65 °C, and speed of the hybridization rotator was set to rotate at 18 rpm for 40 h. Slides were washed with wash buffer after hybridization, dried and scanned using G2505B DNA microarray scanner (Agilent Technology, USA) with Sure Scan High resolution technology (Methodology detailed in Supplementary Methods).

2.3. Microarray data analysis

Agilent Feature Extraction software (v 10.5.1.1) was used to extract background corrected loess normalized green and red processed signal intensities within the array. The subsequent quality analysis was performed using R v2.15 with Bioconductor package Limma. Further, all the statistical and downstream clustering analysis was performed using Gene Spring (V11.5) software. Intra-group quantile normalization for four hybridized slides within BT and AT groups was applied to identify statistically significant probes within BT and AT group with fold change (FC) of ≥1.5 by one sample t-test after Benjamini Hochberg (B–H) false discovery rate (FDR) correction at p < 0.05 [27]. We later performed inter-group analysis to obtain statistically significant differentially methylated probes between BT vs AT (treatment-associated DNA methylation signatures (TADS) and BT vs HC (asthma-associated DNA methylation signatures (AADS). It initially involved quantile normalization of all the twelve arrays (four arrays each from BT, AT and HC), followed by identification of differentially methylated CpG probes by paired t-test with FC ≥ 1.5. We have used p < 0.05 for determining statistically significant differentially methylated probes. B–H FDR correction was further used to obtain FDR (≤0.1) adjusted p-values for differentially methylated probes. The differentially methylated probes from BT vs AT and BT vs HC inter-group analysis were further used to draw hierarchical clustering and principal component analysis (PCA).

2.4. Genomic regulation analysis, functional and biological evaluation of differentially methylated signatures

The sequence of probe region from TADS and AADS were annotated using Galaxy genome browser and Agilent e-array technology file based on its overlapping to specific genomic features as reported earlier [27]. Using RefSeq gene coordinates (Hg18), the probes were mapped to 5′-UTR and 3′-UTR regions. The chromosomal coordinates of the 199,399 probes were mapped on the basis of gene region classified as an upstream (−10 Kb to -5Kb), promoter (-5 Kb to TSS), 5′-UTR, gene body, 3′-UTR, downstream and intergenic (probe sequence overlap neither promoter nor gene body), CpG islands and CpG shores [27]. Enrichment analysis (Chi-square test, p < 0.05) for probes in TADS and AADS was performed for genomic elements, histone marks, regulatory regions and twenty transcription factor binding sites using EpiExplorer [28]. Functional gene enrichment analysis for Gene Ontology terms was performed and the analysis was restricted to biological processes category and KEGG pathway using Database for Annotation, Visualization and Integrated Discovery (DAVID) with default parameters [29]. Fold enrichment greater than 1.3 for biological process clusters was considered significant. Terms for biological processes with enrichment significance of p < 0.05 were further used in REVIGO (http://revigo.irb.hr) for the removal of redundant GO terms and data visualization [30]. Pathway analysis was performed using WEB-based GEne SeT AnaLysis Toolkit [31]^. Each list was analysed in three databases (KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway, Wikipathways and Pathway Commons. Further analysis involved selection of genes enriched in at least two databases and visualised in GeneMANIA [32] using default parameters for various functional and molecular interaction analysis and manually grouped genes based on the enriched pathways. Cytoscape 3.6.0 [33] with ClueGO v2.5.0 plugin [34] was used for immune system processes enrichment using default parameters. Genes with immune system-related functions were downloaded from The Immunology Database and Analysis Portal (ImmPort) system gene list given in InnateDB platform [35]. Network analysis was performed in InnateDB platform using Entrez ID of the genes of interest with default setting followed by opting for ‘Return Allergy and Asthma interactions only’ criteria in the webform. Webtools such as Heatmapper (with default settings) and ValcanoseR (log₂FC methylation (>±0.6) against (-log₁₀) p-value B–H corrected (<0.05) were used for data visualization of 28 genes showing consistent DNA methylation between TADS and AADS [36,37].

2.5. Evaluation of consistent methylation signatures in after treated and healthy control, compared to bronchial asthmatic patients

We hypothesized that Ayurveda intervention of six months mediates its beneficial effects on bronchial asthmatics via DNA methylation changes. To evaluate our hypothesis, we initially identified the genes showing consistent methylation status in AT (from BT vs AT) and HC (from BT vs HC) groups. We further compared for similar enrichment of these common genes either methylated or unmethylated in between AT group and HC group in various analyses. This was performed to understand the methylation status in healthy participants and relate to AT group and assess the possible role of DNA methylation in mediating the potential benefits of Ayurveda intervention in the BT group.

2.6. Statistical analysis

The clinical parameters such as age, gender, absolute eosinophil count, and spirometry reading were tabulated for all the participants in the present study (Supplementary Table S1). To evaluate the differences among continuous variables, paired t-test were performed wherever applicable. Chi-square (χ2) test was applied for categorical variables. We used a two-sided level of significance for all statistical tests. A p-value of <0.05 was considered statistically significant for all tests. Statistical analysis was performed using GraphPad Prism 5 and SPSS version 16.

