Homeoprotein OTX1 and OTX2 involvement in rat myenteric neuron adaptation after DNBS-induced colitis

Animals

Male Sprague–Dawley rats (weight 250–300 g, Envigo, Udine, Italy), were handled following principles of good laboratory animal care, in accordance with specific national and international laws and regulation. Animals were maintained at a regular 12/12 h light/dark cycle, under controlled environmental conditions (temperature 22 ± 2 °C; relative humidity 60–70%), with free access to a standard laboratory tap water and chow. The protocol was approved by the Animal Care and Use Ethics Committee of the University of Pavia (n. 3/2011).

DNBS-induced colitis

The model of DNBS-induced experimental colitis in rats was chosen since the inflammatory response develops rapidly, reaching the maximum grade in 6 days, and shares many features with the human IBD response (Filpa et al., 2017b). A single dose (30 mg) of 2,4-dinitro-benzene-sulfonic acid (DNBS, ICN Biomedicals, CA, USA), dissolved in 0.25 ml of 50% ethanol, was administered to isofluorane anesthetized rats by means of a polyethylene (PE-60) catheter into the colon 8 cm proximal to the anus. Ethanol, owing to its ability to break the mucosal barrier thus permitting DNBS penetration into the bowel wall, was administered (0.25 ml) to control animals as a vehicle. The dose of DNBS used in the study induces adequate inflammation, without causing unnecessary distress, suffering or mortality to the animals. DNBS-treated and control rats were kept separated during the study. After 6 days, animals were euthanized and the small intestine and distal colon was removed and washed with a physiological Tyrode’s solution (in mM: 137 NaCl; 2.68 KCl; 1.8 CaCl₂·2H₂O; 2 MgCl₂; 0.47 NaH₂PO₄; 11.9 NaHCO₃; 5.6 glucose). Throughout the treatment period, animals were monitored to evaluate possible physiological and behavioral changes (i.e., changes in body weight, respiration, occurrence of diarrhea, alterations of posture and in the appearance of the coat) as marks of suffering and distress.

Assessment of colonic damage

The severity of intestinal inflammation was evaluated macroscopically and histologically. Criteria for the macroscopic evaluation included the presence of adhesions between the colon and other intra-abdominal organs, the extension of hyperemia and macroscopic mucosal damage, the thickening of the colonic wall and the consistency of colonic fecal material, all attributed with a score ranging from 0 (no damage) to 6 (maximum damage) Filpa et al. (2017b).

For the microscopic histological evaluation, standard hematoxylin and eosin (HE) staining was carried out on serial sections (7 μm) of small intestine and colonic segments, from control and DNBS-treated rats, which were first fixed for 24–48 h with 4% formaldehyde in acetate buffer 0.05 M and successively embedded in paraffin. Preparations were observed under a light microscope (Nikon Eclipse Ni; Nikon, Tokyo, Japan) and data were recorded using a DS-5M-L1 digital camera system (Nikon Corporation, Tokyo, Japan).

Myeloperoxidase activity

To evaluate the development of the inflammatory state caused by neutrophil infiltration, myeloperoxidase (MPO) was measured as previously described by Filpa et al. (2017a). Briefly, intestinal segments deprived of the mucosal layer were homogenized (50 mg/mL) with a solution of ice cold potassium phosphate buffer (50 mm, pH 6.0) containing 0.5% hexadecyl trimethylammonium bromide (HTAB). After centrifugation (14,000 rpm, 20 min, 4 °C), an aliquot of the supernatant fraction was mixed with HTAB-phosphate buffer containing O-dianisidine dihydrochloride with hydrogen peroxide before spectrophotometrically recording at 460 nm changes in the rate of absorbance. MPO activity was expressed in units (U)/wet tissue weight, where U defined the amount of enzyme that degrades 1 μmol/min of hydrogen peroxide at 25 °C. Experiments were performed six times for each experimental group.

Immunofluorescence

Paraffin-embedded tissue sections and whole-mount preparations were processed for the immuno-localization of OTX1 and OTX2 on small intestine and colon of control and DNBS-treated rats, as previously described in Bistoletti et al. (2019) and Ceccotti et al. (2018).

