Journal of Gastroenterology and Hepatology (1999) 14, 873–879

Na+ and water absorption and abnormalities of itsmetabolism or supply have been implicated in thepathogenesis of colonic diseases, such as ulcerativecolitis4 and colorectal cancer.5

One of the key functions of the colonic epithelium isto act as an effective barrier between the contents of thecolonic lumen and the internal milieu. Macromoleculesthat permeate the epithelial barrier do so mainly via theparacellular route for which the tight junction is therate-determining structure. Impaired epithelial barrierfunction has been implicated in the aetiology of severaldiseases, including ulcerative colitis6,7 and Crohn’sdisease.8 Despite the pivotal place that barrier function

INTRODUCTION

Short-chain fatty acids (SCFA) are produced in thecolonic lumen by bacterial fermentation of luminal,principally dietary, carbohydrate.1 Their importance tothe maintenance of colonic epithelial health is exempli-fied by the mucosal atrophy that occurs in their absenceand the reversal of such atrophy when SCFA areinstilled into the colonic lumen.2 Of the SCFA, butyrateappears to be the most important. It is the preferredenergy substrate for colonic epithelial cells3 and, byitself, is as effective as a combination of SCFA in revers-ing mucosal atrophy.2 It also plays an important role in

PATHOPHYSIOLOGY OF THE COLONIC MUCOSA, COLITISAND COLON DUPLICATION

Effect of butyrate on paracellular permeability in rat distal colonicmucosa ex vivo

JOHN M MARIADASON,* DIANA KILIAS,* ANTHONY CATTO-SMITH† AND PETER R GIBSON*

*University of Melbourne Department of Medicine,The Royal Melbourne Hospital and †Department of Gastroenterology,The Royal Children’s Hospital,Victoria, Australia

AbstractBackground and Aims:The effects of butyrate on colonic epithelial barrier function are poorly under-stood. The aim of this study was to examine the short-term effects of butyrate on paracellular perme-ability of rat distal colonic epithelium.Methods: Mucosa mounted in Ussing chambers was treated with butyrate (1–10 mmol/L) for 4 h.Transepithelial conductance, [51Cr]-EDTA flux, mucosal brush border hydrolase activity and epithelialkinetics, using proliferating cell nuclear antigen (PCNA) staining, were measured.Results: On exposure to butyrate (10 mmol/L, but not 1 or 5 mmol/L), transepithelial conductance was65 ± 2% higher (mean ± SEM; n = 8, P < 0.05, paired t-test) and the rate coefficient for [51Cr]-EDTA fluxwas 65 ± 25% higher (P = 0.03) than those of control tissue. Histologically, the epithelium exhibited nosigns of injury, but butyrate-treated tissue exhibited interstitial oedema consistent with water uptake inassociation with butyrate absorption. Butyrate caused a reduction in crypt column height to 30.6 ± 1.6cells from 33.4 ± 1.8 cells in controls (n = 10, P < 0.03), but the number of cells per crypt column stain-ing with PCNA was unchanged. Butyrate significantly reduced the mucosal activities of alkaline phos-phatase by 40 ± 16%, maltase by 54 ± 12% and dipeptidyl peptidase IV by 41 ± 14%.Conclusions: Acute exposure to butyrate increased paracellular permeability in rat distal colon. Themechanism involved may relate to the loss of differentiated surface epithelial cells, or as a physiologi-cal response to Na+-coupled butyrate uptake.© 1999 Blackwell Science Asia Pty Ltd

Key words: butyrate, cell proliferation, colon, epithelium, permeability.

Correspondence: Peter R Gibson, Department of Medicine, The Royal Melbourne Hospital, Victoria 3050, Australia. Email:<p.gibson@medicine.unimelb.edu.au>

Accepted for publication 29 March 1999.

has in colonic physiology, the influence of butyrate onsuch a function is incompletely defined.