3. Results

3.1. Genome-wide DNA methylation profiling identifies differential methylated signatures in bronchial asthmatics, compared to Ayurveda treated bronchial asthmatics and healthy control

Individuals subjected to Ayurveda intervention displayed improvement in health as published earlier [8]. The overall study design has been summarized in Fig. 1. Participant characteristics (n = 40) from each group (BT, AT and HC) used for the DNA methylation study is presented in Table S1. Pooled DNA strategy was employed, due low amount of peripheral blood DNA available, to determine the genome-wide DNA methylation changes [38,39]. We performed array-based profiling of reference-independent methylation status (aPRIMES) coupled to microarray for genome-wide methylation analysis [26] as this method is reported to reduce the impact of DNA methylation bias due to cell-specificity in heterogeneous cell population of peripheral blood and inter-individual variation [26]. The microarray data analysis scheme employed in the study are illustrated in Supplementary file 1 and Supplementary Fig. S1. In the first section, intra-group DNA methylation analysis was performed to determine probes with consistent DNA methylation status within BT, AT and HC groups. Quantile normalization and one sample t-test (four arrays each for BT, AT and HC groups; Supplementary file 1, Supplementary Fig. S1A) identified 108,099 and 49,467 probes showing significant (p < 0.05 after Benjamini-Hochberg FDR correction) intra-group consistency in methylation with FC ≥ 1.5 in BT and AT respectively. The degree of overlap of the probe region between the BT and the AT group from DNA methylation microarrays are summarized in Supplementary Fig. S2. The intra-group DNA methylation analysis resulted in identifying a total of 113,609 statistically significant probes (p < 0.05 after B–H FDR correction) (Supplementary Fig. S2) common and unique to BT and AT groups.

In the second section, without considering the biological conditions, all 12 arrays (four arrays each from BT, AT and HC) were subjected to quantile normalization to identify highly differentially methylated regions by paired t-test analysis (Supplementary Fig. S1B). To find differentially methylated regions post Ayurveda treatment in bronchial asthmatics, we further limited the paired t-test analysis (BT vs AT and BT vs HC) to 113,609 probes, showing significant intra-group consistency in methylation within BT and AT (Supplementary Fig. S1A). We identified 4820 TADS and 11,643 AADS – from the probes significantly differentially methylated [FDR B–H (≤0.1) adjusted] with a FC ≥ ±1.5 from the BT vs AT and BT vs HC comparisons respectively.

Hierarchical clustering for 4820 TADS performed by Ward's linkage and Euclidian distance metric functions displayed distinct regions specific to AT and BT groups respectively (Supplementary Fig. S3A). Principal component analysis (PCA) using 4820 TADS (Supplementary Fig. S3B) and 11,643 AADS (Supplementary Fig. S3D) differentiated BT from AT groups (Supplementary Fig. S3B) and HC groups (Supplementary Fig. S3D), indicating a substantial difference in DNA methylation profiles between HC, BT and AT groups.

3.2. Genomic distribution of TADS and AADS

Our analysis identified that TADS consisted of 2644 BT methylated/AT unmethylated probes (55%, representing 1884 genes), henceforth referred to as treatment-unmethylated signatures, while 2176 were BT unmethylated/AT methylated probes referred to as treatment-methylated signatures (Fig. 2A). Sequence annotation for 199,399 probes in Agilent 244 K CpG island microarray chip (Agilent Technologies, USA), mapped 87% (165,826) of probes to CpG islands and the rest 13% (33,573) to CpG shores. Out of 4820 TADS, 3882 probes were located in the 3377 CpG islands (78%) and 938 probes were located in CpG shores (22%) (Fig. 2B). The TADS mapped to their nearest protein-coding genes showed that 50% of the TADS resided within the gene body, 3% in upstream region (−10 Kb to −5 Kb) and 24% were located in promoters (−5 Kb to transcription start site (TSS)) (Fig. 2C). Within these promoter-associated TADS, 26% were directly mapped to the TSS of known protein-coding genes and 54% were mapped to region −200 bp to −2000 bp upstream to TSS (Fig. 2D). We observed treatment-unmethylated probes in gene body and upstream region were significantly (Chi-square t test, p < 0.05) overrepresented and decreased respectively, whereas in upstream region, 5′UTR and 3′UTR regions, treatment-methylated signatures were significantly overrepresented (Chi-square t-test, p < 0.05) (Fig. 2E). We analysed the distribution of treatment-unmethylated and methylated signatures with respect to distance and upstream or downstream position to transcription start site (TSS). Enrichment analysis using Chi-square t-test was found to show significant (p < 0.01) underrepresentation for treatment-unmethylated signatures within TSS-5Kb region (Fig. 2F). Moreover, treatment-unmethylated signatures were more frequently overrepresented (Chi-square t test, p < 0.0001) in regions annotated to 5–50 Kb and 50–500 Kb absolute distance to TSS (Fig. 2F). Whereas treatment-methylated signatures were decreased in regions annotated to 5-50 Kb absolute distance to TSS (Chi-square t test, p < 0.05) (Fig. 2F). Similarly, we performed chromosome enrichment analysis (TADS and AADS), genomic distribution and enrichment analysis for AADS and the results are presented in Supplementary File 1 & Supplementary Figs. S4 and S5.