Paraffin sections

Seven μm paraffin cross sections incubated with CD45, as a marker of inflammatory infiltration, or either with OTX1 or OTX2 with the pan neuronal marker HUC/D, for double staining. Optimal dilutions of antibodies are reported in Table 1. Coverslips were mounted with Citifluor mounting medium and then observed with fluorescent microscopy (Nikon Instruments, Melville, NY, USA).

Whole-mount preparations

Whole-mount preparation immunolabeling was performed on the longitudinal muscle with the attached myenteric plexus (LMMP) proceeding with a double-staining with optimally diluted primary and secondary antibodies (Table 1). Preparations were mounted onto glass slides with a mounting medium (Vectashield with DAPI; Vector Lab, Burlingame, CA, USA) and analyzed with Image J NIH image software (http://imagej.nih.gov/ij). Total neuron number per ganglion area was expressed as the ratio between the number of neurons positive for HuC/D and the total ganglion area (µm²). Areas of 15–19 myenteric ganglia from LMMP preparations of 5 vehicle-treated CTR and 5 DNBS-treated animals were measured at 40× magnification by tracing boundaries around stained cell somas (HuC/D) (Filpa et al., 2017a). Neuronal cell body area was measured with Image J. The number of OTX1 or OTX2 immunoreactive neurons that co-localized with HuC/D were counted from a total of 10–15 ganglia (5 animals for each experimental group). The proportion of OTX1 and OTX2 immunoreactive myenteric neurons, was expressed as percentage of the total number of HuC/D positive neurons (Filpa et al., 2017a). To evaluate negative controls and interference control staining, both primary and secondary antibodies were omitted, incubating the colonic whole-mounts with non-immune serum from the same species in which the primary antibodies were raised. In all these conditions, no specific signal was detected. Pictures from preparations, collected with a Leica TCS SP5 confocal laser scanning system (Leica Microsystems GmbH, Wetzlar, Germany) were then processed with Adobe-Photoshop CS6S software.

Real time quantitative RT-PCR

To evaluate the influence of DNBS-induced colitis on OTX1, OTX2, TNFα, pro-IL1β, IL6, HIF1α, VEGFα, mRNA levels, total RNA from rat small intestine and colon LMMPs was extracted with TRIzol (Invitrogen, Carlsbad, CA, USA) and treated with DNase I (DNase Free, Ambion) to remove possible traces of contaminating DNA. 2.5 μg of total RNA was retrotranscribed using the High Capacity cDNA synthesis kit (Applied Biosystems, Life Technologies, Grand Island, NY, USA). Quantitative RT-PCR was performed on the Abi Prism 7000 real-time thermocycler (Applied Biosystems, Foster City, CA, USA) with Power Sybr Green Universal PCR Master Mix (Applied Biosystems, Foster City, CA, USA) following manufacturer’s instructions. Primers were designed to have a similar amplicon size and similar amplification efficiency as required for the utilization of the 2^−ΔΔCt method to compare gene expression (Bin et al., 2018), using Primer Express software (Applied Biosystems, Foster City, CA, USA) on the basis of available sequences deposited in public database (Table 2). For quantitative RT-PCR a final concentration of 500 nm for each primer was used. Experiments were performed at least 7 times for each experimental group. 2^−ΔΔCt values obtained from the comparison between normalized Ct values of DNBS-treated samples with those obtained from control samples were used to evaluate the effect of DNBS-induced colitis on OTX1 and OTX2 mRNA levels in the small intestine and colon.

Western immunoblot analysis

Purified membrane fractions obtained after successive centrifugations of homogenized intestinal LMMPs preparations were used to analyze OTX1 and OTX2 protein level as described elsewhere (Giaroni et al., 2011). Membrane incubation with OTX1 and OTX2 primary antibodies was performed by incubation with a horseradish peroxidase-conjugated secondary antisera (Table 1). The signal of antibody/substrate complex was visualized by chemiluminescence using an enhanced chemiluminescence kit (ECL advance Amersham Pharmacia Biotech, Cologno Monzese, Italy), and then evaluated by densitometric analysis using Image J NIH image software. The effect of DNBS-induced colitis on OTX1 and OTX2 protein levels was expressed as the percentage variation of the optical density (expressed in arbitrary units) of OTX1 and OTX2 signals normalized to the respective β-actin, used as protein loading control, in LMMP preparations obtained from DNBS-treated animals compared to controls. Experiments were performed at least five times for each experimental group. Specificity of OTX1 and OTX2 primary antibodies was evaluated by testing their selectivity in spontaneous immortalized human Müller cell line (MIO-M1) and in rat hippocampus, respectively (data not shown) (Filpa et al., 2017a). Negative controls were performed by omitting the primary antibody.