Atrophic colonic mucosa is associated with impairedparacellular permeability9 and it is presumed, althoughnot documented, that SCFA improve that deficiency byreversing the atrophy. However, the effects of butyrateon paracellular permeability in a non-atrophic colon invivo are unknown. Butyrate improves barrier functionwhen applied to in vitro models of colonic epithelium,such as Caco-2 cell monolayers, by reducing paracellu-lar permeability10 and by promoting restitution ofwounds.11 Elevating luminal butyrate concentrations bydietary means, such as the ingestion of wheat bran,improves paracellular permeability of distal colon innormal rats but it is uncertain whether SCFA or otherluminal changes are responsible for the improve-ment.12 One approach to help define the effects ofbutyrate on the permeability of the colonic epitheliumis to apply butyrate to colonic mucosa in organ culture,mounted in Ussing chambers.

The present study, therefore, aimed to examine theeffect of butyrate on paracellular permeability of normalrat distal colon by short-term exposure of butyrate tothe luminal surface of the epithelium of distal colonicmucosa mounted in Ussing chambers. The distal colonwas studied because this is the region that is exposed tohighly variable and potentially deficient luminal con-centrations of butyrate.The distal colon and rectum arealso the principal sites of epithelial diseases, such asulcerative colitis and cancer. As butyrate also potentiallyinfluences the turnover of colonic epithelial cells,13 therelationship of epithelial kinetics and differentiationstatus to changes in paracellular permeability were alsoexamined.

METHODS

Tissue studied

Male Sprague–Dawley rats, weighing 260–280 g andfed on rat chow, were anaesthetized with nembutal(Boehringer Ingelheim, Artarmon, NSW, Australia),and the distal colon was removed. The animals werethen killed. The tissue was stripped of the underlyingmuscle layers. The most distal 2 cm was prepared formounting in Ussing chambers. The protocol wasapproved by the Hospital Campus Animal EthicsCommittee of The Royal Melbourne Hospital on 1 Sep-tember 1994. The experiment was conducted in accor-dance with the guidelines laid down in the AustralianCode of Practice for the Care and Use of Animals forScientific Purposes in a registered animal facility.

Measurement of paracellular permeability

The stripped colonic mucosa was mounted in 0.76 cm2

Ussing chambers (CSIRO,Victoria, Australia). Separatereservoirs, each containing 15 mL Krebs buffer at 37°Cand pH 7.4, bathed the apical and basolateral surfaces.The buffer contained (in mmol/L) Na, 140; K, 10; Mg,

874 JM Mariadason et al.

1.1; Ca, 1.25; Cl, 127.7; H2PO4, 2; HEPES, 25; andglucose, 10 (BDH Chemicals, Poole, UK). Oxygen wasbubbled through the chambers to oxygenate the tissue.The samples were allowed to equilibrate for 30 minprior to measurement of paracellular permeability,which was measured by two methods.

Transepithelial conductanceThe spontaneous potential difference (PD) across theepithelium was determined and the tissue clamped atzero voltage by introducing an appropriate short-circuitcurrent (Isc) with an automatic voltage clamp (DVC1000; World Precision Instruments, New Haven, CT,USA). The Isc was continuously monitored and PDmeasured by briefly removing the voltage clamp for5–10 s every 15 min.Transepithelial tissue conductancewas calculated according to Ohm’s law and expressedas mS/cm2.

Transepithelial [ 51Cr]-EDTA flux[51Cr]-EDTA (2 MBq; Australian Radioisotopes, NSW,Australia) was added to the apical reservoir immediatelyfollowing the mounting of tissue in the Ussing chamberand the measurement of the first electrical indices.Serial aliquots of 0.5 and 2 mL were taken at regularintervals from the apical and basolateral reservoirs,respectively. The differing volumes related to the rela-tive radioactive counts in the two chambers. Radioac-tivity in those aliquots was determined by gammacounting and calculations of rate coefficients were performed as described by Dawson.14 Rate coefficientswere expressed as mL/cm2 per min.