Fig. 2.

Characterization of treatment-associated DNA methylation (TADS) signatures by genomic distribution and enrichment analysis. (A) Among the 4820 differentially methylated probes, 2644 probes (55%; representing 1884 genes) were treatment unmethylated signatures (methylated in BT group and unmethylated in AT); and 2176 probes (45%; representing 1546 genes) were treatment-methylated signatures (unmethylated in BT and methylated in AT group). (B) Distribution of TADS in regions in relation to CpG islands. (C) Overall distribution of the TADS in genomic elements. (D) Distribution of the promoter probes in 4820 TADS around TSS and upstream region from the start of the gene. (E) Frequency of TADS in different genic regions categorized based on the sequence overlap and distance from TSS. (F) Frequency of treatment-unmethylated signatures and treatment-methylated signatures categorized according to absolute distance to the nearest TSS of gene.

3.3. TADS and AADS enrichment analysis for various genome regulatory regions

We analysed the chromosomal coordinates of methylated and unmethylated probes located in regulatory regions such as promoter and gene body in TADS and AADS overlapping the regions for any particular histone states, genome regulatory roles and twenty transcription factor (TF) binding sites using the EpiExplorer platform [28]. We performed our analysis by checking for regions showing strong overlap (at least 50%) with option – “any” for tissue, out of the annotated nine cell types available online as GM12878 cell line (lymphoblastoid) data was not consistently available for all twenty TF binding sites and histone mark. For comparison, we used chromosomal coordinates from the promoter and gene body in the Agilent array and obtained overlapping of the various features of our interest.

H3K4me1 marks-active transcriptional enhancers, H3K9ac marks - open chromatin/active promoter, H3K27ac-enhancer-associated, H3K4me2 mark-poised and active enhancer, and H3K4me3 marks-poised and active promoter marks were enriched for TADS in gene bodies (Chi-square t-test, p < 0.05) (Fig. 3B). When analysed for enrichment for various regulatory regions we observed that treatment-methylated signatures were significantly (Chi-square t-test, p < 0.05) depleted for weakly transcribed regions and conserved regions, while treatment-unmethylated regions were enriched for conserved regions (Fig. 3C). Furthermore, we observed that regions overlapping to heterochromatin, polycomb-repressed weakly transcribed, conserved regions and DNaseI sites were enriched for unmethylated signatures (Fig. 3D). Treatment-methylated signatures were significantly (Chi-square t-test, p < 0.05) enriched in weak promoters and depleted in heterochromatin, weakly transcribed and conserved regions (Fig. 3D).

Fig. 3.

Regions from treatment-associated DNA methylation signatures (TADS, 4820) overlapping to various histone state maps, genome regulatory feature and transcription factor using EpiExplorer platform. Data was generated by comparing the overlap of treatment-methylated and treatment-unmethylated signatures belonging to promoter and gene body TADS against all promoter and gene body features in Agilent 244 K CpG array. Overlap of TADS in (A) promoter and (B) gene body with various histone state maps. Overlap of TADS in (C) promoter and (D) gene body with various genome regulatory features. Overlap of TADS in (E) promoter and (F) gene body with twenty transcription factor binding sites. Chi-square test was used to calculate statistical significance; and significance of p < 0.05 is indicated by an asterisk. Two asterisk denotes statistical significance of p < 0.01 and three asterisk denotes p < 0.0001 by the chi-square test for comparisons.

Investigation of overlapping in the region for 20 transcription factor (TF) binding sites in both promoter and gene bodies was performed for both TADS and AADS. We found that promoter regions for TR4, YY1, GTF2B and ZNF263 sites were enriched for treatment-methylated signatures (Chi-square t-test, p < 0.05) (Fig. 3E), while c-FOS and NF-kB binding sites were enriched for unmethylated signatures (Chi-square t-test, p < 0.05) (Fig. 3E). In gene bodies, we observed that treatment-methylated signatures showed enrichment for TR4, ATF3, YY1, MAX, c-JUN, c-MYC, E2F4, E2F6, NF-kB, GTF2B, SETDB1, ZNF263, and TFIIIC110 binding sites (Chi-square t-test, p < 0.05) (Fig. 3F) and unmethylated for CTCF sites (Chi-square t-test, p < 0.0001) (Fig. 3E). Such similar analysis was performed for AADS and the results are presented in Supplementary File 1 & Supplementary Fig. S6.