Statistical analysis

All data are expressed as mean ± S.E.M. Statistical significance was calculated by either Student’s t test or by one-way ANOVA with Tukey’s post hoc test, where appropriate, using GraphPad Prism (version 5.3; GraphPad Software, San Diego, CA, USA). Differences among groups were considered significant when P value were 0.05 or lower.

Assessment of colitis

Histological assessment

A further set of experiments was carried out on intestinal cross sections to evaluate the impact of DNBS–induced colitis on the architecture of the rat small intestine and distal colon by means of standard HE staining, and the degree of inflammatory infiltrate by CD45 immunofluorescent staining. Small intestine and distal colon cross-sections obtained from control animals showed normal histological features, with compact myenteric ganglia, formed by healthy neuron and glial cells, laying between the circular and longitudinal muscle of the muscularis propria (Figs. 1A, 1B, 1K and 1L). CD45 staining, indicative of inflammatory cell infiltration, was almost negligible in both small intestine and colon control cross sections (Figs. 1C and 1M). At day 6 after DNBS treatment, small intestine specimens did not show prominent histological abnormalities (Figs. 1F and 1G). In contrast, both mucosa and serosal epithelium of the distal colon obtained from DNBS-treated rats, displayed morphological abnormalities (Figs. 1P–1R). The colonic mucosal surface was irregular and crypt architecture was profoundly altered, the muscularis mucosae, submucosa and muscularis propria layers were thickened (Fig. 1P). Prominent spaces between smooth muscle cells and leukocyte infiltration were also observed (Fig. 1Q). In the small intestine of DNBS-treated animals myenteric neurons had a normal morphology comparable to that observed in control preparations as shown Fig. 1G. In contrast, colonic myenteric ganglia underwent important degenerative changes, with neurons displaying cytoplasm vacuolization and irregular nuclear and cellular membrane. Large spaces between muscle cells were also evident (Fig. 1R). After DNBS-treatment, CD45 staining significantly increased both in the small intestine and in the colon (Figs. 1H and 1S).

We then focused our histological investigations on myenteric ganglia by carrying out immunofluorescence staining of small intestine and distal colon LMMP whole-mount preparations with the pan neuronal marker HuC/D. In this set of experiments changes in myenteric neuron number were observed in both gastrointestinal regions after DNBS treatment (Figs. 2A–2F). In particular, the number of myenteric neurons staining for HuC/D was significantly reduced (P < 0.05, by one way-ANOVA with Tukey’s post hoc test) with respect to values obtained in respective control preparations in both regions (Fig. 2E). Myenteric neuron number was reduced to a significantly greater extent (P < 0.05, by one way-ANOVA with Tukey’s post hoc test) in the colon than in the small intestine (Fig. 2E). In addition, in the colon, but not in the small intestine, DNBS-induced inflammation was associated with a significant reduction of myenteric neuron soma area (P < 0.001, by one way-ANOVA with Tukey’s post hoc test) compared to control preparations (Fig. 2F). As expected, in accordance with previous studies, in control preparations, the number of myenteric neurons was significantly lower in the colon (P < 0.05, by one way-ANOVA with Tukey’s post hoc test) than in the small intestine (Fig. 2E). Additionally, in the distal colon the neuronal soma area was significantly larger (P < 0.05, by one way-ANOVA with Tukey’s post hoc test) than in the small intestine (Fig. 2F). Overall, these observations suggest that after DNBS-induced colitis, major histological changes involving myenteric ganglia, occur in the distal colon. However, also the small intestine may be influenced by the injury, as indicated by the increased infiltration of inflammatory CD45⁺ cells and by the reduction of myenteric neuron number.