Measurement of cell proliferation

Cell proliferation was determined by the proliferatingcell nuclear antigen (PCNA) method. At the comple-tion of the experimental period, tissue was fixed inmethacarn (methanol: chloroform: acetic acid, 6:3:1)for 1 h, dehydrated in three changes of chloroform overa period of 3 h and vacuum embedded in three or fourchanges of paraffin wax over a period of 1 h. Serial sec-tions (2 mm) were cut and mounted in sequence onnumbered slides. Slides were stained with a 1:1000dilution of a monoclonal mouse antihuman PCNA(Sigma-Aldrich, St Louis, MO, USA), washed andstained with anti-rat immunoglobulin (Ig) G (1:40;Sigma-Aldrich). After 30 min, sections were washedagain and stained for a further 30 min with a 1:40 dilu-tion of rabbit horseradish peroxidase antiperoxidase(Sigma-Aldrich). Between reagent changes, sectionswere washed for 10 min in three changes of Tris buffer.

Kinetic indices in each tissue sample were evaluatedin 10 well-orientated, longitudinally cut crypt columnsas previously described and validated.15 The PCNA-positive cells were identified as strongly stained cells.Cell proliferation was determined by counting thenumber of PCNA-positive cells per crypt column.Crypt column height (CCH) was determined by count-

ing the number of cells comprising a crypt column.Thelabelling index (LI) was calculated by dividing thenumber of PCNA-positive cells by the total number ofcells in the counted crypts.

Enzymic analyses

Mucosa was stored at – 20°C in 0.5–1 mL mannitolbuffer (50 mmol/L D-mannitol and 2 mmol/L trizmabase in dH2O, pH 7.4). Prior to assaying, the tissue washomogenized on ice using a Polytron homogenizer(Kinematica AG, Switzerland) and triton-X 100 (BDHChemicals, Poole, UK) was added to a final concentra-tion of 0.1%.

Alkaline phosphatase activity was measured spec-trophotometrically in cell homogenates at 37°C usingp-nitrophenyl phosphate (Sigma-Aldrich) as sub-strate.16 Maltase activity was determined according to the method of Messer and Dahlqvist17 with maltose(< 0.05% glucose as a contaminant; Sigma-Aldrich) assubstrate. Dipeptidyl peptidase IV (DPPIV) activity wasdetermined according to the method of Maroux et al.,18

with glycyl-L-proline-p-nitroanilide (Sigma-Aldrich) assubstrate. Activities were expressed relative to theprotein content of homogenates, as determined spec-trophotometrically using bovine g-globulin as a stan-dard.19 The coefficient of variation of all assays was < 10%.

Statistical methods

Group data from all experiments were expressed asmean ± SEM. For transepithelial conductance andEDTA flux, data over time were expressed as an area-under-the-curve summary measure. Experimental andcontrol groups were compared using a two-sided pairedt-test. A P value £ 0.05 was considered statistically sig-nificant in all cases. All statistical analyses were per-formed using Minitab for Windows, version 9.2(Minitab Inc, State College, PA, USA).

RESULTS

Effect of butyrate on paracellularpermeability

Transepithelial conductance (Fig. 1) in untreated(control) colonic mucosa was constant over the 240 minexperimental period.Transepithelial conductance was amean 66% higher (P = 0.03) across mucosa exposed to10 mmol/L butyrate compared with control mucosa,while concentrations of 1 and 5 mmol/L butyrate had minimal effects (Figs 1,2). Conductance progres-sively increased over the 4 h-experimental period andwas significantly greater than in control tissue at all timepoints on and after 180 min (Fig. 1). In parallel withtransepithelial conductance, the apical-to-basolateralflux of [51Cr]-EDTA was a mean 65% higher in mucosa exposed to 10 mmol/L butyrate with a rate

Butyrate and colonic epithelial permeability 875

coefficient of 24.7 ± 4.1 mL/cm2 per min compared with17.2 ± 2.3 mL/cm2 per min in controls (P = 0.03). Nochanges of either index occurred at concentrations of 1 and 5 mmol/L (Fig. 2).

