Effect of topical butyrate on rectal epithelial kinetics and by dfsiopmhy6


									Clinical Science (1998) 94, 671–676 (Printed in Great Britain)                                                                    671

Effect of topical butyrate on rectal epithelial kinetics and
mucosal enzyme activities
Peter R. GIBSON, Diana KILIAS, Ourania ROSELLA, Jenny M. DAY, Margaret ABBOTT,
Caroline F. FINCH and Graeme P. YOUNG
University of Melbourne Department of Medicine, The Royal Melbourne Hospital, Victoria 3050, Australia

                                                                            to butyrate but the response often does not parallel
1. This study aimed to determine the effect of luminal                       that found in colon cancer cell lines. For example,
butyrate on proliferative kinetics, a differentiation                        butyrate has no effect on the rate of DNA synthesis
marker (alkaline phosphatase), and a molecule that                          and does not promote differentiation in isolated
controls cell-substratum adhesion (urokinase) in histo-                     normal cells [6] but suppresses it in cells from cancer-
logically normal human rectal mucosa.                                       bearing colons [7]. Butyrate also inhibits the pro-
2. Ten subjects with a colonoscopically normal colon                        duction of the neutral protease, urokinase (u-PA) [8],
(seven had previous adenomas) were given either bu-                         its receptor (u-PAR) [9], and interleukin-8 [10]. How-
tyrate or saline enemas for 4 days in a double-blind                        ever, it is uncertain whether the effects of butyrate on
cross-over manner. Rectal biopsies were taken before                        normal cells in vitro also occur in the normal colonic
and after each course of enemas. Epithelial proliferative                   epithelium when exposed to butyrate in vivo. Extra-
kinetics were measured immunohistochemically using                          polation from the isolated cell system must be made
antibodies to proliferating cell nuclear antigen. Uro-                      cautiously since normal epithelial–mesenchymal re-
kinase and alkaline phosphatase activities were                             lationships are disrupted and the trauma of their
measured spectrophotometrically in biopsy homo-                             isolation appears to markedly induce the synthesis and
genates.                                                                    secretion of factors such as u-PA [8], u-PAR [9] and
3. Both saline and butyrate enemas were well tolerated                      interleukin-8 [10].
and induced no histological change except for a sig-                           In several studies, butyrate has been instilled lumi-
nificant increase in crypt length (P 0.05). The number                       nally in rat large bowel which has been starved of
of proliferating cells per crypt also increased sig-                        carbohydrate fermentation products by unphysio-
nificantly after butyrate (P l 0.018).                                       logical manoeuvres such as the use of an elemental diet
4. Compared with saline enemas, butyrate did not affect                      or diversion of the faecal stream away from that
kinetic indices nor alkaline phosphatase activities. How-                   segment. These have consistently shown butyrate’s
ever, mucosal urokinase activities were significantly                        trophic effect on colonic epithelium [11,12] although
lower in butyrate-treated patients (9.5p2.0 i.u./g) than                    the effects on mucosal hydrolase or u-PA activities
in saline-treated patients (12.8p2.0 i.u./g ; P l 0.045).                   have not been reported. Under physiological con-
5. Delivering of extra butyrate to the distal colon in                      ditions, however, greater degrees of luminal fermen-
healthy subjects may stabilize cell-substratum adhesion                     tation do occur and mucosal atrophy is not seen.
in surface epithelium and therefore offer a potential                        Modulating luminal butyrate concentrations more
mechanism by which elevating distal colonic luminal                         than 10-fold by varying the amount and\or type of
butyrate concentrations might be beneficial in patients                      dietary fibre ingested had, in rats, only small effects, if
with colitis or hyperproliferative large bowel epithelium.                  any, on colonic epithelial proliferative kinetics and
                                                                            hydrolase expression, but significantly altered mucosal
                                                                            u-PA activity ; the lowest activities were found in the
                                                                            rats with the highest faecal butyrate concentrations
                                                                            (the rats fed unprocessed wheat bran) [13]. However,
                                                                            these effects do not necessarily specifically reflect
   Fermentation of carbohydrates in the large bowel                         differences in butyrate concentration since multiple
lumen yields large amounts of the short-chain fatty                         other luminal characteristics, such as concentrations
acid, butyrate [1]. Butyrate plays essential roles in                       of other short-chain fatty acids, pH and the spectrum
providing energy for colonic epithelium [2] and in                          of the bacterial flora, are also affected by the diets.
promoting sodium and water absorption [3]. Butyrate                            The effect of direct instillation of butyrate into large
has other powerful effects on cell lines in vitro,                           bowel lumen on epithelial and mucosal characteristics
including inhibition of cell growth, induction of                           in humans has only been studied in patients with active
differentiated phenotype and apoptosis, and the stimu-                       ulcerative colitis [14]. Butyrate enemas used twice daily
lation or inhibition of the synthesis of a large variety of                 changed the distribution of proliferating cells and
proteins [4,5]. Isolated colonic crypt cells also respond                   favoured reduced proliferative activity. Since inflam-

