Pentoxifylline decreases post-operative intra-abdominal adhesion formation in an animal model

Animals

Male BALB/c mice weighing 25–30 g (Charles River Laboratories, BioLasco, Taiwan) were maintained in a temperature and light-controlled room (12-hour light/dark cycle) and allowed free access to water and food. The experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUCs) of the National Taiwan University Hospital (approval ID: 20120285). All the protocols are in adherence to the guidelines established in the Guide for the Care and Use of Laboratory Animals of the National Health Research Institutes.

Experimental design

Seventy BALB/c mice were randomly divided into three different groups: (1) sham group without peritoneal adhesion model (sham group, n = 8); (2) peritoneal adhesion model (PA group, n = 31); (3) peritoneal adhesion model with PTX (Trental) (100 mg/kg/day i.p. once daily) administration was started on preoperative day 2 and continued daily (PA + PTX group, n = 31). We used a slightly modified standard adhesion model (Oncel et al., 2001). The details for peritoneal adhesion model had been reported before (Lee et al., 2016).

Briefly, mice were anesthetized using 2% isoflurane in oxygen. The abdomen was then shaved and disinfected with povidone iodine. A 4 cm median laparotomy was performed to gain access to the abdominal cavity. In the peritoneal adhesion model (PA and PA + PTX group), mice were pooled and randomly underwent surgery. The cecum was gently externalized and abraded with 20 vertically reciprocal movements of dry gauze. The right abdominal sidewall was rubbed more aggressively than the cecum until punctate bleeding was observed. The injury sites were cleaned with physiological salt solution and covered the gauze, making sure that there was no active bleeding. The cecum was then placed back into the abdominal cavity and the surgical wound was sutured. For the sham group, only open laparotomy and closure was conducted and there is no abrading of cecum and abdominal wall.

The PTX group received 100 mg/kg of PTX from the left abdominal cavity, whereas the other two groups (group sham and PA) received 0.125 ml of physiological saline solution. For preventing postoperative pain, buprenorphine (0.05 mg/kg s.c., twice daily) was administered during the two postoperative days. Mice were placed under a warming lamp and observed until they recovered fully from anesthesia. Mice were monitored daily for signs of wound infection and general health condition periodically until three or seven days after surgery.

Adhesion score

Mice were euthanized on postoperative day 3 and day 7. The abdominal cavity was opened via a U-shaped incision. The adhesion score was evaluated and was performed by an observer blinded treatment, using the Lauder scoring system (Lauder et al., 2011). The adhesions were graded in a blinded fashion using the classification system described (Table 1).

Histology staining

Tissue samples from the parietal peritoneum, liver, and mesentery were collected after euthanasia. For histological staining, tissues were fixed in 10% neutral-buffered formalin (NBF), paraffin embedded, thinly sectioned. Tissue sections of 4–5 µm thickness were prepared for staining. After deparaffinization and rehydration, sections were counterstained in Gill’s hematoxylin (Sigma, St Louis, MO, USA) and for 5 min, cleared in 0.1% acid alcohol for 30 s, and rinsed in tap water, then stained in eosin (Sigma) for 2 min, cleared in 95% alcohol, and rinsed in 70% alcohol to remove the staining solution, dehydrated, and mounted for histologic assessment.

Picrosirius red staining

Picrosirius red staining was used to compare collagen and fibrosis in tissues between different groups. Peritoneal sections (4–5 µm) were deparaffinized, rehydrated and then subjection to counterstaining in Gill’s hematoxylin (Sigma) for 5 min, cleared in 0.1% acid alcohol for 30 s, and rinsed in running tap water. Then they were stained in Picrosirius Red Stain kit (Polysciences, Inc., Warrington, PA, USA). Subsequently, sections were dehydrated and mounted for assessment.

Tissue plasminogen activator

Peritoneal tissues were prepared by grinding on ice in radioimmunoprecipitation assay buffer (RIPA buffer) with protease inhibitor cocktail (Sigma). After the samples were centrifuged at 12,000 g for 15 min at 4 °C, supernatants were aspirated and placed in new tubes. Samples were analyzed for total antigen concentration of tPA, using commercially available ELISA kits from Molecular Innovations (Molecular Innovations, Novi, MI, USA). Total protein content was determined by Bradford assay (Sigma).