Effect of butyrate on epithelial cell kinetics

The effects of butyrate on epithelial cell kinetics areshown in Table 1. Over the 240 min of culture, the CCHfell slightly but insignificantly (P = 0.21) by a mean 5%in control tissue compared with that prior to culture.However, the CCH in butyrate-treated mucosa fell sig-nificantly by a mean 13% compared with the CCHprior to culture (P = 0.004) and was 8% less than thatof control tissue at the end of culture (P = 0.03). No dif-ferences in the number of cells per crypt column stain-ing positively with PCNA or in the labelling index wereobserved between control and butyrate-treated mucosaor in control mucosa between the start and completionof the experimental period.

Effect of butyrate on mucosal hydrolase activities

As shown in Fig. 3, mucosal hydrolase activities werelower in tissue treated with 10 mmol/L butyrate than incontrol tissue at the end of the 240-min culture period

Figure 1 The time course of transepithelial conductance(measured in siemen/cm2) across rat distal colonic mucosa inthe presence (j) or absence (r) of 10 mmol/L butyrate. Aftera 30-min equilibration period, 10 mmol/L butyrate was addedto the apical reservoir of Ussing chambers (shown as timezero). Transepithelial conductance was then monitored overthe following 240 min and compared with that from chamberswithout butyrate added (control). Values shown are mean ±SEM for eight experiments. Transepithelial conductance wassignificantly higher in butyrate-exposed tissue at 120, 180 and240 min. *P < 0.05, paired t-test.

(n = 8). Alkaline phosphatase activities in mucosaexposed to butyrate fell by 40 ± 16% (P = 0.02), maltaseactivities by 54 ± 12% (P = 0.03) and DPPIV activitiesby 36 ± 14% (P = 0.03) compared with those in controlmucosa.

Effect of butyrate on tissue morphology

Sections of colonic mucosa were taken before cultureand after 240 min of culture in control conditions orwith 10 mmol/L butyrate. Typical findings are shown in Fig. 4. There were no signs of epithelial necrosis orerosions or of distortion of crypt architecture in anysection. However, mucosa exposed to 10 mmol/Lbutyrate consistently showed interstitial oedema with

876 JM Mariadason et al.

greater separation of crypts and reduced density oflamina propria cells.

DISCUSSION

The use of Ussing chambers in the study of intactcolonic epithelial function has the distinct advantage ofan in-built measure, in the form of transepithelial con-ductance, of the viability of the epithelium undercontrol conditions, but is limited by the duration ofsuch viability. In the present study, viability was main-tained in all untreated mucosa studied over 4 h and themorphology of the tissue confirmed the epithelium tobe intact. However, beyond that time period, conduc-tance became more variable (pers. obs.) limiting theexperimental period to 4 h. Butyrate had no effect onparacellular permeability at concentrations of 1 and 5 mmol/L but caused marked changes at 10 mmol/Lwhere it induced 65 and 66% increases in transepithe-lial conductance and [51Cr]-EDTA flux, respectively.This effect was evident within 120 min of the additionof butyrate and its magnitude progressively increasedover the experimental period. These observations areconsistent with those recently reported by Luciano et al.who demonstrated an approximately two-fold increasein transepithelial conductance in the proximal colon of guinea pig after 90-min exposure to 10 mmol/Lbutyrate.20

There are three possible mechanisms for the effectsof butyrate on paracellular permeability observed in thepresent study. First, butyrate might be inducing epithe-lial cell necrosis via non-specific toxicity. Cells in vitrogenerally have a poor tolerance of butyrate concentra-tions greater than 2 mmol/L. This is manifested byincreased leakage of lactate dehydrogenase from thecells10,11,21 and by loss of functional competence of thecells.10,11 In contrast, the colonic mucosa can tolerate

Figure 2 Concentration-dependent effects of butyrate onparacellular permeability in rat distal colonic mucosa.Butyrate was added to the apical reservoir of rat distal colonicmucosa mounted in Ussing chambers and (a) transepithelialconductance and (b) [51Cr]-EDTA flux monitored over thefollowing 240 min. Values shown are mean + SEM of the per-centage change of the area-under-the-curve summary measurerelative to the control for the duration of the butyrate expo-sure, and represent five independent experiments for 1 ( )and 5 mmol/L ( ) butyrate, and eight for 10 mmol/L butyrate(j). *P = 0.03 compared to control, paired t-test.

