1, 2-dimethylhydrazine (LOT: A0251168) was purchased from Acros organics, NJ, United States. All other chemicals used were of analytical grade and procured from Hi-media Laboratories, Mumbai, India, else otherwise stated in the text.

The male albino wistar rats (120–140 g) were obtained from the central animal house facility and maintained under standard laboratory conditions, temperature (25 ± 1°C) with 12 h light/dark cycle. Animals were fed with standard animal diet [Chickpeas (Cicer arietinum) (30%); DRB (protein and calcium) (10%); Table salt (2%); Husk (20%); (10%); Corn (8%); Refined oil (20%)], and water ad libitum.

The study was approved by the institutional animal ethics committee of Sam Higginbottom University of Agriculture, Technology and Science – A Deemed University (approval no. IAEC/SHIATS/PA16III/SJPG17).

Animals were randomly chosen and grouped in ten groups of eight animals each. Carcinogenicity was induced by weekly injection of DMH (30 mg/kg, s.c) and taVNS was applied at various intensities through the auricular chonchal region.

For taVNS, animals were stimulated with specific stimulating parameters, including pulse width (ms), frequency (Hz), voltage (v) and duration (min). Stimulation started from the 4th day of DMH administration with the exception of taVNS control and dummy VNS groups.

Animal nerve stimulating electrode (MLA0320, AD Instruments, Australia), consisting of two exposed gold-plated platinum hook electrodes with rounded ends, was used for taVNS. Each electrode had a 2 mm diameter and electrodes were separated from one other by 3 mm.

The animals were anesthetized using a ketamine hydrochloride (100 mg/kg, i.m.) and diazepam (5 mg/kg, i.m.) combination followed by mounting on a wax tray, and electrodes were placed on the auricular chonchal region for stimulation (Rong et al., 2012; Li et al., 2014) (Figure 1A). The electrode leads were connected with power lab system 2/26 (AD Instruments, Australia) for stimulations. ECG and HRV were recorded (Power lab system 2/26, AD Instruments Australia) during stimulation for supervision of hemodynamic variations. To improve electronic conduction, saline was applied on the skin surface. The stimulation started from the 4th day after DMH treatment and continued weekly up to 6 weeks along with DMH. The stimulation parameters are detailed in Supplementary Table 1 (Kong et al., 2012; Li et al., 2014).

On the 42nd day, animals were recorded for ECG and HRV. On the 43rd day blood was collected from retro-orbital plexus and incubated (37°C, 1 h) followed by centrifugation to collect serum. Afterward, the animals were sacrificed under light ether anesthesia. The colon tissue of animals was collected by securing both ends with a surgical suture to prevent drainage of the colonic content (Figure 1B).

The positive and negative gold-plated platinum hook electrodes were placed on the skin on the left and right side of the thorax and the neutral electrode on the skin of the peritoneal region of anesthetized rats. The other ends of the electrodes were connected with power channel (ML-826) and Bio-amplifier (ML-136) for the conversion of analog signals into digital signals. The record ECG signals were analyzed offline, using Lab Chart Pro-8 software. The obtained recording graph was used to calculate ECG and HRV parameters with Lab Chart Pro-8 software. Analysis of hemodynamic raw data started with ensuring the correctness of R wave followed by calculation of heart rate and other ECG parameters. Similarly, the time and frequency-domain parameters for HRV were calculated from the ECG signals using Lab Chart Pro-8 software, following the method elaborated elsewhere (Shintaku et al., 2014; Xiong et al., 2016) (Figure 1B).

Weight variation within treatment groups was observed and calculated using the formula I (Mishra et al., 2016).

The colonic content was collected and evaluated for pH using pen type pH meter (Hanna Instruments, HI98107). Total acidity was calculated using a previously described procedure (Giri et al., 2015; Kumar et al., 2016).

Longitudinally opened colonic tissue was washed with normal saline and fixed by placing it between two layers of whatman paper for 24 h in 10% formalin. Subsequently, tissue was stained with 2% methylene blue and visualized through the light microscope. The ACF was visualized and counted in five independent frames using the method elaborated previously by our laboratory (Bird, 1987; Mishra et al., 2016).