Immunohistochemistry

Formalin-fixed and paraffin-embedded peritoneal tissue was sectioned at 4–5 µm and then subjected to double immunostaining. Briefly, sections were deparaffinized, rehydrated and endogenous peroxidase activity was quenched by 3% hydrogen peroxide (H₂O₂) for 10 min. Sections were subjected to antigen retrieval performed in pH 6.0 citrate buffer using a microwave oven for 15 min. Blocking of non-specific binding was done by incubation with 2.5% horse serum at room temperature for 30 min. Sections were incubated with primary antibodies, rabbit anti-ki67 (1:200, Abcam, Cambridge, MA, USA), rat anti-F4/80 (1:200, Abcam) or rabbit anti-FSP-1 (1:200, Abcam) overnight at 4 °C. After washing with Tris-buffered saline (TBS, pH7.4), sections were incubated using the ImmPRESS AP anti-rabbit polymer reagent (Vector Laboratories, Burlingame, CA, USA) for 30 min at room temperature. Positive signals resulted in blue nuclear staining with the VECTOR Blue kit (Vector Laboratories). After washing and blocking again, the sections were incubated with goat anti-CD31 (1:500, R&D System, Minneapolis, MN, USA) for 1 h at room temperature. After washing, ImmPRESS HRP anti-goat polymer reagent (Vector Laboratories) was used for 30 min at room temperature. Positive reactions for endothelial cells resulted in brown red staining with the NovaRed substrate kit (Vector Laboratories). Sections were examined by light microscopy (Nikon Instruments, Nikon Corporation, Tokyo, Japan).

Immunofluorescence staining

Peritoneal tissue sections (4–5 µm) were performed for double immunofluorescence staining. Briefly, sections were deparaffinized, rehydrated and were treated with 0.3% H₂O₂ for 10 min to block endogenous peroxide activity and boiled in pH 6.0 citrate buffer using a microwave oven for 15 min. Sections were subsequently incubated with 5% Donkey serum for 20 min at room temperature. Sections were incubated with rabbit anti-cytokeratin 18 (CK18) (1:200, Enogene, New York, NY, USA) overnight at 4 °C, washed in phosphate-buffered saline (PBS, pH7.4) and incubated using donkey anti-rabbit DyLight 488 antibody (Thermo Scientific, Rockford, IL, USA) and Cy3-conjugated mouse anti-α-smooth muscle actin (α-SMA) (Sigma) and then mounted and subjected to fluorescence microscopy (Leica DMRA, Leica Microsystems, Wetzlar, Germany). Images were recorded at x100, x200 and x400 magnification of light microscopy, and were then digitalized and analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA).

Statistical analysis

Normal distributed continuous data were expressed as mean ± standard error (SE). For non-parametric data, results were expressed as median ± interquartile range (IQR). The difference between continuous variables were evaluated using one-way ANOVA when data distribution was normal and a Mann–Whitney test was used for non-normal distributed continuous data. A p value of less than 0.05 was considered statistically significant. The statistical analyses were performed with GraphPad Prism (version 6.0, GraphPad Software, Inc., La Jolla, CA, USA).

Deaths of animal

Surgical procedures were successfully completed on 69 animals, except for one mouse in the sham group, which died due to anesthesia-related complications before the commencement of surgery. One mouse in the PA + PTX group died during recovery from anesthesia. Four mice died within 48 h of surgery and were excluded from the study (PA and PA + PTX group, n = 2/group). Two mice in the PA group were excluded due to severe distress, according to three criteria in Health Evaluation of Experimental Laboratory Mice: very rough hair coat, hunched, and not eating or drinking. No mice in the PA + PTX treated group incurred any life-threatening side effects or deaths at 48 h after surgery, which would have lead to exclusion from the study. Therefore, a total of eight animals were excluded from the study.

Mice were euthanatized for the planned experiments on postoperative day 3 (PA group, n = 14; PA + PTX group, n = 13) and day 7 (sham group, n = 7; PA group, n = 13; PA + PTX group, n = 15).

Pentoxifylline treatment reduces adhesion score

Total adhesion scores data was examined and plotted for post-operative day 3 and 7 (Fig. 1). We observed that the sham group, which had not undergone the adhesion model, had a significantly lower adhesion score than the animals that had undergone adhesion model, as expected. The PA + PTX group (median, 1.00; IQR, 0.50–2.00) had a significantly lower adhesion score than the PA group (median, 2.00; IQR, 2.00–3.00) on day 3 (p < 0.05). On day 7, mice treated with PTX (median, 3.00; IQR, 0.00–3.00) still had lower adhesion scores than the PA group (median, 3.00; IQR, 3.00–4.50) (p < 0.05).

Pentoxifylline treatment inhibits collagen deposition

We used the HE staining to compare changes in peritoneal structure (Fig. 2A). In general, there was increased thickness of the submesothelial layer on day 3, and the adhesion score also increased. We observed that the sham group had the thin submesothelial layer as demonstrated in Fig. 2A. The severe adhesion and thick submesothelial layer were observed, as well as the increased cellularity in the PA group. In contrast, the PA + PTX group had less peritoneal submesothelial thickness and adhesion severity as compared with the PA group.

We further used Picrosirius red staining to assess the quality of collagen fiber in peritoneal adhesion (Fig. 2B). Compared with the PA group, the PA + PTX group had less collagen deposition and thickness of the abdominal adhesions. Our data suggested that PTX could decrease collagen deposition during adhesion formation.

