INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ENDOTHELIN-1 (ET-1) is an endothelium-derived vasoconstrictor (9, 32) and consists of 21 amino acid residues with two intramolecular disulfide bonds. Ulcerative colitis and Crohn's disease are the two major disease entities of chronic relapsing inflammatory bowel disease. High ET-1 immunoreactivity has been shown in lamina propria and submucosa of patients with Crohn's disease and ulcerative colitis (19), and it has been suggested that local endothelin production by inflammatory cells may contribute to vasculitis in chronic inflammatory bowel disease by inducing intestinal ischemia through vasoconstriction (19). ET-1 also induces impairment of mucosal microvascular perfusion through activation of ET_A receptors, causing significant tissue injury in the rat small intestine (17). The precise immunopathological events occurring in appendicitis are not completely known, and many appendixes removed for suspected appendicitis are subsequently classified as normal by conventional histological staining (30). There is an uncommon, enigmatic chronic appendicitis that shares histological features with typical Crohn's disease, but it presents as appendiceal disease. This unusual Crohn's diseaselike appendicitis has been called "Crohn's disease of the appendix" or "granulomatous appendicitis" (5, 10).

Local ET production by inflammatory cells may therefore be implicated in pathogenesis of appendicitis. In this study, we investigated the presence, distribution, and phenotype of ET-1⁺ cells and prepro-ET-1 (pp-ET-1) mRNA-expressing cells in human appendixes, using immunohistochemistry (IHC) and a nonradioactive in situ hybridization (ISH) method, respectively.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Patients and samples. Thirty-three appendixes removed surgically from patients with a clinical diagnosis of acute appendicitis were studied. Twenty-six were subsequently classified as histologically normal, and seven were diagnosed as acute appendicitis. All studied subjects, according to their clinical history, were not suffering of peritonitis, hypertension, pulmonary hypertension, cirrhosis, renal failure, acute myocardial infarction, peptic ulcer, diabetes mellitus, or atherosclerosis. Three normal appendixes, removed from patients with colon cancer, served as control. Appendixes were divided into at least three segments from tip to base, fixed in Bouin's or formalin solution, and embedded in paraffin. Tissue sections (6-7 µm) were deparaffinized and washed with 0.1 M PBS (pH 7.4). ECV304, a human endothelial cell line, was used as a positive control. For ISH, the cells were harvested, washed in 0.1 M PBS, fixed in 4% paraformaldehyde for 30 min, washed in PBS, mixed with molten agar (2.5% in distilled water), and allowed to solidify. The cell bag was then routinely processed to produce a cell block from which paraffin sections were cut. For IHC, the cells, grown on chamber slides (Nalge Nunc, Naperville, IL), were fixed in 4% paraformaldehyde/5% acetic acid and permeabilized with 0.2% Triton X-100. Histological examination of hematoxylin and eosin stained appendix sections was carried out by the same histopathologist.

Probe preparation. The ET-1 probe was prepared according to the procedure described by Klein et al. (13), with minor modifications. Total RNA was extracted from 6 × 10⁶ ECV304 (27), using "RNAeasy mini kit" (Quiagen, Hilden, Germany). Its integrity was checked on 1.6% (wt/vol) agarose gel, and the concentration was determined spectrophotometrically. To obtain cDNA, 5 µg of the extracted RNA were used as a template in the RT reaction, which was performed at 37°C for 30 min in a final volume of 25 µl and the ET-1 antisense specific primer 5'-GCT CTC TGG AGG GCT TGC-3'. Then, the cDNA of ET-1 was amplified in a PCR, using 200 ng of the sense primer 5'-CAG TTT GAA CGG GAG GTT TTT-3' and of the antisense primer used for the RT reaction. Amplification was performed in a thermal cycler (PCR Sprint, Hybaid, UK) under the following conditions: 94°C for 5 min, then 30 cycles at 93°C for 30 s, 56°C for 45 s, 72°C for 1 min, and finally 72°C for 5 min. The 645-bp ET-1 amplicon obtained was then labeled with digoxigenin (Dig) in a repeated PCR with the same ET-1 primers under the same conditions as before, apart from the addition of 70 µM Dig-labeled dUTP (Dig-11-dUTP, Boehringer Mannheim, Mannheim, Germany).