Figure 3 The effect of butyrate on mucosal brush borderhydrolase activities. Rat distal colonic mucosa mounted in Ussing chambers was treated with (j) or without (h) 10 mmol/L butyrate for 240 min. At the completion of theexperimental period, control and butyrate-treated mucosawere harvested and alkaline phosphatase (ALP), maltase anddipeptidyl peptidase IV (DPPIV) activities were measured inmucosal homogenates. Values shown are mean + SEM foreight experiments. *P < 0.05 compared to appropriate control;paired t-test.

Table 1 Effect of butyrate on epithelial kinetics in distalcolonic mucosa before and after culture in Ussing chamberswith or without exposure to butyrate for 240 min

Following culture:

Prior to Butyrate culture Control (10 mmol/L)

Crypt column height 35.1 ± 0.7 33.4 ± 1.8 30.2 ± 1.7*PCNA+ cells/crypt 5.2 ± 0.4 5.4 ± 0.7 5.3 ± 0.6

columnLabelling index 15.8 ± 1.1 16.9 ± 2.5 18.0 ± 2.4

PCNA+, proliferating cell nuclear antigen positive.The dataare shown as mean ± SEM of eight experiments. * P < 0.05compared to control and prior to culture (paired t-test).

high concentrations (> 20 mmol/L) in vivo without evi-dence of toxicity. In the present ex vivo model, toxicityof butyrate seems less likely, especially as histologicalevidence of cell necrosis and epithelial erosions were notobserved.

The second possibility is that butyrate induces loss ofepithelial cells that share the most efficient tight junc-tions. Butyrate led to approximately 8% fewer cells percrypt column, indicating a reduction in the total cellpopulation of around 8%.This was not due to a reducedrate of cell proliferation as the number of cells stainingpositively with PCNA in each crypt was unchanged.This observation might superficially appear at odds witha previous report that butyrate increases proliferation incolonic mucosa in organ culture over a similar timeperiod.22 However, the only kinetic index shown in thatstudy was the labelling index, which was increased, asours tended to be. As demonstrated in the presentstudy, such an increase is due to reduced cell popula-tion (the denominator of the LI) rather than anincreased birth rate (the numerator of the LI). Thereduced cell population was, therefore, due to increasedloss of cells from the epithelial compartment. The con-siderable reduction of brush border hydrolase activitiessuggested that the cells strongly expressing those hydro-lases, that is, cells present in the surface compartmentand neck of the crypts,23,24 were dying. As no histolog-ical evidence for cell necrosis was found (as discussedabove) and cell shedding is unlikely in a static environ-ment, the most probable mechanism of cell death wasby apoptosis of differentiated cells in the surface com-partment. This view is further supported by the obser-vations that apoptosis is rapidly induced in the surfaceepithelial compartment when colonic mucosa is

Butyrate and colonic epithelial permeability 877

removed from its blood supply25 and that butyrateinduces apoptosis in colon cancer cell lines.26 As thesurface epithelium exhibits maximal expression of thetight junction protein ZO-1 and occludin (J Mariada-son, unpubl. obs., 1999), increased loss of the surfaceepithelial cells without loss of epithelial continuity, maypotentially lead to reduced tight junction function.

Recent observations in guinea pig proximal colonmounted in Ussing chambers do not support thissecond hypothesis. Butyrate at the same concentrationused in the present study (10 mmol/L) was shown toincrease rather than decrease crypt height and to inhibitapoptosis.20 The withdrawal of butyrate led to a rapidrise in the expression of the pro-apoptotic Bax protein.27

Species differences may underlie such a paradoxicaleffect of butyrate. Nevertheless, butyrate also doubledtransepithelial conductance in this guinea pig model,suggesting that the loss of differentiated cells is anunlikely complete explanation for the reduction of para-cellular permeability induced by butyrate.