Key words: alkaline phosphatase, butyrate, cell proliferation, colonic epithelium, urokinase.
Abbreviations: PCNA, proliferating cell nuclear antigen; u-PA, urokinase; u-PAR, urokinase receptor.
Correspondence: Dr Peter R. Gibson.
672                                               P. R. Gibson and others

matory activity was also reduced, it is uncertain               corrected to 290 mOsmol\l using normal saline and
whether the effects were secondary to butyrate itself or         water for irrigation. Saline enemas comprised isotonic
to butyrate’s effect on reducing inflammation. The aim            sodium chloride at pH 7.0. The total volume of each
of the present study then was to use a similar regimen          enema was 60 ml.
of butyrate enemas in human subjects in whom rectal
histology was normal and to measure their effects on             Epithelial proliferative kinetics
rectal epithelial proliferative kinetics, a differentiation
marker (alkaline phosphatase), and the activity of u-              Thin sections of paraffin-mounted, Methacarn-fixed
PA, a molecule that controls cell-substratum adhesion.          biopsies were cut (2 µm thick) and stained using an
                                                                immunoperoxidase technique with antibodies to the
METHODS                                                         proliferating cell nuclear antigen (PCNA, Sigma,
                                                                Castle Hill, NSW, Australia) as recently described and
Subjects studied                                                validated by us [15]. Briefly, longitudinally cut crypts
  Ten subjects were studied. Their ages ranged from             were identified and the number of cells in each crypt
42 to 70 (mean 56) years and six were female. The               column counted. The number of crypt columns ex-
subjects had had a colonoscopy within the previous 12           amined ranged from 28 to 58 (mean l 40). Cells
months which had revealed the large bowel free of               staining strongly positive for PCNA were identified
adenomas or other lesions. Seven of the subjects had            and their position in the crypt column recorded. From
previous adenomatous polyps removed, one had pre-               these data, the following indices were determined :
vious hyperplastic polyps, and two had diverticular             crypt column height, the number of PCNA-positive
disease alone. They all gave written informed consent           cells per crypt, the labelling index calculated by
before commencement of the study and the protocol               dividing the number of positive cells by the total
was approved by the Board of Medical Research and               number of cells in the crypt columns (mean 3480 cells
Ethics Committee of The Royal Melbourne Hospital.               counted), and the distribution of positive cells ac-
                                                                cording to five equal quintiles, quintile 1 being the base
Protocol                                                        of the crypt, and quintile 5 at the surface end of the
                                                                crypt, expressed as percentage of positive cells in that
   The subjects were instructed to ingest a diet low in         quintile. All sections were evaluated by one experi-
fibre and resistant starch for the duration of the study.        enced scientist who was blinded to the timing of the
Three days after starting the diet, rectal biopsies were        biopsy. The performance of the latter in terms of
taken 5–7 cm from the anal verge without sigmoid-               variances of component errors in cell counting has
oscopy, using standard colonoscopic biopsy forceps.             recently been described in detail [15].
The patients were then randomized to receive an
enema of either normal saline or sodium butyrate (for           Enzymic activities
composition, see below) twice daily for 4 days. Neither
the patient nor the investigator was aware of the                  Before assay, mucosal samples in Tris–mannitol
nature of the enema. The enemas were self-admini-               buffer were thawed and mechanically homogenized at
stered. At the end of that time a further set of rectal         4 mC. Triton X-100 was added to a final concentration
biopsies were taken in the morning approximately 2 h            of 0.1 %. Myeloperoxidase activity was measured
after the last enema. The subjects then crossed over to         within 5 min of homogenization spectrophoto-
the other enema (saline or butyrate) which was also             metrically using guaiacol as substrate [16]. The results
self-administered twice daily for 4 days. Biopsies were         were expressed as change in absorbance ‘ units ’\min
taken at a similar time of the day about 2 h after the          per mg of protein of the homogenate. Alkaline
last enema. The subjects also filled out a diary card            phosphatase activities were assayed in aliquots of the
stating the times of day the enemas were administered           homogenate spectrophotometrically using p-nitro-
and when the bowels opened as well as any adverse               phenol as substrate as previously described [17]. u-PA
events or other effects experienced during the treat-            activity was measured in mucosal homogenates by the
ment. Six patients received the saline enema first and           colorimetric method of Coleman and Green [18]. The
four the butyrate enema first.                                   assay has previously been shown to be specific for u-
   Four biopsies were taken on each occasion. Two               PA and the results not affected by the addition of
were placed immediately in Methacarn fixative and,               antibody specific for tissue plasminogen activator [19].
after 1 h, transferred to 100 % ethanol and stored at           Because of the presence of plasminogen in the assay,
k20 mC. The other two biopsies were placed in 1 ml of           the activities of both pro-u-PA and u-PA are
ice-cold Tris–mannitol buffer (50 mM -mannitol,                 measured. Enzyme activities were expressed relative to
2 mM Trizma base in dH O, pH 7.4) and stored at                 the mucosal protein content which was measured
                            #                                   using bovine γ-globulin as standard [20].
k20 mC. The biopsies were coded in order to blind the
investigator to the timing of the biopsy.
                                                                Statistical evaluation
Enema composition
                                                                  Results are expressed as meanspS.E.M. unless
  Butyrate enemas comprised sodium n-butyrate at a              otherwise stated. Paired data were compared using a
concentration of 80 mM at pH 7.0. Osmolality was                two-tailed paired t-test. Because of the study’s cross-
                                                    Topical butyrate and the rectum                                                           673