Biomolecular assessment

The occurrence of an inflammatory response in LMMP preparations, on site and distantly from the injury, was further investigated by evaluating the levels of the mRNA of different cytokines, which are considered standard markers of intestinal inflammation, and by measuring MPO activity. In both intestinal regions, after DNBS treatment, a significant enhancement of the expression of inflammatory cytokines TNFα (P < 0.001 in both regions, by one-way ANOVA with Tukey’s post hoc test), pro-IL1β (P < 0.01 in the ileum and P < 0.001 in the colon, by one way-ANOVA with Tukey’s post hoc test), IL6 (P < 0.001 in both regions by one-way ANOVA with Tukey’s post hoc test), and of HIF-1α (P < 0.05 in both regions, by one-way ANOVA with Tukey’s post hoc test) and VEGFα (P < 0.01 in both regions, by one-way ANOVA with Tukey’s post hoc test), was evidenced with respect to control animals (Figs. 3A–3E). As regards the pro-inflammatory cytokines, enhancement of TNFα mRNA was similar in the colon and in the small intestine, while pro-IL1β and IL-6 mRNA levels were significantly higher in the colon with respect to the small intestine (P < 0.05 and P < 0.001, respectively, by one-way ANOVA with Tukey’s post hoc test) (Figs. 1A–1C). HIF1α and VEGFα mRNA levels after DNBS treatment was not significantly different in the small intestine with respect to the distal colon (Figs. 1D and 1E).

MPO activity significantly increased in rat small intestine (P < 0.001, by one way-ANOVA with Tukey’s post hoc test) and colonic (P < 0.001, by one way-ANOVA with Tukey’s post hoc test) segments compared to control animals, suggesting the occurrence of inflammation-induced neutrophil infiltration on site and distally from the injury (Fig. 1F).

Distribution of OTX1 immunoreactivity in small intestine and colon LMMP whole-mount preparations from normal and DNBS-treated animals

In view of the adaptive changes observed after DNBS-induced inflammation in the rat myenteric plexus, we subsequently evaluated if this adaptation encompasses alterations in the distribution of homeobox proteins involved neuronal plasticity, OTX1 and OTX2, in cross sections and LMMP whole-mounts of the small intestine and distal colon. In cross-sections of small intestine and distal colon obtained from vehicle-treated control rats, myenteric ganglia displayed faint OTX1 immunoreactivity (Figs. 1D and 1N). In control LMMP whole-mount preparations OTX1 antibody stained few myenteric neurons in both the small intestine (3.79 ± 1.00%, n = 15) and distal colon (4.77 ± 1.43%, n = 15) (Fig. 4A). In both intestinal regions, OTX1 antibody stained the soma and nucleus of large and medium size myenteric neurons with either a round or an ovoidal shape (Figs. 5A–5C and 5G–5I). In small intestine and distal colon cross-sections obtained from DNBS-treated animals, a significant increase of OTX1 staining was observed within the myenteric plexus with respect to control preparations (Figs. 1I and 1T). After DNBS treatment, a significant increase in the number of OTX1 immunoreactive myenteric neurons was observed in whole-mount preparations of both small intestine (9.86 ± 0.86%, n = 15, P < 0.0001, by one-way ANOVA with Tukey’s post hoc test) and colon (40.25 ± 1.2%, n = 15 P < 0.0001, by one-way ANOVA with Tukey’s post hoc test) with respect to control preparations (Figs. 4A, 5D–5F and 5J–5L). In DNBS-treated colonic whole mount preparations, the percentage of OTX1 immunoreactive myenteric neurons was significantly higher than in the small intestine (Fig. 4A). In LMMP whole-mount preparations of both intestinal regions, OTX1 antibody co-localized with the glial marker S100β, and co-staining was particularly intense after DNBS-induced inflammation (Figs. 5M–5R). In small intestine and colonic LMMPs obtained from control and DNBS-treated animals, myenteric neurons staining for OTX1 were also immunopositive for iNOS, as shown in Figs. 6A–6L. Both in control and DNBS-treated LMMP whole-mount preparations, OTX1 immunostaining did not co-localize with nNOS (Figs. 6M–6O). These results suggest that OTX1 is expressed both in myenteric neurons, where it is highly co-expressed with iNOS, and in enteric glial cells of the rat small intestine and distal colon. After DNBS-induced colitis the number of myenteric neurons expressing OTX1 is highly upregulated both on site and distantly.