The third possibility is that the increased permeabil-ity represents a physiological response to the uptake ofbutyrate. The major mechanism of butyrate uptake bycolonic epithelial cells in vivo is by passive diffusion ofits protonated form. Once taken up, it dissociates withinthe cytoplasm releasing a proton. The reduction inintracellular pH that results is compensated by theextrusion of H+ in exchange for Na+, via the Na+/H+

antiporter.28,29 Furthermore, the coupling of Na+

absorption with butyrate uptake has been shown toinduce cell swelling30 and, in some cases, invoke a reg-ulatory volume decrease response.31 The movement of Na+ into the lamina propria will cause water dif-fusion across the tight junctions. Butyrate-treated

Figure 4 Effect of butyrate on tissue morphology. Represen-tative histological sections stained with HE of (a) normal ratcolonic mucosa prior to mounting in Ussing chambers; (b)control rat colonic mucosa after incubation in Ussing chambersfor 240 min; and (c) rat colonic mucosa after treatment with 10 mmol/L butyrate for 240 min, showing greater separation ofcrypts and more sparse cellular density in the lamina propria,which is indicative of interstitial oedema.

colonic mucosa, in this study, exhibited interstitialoedema, which supports the uptake of fluid from theapical compartment.

Similar physiological events have been implicated inassociation with the uptake of monosaccharides, aminoacids and medium-chain fatty acids in increasing para-celllar permeability in the small intestinal epitheliumusing Ussing chambers, impedance analysis or in vivoperfusion techniques.32–38 The uptake of glucose andamino acids is coupled to Na+. Sodium-coupled solutetransport increases fluid uptake by the cells, leading tocell swelling.39 The cells compensate by undergoing aregulatory volume decrease, which has been shown tocause an elevation of the intracellular Ca2+ concentra-tion in certain epithelia.40 Turner and Madara havehypothesized that elevated intracellular Ca2+ concentra-tions may activate myosin light chain kinase, which, inturn, phosphorylates myosin regulatory light chain.41

The resulting increase in cytoskeletal tension may leadto increased tight junction permeability.41 In addition,water movement across the tight junctions causes a‘solvent drag’ effect, which itself may cause increasedparacellular permeability.39 It is reasonable, therefore,to speculate that the butyrate-induced increase in para-cellular permeability may occur via a similar mechanismto that proposed for glucose and amino acids. Further-more, transient opening of the tight junctions inresponse to high luminal concentrations of SCFA mightenable additional harvest of luminal SCFA via solventdrag through the paracellular route, as previously sug-gested to occur in guinea pig proximal colon.42

The effect of butyrate on paracellular permeability inthis ex vivo model of colonic epithelium is in contrastto its effect on paracellular permeability in Caco-2 cellsin vitro. However, the latter effect requires 2–3 days tofully manifest itself and requires RNA and protein syn-thesis.10 The short time course involved makes itunlikely that gene transcription and protein synthesisare involved in this ex vivo effect.

In conclusion, this study of the short-term effects ofdirect application of butyrate to distal colonic mucosahas highlighted the complexity of butyrate’s potentialeffects on epithelial barrier function in vivo. In additionto its ability to tighten the tight junction via new proteinsynthesis (as demonstrated in vitro), it may open tightjunctions in response to physiological changes associ-ated with its uptake or by the loss of functionally effi-cient surface epithelial cells.

REFERENCES

1 Fleming SE, Arce DS. Volatile fatty acids: Their produc-tion, absorption, utilization, and roles in human health.Clin. Gastroenterol. 1986; 15: 787–814.

2 Kripke SA, Fox AD, Berman JM, Settle RG, Rombeau JL.Stimulation of intestinal mucosal growth with intracolonicinfusion of short-chain fatty acids. J. Parent. Ent. Nutr.1989; 13: 109–16.

3 Roediger WEW. Utilization of nutrients by isolated epithelial cells of the rat colon. Gastroenterology 1982; 83:424–9.

878 JM Mariadason et al.

4 Roediger WEW. The colonic epithelium in ulcerativecolitis: An energy deficiency disease? Lancet 1980; ii:712–15.

5 Young GP, Gibson PR. Butyrate and the human cancercell. In: Cummings JH, Rombeau JL,Takashi S, eds. Phys-iological and Clinical Aspects of Short-chain Fatty Acids.Cambridge: Cambridge University Press, 1995; 319–35.