over design, the order in which the enemas were
applied was taken into consideration but found to
have no influence on the results. The number of bowel
actions over the 4-day period during saline or butyrate
enema treatment was compared using the Wilcoxon
signed rank test. A P value of 0.05 or less was
considered statistically significant.

Compliance to and tolerability of butyrate and saline
   All subjects were able to retain the enemas for at
least 1 h. Compliance, as assessed by daily diary cards,
was very good with only one subject failing to use the
final butyrate enema due to spillage. Compared with                          Figure 2 Distribution of epithelial cells staining with
their normal bowel habit, all but one subject ex-                           PCNA within the crypt in rectal mucosa before (‘Pre’)
perienced reduced frequency of bowel actions after                          and after 4 days of enemas containing either saline or
commencement of the low fibre diet. Over the 4-day                           butyrate
treatment period, the median (range) number of times                        The crypt was divided into five equal parts, quintile 1 being at the
the subjects opened their bowels was 7 (1–11) during                        base of the crypt. The results are expressed as the mean of the
use of saline enemas, which was significantly more                           percentage of stained cells in that quintile. There were no
frequent than that during butyrate enema treatment [4                       significant differences across the groups.
(2–7) ; P l 0.043]. Only one patient reported an
adverse event, comprising the sensation of being                         use of enemas per se (irrespective of whether they
‘ bound up ’ together with tiredness during butyrate                     contained saline or butyrate) resulted in significantly
enema therapy only. The number of bowel actions                          longer crypts. Likewise, the number of PCNA-positive
during butyrate enema therapy in this patient was 4                      cells per crypt tended to be greater after the use of
compared with 10 during saline enema therapy.                            enemas (Figure 1B), but this only reached statistical
                                                                         significance in association with butyrate enemas.
                                                                         However, the labelling index was unaffected (Figure
Effect on proliferative kinetics
                                                                         1C) and the distribution of PCNA-positive cells across
   Histology of all biopsies was normal and no features                  the 5 quintiles was similar (Figure 2). When the effect
of inflammation were seen. As shown in Figure 1A, the                     of butyrate was compared with that of saline enemas,

  Figure 1 Epithelial proliferative kinetics in rectal mucosa before (‘Pre’) and after 4 days of enemas containing either saline or

  Indices measured were (A) the crypt column height (CCH), (B) the number of PCNA-positive cells per crypt column, and (C) the labelling
  index (number of PCNA-positive cells per 100 crypt cells). The results are expressed as meansjS.E.M. and the P values shown derive from
  paired t-tests.
674                                                           P. R. Gibson and others

                                                                            butyrate enemas compared with those after treatment
                                                                            with saline enemas (P l 0.045). u-PA activities after
                                                                            butyrate treatment also tended to be lower than those
                                                                            before enema therapy but failed to reach statistical
                                                                            significance (P l 0.074). Saline enemas had no effect
                                                                            on u-PA activity compared with pretreatment levels (P
                                                                            l 0.61) The order in which the enemas were given had
                                                                            no influence upon the results.

