In the present study we assessed surface hydrophobicity of gastrointestinal mucosa as an index of the functional integrity of the mucus layer, a layer that acts as a "closing seal"[22
] of the intestinal barrier. This bio-physical characteristic can be studied by measuring contact angles formed by sessile water droplets placed on freshly collected biopsies. The interrelation between contact angle and surface energy assumes that the surface on which contact angle is measured is smooth and homogenous [23
], and this is obviously not the case for mucosal biopsies. Furthermore, drying, presence of debris and trauma of biopsy collection may cause non-physiological changes. Although these observations are a matter of concern, drying has been reported by Hills [24
] to provide a more conservative estimate of the energy reduction that would occur if the mucus remained in the physiological hydrated state. Our validation study confirms that contact angle increases during mucosal drying to a point when it remains stable and reproducible up to at least 60 min following biopsy collection.
The main aspect of our study is that we assessed the hydrophobicity of the duodenal mucus barrier in CD patients studied at diagnosis and in CD patients on GFD. This assessment was prompted by a large body of evidence indicating that intestinal permeability to macromolecules is increased in CD patients [6
]. This phenomenon has been extensively studied in relation to the epithelial component of the intestinal barrier, and shown to involve altered regulation of the tight junctions that provide contiguity of epithelial cells and regulate the traffic trough the paracellular pathway. Zonulin has been identified as a protein with capacity to disassemble TJ [7
] as an early event in the development of CD [27
] favouring the passage of gliadin toxic fragments via the paracellular pathway from the intestinal lumen to the submucosa. Although the validity of this model has been questioned [28
], the pathogenetic role of increased permeability in development of CD is firmly established.
Our study clearly indicates that the mucus barrier is altered in CD, and that this defect is not restored to normal during GFD as indicated by the observation that hdyrophobicity of duodenal mucosa is lower in CD patients than in control subjects, and that this abnormality persists during GFD. These differences cannot be accounted by artefacts due to Helicobacter pylori infection and by differences in age, as reported by others for gastric biopsies [29
], because there was no relationship in our patients and in controls between these parameters and duodenal CA. Furthermore, the younger age of our CD compared with controls should point, if anything, to a higher CA, an opposite effect of what we have found.
The reduced hydrophobicity is specific for duodenal mucosa in CD, as indicated by the finding that hydrophobicity of gastric mucosa was similar in CD patients and in controls.
The mechanism involved in this bio-physical alteration of the duodenal mucus barrier is unclear. Previous studies have reported alteration of the structure of duodenal mucus in CD patients associated with defect of the trefoil factors family [30
] and with the structure and secretory pattern of mucus glycoprotein [31
], defects that may affect stability of mucus. We could not further investigate the mechanism involved in reduced hydrophobicity of duodenal mucus barrier, but it is clear from our results a relationship with the anatomical changes of the mucosa, independently of clinical diagnosis. This is suggested by our finding that by plotting together results of duodenal CA obtained in control subjects with those obtained in CD patients off and on GFD, there was a clear-cut dependency of reduced mucus hydrophobicity on severity of duodenal histopathology (Figure ). This was irrespective of clinical diagnosis because the group with normal histology comprised CD patients on GFD in addition to control subjects, and the group with Marsh-1-2 lesion comprised normal subjects in addition to CD patient on or off GFD. On the basis of our results we cannot firmly establish the nature of this relationship, and although a cause-effect mechanism cannot be proven it seems to us plausible because of the independence of the aetiology of mucosal damage.
To what extent the alteration in mucus barrier contributes to the pathogenesis of CD, and whether this alteration is a primary defect is totally unclear but it fits a paradigm for initiation of autoimmune diseases proposed by Arrieta et al [5
]. According to this model, undigested gliadin fractions are kept separate from sub-epithelial immune system by a competent intestinal barrier. Disease activation occurs when factors with capacity of altering the intestinal barrier integrity favour paracellular transit of toxic fractions of gliadin, and these initiating factors include drugs such as NSAIDs or infections that are all well known factors for their capacity to disrupt the mucus barrier [13
]. The late onset of CD in genetically predisposed subjects fits with this hypothesis on the role of initiating factors.
Another aspect of our study is that we looked at regional differences of hydrophobicity of intestinal mucosa in controls and in CD, and found marked differences in different intestinal regions. These differences are qualitatively similar to those reported by Spychal et al [21
] in humans showing high hydrophobicity in gastric and in rectal mucosa, with lower values in the duodenum and in the ileum. This regional difference is common to many mammalian species [10
] although slight species-related differences occur. By contrast with results in the animal, in our study mucosal hydrophobicity was lower in the distal ileum than in the duodenum. No other study to our knowledge has reported results for hydrophobicity of distal ileum in humans, but the tendency for hydrophobicity to be reduced from proximal to distal duodenum reported by Spychal [21
] in healthy subjects is consistent with our results. The low hydrophobicity in the oesophagus observed in our study is also novel, and explains the susceptibility of oesophageal mucosa to refluxed gastric acid because hydrophobicity is important in repelling the diffusion of hydrogenions [2
] from coming in contact with the epithelial cells.