Scanning electron microscope (JEOL-JSM-6490LV) was used to study the morphological changes on the colonic mucosa. Samples were prepared by fixation of tissues in 2.5% glutaraldehyde for 6 h at 4°C. Tissues were washed with 0.1 M phosphate buffer; post fixed and dehydrated using the method described elsewhere. The tissues were mounted on the aluminum stub with adhesive tape and visualized at magnification of 500X and 2000X (Cooke et al., 1984; Sammi et al., 2018).

Hematoxylin and eosin staining was used to study the histopathology of colonic tissues. The tissues were fixed overnight in paraformaldehyde with subsequent dehydration using isopropanol and xylene. Dehydrated tissues were fixed in paraffin wax and sectioned (0.5 mm) using a microtome. Sections were stained with H&E, and visualized under the digital biological microscope at a magnification of 40X (N120, BR-Biochem Life Sciences, New Delhi, India) (Serafino et al., 2014). Quantification of H&E staining and SEM was performed by using Image J software (Jensen, 2013).

The colon tissues (10% w/v) were homogenized in 0.15% KCl and centrifuged at 10,000 rpm. The supernatant was used for biochemical estimations for TBAR’s, SOD, catalase, GSH, and PC using previously established methods at our laboratory (Gupta et al., 2014; Kumar et al., 2014; Raj et al., 2014; Gautam et al., 2016).

The colon tissue was lysed in the RIPA lysis buffer to obtain total protein lysate and quantified using Bradford’s reagent. The proteins were resolved through 12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membrane (IPVH 00010 Millipore, Bedford, MA United States) using a method established in our laboratory (Roy et al., 2017).

Following the transfer, the membrane was blocked with solution of bovine serum albumin and skimmed milk in TBST (3%) for 3 h. The blocked membrane was incubated overnight with primary antibodies for Bcl-xl (MA-5-15142), B cell lymphoma- 2 (Bcl-2) (SC-7382), BAX (SC-23959), BAD (SC-8044), VDAC (SC-390996), cytochrome-c (SC-13561), Apaf-1 (SC-65891), procaspase-9 (SC 73548), HMGB-1 (SC-56698), α7nAchR (SC-5544), NFκBp65 (MA5-1616), and TNF-α (SC-1350) at 4°C. After incubation with primary antibody, the membrane was washed thrice with TBST and once with TBS. Subsequently, the membrane was incubated with the HRP conjugated secondary antibody, including anti-rabbit (SC-2030), anti-goat (SC-2020), anti-mouse (31430, Pierce Thermo Scientific, United States) (1:5000 dilutions) for 3 h.

The membrane was developed by using enhanced chemiluminescence substrate (Western Bright ECL HRP substrate, Advansta, Melanopark, CA, United States) in gel dock system after three and single washings with TBST and TBS, respectively. The protein bands were quantified by densiometric digital analysis using Image J software. β-actin (MA5-15739-HRP) was used as a standard reference (Roy et al., 2017).

Primer quest, a tool from the Integrated DNA Technologies website¹, was used for online primer designing in real time. Size of the amplicon was kept between 100 and 200 base pairs. Guanine-cytosine ratio was kept above 50% and melting temperature was kept between 58 and 62°C. The sequences of forward and reverse primers used for qRT-PCR are detailed in Supplementary Table 2 (Liu et al., 2004).

Total RNA was extracted from colon tissues using trizol reagent according to manufacturer instructions. Tissues were meshed in 250 μl trizol reagent by using micro pestle and volume was made up to 1 ml followed by addition of 200 μl chloroform and vortexing for 5 min.

Tissue suspension was centrifuged at 14,000 rpm, 4°C for 15 min. The aqueous phase was separated and transferred to a fresh vial, with subsequent centrifugation at 14000 rpm for 10 min. The pellets were washed twice with chilled ethanol and dissolved in 15 μl of 1% diethyl polycarbonate water. The samples were quantified through nanodrop (Qua Well Q5000).

1 μg of tissue RNA was used for cDNA synthesis, which was used as a template for qRT-PCR reaction using the method described previously by us. β actin was used as a house keeping control (Roy et al., 2017).

Fluorometric assays for caspase 3 and 8 were performed in amber colored 96 well plate using the protocol provided by the manufacturer. The equal volume of serum taken from all treatment groups was diluted with reaction buffer followed by addition of dithiothriol to a final concentration of 10 mM. 5 μl of IETD-AFC/DEVD-AFC substrate was added to the reaction mixture and incubated for 1 h at 37°C. The formed free AFC levels were measured in a plate reader at 400 nm excitation and 505 emissions. The results were expressed as fluorescence units/mg of protein (Kim et al., 2003; Brown et al., 2015).