ISH. The appendix sections were dewaxed, rehydrated, and fixed in 4% paraformaldehyde in PBS pH 7.4 for 20 min and processed for hybridization as previously described (13). The sections were subsequently incubated with Fab fragments from sheep anti-Dig-alkaline phosphatase (APase) antibodies (150 U/200 µl; Boehringer) diluted 1:100 in PBS-BSA for 1 h at 37°C. APase color development was performed with Red (Vector), and endogen APase was inhibited by the addition of levamisole (Vector) in the substrate solution. Instead, the ECV304 slides were digested with 0.1% pepsin (Sigma, St. Louis, MO) in 0.01 M HCl at 37°C for 10 min. Hybrids were detected by application of, in sequence, mouse anti-Dig antibodies (2 µg/ml; Boehringer), peroxidase (PO)-conjugated rabbit anti-mouse immunoglobulins (13 µg/ml; DAKO, Milan, Italy), and PO-conjugated swine anti-rabbit immunoglobulins (26 µg/ml; DAKO). PO activity was revealed by development in 3-amino-9-ethylcarbazole (AEC; DAKO).

IHC. The EnVision⁺ peroxidase detection system was used in an immunoenzymatic technique (EnVision⁺ labeled polymer, DAKO). The sections were incubated with 6% H₂O₂ in H₂O for 5 min followed by 3% PBS-BSA and then incubated overnight with appropriate primary monoclonal antibodies (at working dilutions listed in Table 1). Immunoreactivity was visualized with AEC or 3,3'-diaminobenzidine (DAB; DAKO).

Combined ISH and IHC. To define the immunophenotype of pp-ET-1 mRNA-expressing cells, we first used ISH combined with IHC in the same section of appendix, using APase and PO as reporter enzymes. Because all of the primary antibodies listed in Table 1 reacted with a cytoplasmic antigen, it was difficult to detect the ET-1 mRNA signal and the antigen specific for immunophenotyping simultaneously in the same cell, even using contrasting colors. After ISH procedure, pp-ET-1 mRNA-expressing cells were counted and photographed, and then the chromogen was removed with acetone without dehydration. For subsequent IHC, lactoferrin and CD68 antigen were detected by goat anti-mouse immunoglobulins (EnVision⁺ peroxidase). The tryptase antigen was detected with mouse anti-human antibody conjugated with APse (Chemicon, Temecula, CA). In this case, APase activity still present after the APase-Red reaction for ISH was inactivated by incubating the slides in 0.01 N HCl for 10 min at room temperature. The photographs obtained were printed on transparent plastic sheets, and the area with pp-ET-1 mRNA⁺ cells was superimposed on the areas immunostained for polymorphonuclear neutrophils (PMNs), mast cells (MCs), and macrophages (MØs).

ISH and IHC controls. Four negative control procedures were performed to assess the specificity of the ISH signal. 1) Before ISH, tissue sections were treated overnight with 300 µg/ml ribonuclease A (Boehringer) and 300 IU ribonuclease T₁ (Boehringer) in Tris · HCl at 37°C. Tissue sections were treated; 2) omitting Dig-labeled probe; 3) omitting anti-Dig-APase antibodies; and 4) omitting Dig-labeled probe and anti-Dig-APase antibodies. After ISH, controls for IHC (in the same section) included omission of primary antibody and/or secondary antibody. In addition, ECV304, a human endothelial cell line, was used as a positive control.

Quantification of stained cells. The entire cross section of appendix was examined to evaluate the localization and relative frequencies of stained cells in the different compartments of the organ. All positive stained cells were counted (objective ×40) by two investigators who were unaware of the clinical groups. Results were presented as the number of stained cells per section, which were estimated semiquantitatively (+ = 1-5 cells, ++ = 6-20 cells, +++ = 21-50 cells, ++++ = 51-100 cells, +++++ = >100). The average tissue area of sections, measured by a computerized image analyzer (National Institutes of Health Image), was 2.5 × 10⁷ ± 8% µm².