3.4. Gene ontology analysis identifies consistent enrichment for transcription term in biological processes category between TADS and AADS

We performed analysis for gene ontology-biological processes to identify the most prominent enriched functions associated with the genes from treatment-methylated (in promoter regions: 469 genes; gene bodies: 831 genes) and treatment-unmethylated signatures (promoter regions: 498 genes; gene bodies: 1147 genes) for TADS, and asthma-methylated (in promoter regions: 713 genes; gene bodies: 2194 genes) and asthma-unmethylated signatures (in promoter regions: 1120 genes; gene bodies: 1759 genes) for AADS. Detailed information of terms enriched biological processes, related fold enrichment and the enriched genes belonging to various clusters (Fig. 4.) for eight gene list have been presented in the supplementary file 2 and 3. Terms with fold enrichment greater than 1.3 were considered to be significant. Positive regulation of transcription from RNA polymerase II promoter was one of the top terms showing significant enrichment of genes for treatment-unmethylated signatures in promoter and gene body (Fig. 4A and C, Supplementary file 2). Furthermore, in promoter and gene body treatment-methylated signatures showed significant enrichment for top term such as transcription from RNA polymerase II promoter (Fig. 4B and D, Supplementary file 2). Similarly, we performed enrichment analysis for AADS, and the results are presented in Fig. 4E–H and Supplementary File 3.

Fig. 4.

Significantly enriched terms for Gene ontology (GO) category -biological processes visualised using REVIGO (http://revigo.irb.hr/). Terms enriched (p < 0.05) in DAVID for biological processes were further summarized and presented in the scatter plot after removing the redundant terms using in semantic similarities by pairwise distance matrix using REVIGO. Terms enriched for biological processes category in (A) treatment-unmethylated signatures in promoter, (B) treatment-methylated signatures in promoter, (C) treatment-unmethylated signatures in gene body, (D) treatment-methylated signatures in gene body, (E) asthma-methylated signatures in promoter, (F) asthma-unmethylated signatures in promoter (G) asthma-methylated signatures in gene body, (H) asthma-unmethylated signatures in gene body. Colour of the circle indicates significance of the GO terms in the form of log10 (p-value), and GO terms are grouped together based on semantic-similarity. The gene count, gene symbol, p-value, fold enrichment is presented in supplementary file 2 and supplementary file 3.

Altogether gene ontology analysis for both TADS and AADS located in promoter and gene body region indicated that biological processes related terms such as 1) development biology, 2) transcription, 3) negative or positive regulation of various cellular processes, 4) signaling pathways, 5) various cellular response, 6) ion-transport, 7) protein-phosphorylation, modification, stabilization and 8) cell proliferation/cell cycle and actomyosin structure organization were enriched majorly (Fig. 4).

3.5. Neurotrophin TRK receptor signaling pathway was consistently differentially methylated in after-treatment bronchial asthmatics and healthy control participants, compared to before-treatment bronchial asthmatics

We performed pathway analysis for asthma-methylated (comprising 3123 genes) and asthma-unmethylated signatures (3262 genes) from AADS, treatment-methylated (1546 genes) and treatment-unmethylated signatures (1884 genes) from TADS to identify the genes with common methylation status with respect to the presence of bronchial asthma and remission of disease in response Ayurveda intervention (Supplementary Fig. S7). Pathway analysis were generated for treatment unmethylated signatures, asthma methylated and unmethylated signatures (p < 0.01, BH correction).

In treatment-unmethylated signatures, 98 genes were enriched in the pathway analysis, and network analysis showed that 40.41% these genes displayed physical interaction, 34.76% displayed co-expression (Fig. 5A, Table S2).

Fig. 5.

Functional analysis for genes in top seven pathways, epigenetically regulated by DNA methylation. (A) Network analysis for 98 genes in treatment-unmethylated signatures (B) network analysis for 188 genes in asthma-methylated signatures. Colour representations for various networks: purple co-expression, olive green shared protein domain, pink physical interaction, light blue co-localization, and orange predicted. Genes with striped circles were given as queries; genes in circles without stripes were included as attributes (default parameter) by GeneMANIA web tool. Functional network for genes in top seven pathways has been presented within the gene circles.

We found 188 genes from asthma-methylated signatures agreeing with the pathway analysis criteria; among these, 37.37% of genes displayed physical interactions and 32.02% enriched in the co-expression network (Fig. 5B, Table S2). Pathways such as neurotrophin TRK receptor signaling pathway, neurotrophin signaling pathway, ERBB signaling pathway and epidermal growth factor receptor signaling pathway were commonly enriched between treatment-unmethylated signatures and asthma-methylated signatures. Network analysis for 124 asthma-unmethylated signatures 124 genes is summarized in Supplementary Fig. S8, Table S2.

In case of asthma-unmethylated signatures, we found that 124 genes agreed with the pathway analysis criteria, and network analysis showed that 43.92% of genes displayed co-expression, 31.55% of these genes displayed physical interactions network (Supplementary Fig. S8, Table S2). Pathways such as neurotrophin TRK receptor signaling pathway, neurotrophin signaling pathway, ERBB signaling pathway and epidermal growth factor receptor signaling pathway were commonly enriched between treatment-unmethylated signatures and asthma-methylated signatures.