Results are presented as mean ± SD and analyzed by one-way ANOVA followed by Bonferroni test for the possible significance identification between the various groups. ^ap < 0.05, ^bp < 0.01, ^cp < 0.001 were considered statistically significant. Statistical analysis was performed using Graph Pad Prism software (5.02).

Electrocardiogram and HRV were analyzed using Metabo Analyst. Metabo Analyst is a web-based data analysis tool. In this, obtained results are in tabular as well as in the graphical format, easy to understand. Metabo Analyst can be used for various analyses including biomarker analysis and pathway analysis but is here used for statistical analysis. Raw data was saved in the comma separated values and uploaded as spectral bins followed by data integrity check, data filtering, and normalization of data (pareto scaling). Finally, multivariate analysis of data was performed using principal component analysis and partial least squares – discriminant analysis method (Xia et al., 2009).

The DMH treatment was very well manifested for increase in RR interval (0.20 ± 0.01s), heart rate (380.1 ± 29.96 bpm), QRS interval (0.02 ± 0.001 s), P(0.04 ± 0.003 mv), and Q (0.20 ± 0.01 mv) wave amplitude along with decrease in QT interval (0.07 ± 0.007 s), JT interval (0.05 ± 0.003 s), QTc complex (0.15 ± 0.01 s) and R wave amplitude (1.35 ± 0.12 mv) (Figure 1C, Supplementary Table 3, and Supplementary Figure 2). The principal component analysis and partial least squares – discriminant analysis of the ECG signals revealed that taVNS 3, 4, and 5 could favorably regulate the ECG signals toward control. The HRV analysis revealed significant curtailment of average RR (157.80 ± 9.15 ms), median RR (158.8 ± 10.26 ms), low frequency (1.2 ± 0.10 μs²), high frequency (9.10 ± 0.81 μs²), and low frequency/high frequency (0.13 ± 0.01) in DMH treated animals (Supplementary Table 4 and Supplementary Figure 3). taVNS 2 and 3 could favorably curtail the deleterious effects of DMH (Figure 1D). When compared with standard chemotherapy, treatment with taVNS more favorably restored hemodynamic changes toward normal control (Supplementary Figures 2, 3).

When scrutinized on the account of physiological parameters, DMH treatment showed significant decrease in weight (-14.17 ± 1.32%) (Supplementary Figure 1), non-consequential decrease in pH (6.31 ± 0.56), with increase in total acidity (127.96 ± 8.6 mEql^-1) and ACF (56 ± 2.29NoS). taVNS regulated the weight variation and pH significantly. The total acidity and ACF count were favorably regulated more profoundly by taVNS3. When compared with standard chemotherapy, taVNS resulted in better regulation of weight variation (all taVNS groups), pH (taVNS 1 and 2) and total acidity (taVNS 1) toward control (Table 1, Supplementary Table 5, and Supplementary Figure 1). The taVNS treatment could not regulate the ACF when compared with standard chemotherapy against DMH induced colon carcinogenesis (Table 1).

When accounted for the oxidative stress markers, the DMH treatment upsurged TBARs (187.82 ± 6.78 nM of MDA/μg of protein) and PC (106.68 ± 10.82 nM/ml unit) in comparison to control. The SOD (0.004 ± 0.001 unit of SOD/mg of protein) and catalase (1.67 ± 0.086 nM of H₂O₂/min/mg of protein) levels were down-regulated with a significant increase in GSH (0.68 ± 0.07 mg %) in DMH treated animals. taVNS 2 and 3 embarked upon a more favorable regulation for TBARs and PC, respectively. taVNS 2 and 3 regulated the GSH, SOD and catalase as well (Table 2). When compared to the effects of taVNS and standard chemotherapy against DMH induced colon carcinogenesis, taVNS was more effective in regulating the TBARs (all taVNS treatments), catalase (all taVNS treatments), and GSH (taVNS 4).

Treatment with taVNS resulted in positive regulation of PC and GSH against DMH control but was comparatively less effective as standard chemotherapy.