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Immunohistochemical localization of ET-1+ cells in appendix. Figure 1D' shows the positive control, a human endothelial cell line, for immunocytochemical detection of ET-1. Blood and lymphatic endothelial cells, epithelial cells of the crypts and lining of the appendix lumen, lymphoid follicles, muscularis mucosae and the two muscle layers, and myenteric and submucous plexuses did not immunostain. On the contrary, stromal cells with immunocytochemical staining of mature ET-1 in the cytoplasm were detected (Fig. 1), but their distributions and relative frequencies differed in the three clinical groups (Table 2).

View larger version (94K): [in this window] [in a new window]   Fig. 1.   Immunohistochemical localization of endothelin-1 (ET-1)+ cells in noncounterstained cross sections of appendix mucosa. A': normal appendix from patient with colon cancer; B': histologically normal appendix from patient with clinical diagnosis of appendicitis; C': histologically inflamed appendix; A, B, C: specular cross sections treated omitting primary antibody; C'': submucosa, muscle layers, and serosa of section C'; C''': submucosa with several ET-1+ cells in blood vessels. Substrate: 3-amino-9-ethylcarbazole. Original magnification: ×200. D, D': immunocytochemical detection of ET-1 in human endothelial cell line ECV304, used as positive control. D: cells treated omitting primary antibody and counterstained with hematoxylin; D': ET-1+ cells; substrate: 3,3'-diaminobenzidine. Original magnification: ×1,200.

Localization of pp-ET-1 mRNA-expressing cells in appendix by ISH. To determine whether these ET-1⁺ stromal cells produced ET-1 or took it up from the surrounding tissue, we used ISH to detect pp-ET-1 mRNA-expressing cells in the same three groups of appendixes. The intrinsic labeling of pp-ET-1-specific PCR product with Dig-11-dUTP was evaluated by agarose electrophoresis (Fig. 2). The distributions and relative frequencies of pp-ET-1 mRNA-expressing cells were similar to those of mature peptide (Fig. 3 and Table 2). Figure 3F shows the positive control for ISH.

View larger version (41K): [in this window] [in a new window]   Fig. 2.   Repeated PCR with 645-bp prepro-ET-1 (pp-ET-1) PCR product in absence and presence of digoxigenin (Dig)-11-dUTP. Because Dig-labeled UTP counts for 2 nucleotides, the molecular weight of the Dig-labeled PCR fragment is higher. The marker contains 0.39 µg of 100 bp DNA ladder (Promega).

View larger version (122K): [in this window] [in a new window]   Fig. 3.   Localization by in situ hybridization (ISH) of pp-ET-1 mRNA-expressing cells in noncounterstained cross sections. Hybridization was performed on histologically normal appendixes (same subgroup as Fig. 1B') from patients with clinical diagnosis of appendicitis. A: pp-ET-1 mRNA-expressing cells in mucosa; A': muscle layers. Four negative control procedures were performed to assess specificity of ISH signal. Tissue sections were treated: omitting Dig-labeled probe (B); with RNase (C); omitting anti-Dig-APase antibodies (D); omitting Dig-labeled probe and anti-Dig-APase antibodies (E). Substrate: Red. Original magnification: ×200. F, G: ISH on ECV304, a human endothelial cell line used as positive control. pp-ET-1 mRNA-expressing ECV304 (F). Negative control: ECV304 were treated omitting Dig-labeled probe (G). Substrate: AEC. Original magnification: ×1,200.