3.6. Identification of functional immune-related genes with consistent methylation status between Ayurveda treated bronchial asthmatics and healthy control individuals

We performed a Venny analysis on 1884 unmethylated and 1546 methylated genes in AT group from TADS and 3123 unmethylated and 3262 methylated genes in the HC group from AADS. This identified 818 genes (Supplementary Fig. S7A) and 592 genes (Supplementary Fig. S7B) showing consistent unmethylation and methylation status respectively between AT and HC groups. We performed gene ontology analysis for immune system processes for these 818 unmethylated and 592 genes methylated genes between AT and HC groups (Supplementary file 1, Supplementary Fig. S9, Table S3). Terms negative regulation of monocyte differentiation (50% of the associated genes, term corrected p-value = 0.02) and negative regulation of myeloid cell differentiation (10.7% of the associated genes, term corrected p-value = 0.017) were significantly enriched for 818 unmethylated and 592 methylated genes respectively between AT and HC group (Supplementary Figs. S9A and B). KEGG analysis resulted in enrichment of signaling pathways regulating pluripotency of stem cells (fold enrichment = 3.1, p-value = 0.01 after Bonferroni correction) and axon guidance (fold enrichment = 3, p-value = 0.04 after Bonferroni correction) in 818 unmethylated genes (Supplementary Fig. S9C). Similarly, progesterone-mediated oocyte maturation (fold enrichment = 3.9, p-value = 0.3 after Bonferroni correction) was shown to be the top enriched pathway in 592 methylated genes between AT and HC group (Supplementary Fig. S9D).

We used 4723 immune gene list obtained from the ImmPort system available from InnateDB web tool [35] for identification of immune-related genes from both TADS and AADS having at least two or more probes located in promoter or 5′UTR region. The scheme used for narrowing down the list and identify consistently unmethylated and methylated genes in both TADS and AADS is shown in Supplementary Fig. S10.

List of over 100 differential methylated immune-related genes identified in TADS and AADS is given in Supplementary file 4 and their enrichment for gene ontology-biological processes is given in Supplementary Fig. S11 and Supplementary Tables 5 and 6 Within these 28 genes were showing consistent methylation status BT group in comparison to AT and HC groups respectively (Supplementary file 5, Supplementary Fig. S12). Top genes among these identified in TADS and AADS is given in Table 1. Gene information, gene function, gene ontology for molecular function, cellular component and biological processes and the probe information of the 28 genes in TADS and AADS has been given supplementary file 5.

Table 1.

Top immune function related genes commonly identified in TADS and AADS.

Gene/Probe ID	Full name	p-value	FDR-B-H corrected	Description	Methylation status	Gene ontology	Biological process
Treatment associated DNA methylation signatures (4820)
GP1BB/A_17_P11430476	glycoprotein Ib (platelet), beta polypeptide	6.02E-04	0.00091	5′UTR	Methylated in BT group compared to AT	GO:0004888|transmembrane signaling receptor activity	GO:0007166|cell surface receptor signaling pathway
NKX2-2/A_17_P11110617	NK2 homeobox 2	9.33E-04	0.00139	Promoter	Methylated in BT group compared to AT	GO:0001159|core promoter proximal region DNA binding	GO:0003326|pancreatic A cell fate commitment
CXCR4/A_17_P01498754	chemokine (C-X-C motif) receptor 4	0.00142	0.00205	Promoter	Methylated in BT group compared to AT	GO:0000187|activation of MAPK activity	GO:0001569|patterning of blood vessels|
MLL5/A_17_P05663637	myeloid/lymphoid or mixed-lineage leukemia 5 (trithorax homolog, Drosophila)	0.00338	0.00520	Promoter	Unmethylated in BT group compared to AT	GO:0002446|neutrophil mediated immunity	GO:0003713|transcription coactivator activity
JARID2/A_17_P04573234	jumonji, AT rich interactive domain 2	0.00758	0.01149	Promoter	Methylated in BT group compared to AT	GO:0000122|negative regulation of transcription from RNA polymerase II promoter	GO:0001889|liver development
Asthma associated DNA methylation signatures (4820)
HOXA5/A_17_P05377907	homeobox A5	0.00055	0.00097	Promoter	Methylated in BT group compared to HC	GO:0001501|skeletal system development	GO:0002009| morphogenesis of an epithelium
FOXD1/A_17_P04013266	forkhead box D1	0.00123	0.00236	Promoter	Unmethylated in BT group compared to HC	GO:0000978|RNA polymerase II core promoter proximal region sequence-specific DNA binding	GO:0003677|DNA binding
PCDHGA12/A_17_P04327634	protocadherin gamma subfamily A, 12	0.00164	0.00316	5′UTR	Methylated in BT group compared to HC	GO:0005509|calcium ion binding	GO:0016,020|membrane
UCN/A_17_P15228412	urocortin	0.001956	0.003701795	Promoter	Methylated in BT group compared to HC	GO:0001664|G-protein coupled receptor binding	GO:0005179|hormone activity
GP1BB/A_17_P17255222	glycoprotein Ib (platelet), beta polypeptide	0.002835	0.005376621	5′UTR	Methylated in BT group compared to HC	GO:0004888|transmembrane signaling receptor activity	GO:0007166|cell surface receptor signaling pathway

Further, these identified 28 immune-related genes were used to generate network analysis. Network analysis results showed that majority (64.45%) of the genes displayed enrichment for co-expression network. Additional network analysis in InnateDB [35] gave an output of physical association relevant in allergy and asthma (Supplementary file 5).