The colonic tissue was evaluated morphologically through methylene blue staining, H&E staining, and SEM. After methylene blue staining of the colonic mucosa of DMH treated animals, dark-blue stained cells were visualized representing ACF. The ACF in the colonic mucosa of the DMH treated animals was also evident through H&E staining and SEM. The SEM images of DMH treated animals were besides apparent for small neoplastic lesions (Figure 2 – XXXIII). taVNS 3 and 4 curtailed the ACF formation and normalized the histopathological architecture more favorably. The formations of neoplastic lesions were also subsided after the taVNS application in general and profoundly by taVNS 3 and 4 (Figure 2, Supplementary Figure 4, and Supplementary Table 6).

Proteins extracted from individual groups were subjected to immunoblotting assay for anti-apoptotic (Bcl-2 and Bcl-xl), pro-apoptotic (BAX and BAD) proteins along with downstream markers of the apoptotic pathway (VDAC, cytochrome-c, Apaf-1 and procaspase-9).

After treatment with DMH, the expression of anti-apoptotic proteins (Bcl-2 and Bcl-xl) was increased. The expression of pro-apoptotic protein BAX was decreased while the BAD was increased. The taVNS application resulted in down-regulation of Bcl-2, with mixed effects on expression of Bcl-xl was noted. taVNS 1, 2 and 3 resulted in Bcl-xl up-regulation while taVNS 4 and 5 resulted in down-regulation. In response to DMH treatment, expression of BAX decreased while BAD increased. taVNS resulted in down-regulation of BAX and up-regulation of BAD. All in all, taVNS was recorded to have more favorable regulation of pro-and anti-apoptotic proteins, when compared to DMH treated animals.

1, 2-dimethylhydrazine administration was also evident for increased expression for VDAC, Apaf-1 and procaspase-9 with the concomitant decrease in cytochrome-c expression.

Transcutaneous vagus nerve stimulation down-regulated the VDAC (taVNS 1), Apaf-1 and procaspase-9 (except taVNS 4) expression (Figure 3A). taVNS favorably regulated the expression of VDAC (taVNS 1), Apaf-1 (taVNS 1 and 2), and procaspase-9 (except taVNS 4) (Figure 3A). DMH treatment was also evident for decreased levels of caspase 3 and 8, which were up-regulated after taVNS application (Figure 3C,D).

1, 2-dimethylhydrazine treatment decreased the α7nAchR expression with up-surged expression of NFκBp65, TNF-α, and HMGB1. Majority of taVNS treatments down-regulated the expression for NFκBp65 (except taVNS3), TNF-α, HMGB-1and up-regulated the α7nAchR expression (except taVNS 3) (Figure 4A). On considering standard chemotherapy, the majority of taVNS treatments resulted in better responses in regulation of CAP associated proteins with the exception of α7nAchR.

The fold change of gene expressions for a particular protein was validated through qRT-PCR. The qRT-PCR assay was used for the affirmation of genetic phenotypes for the protein markers associated with mitochondrial apoptotic pathway and CAP. Results of qRT-PCR were found in line with immunoblotting assay. Significant fold change of gene expression was noted in response to taVNS application. When considering mitochondrial apoptotic pathway, gene expression and protein expression of Bcl-2 and BAX were decreased and BAD expression was increased after taVNS application. taVNS exerted mixed effects for gene expression of Bcl-xl, taVNS 1,2 and 3 up-regulated while taVNS 4 and 5 down-regulated the expression of Bcl-xl.

All in all, taVNS positively modulated the gene expression for pro-and anti-apoptotic genes as compared to DMH application. DMH treatment was evident for increased expression for VDAC, Apaf-1, and procaspase-9 with collateral decreased cytochrome-c gene expression. qRT-PCR affirmed down-regulation of VDAC (taVNS 1), Apaf-1 and procaspase-9 with concomitant up-regulation of cytochrome-c gene expression in response to taVNS treatment (Figure 3B). qRT-PCR studies validated the effects of taVNS on CAP associated proteins. Toxicant treatment decreased the α7nAchR gene expression along with increased gene expression of NFκBp65, TNF-α, and HMGB-1. All taVNS treatments decreased the expression for NFκBp65, TNF-α, HMGB1 and increased the α7nAchR expression (except taVNS 3) (Figure 4B).