Phenotypic characterization of pp-ET-1 mRNA-expressing cells. Because PMN infiltration and increased mononuclear cell numbers are a feature of acute appendicitis, and because MØs (6), enteric MCs (15), and neurons (11) are reported to express ET-1, we evaluated the distribution of these cell types using the monoclonal antibodies listed in Table 1. Figure 4 shows the mucosal distribution (between the base of the crypts and the muscularis mucosae) of PMNs (Fig. 4A), MØs (Fig. 4B), and MCs (Fig. 4C) in serial sections of the same specimen from one of the subset of six histologically normal appendixes. Moreover, this subset of appendixes showed more PMNs, MØs, and MCs compared with 20 histologically normal appendixes, with similar distribution to ET-1⁺ cells reported in Table 2.

View larger version (90K): [in this window] [in a new window]   Fig. 4.   Mucosal distribution of neutrophils, macrophages, and mast cells. All images have been obtained in noncounterstained serial cross sections of the same specimen from a histologically normal appendix. Positive cells for lactoferrin (A; neutrophils, substrate: AEC); CD68 (B; macrophages, substrate: AEC); and tryptase (C; mast cells, substrate: Red). D: negative control cross section treated omitting primary antibody. Original magnification: ×200.

View larger version (124K): [in this window] [in a new window]   Fig. 5.   Phenotypic characterization of pp-ET-1 mRNA-expressing cells. All images have been obtained in noncounterstained cross sections of a histologically normal appendix (same subgroup as Figs. 1B' and 3). A: mucosal distribution of pp-ET-1 mRNA-expressing cells. After removal of chromogen (Red), the same section was treated to detect mast cells (B) by immunohistochemistry (IHC), with Red as substrate. Arrows indicate positive cells to combined in situ hybridization-IHC with alkaline phosphatase reporter enzyme (see MATERIALS AND METHODS). Original magnification: ×200.

View larger version (144K): [in this window] [in a new window]   Fig. 6.   Phenotypic characterization of pp-ET-1 mRNA-expressing cells. All images have been obtained in noncounterstained cross sections of a histologically normal appendix (same subgroup as Figs. 1B' and 3). A: mucosal distribution of pp-ET-1 mRNA-expressing cells. After removal of chromogen (Red), the same section was treated to detect neutrophils (B) by IHC, with AEC as substrate. Arrows indicate positive cells to combined in situ hybridization-IHC with alkaline phosphatase and peroxidase reporter enzymes (see MATERIALS AND METHODS). Original magnification: ×200.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In this study, we have shown the presence of ET-1-positive stromal cells and pp-ET-1 mRNA-expressing cells and demonstrated their different distributions and relative frequencies in normal control appendix, histologically normal appendix, and inflamed appendix. In particular, a subgroup of histologically normal appendixes from patients with a clinical diagnosis of acute appendicitis had an increased number of ET-1-positive stromal cells and elevated pp-ET-1 mRNA expression, similar to the inflamed appendix. However, in this subgroup of histologically normal appendixes, there were more immunoreactive cells in the lamina propria than in the submucosa and muscularis, whereas the converse was true of inflamed appendix (more immunoreactive cells in the muscularis and submucosa than in the lamina propria). The lower incidence of positive cells in the mucosa of inflamed appendixes may be attributed to partial or complete destruction of the lamina propria by the inflammatory process.

Furthermore, our findings show clearly that the stromal pp-ET-1 mRNA-expressing cells are PMNs (40%) and MCs (25%). The ability of MCs and MØs but not of PMNs to synthesize and secrete ET-1 is well known (6, 15). Neutrophils are currently only regarded as being involved in the cleavage of exogenous big ET-1 to the biologically active peptide ET-1 (12, 26) and in the degradation of endothelin through release of proteases (12, 24). Identification of appendix PMNs as ET-1-producing cells is in line with our recent results that human activated PMNs, isolated from venous blood of normal donors, expressed pp-ET-1 mRNA and secreted appreciable levels of the mature ET-1 in the presence of lipopolysaccharide (2).

