The study included endoscopic biopsies from 86 patients of which 41 were women. The median age was 64 years (range 20 – 98). All patients were referred to colonoscopy for examination on suspicion of colonic neoplasia. The included subjects were without pathological findings at colonoscopy (controls) except for diverticulosis in the left part of the colon in 22 cases (D-patients).

Our study-protocol was approved by the Scientific Ethical Committee for Copenhagen (KA 97161) and Frederiksberg Counties (KF01-232/03) and conducted in accordance with the Declaration of Helsinki V. All patients gave written informed consent. All signs of disease in colon were noted. The patients' medication at the time of examination was noted.

Five biopsies were obtained from the sigmoid part of the colon (30 cm aborally to the anus on retraction of the endoscope) using a standard biopsy forceps (Boston Scientific, Denmark). The biopsies were taken from macroscopically normal appearing mucosa, not from the diverticuli per se.

Biopsies were transported in ice-cold bicarbonate-Ringer solution to the laboratory. They were mounted within 30 min in MUAS chambers, which uses constant air suction to fixate biopsies 9. Mounting was carried out at 10 times magnification by means of a stereomicroscope to secure mucosa-serosa orientation and proper fixation. The exposed tissue area was 5 mm². Both sides of the tissue were bathed with a bicarbonate-Ringer solution containing the following (in mM): 140 Na⁺, 4 K⁺, 121 Cl^-, 1 Ca²⁺, 0.5 Mg²⁺, 0.5 SO₄^2-, and 25 HCO₃^-, oxygenated with 95 % O₂/5 % CO₂, circulated by gas-lift. Media at the serosal side were further added 11 mM D-glucose and 11 mM D-sorbitol at the mucosal side. Temperature was maintained at 37°C by water jackets. Short-circuit current (SCC) measured in μA·cm^-2 and slope conductance (G) in mS·cm^-2 were recorded continuously by an automated voltage-clamp device 9. Correction for the resistance in solutions was performed immediately before each new tissue was mounted.

The height of the suction sleeve in a MUAS chamber is of importance to mounting of tissue. For biopsies to stay in place, we found that the height of the sleeve should be reduced to 40 μm to fit the colon biopsies compared to 50 μm for duodenal biopsies 9.

Experiments were performed after an equilibration period of 15 min. In order to evaluate tissue viability and transport capacity, application of various stimulators (serotonin, 5-hydroxytryptamine, 5-HT; forskolin; theophylline; carbachol) and inhibitors (amiloride; indometacin; bumetanide; ouabain) of epithelial ion transport were added to the serosal bathing solution, except for amiloride which was added to the mucosal solution and for indometacin added to both sides.

All drugs were purchased from Sigma (Vallensbaek Strand, Denmark) except for bumetanide, which was a gift from Leo Pharmaceuticals, Denmark.

Data are presented as median and range (minimum-maximum) followed by (N = number of patients, n = number of biopsies). Mann-Whitney rank sum test was used for statstical analysis according to results of normality and equal variance tests (SigmaStat 2.0 for Windows, SPSS Inc., USA). P < 0.05 was considered significant. The effect of compounds was defined as data before compared to data after application of the compounds.

In a supplementary data-sheet we present a vertical point plot of SCC and G for the individual patients, additional file 1.

Protocols were blinded to the examiner. Fixation was performed in 4 % buffered paraformaldehyde after taking biopsies and after experiments in the MUAS chamber. The fixed tissue samples were then dehydrated and embedded in paraffin and cut into 10 μm sections. Sections were stained with hematoxylin/periodic acid Schiff, examined and photographed using a Leitz Ortoplan microscope (Wetzlar, Germany) fitted with a cooled camera (Evolution MP, MediaCybernetics, Wokingham, Berkshire, UK).

Thirty-seven patients were under current medical treatment for arterial hypertension, 21 for mental or other neurological disorders, 10 for ischaemic heart disease, 9 for asthma or chronic obstructive lung disease, 9 for diabetes mellitus, and 2 for thyroid disease. 37 patients did not take any medication at all and were equally distributed in control and D-patient groups. Median systolic blood pressure was 140 mmHg (range 188 – 98), median diastolic blood pressure was 82 mmHg (range 112 – 72) and median heart rate was 78 (range 48 – 110), with no significant difference between controls and D-patients (p = 0.734, p = 0.805, p = 0.242, respectively). No complications due to endoscopy were reported.

Biopsies were excluded based on unstable SCC and G or on no response to any of the secretagogues. Of 5 biopsies from the same patient at least 3 were viable and (at least) 2 of these 3 gave stable measurements leaving us with a success-rate of about 40 %.