To visualize these similar methylation pattern in AT with that of HC group, both when compared to BT we have generated heatmap for 28 genes that are consistently differentially methylated in TADS and AADS. For the same set of genes, we can observe comparable methylation pattern for AT group data (Supplemental Fig. 12) from TADS and HC group data from AADS, against a constant BT group. Furthermore, we visualized these genes in volcano plot and observed that common set of genes were significantly hypermethylated in BT group compared to AT and HC groups. Similar observation can be made for common set of genes significantly hypomethylated in BT group compared to AT and HC groups (Supplemental Fig. 13). This substantiates our hypothesis and clearly suggests the possible role of Ayurveda intervention mediating its therapeutic effect via methylation in bronchial asthmatics.

4. Discussion

The present study was undertaken to understand the role of DNA methylation in bronchial asthma and to delineate Ayurveda treatment induced DNA methylation signatures using a genome-wide methylation profiling approach in peripheral blood DNA of bronchial asthmatics. The primary finding of this study is that we have identified 28 immediate-early response and immune-regulatory genes (Supplementary Figs. S12 and 13) showing consistent methylation status between AT and HC groups. Furthermore, we found that the neurotrophin TRK receptor signaling pathway, epidermal growth factor receptor signaling pathway and ERBB signaling pathway was significantly enriched between treatment-unmethylated signatures and asthma-methylated signatures. This is the first report of a comprehensive study in bronchial asthmatic individuals upon Ayurveda. Results presented here suggest that these genes and identified pathways are potentially important in mediating the improved health and bronchial asthma remission reported earlier in this study [8].

Role of epigenetics and lifestyle-related effect has been investigated in some studies [40]. Several asthma-animal model [41,42] and human studies [43,44] have demonstrated the effect of various herbal plants and formulation in mediating the anti-asthmatic effects. Additionally, Ayurvedic compounds or formulations have been reported to provide improved resistance against asthma by modulating the Th2 cells or immune system [[45], [46], [47]] and herbs and spices mediated their anti-inflammatory effect by inducing epigenetic modifications [44]. Considering such studies, our work is an attempt to find epigenetic role played by classical Ayurvedic treatment along with diet for bronchial asthmatics and its implications in managing the disease. Asthma is chronic inflammatory disorder of airways known to be caused by environmental factors and genetics integration mediated by epigenetics [48]. Several cytokines are known to mediate their effect and drive the asthma disease [49]. Furthermore, IL-3, TIGIT and RUNX3 were found to be hypomethylated in inner city asthmatic children compared to control [50]. GATA3 TF, known for role in Th2 cell cytokine production was reported to be regulated by DNA methylation at birth in children and was associated with reduced risk for asthma at ages 3 [51]. Role of several genes is found to be associated with DNA methylation in asthma and environmental exposure such as air pollution, cigarette smoke, chemical substances and in utero diet [52]. Considering such environment–epigenetics interactions in inflammatory disorder, we hypothesized that upon Ayurveda intervention to bronchial asthmatics, the improvement in clinical outcome may be due to altered gene expression, especially in immune response genes resulting from changes in the DNA methylation pattern. To test and validate our hypothesis we have used aPRIMEs approach, a reference independent microarray hybridization for DNA methylation profiling. Due to low concentration of DNA, we restricted our DNA methylation sample size to 40 per condition i.e., BT, AT and HC samples. Sample pooling strategies are often used in such scenario. Opting for pooling strategy enabled us to use less DNA, minimize the experiment variability between samples, increase statistical power and reduce the cost of experiment [38]. Earlier report showed that for EWAS when using DNA pool strategy, to achieve 95% statistical power and a 10⁻⁶ significance level, considering Cohen's d effect sizes of 1.5, minimum sample size required is 43/condition [38]. Furthermore, Serdar et al., 2021 showed the relationship between effect size and sample size using few examples and suggested sample size of 34 for clinical study with power of 0.8 and effect size of 1 [53]. In our study, sample size used for DNA methylation analysis per treatment group was 40, slightly less than that what was reported in Gallego-Fabrega C et al., 2015 for pooled sample EWAS study. Furthermore, we used a minimum of 10 samples for pooling to ensure equal and accurately measured DNA from each sample for a pool construction. Such precaution was taken to ensure actual representation of methylation from each sample used in the pool and minimize technical errors that may result inappropriate estimation of methylation levels [54,55]. We split our samples according to gender during pooling as well as in microarray hybridization to eliminate the effect caused by dosage compensation in female samples achieved by X-chromosome inactivation via DNA methylation and other epigenetic modifications [56]. Furthermore, as Ayurveda follows a holistic approach to treatment our focus in this study was on the identification of subtle DNA methylation variations attributed to the Ayurveda treatment.