ET-1 immunoreactivity was not detected in the endothelium of blood and lymphatic vessels using two different primary monoclonal antibodies. This result is coherent with previous data in human colon (11) and probably reflects the tissue-processing method used. In fact, the human endothelial cell line, grown on chamber slides and fixed in paraformaldehyde/acetic acid, showed immunopositivity for ET-1 (Fig. 1D'). Besides, the same cells, fixed in paraformaldehyde, mixed with molten agar, and then embedded in paraffin express pp-ET-1 mRNA (Fig. 3F). Nevertheless, our processing method with Bouin's solution was optimal for demonstration of ET-1 immunoreactivity in inflammatory cells of the appendix. However, we cannot entirely exclude genetic predisposition and influences of appendix microenvironment. Indeed, regional differences in vascular endothelium structure, function, antigenicity, receptors, mRNA, and protein expression have been described (1, 23).

The results of the present study raise the question of the meaning of local ET-1 production by inflammatory cells and its possible role in the pathogenesis of appendicitis. Adeguate microvascular blood flow in the mucosa is considered essential to maintain mucosal integrity, and this blood flow is reduced during sepsis (7, 29) or hemorrage, leading to tissue damage. Elevated plasma levels of ET-1 in clinical septic conditions (25) suggest that ET-1 may play a role in the pathogenesis of various microvascular disorders (31) and intestinal mucosal injury. In fact, local intra-arterial infusion of low doses of ET-1 induces extensive ulceration and hemorrhagic damage of the gastrointestinal mucosa (18). Furthermore, increased ET-1 production and its receptor, ET_A, are involved in the pathogenesis of endotoxin-induced intestinal mucosal damage (18) and high ET-1 immunoreactivity has been shown in lamina propria and submucosa of patients with Crohn's disease and ulcerative colitis (19).

During tissue ischemia and reperfusion (I/R), granulocyte infiltration and mucosal dysfunction are coexisting phenomena, and the role of neutrophils in microvascular injury has been described (8). Circulating PMN depletion and prevention of PMN adherence significantly attenuate I/R-induced microvascular injury (8), and depletion of extravascular mucosal PMN greatly attenuates mucosal dysfunction (14). Thus these findings suggest that PMNs mediate cell injury, exacerbating ischemic damage by plugging the microvasculature and releasing undefined cytotoxic substances (8). One of these mediators could be ET-1, and several lines of evidence are consistent with this hypothesis. First, activation of PMNs plays a central role not only in response to invading pathogens, but also as an essential component of the reaction to various forms of organ injury (3). PMNs are activated by LPS through CD14, a glycosylphosphatidylinositol-anchored protein (16), and LPS-activated PMNs produce ET-1 (2). IFN- and TNF- increase CD14 expression on PMNs and enhance LPS binding to PMNs (28). Second, a subgroup of histologically normal appendixes from patients with a clinical diagnosis of acute appendicitis have abnormal TNF- (30) and IFN- (20) expression. Although PMNs are regarded as terminally differentiated short-lived cells, they release cytokines such TNF- (4) and IFN- (33). Third, a recent study suggests that infusion of ET-1 into the superior mesenteric artery causes dose-dependent infiltration of PMNs, mucosal dysfunction in rat small intestine, and neutropenia attenuates ET-1-induced increase in mucosal permeability (21). Finally, ET-receptor blockers reduce significantly I/R-induced intestinal mucosal injury and PMN infiltration (22), indicating that tissue PMNs are important mediators of ET-1-induced intestinal damage.

In conclusion, our results demonstrate the presence of pp-ET-1 mRNA and its mature peptide in PMNs, MCs, and rare MØs, with different distributions and relative frequencies in normal control appendix, histologically normal appendix, and inflamed appendix. In particular, a subgroup of histologically normal appendixes from patients with a clinical diagnosis of acute appendicitis had many ET-1-positive cells and high pp-ET-1 mRNA expression, similar to inflamed appendix. This last result is in line with reports on the existence of a subgroup of histologically normal appendixes with abnormal cytokine expression (20, 30). We therefore suggest that local production of ET-1 by neutrophils and other inflammatory cells could be a molecular sign of focal inflammation in microscopically normal appendixes and that ET-1 could be implicated with other cytokines, such as TNF-, IL-2, and IFN-, in the pathogenesis of appendicitis by inducing appendiceal ischemia through vasoconstriction.