After an equilibration period in the chamber, basal SCC was 43.8 μA·cm^-2 (64, 120) for controls and 59.3 μA·cm^-2 (22, 40) for D-patients with no significant difference (p = 0.106), table 1. Results for G are shown as well in table 1 and were also not significantly different. During basal conditions, these parameters remained stable for more than 2 hrs. Furthermore, reproducible SCC-responses could be obtained for up to 2 hrs after mounting for all the tested stimulators and inhibitors of ion transport. In the majority of experiments, a slight progressive increase in SCC and G appeared within the 3 hrs and after 6–8 hrs a more than 100 % increase in SCC and G was observed (data not shown).

Histological assessments were performed for the extent of tissue and edge damage and the thickness of biopsies. The damage found in each biopsy was denoted by a severity score, as previously described 10. When examined in the stereomicroscope, it was noticed that specimens exposed in the MUAS chamber as compared with controls in most cases showed very little tissue damage. The damage seemed to originate from the biopsy forceps more than the chamber, because there was no difference in epithelial damage in the experimental biopsies in comparison with the control biopsies apart from possible minor indications of edge damage. The lack of damage to the surface epithelium following MUAS chamber exposure was confirmed at histological examination. The depth of biopsies varied somewhat and always included the surface epithelium and entire lamina propria. Several biopsies also included the lamina muscularis mucosae and some had parts of the submucosal layer preserved. Two examples without lamina muscularis are shown in figure 9. No difference in histology was detectable for D-patients as compared to controls. In particular, no signs of inflammation were detected in the biopsies from D-patients. When basal electrical parameters were compared with the extent of epithelial damage found in the biopsies on histological examination, there was no clear correlation between the different epithelial damage scores from biopsies scored 0 or 1 vs. biopsies scored 2 or 3.

The study provides information about the electropathophysiological characteristics of diverticulosis in the sigmoid part of the colon. We hypothesized the existence of changes in epithelial ion transport functions in D-patients similar to those found in smooth muscle cells. However, this study did not demonstrate such changes, suggesting that epithelial ion transport and probably the neuroepithelial integrity is not altered in patients with diverculosis, table 1, 2, and 3.

The etiology of colonic diverticulosis is unknown. It is assumed that a diet low on fibres is related to colonic diverticulosis. However, there does not seem to be a clear relation between increased intracolonic pressure and the appearance of divertculosis in the sigmoid part of the colon 11. A recent study suggested an altered neuromuscular function as patients with diverticulosis demonstrated a decrease in acetylcholine transferase activity, an up-regulation of muscarinic M₃ receptors and an increase in reactivity to exogenous acetylcholine 5 in smooth muscle cells of the sigmoid part of the colon. This hypersensitivity to acetylcholine could result from decreased cholinergic innervation 5. In the present study there is no indication of such an altered muscarinic receptor sensitivity, table 2.

The Ussing chamber technique is widely used to characterize epithelial ion transport in the gut. The conventional Ussing chamber has been of limited use for human studies because of limited availability of tissue samples of adequate size, i.e. surgical specimens. Endoscopically obtained tissue samples are desirable due to their wide availability and the tissues are probably less physiologically altered from its original conditions because of less surgical stress and shorter time at ambient temperature than in surgical specimens, that are often without perfusion for some time before the tissue is placed in cold bicarbonate-Ringer. Biopsies are more readily available for a wider array of diseases. Furthermore the patients can be matched with healthy controls not undergoing surgery or even healthy volunteers. Accordingly, the Ussing chamber technique has been refined for the study of small specimens obtained during endoscopy using capsule systems or large size biopsy forceps 91213141516. In human biopsy studies the epithelial area in different chambers usually varies from 3 to 5 mm² 121314, but ranges from 1 mm² 17 up to 13.2 mm²15.

Reproducible measurements require a high sensitivity of the equipment and great care in the phase of mounting. Various principles of mounting biopsies in the modified Ussing chambers have been tried. The use of mucosal discs, gluing of specimens, and placement on filter paper has been employed for fixation of specimens. In worry these techniques could cause significant mechanical or chemical edge damage, increase the thickness of unstirred layers and ultimately interfere with the function.

The MUAS chamber has been developed and evaluated for the duodenum for use with (human) forceps for small biopsy specimens 9. The MUAS chamber has proven its value in functional characterization of muscarinic, prostanoid and serotonin receptors in human duodenum 101819. The present study suggests that the MUAS technique can be adapted for the study of epithelial ion transport in human endoscopic colonic specimens. The MUAS technique includes utilization of steady air suction in an easy manageable manner. The tissue is kept in place by air-suction making it unnecessary to surround the tissue with any kind of film (mounting the biopsies between polyesterfilms adds to the unstirred layer). The MUAS technique is fast, simple, easy to use, with minimal loss of tissue, associated with only minor degrees of edge damage, a high degree of viability, and reliable responses to various secretagogues and inhibitors for more than 2 hrs, and specimens can be easily changed for another specimen in the chamber setup. The mounting principle with air suction provides a stretch to the tissue, which ensures an optimal exposed area without causing damage to the tissue as evaluated by histology.