aPRIMEs approach was employed with the rationale of evading inter-individual variation and tissue-specific bias in DNA methylation as blood consists of heterogeneous cell population [26]. Upon intra-group analysis we could observe that upon Ayurveda intervention the number of probes showing consistency in methylation within AT group reduced to <50% of the probes (upon B–H FDR correction) in comparison to the BT group (Supplemental Figs. 1 and 2). This itself clearly indicates the effect of Ayurveda intervention in mediating the DNA methylation change at various loci in the genome. Intra-group analysis helped us to narrow down the number of probes to 113,609, showing DNA methylation consistency within BT and AT group, which were further used to identify the differentially methylated probes in inter-group analysis (Supplemental Fig. 1). Microarray analysis resulted in identifying 4820 (TADS) and 11,643 (AADS) differentially methylated probes with a FC ≥ ±1.5 from the BT vs AT and BT vs HC comparisons respectively. We have used p < 0.05 for determining statistically significant differentially methylated probes. B–H FDR correction was further used to obtain FDR (≤0.1) adjusted p-values for differentially methylated probes.

To elucidate the functional relevance and biological relationship of the TADS and AADS, we performed various enrichment analysis. Upon evaluating the enrichment for regulatory regions in our study, we observed that H3K27ac in gene body region was significantly enriched for treatment-unmethylated signatures, whereas it was significantly enriched for asthma-unmethylated signature suggesting possible loss of DNA methylation in AT group upon Ayurveda treatment to bronchial asthmatics in comparison to HC (Fig. 3B and Supplemental Fig. 6B). Such enrichment can also be observed in probes region overlapping to promoter conserved region (Fig. 3C and Supplemental Fig. 6D) and not in gene body conserved region (Fig. 3D and Supplemental Fig. 6D) suggesting possible loss of methylation in promoter conserved region alone upon Ayurveda treatment. Earlier studies have also reported enrichment of H3K27ac histone modification, an enhancer-associated mark for actively expressing genes [57] in non-coding (intergenic) region of bronchial epithelial cells (BEC) [58] and asthmatic CD4+ T-cells [59]. Conserved regions were also reported to be predominantly differentially methylated [60] and enriched as blocks of conserved regions in first intron with increased active regulatory chromatin marks [61]. Furthermore, we analysed TF binding site for NF-kB promoter (Fig. 3E and Supplemental Fig. 6E) and CTFC gene body (Fig. 3F and Supplemental Fig. 6F) overlapping regions. Results showed treatment-unmethylated signature being enriched in TADS and asthma-unmethylated signatures being enriched in AADS for both NF-kB promoter and CTFC promoter binding site further indicating loss of methylation upon Ayurveda treatment. NF-kB plays important role in regulating the expression of inflammatory genes with increased signaling in asthma patients [62] and shown to be functionally regulated by DNA methylation [63]. Similarly, CTCF binding sites are known to be regulated by asthma-associated alleles [64] and DNA methylation, thus making CTCF TF sensitive to binding to its target gene DNA [65,66]. Enrichment analysis results on regulatory regions from our study are preliminary and would require future validation studies on DNA from whole blood and further purified cellular population to confirm the relevance and specificity of our findings in bronchial asthmatics.

In pathway analysis more than 30% of genes in treatment-unmethylated (Fig. 5A, Table S2), asthma-methylated (Fig. 5B, Table S2) and asthma-unmethylated signatures (Supplementary Fig. S8, Table S2) were enriched for physical interaction and co-expression network, further indicating their possible direct relationship in either in protein complex, signaling pathways and and/or other cellular processes. Differential methylation of genes in enriched pathways such as neurotrophin TRK receptor signaling pathway [67], ERBB/epidermal growth factor receptor signaling pathways [68,69] were earlier reported being associated with asthma, which further substantiated the significance of these pathways in bronchial asthma and Ayurveda treatment response.

We have also identified over 100 immediate-response and immune-related genes from TADS and AADS displaying differential methylation in at least two or more probes within promoter or 5′UTR (Supplementary Fig. S11, Tables S5 and 6 and Supplementary file 4), as methylation in these CpG rich regions are in general associated with suppressed gene expression [70] and related functional consequence. Among these we found that 28 genes were commonly differentially methylated between TADS and AADS (Supplementary Fig. S12, Supplementary file 5). On GeneMANIA analysis we found that all these candidates displayed direct relationships via physical interaction (Fig. 6, Supplementary file 5). These genes majorly belonged to terms TF activity (FOXD1, FOXD2, GATA6, HOXA3, HOXA5, MZF1, NFATC1, NKX2-2, NKX2-3, RUNX1, KLF11) and few are already known to drive the inflammatory genes [[71], [72], [73], [74], [75]], symptoms [76,77] in asthma. Allergy and asthma interactions related to genes NFATC1 and RUNX1 (Supplementary file 5) were enriched in network analysis of the InnateDB platform using 28 common differentially methylated immune genes between TADS and AADS.