Pentoxifylline treatment increased tPA level

The tPA protein levels in the peritoneum of mice were perceived to be measured and plotted on post-operative day 3 and 7 (Fig. 3). We observed that the sham group of mice had the lowest tPA protein level throughout the study period (p < 0.001). Those mice treated with PTX had higher tPA protein levels than untreated mice (the sham and PA groups). There was significant difference between the PA + PTX group and the PA group on day 7 (0.365 ± 0.024 vs. 0.193 ± 0.03, p < 0.001).

Pentoxifylline treatment reduced angiogenesis

We performed immunohistochemical staining assay to analyze the status of angiogenesis during peritoneal repair. Proliferating endothelial cells were identified as those cells with cytoplasmic CD31 staining and nuclear Ki67 staining (Fig. 4A, arrows). Cells that stained positive for CD31 but without nuclear staining for Ki67 were scored as nonproliferating endothelial cells. We examined the effect of endothelial cell proliferation in peritoneum was quantified by measuring the number of ki67⁺ CD31⁺ cell at the site of adhesion (Fig. 4B). Our results showed that the PA group had significantly higher proliferating endothelial cells compared to both the sham and PA + PTX groups on day 3 and day 7 (all p < 0.001). We also observed that the ki67⁺ CD31⁺ proliferating endothelial cell count decreased substantially over time (p < 0.001) in the PA group. Also, we examined the angiogenic effect in peritoneum by measuring the area of CD31⁺ microvessel at the site of adhesion (Fig. 4C). We also observed that the PA group had a vessel coverage (the percentage of area covered by CD31⁺ per field) of 1.63% at day 3, and 3.63% at day 7, respectively. On post-operative day 7, the PA group demonstrated an increase in CD31 expression compared to day 3 and indicated that blood vessel formation was significantly more prominent in the PA group compared to the PA + PTX groups (p < 0.001). Thus, PTX can significantly suppress angiogenesis during peritoneal repair.

Pentoxifylline treatment reduced inflammation

Inflammation, an important component both in normal and pathological healing, is a protective response to tissue injury, designed for removal of the causative agent and restoration of tissue structure and function. We performed immunohistochemical staining assay to analyze the infiltration of macrophage during peritoneal repair. F4/80 is a macrophage-specific marker in mice. As shown in Fig. 5A, a large number of F4/80⁺-expressed cells were observed in the PA group. Quantification analysis of the IHC image revealed significantly increased expression of F4/80⁺ cells on day 3 (0.80% ± 0.10%), and highest expression on day 7 (2.56% ± 0.22%) in the PA group as compared to the sham group (p < 0.01 for both day 3 and day 7) or the PA + PTX group (1.37% ± 0.28%, p < 0.01 on day 7, Fig. 5B). The PA + PTX group had significantly lower expression of F4/80⁺ as compared with the PA group.

Pentoxifylline treatment reduced the expression of fibrosis marker FSP-1

FSP-1, also known as fibroblast-specific protein 1 (FSP1), belongs to the S100 superfamily of cytoplasmic calcium-binding proteins and can be expressed by cell of mesenchymal origin or fibroblastic phenotype. This protein is reported to be specific for fibroblasts and plays a causal role in EMT. As shown in Figs. 6A and 6B, a large number of FSP-1⁺-expressed cells were observed in the PA group. Quantification analysis of the IHC image revealed significantly increased expression of FSP-1⁺ on day 3 (5.37% ± 1.03%), and highest expression on day 7 (11.26% ± 1.66%) in the PA group compared to the PA + PTX group (p < 0.05 for both day 3 and day 7, Fig. 6C). Consistently, we found the mice treated with PTX had significantly reduced expression of FSP⁺ as compared with the PA group.

Pentoxifylline treatment reduced the expression of fibrosis marker α-SMA

As a response to injury, mesothelium undergoes a change called mesothelial-to-mesenchymal transition (MMT). Thus, we further performed double immunofluorescence staining for CK18 and α-SMA for peritoneal injury (Figs. 6A and 6B). Many studies have demonstrated that mesenchymal cell markers, including α-SMA, are proposed as indicators of EMT (Margetts et al., 2005). Cytokeratin (CK) are structural marker proteins specific for epithelial cells, and CK18 is highly expressed in mesothelial cells. α-SMA has become the most reliable marker of myofibroblasts. Figure 7A is consistent with previous studies, where the PA group observed a few CK18⁺α-SMA⁺ double-positive cells appear first in the mesothelial monolayer and later in the reorganized submesothelial matrix. We examined the extent of accumulation of myofibroblasts in peritoneum was quantified by assessing the percentage α-SMA⁺ cells (Fig. 7C). Our result showed that the PA group had significantly increased α-SMA⁺ expression at day 3 (3.48% ± 1.28%), and highest expression at day 7 (13.71% ± 1.40%) compared with the PA + PTX group (p < 0.05 for day 3 and p < 0.01 for day 7, respectively). PTX significantly attenuated thickening of fibrotic peritoneum, and accumulation of α SMA⁺ myofibroblasts in peritoneum after injury.