6 Almer S, Franzen L, Olaison G, Smedh K, Strom M.Increased absorption of polyethylene glycol 600 depositedin the colon in active ulcerative colitis. Gut 1993; 34:509–13.

7 O’Morain CA, Abelow AC, Chervu LR, Fleischner GM,Das KM. Chromium 51-ethylenediaminetetraacetate test:A useful test in the assessment of inflammatory boweldisease. J. Lab. Clin. Med. 1986; 108: 430–5.

8 Hollander D, Vadheim C, Brettholz E, Pattersen GM,Delahunty T, Rotter JI. Increased intestinal permeabilityin patients with Crohn’s disease and their relatives. Ann.Intern. Med. 1986; 105: 883–5.

9 Wang Q, Xiang-Dong W, Jeppsson B, Andersson R, Karls-son B, Westrom B. Influence of colostomy on in vivo andin vitro permeability in the rat colon. Dis. Colon Rectum1996; 39: 663–70.

10 Mariadason JM, Barkla DH, Gibson PR. Effect of short-chain fatty acids on paracellular permeability in Caco-2intestinal epithelium model. Am. J. Physiol. 1997; 272:G705–12.

11 Wilson AJ, Gibson PR. Short-chain fatty acids promotethe migration of colonic epithelial cells in vitro. Gastroen-terology 1997; 113: 487–96.

12 Mariadason JM, Catto-Smith A, Gibson PR. Modulationof distal colonic epithelial barrier function by dietary fibrein normal rats. Gut 1999; 44: 394–9.

13 Gibson PR, Anderson RP, Mariadason JM,Wilson AJ.Theprotective role of the epithelium of the small intestine andcolon. Inflamm. Bowel Dis. 1996; 2: 279–302.

14 Dawson DC. Intestinal absorption and secretion. In:Schultz SG, Field M, Frizzel RA, eds. Principles of Mem-brane Transport. Handbook of Physiology. Bethesda: Ameri-can Physiological Society, 1991; 1–44.

15 Macrae FA, Kilias D, Sharpe K, Hughes NR,Young GP,MacLennan R. Rectal epithelial cell proliferation: Com-parison of errors in measurement with intersubject vari-ance. J. Cell. Biochem. 1994; 19 (Suppl.): 84–90.

16 Young GP, Rose IS, Cropper S, Seetharam S, Alpers DH.Hepatic clearance of rat plasma intestinal alkaline phos-phatase. Am. J. Physiol. 1984; 247: G419–26.

17 Messer M, Dahlqvist A. A one-step ultramicro method forthe assay of intestinal disaccharidases. Anal. Biochem.1966; 14: 376–92.

18 Maroux S, Louvard D, Baratti J.The aminopeptidase fromhog intestinal brush border. Biochem. Biophys. Acta 1973;321: 282–95.

19 Bradford M. A rapid and sensitive method for the quan-titation of microgram quantities of protein utilizing theprinciple of protein dye binding. Anal. Biochem. 1976; 72:248–54.

20 Luciano L, Hass R, Busche R, Engelhardt WV, Reale E.Withdrawal of butyrate from the colonic mucosa triggers‘mass apoptosis’ primarily in the Go/G1 phase of the cellcycle. Cell Tissue Res. 1996; 286: 81–92.

21 Gibson PR, Moeller I, Kagelari O, Folino M, Young GP.Contrasting effects of butyrate on the expression of

phenotypic markers of differentiation in neoplastic andnon-neoplastic colonic epithelial cells in vitro. J. Gastroen-terol. Hepatol. 1992; 7: 165–72.

22 Scheppach W, Bartram P, Richter A et al. Effect of short-chain fatty acids on the human colonic mucosa in vitro. J.Parenter. Enter. Nutr. 1992; 16: 43–8.

23 Bell C,Williams C. Histochemical demonstration of alka-line phosphatase in human large intestine, normal and dis-eased. Histochemistry 1979; 60: 85–9.