Fig. 6.

Network analysis for 28 immune system-related functions by GeneMANIA web tool. Colour representations for various networks: purple – co-expression for 64.45% of the genes, olive green - shared protein domain for 7.27% of the genes, pink – physical interaction – 11.76% of the genes, light green – genetic interactions for 3.45% of genes and light blue – co-localization for 9.13% of genes. Genes with striped circles were given as queries; genes in circles without stripes were included as attributes (default parameter) by GeneMANIA web tool.

These genes along with KLF11, FOXD1, CXCR4, UCN, JARID2 was significantly methylated in BT and GATA6, PTGER4, PANX1, SOX9 and TRPM7 was found being significantly unmethylated in BT in comparison to AT and HC group (Supplementary Fig. S13, Supplementary file 5). Earlier gene expression studies [71,72], DNA methylation studies [78,79] and animal studies have demonstrated the key role of these genes in asthmatic inflammation [75,[80], [81], [82], [83], [84], [85], [86]].

Collectively, our preliminary results were consistent with our hypothesis and the treatment followed here belongs to Ayurveda classical text that has been practiced for over centuries and there has been no known adverse events reported with this Ayurvedic treatment for asthma. Hence, any adverse implication due to treatment induced changes in conserved or tissue specific methylation is a rare possibility. However, there are several limitations in our study. Firstly, Ayurveda intervention happened on mild-moderate asthmatics (bronchial asthma) and therefore, we do not know the possible effect on methylation status for the observed TADS in increased severity. Secondly, we have used peripheral blood for genome-wide DNA methylation profiling, which may add on to the complexity of the data by introducing a bias based on the proportion of various cell types. We have tried to meet the problems associated with this limitation by using a pooling strategy and aPRIMES approach to enrich fragments that show homogeneous (average) status in methylation within the groups, and aid in reducing the bias introduced by inter-individual variations, tissue-specific methylation, sampling handling and experimental variations.

Thirdly, the use of nasal epithelial/airway cells or lung fluids would have been appropriate for DNA methylation profiling in asthma. However, obtaining such samples may be possible from adults, but difficult from children. Blood being an easily accessible biological sample, can be used as a surrogate sample in asthma as immunological alterations often pave the way for asthma pathophysiology and may reflect asthma disease related DNA methylation changes [27]. Several recent studies have used whole blood as the samples for identification of DNA methylation changes in asthma, replicated results in asthma-relevant tissue, indicating the presence of cross-tissue epigenetic effect [52,[87], [88], [89], [90]]. Furthermore, Ayurveda follows a holistic therapeutic approach and treatment to bronchial asthmatics in our study does not target any specific type of cell in the diseased body, but it works on the aggravated dosha and aids in remission of the disease [8].

Lastly, one of the limitations of the study is lack for confirmatory proof of the observations at expression levels. There were several challenges including expression of target genes in the blood cells itself is uncertain and also can be cell type specific. However, we intend take up these studies in future.

5. Conclusion

As a proof of concept, we have here shown that bronchial asthmatics display changes in DNA methylation upon Ayurveda treatment, identified various enriched genes, pathways that may be epigenetically regulated by DNA methylation. Furthermore, our findings are preliminary and further mechanistic studies are required to investigate and elucidate the role of these candidate genes regulated by DNA methylation in the pathogenesis, prognostics, diagnostics and therapeutics of asthma.

Author contributions

Smitha Bhat: Methodology, Investigation, Data curation, Formal analysis, Validation, Visualization, Writing - original draft. Harish Rotti: Investigation, Supervision, Writing - review & editing. Keshava Prasad: Visualization, Writing - review & editing. Shama Prasada Kabekkodu: Supervision, Writing - review & editing. Abdul Vahab Saadi: Writing - review & editing. Sushma P. Shenoy: Investigation, Writing - review & editing. Kalpana S Joshi: Conceptualization, Methodology, Investigation, Writing - review & editing. Tanuja M Nesari: Conceptualization, Methodology, Investigation, Data curation, Writing - review & editing. Sushant A Shengule: Investigation, Writing - review & editing. Amrish P Dedge: Investigation, Data curation, Writing - review & editing. Maithili S Gadgil: Investigation, Writing - review & editing. Vikram R Dhumal: Investigation, Data curation, Writing - review & editing. Sundeep Salvi: Investigation, Writing - review & editing. Kapaettu Satyamoorthy: Conceptualization, Methodology, Supervision, Validation, Funding acquisition, Resources, Writing - review & editing.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors wish to thank Manipal Academy of Higher Education, DST-FIST for infrastructure facilities and provided for this study.

Source of funding

This work was funded by Department of Science and Technology (DST), New Delhi, Government of India, Grant number-Prn.SA/ADV/Ayurveda/2010.

Appendix A. Supplementary data

The following are the Supplementary data to this article.