In the present study, basal electrical parameters varied over a wide range. Similar large variability has been noticed in other studies on human specimens from colon 820, table 1.

SCC and G were stable for more than 2 hrs, which is also consistent with another study using a different type of modified Ussing chamber 16. Despite a subsequently progressive increase in SCC and G, reproducible responses could be obtained for more than 6 hrs to theophylline and for up to 4 hrs for 5-HT. This is somewhat surprising, but partially in agreement with a Ussing chamber study on human ileum, where transmucosal glucose fluxes were stable for 4 hrs despite progressive changes in SCC plus signs of epithelial histological changes after only 2 hrs 21.

In another study by Mullin and co-workers on histological normal surgical specimens from the left part of colon in patients with diverticular disease, SCC dropped initially from 250 to 50 μA·cm² and further down to 19 μA·cm² after 30 min – with large variations 22.

For the distal colon pronounced variations in resistance from 52 to 220 Ω·cm² has been reported 8. Explanations for this large variability might be differences in the biopsy and mounting techniques, resulting in varying thickness of the samples and varying degrees of tissue damage, or that specimens are taken from different locations in the same region suggesting that true functional differences exist even within relative small distances in the intestine. Both these explanations can be corroborated by observations quoted in the literature 122324. Also the specimen size seems of importance since large-sized (50 – 70 mm²) surgical specimens exhibits higher resistance (31 – 312 Ω·cm²) 25262728 compared to small-sized (0.65 – 5 mm²) biopsy specimens (12 – 30 Ω·cm²) in this study and others 162429.

Parallel changes in SCC and G was observed. With the (electronic) circuitry of the tissue, while measuring SCC across epithelia, a generator as the Na-pump is located in the basolateral membrane and the luminal membrane represents a serial conductance as does the internal conductance of the pump. Contrary, neighbor conductances to the pump in the basolateral membrane and conductances in paracellular pathways will act as parallel conductances. Hence, an increase in the luminal conductance will increase the SCC, while a change in parallel conductances will have no effect on measured transmural SCC.

Finally comparison of studies is hampered due to different levels in calcium concentrations in chamber bathing solutions varying from 0.5 – 3 mM, which is likely to be of importance as calcium regulates tight junction permeability 30.

Edge-damage in the MUAS chamber have been discussed earlier and does not seem to be an explanation for the rather low resistances observed in the present study 9. However, it should be recognized, that in case, as we did, a careful solution correction is performed immediately prior to tissue insertion, low tissue resistance as such does not affect the measured SCC as long as the transmural potential difference is zero. Correct short-circuiting eliminates shunt currents both in the tissue as well as through possible edge-damage pathways.

We conclude that colonic biopsies have much lower resistance than for instance surgical preparations for the Ussing chamber and in our study on colonic biopsies resistances are exceptionally low probably due to low calcium concentration in the media.

The applied secretagogues and inhibitors all induced expected changes in SCC and was accompanied by changes in G.

For these effects it is assumed that, 5-HT activates various serotonergic receptors 31, carbachol muscarinic receptors 32, and forskolin adenylyl cyclase 33. Theophylline inactivates phosphodiesterase 34, amiloride blocks electrogenic epithelial sodium channels 35, indometacin inhibits cyclooxygenase-enzymes 36, bumetanide inhibits sodium-potassium-chloride-co-transporter 37, while ouabain inhibits sodium-potassium pump 38.

The ionic basis for the observed changes in SCC and G were not investigated further, but are most likely due to opening and closing of ion channels and transporters for Cl^-, HCO₃^-, Na⁺ and K⁺, as demonstrated previously in various gut segments and species 39. Another possibility for the observed SCC changes could be a depletion of salt or a build-up of an osmotic gradient in the lateral intercellular space with a closure or an opening of the space as seen for leaky epithelia 40. However, these phenomena are unlikely during short circuit current voltage clamp to zero, which prevents the removal or built-up of salt in the paracellular pathway. Because we correct for the solution resistance just before every measurement, our tissues are well short circuited, and we can discard a closure of the intercellular space as an explanation for the observed fall in conductance.

It is a drawback of the MUAS chamber technique that only 40% of biopsies can be used, based on the criteria of stable electrical parameters and expected responses to stimulators of ion secretion. Similarly the success-rate for duodenal specimens is reduced 9. Reasons for the relatively low "success-rate" are still unclear. We have tried to change the experimental process in many ways (e.g. mechanical properties of the chamber and air-suction system, the bathing solutions, temperature etc.) without being able to improve the "success-rate". Some of the biopsies are too narrow to fit in the central opening of the discs and some can be mounted but demonstrate no response to stimulators or inhibitors. In some cases histology disclosed the likely cause due to substantial damage to the epithelium, which result from the biopsy forceps. Whether the success-rate using the MUAS chamber technique is different from other techniques for endoscopic biopsies cannot be assessed as such information in not yet available in the literature for other techniques.