24 Young GP, Macrae FA, Gibson PR, Alexeyeff M, White-head RH. Brush border hydrolases in normal and neo-plastic colonic epithelium. J. Gastroenterol. Hepatol. 1992;7: 347–54.

25 Hall PA, Coates PJ, Ansari B, Hopwood D. Regulation ofcell number in the mammalian gastrointestinal tract:Importance of apoptosis. J. Cell Sci. 1994; 107: 3569–77.

26 Hague A, Elder DJE, Hicks DJ, Paraskeva C. Apoptosis incolorectal tumor cells: Induction by the short chain fattyacids butyrate, propionate and acetate and by the bile saltdeoxycholate. Int. J. Cancer 1995; 60: 400–6.

27 Hass R, Busche R, Luciano L, Reale E, Engelhardt WV.Lack of butyrate is associated with induction of Bax andsubsequent apoptosis in the proximal colon of guinea pig.Gastroenterology 1997; 112: 875–81.

28 Rowe WA, Lesho MJ, Montrose MH. Polarized Na+/H+

exchange function is pliable in response to transepithelialgradients of propionate. Proc. Natl Acad. Sci. USA 1994;91: 6166–70.

29 DeSoignie R, Sellin JH. Propionate-initiated changes inintracellular pH in rabbit colonocytes. Gastroenterology1994; 107: 347–56.

30 Sellin J, Shelat H. Short-chain fatty acid (SCFA) volumeregulation in proximal and distal rabbit colon is different.J. Membr. Biol. 1996; 150: 83–8.

31 Diener M, Helmle-Kolb C, Murer H, Scharrer E. Effectof short-chain fatty acids on cell volume and intracellularpH in rat distal colon. Pflugers Arch. 1993; 424: 216–23.

Butyrate and colonic epithelial permeability 879

32 Atisook K, Carlson S, Madara JL. Effects of phlorizin andsodium on glucose-elicited alterations of cell junctions inintestinal epithelia. Am. J. Physiol. 1990; 258: C77–85.

33 Atisook K, Madara JL. An oligopeptide permeates intesti-nal tight junctions at glucose-elicited dilatations. Gas-troenterology 1991; 100: 719–24.

34 Pappenheimer JR, Volpp K. Transmucosal impedence ofsmall intestine: Correlation with transport of sugars andamino acids. Am. J. Physiol. 1992; 263: C480–93.

35 Madara JL, Carlson S. Supraphysiologic L-tryptophanelicits cytoskeletal and macromolecular permeability alter-ations in hamster small intestinal epithelium in vitro. J.Clin. Invest. 1991; 87: 454–62.

36 Sawada T, Ogawa T,Tomita M, Hayashi M, Awazu A. Roleof paracellular pathway in nonelectrolyte permeationacross rat colon epithelium enhanced by sodium caprateand sodium caprylate. Pharm. Res. 1991; 8: 1365–71.

37 Tomita M, Hayashi M, Awazu S. Absorption-enhancingmechanism of sodium caprate and decanoylcarnitine inCaco-2 cells. J. Pharmacol. Exp.Ther. 1995; 272: 739–43.

38 Sandowski DC, Meddings JB. Luminal nutrients altertight-junction permeability in the rat jejunum: An in vivoperfusion model. Can. J. Physiol. Pharmacol. 1993; 71:835–9.

39 Pappenheimer JR. Physiological regulation of transepithe-lial impedance in the intestinal mucosa of rats and ham-sters. J. Membr. Biol. 1987; 100: 137–48.

40 Christensen O. Mediation of cell volume regulation byCa2+ influx through stretch-activated channels. Nature1987; 330: 66–8.

41 Turner JR, Madara JL. Physiological regulation of intesti-nal epithelial tight junctions as a consequence of Na+-coupled nutrient transport. Gastroenterology 1995; 109:1391–6.

42 Luciano L, Reale E, Rechkemmer G, Engelhardt WV.Structure of zonulae occludentes and the permeability ofthe epithelium to short-chain fatty acids in the proximaland distal colon of guinea pig. J. Membr. Biol. 1984; 82:145–56.