Distribution of OTX2 immunoreactivity in small intestine and colon LMMP whole-mount preparations from normal and DNBS-treated animals

In control sections of the rat small intestine and colon, faint OTX2 staining was found in myenteric ganglia (Figs. 1E and 1O). In LMMPs whole-mount preparations obtained from control animals OTX2 specific antibody stained the soma of few myenteric neurons both in the small intestine and in the distal colon (Figs. 7A–7C and 7G–7I, respectively). Some neurons had a large cell body, while other smaller neurons with an ovoidal shape were, prevalently, localized in the periphery of the ganglion. The percentage of OTX2-immunoreactive neurons in control small intestine and distal colon myenteric plexus was 7.07 ± 1.68%, n = 15 and 8.93 ± 2.11%, n = 15, respectively (Fig. 4B). In small intestine and colon cross-sections obtained from DNBS-treated animals, OTX2 staining significantly increased in myenteric ganglia with respect to control preparations (Figs. 1J and 1U). Accordingly, after DNBS treatment, the percentage of OTX2 immunoreactive myenteric neurons in LMMP whole-mount preparations obtained from the small intestine and colon significantly increased (47.95 ± 3.20%, n = 15; 41.47 ± 5.12%, n = 15, respectively P < 0.001, by one-way ANOVA with Tukey’s post hoc test) with respect to control values (Figs. 4B, 7D–7F and 7J–7L). The number of OTX2 immunoreactive neurons was not significantly different in the two regions (Fig. 4B). In small intestine and colonic myenteric ganglion obtained from control and DNBS-treated animals, the majority of OTX2 immunoreactive neurons stained for nNOS (Figs. 8A–8L). There was no evidence of co-staining between OTX2 and iNOS, as shown in Figs. 8M–8O. In small intestine and distal colon LMMPs obtained from control animals 31.94 ± 3.33%, n = 10 and 30.83 ± 10.83%, n = 10, respectively, of OTX2-IR myenteric neurons stained also for OTX1. In both regions, after DNBS treatment, 25.82 ± 5.78%, n = 10 and 25.63 ± 2.17% n = 10, respectively, of OTX2-IR were also immunoreactive for OTX1. These findings suggest that OTX2 is exclusively expressed in myenteric neurons and displays a high co-localization with nNOS. Experimentally-induced colitis upregulates the number of myenteric neurons expressing OTX2 both on site and distantly.

DNBS-induced colitis upregulates OTX1 mRNA and protein in LMMPs of the rat small intestine and distal colon

In order to confirm the immunofluorescence data reflecting an up-regulation of OTX1 and OTX2 in the rat small intestine and colon myenteric plexus after DNBS-induced colitis, we evaluated the levels of mRNA and proteins of both homeobox proteins in LMMP preparations. Rat small intestine and colonic OTX1 mRNA levels obtained in the different experimental groups are described in Fig. 9A. In LMMPs of both regions, DNBS treatment induced a significant increase of OTX1 mRNA with respect to values obtained in control preparations (small intestine: P < 0.05; distal colon: P < 0.001, by one-way ANOVA with Tukey’s post hoc test). OTX1 mRNA enhancement in colonic specimens was significantly higher than in the small intestine (P < 0.001, by one-way ANOVA with Tukey’s post hoc test). In rat small intestine and colonic LMMP preparations the specific OTX1 antibody revealed one band at 37 kDa (Fig. 9B). In both gut regions, after DNBS treatment, OTX1 protein expression significantly increased (small intestine: P < 0.05; distal colon: P < 0.01, by one-way ANOVA with Tukey’s post hoc test). These biomolecular data confirm that an experimentally-induced colitis may up-regulate both the transcript and protein of OTX1 in the myenteric plexus both on site and distantly from the injury.

DNBS-induced colitis upregulates OTX2 mRNA and protein in LMMPs of the rat small intestine and distal colon

OTX2 mRNA levels measured in rat small intestine and colonic LMMPs preparations obtained from control and DNBS-treated animals are described in Fig. 9C. In both regions, DNBS treatment induced a significant increase of OTX2 mRNA with respect to values obtained in control preparations (small intestine: P < 0.05; distal colon: P < 0.001, by one-way ANOVA with Tukey’s post hoc test). OTX2 mRNA levels in small intestine specimens were not significantly different than those obtained in the distal colon. In rat small intestine and colon LMMP preparations the specific OTX2 antibody revealed one band at 32 kDa (Fig. 9D). In both gut regions, OTX2 protein expression significantly increased (P < 0.01, by one-way ANOVA with Tukey’s post hoc test) after DNBS treatment. After DNBS-induced colitis, analogously to OTX1, up-regulation of OTX2 in the rat myenteric plexus was evidenced also for the transcript and protein, by means of qRT-PCR and western immunoblotting.