                                                                               Previous studies of the direct effects of butyrate or a
                                                                            combination of short-chain fatty acids on the colonic
                                                                            mucosa in humans and experimental animals have
                                                                            used one of three models : atrophic colon due to
                                                                            luminal butyrate deficiency after faecal diversion or
                                                                            elemental diet [11,12], hyperproliferating colon due to
                                                                            bile salt injury [21], or inflamed colon [14]. In these
                                                                            models, the control comparator comprises abnormal
                                                                            mucosa and this creates difficulties in interpreting the
                                                                            physiological effects. For example, in butyrate defi-
                                                                            ciency, the mucosa is atrophic mainly because of the
                                                                            dependence of the epithelium on butyrate as an energy
                                                                            source [2]. It is therefore difficult to distinguish effects
                                                                            of enhanced energy supply per se from the effects
                                                                            directly mediated by butyrate. Similarly, in colitis
                                                                            models, distinguishing the effects of the direct action of
                                                                            butyrate from those indirectly mediated by reduction
                                                                            in mucosal inflammation is not readily achieved. An
                                                                            alternative approach is to identify the effects of
                                                                            butyrate in vivo by delivering more butyrate to the
                                                                            distal colon via the ingestion of slowly fermented fibres
                                                                            such as wheat bran [13]. This dietary change is,
                                                                            however, associated with multiple other effects on
                                                                            luminal contents, such as changes in pH and in the
                                                                            concentrations of other short-chain fatty acids. It
                                                                            introduces, therefore, too many variables to be able to
                                                                            confidently dissect the effects of butyrate alone. The
                                                                            approach applied in the present study – to examine the
                                                                            effect of supplementing butyrate delivery to a histo-
  Figure 3 Enzyme activities in rectal mucosa before                        logically normal rectum (albeit mostly in patients with
  (‘Pre’) and after 4 days of enemas containing either saline               previous adenomas) – has not previously been re-
  or butyrate                                                               ported. An attempt was made to reduce luminal
  Indices measured were (A) the crypt alkaline phosphatase ac-
                                                                            butyrate levels by the use of a low residue diet. This did
  tivities and (B) urokinase activities expressed relative to mucosal       constipate most of the subjects but the rectal mucosa
  protein content. The results are shown as meansjS.E.M. and the            showed no histological evidence of energy deficiency
  P value shown derives from a paired t-test.                               since the crypts were of normal length. Thus the
                                                                            experimental design addressed the issue of whether a
                                                                            modest increase in delivery of butyrate to the normal
there was no difference for any of the epithelial kinetic                    rectal mucosa induced changes.
indices.                                                                       Butyrate has been shown to cause a variety of effects
                                                                            on epithelial proliferative kinetics. It suppresses cell
                                                                            proliferation [4] and induces apoptosis [5] in colon
Effect on enzyme activities
                                                                            cancer cell lines. Treatment of normal colonic mucosa
  Myeloperoxidase activity was very low (                                   in organ culture with butyrate results in shortening of
0.8 units\min per mg of protein) in all biopsies which                      crypts with an increased relative number of prolifer-
corresponded to the lack of inflammation evident                             ating cells [22]. In vivo, luminally instilled butyrate
histologically. Alkaline phosphatase activities were                        increases epithelial proliferation and crypt length in
similar across the groups (Figure 3). In contrast, u-PA                     atrophic mucosa [11] but decreases proliferation in
activities differed across the groups (Figure 3). u-PA                       inflamed mucosa [14]. Despite these varied but sub-
activities were significantly lower after treatment with                     stantial effects, rectal butyrate supplementation under
                                               Topical butyrate and the rectum                                         675

the conditions of the present study did not affect                inhibition of histone deacetylase [27]. In cells not
epithelial population kinetics. These findings mimic              dependent upon butyrate for its energy supply and not
butyrate’s effect on isolated normal crypt cell popu-             usually exposed to butyrate, such as cell lines, this
lations in which no change in the rate of DNA                    effect is likely to predominate. In normal colonic
synthesis nor DNA content was observed [6].                      epithelium, histone deacetylase may be less sensitive to
    The use of an enema per se did, however, induce              butyrate-mediated suppression due to chronic ex-
longer crypts, most probably by increasing prolifer-             posure to butyrate. The mechanism by which butyrate
ative activity to support the expanded population. The           inhibits u-PA synthesis is not known but, like alkaline
underlying mechanisms are unclear but may involve a              phosphatase [27], may not be due to inhibition of
‘ cleansing action ’ of either enema in washing away             histone deacetylase. The disparity of findings, there-
toxic substances that might contribute to cell death in          fore, might reflect different mechanisms of action of
the surface epithelium. Reduction of such injury may             butyrate, each having different concentration curves.
lead to prolongation of the life span of cells and                  Secondly, differences in butyrate delivery to the
subsequent lengthening of crypts. Alternatively, the             target cells may be responsible. It is likely that
effect may represent the result of ‘ physical stimulation ’       luminally instilled butyrate is being effectively de-
of the colonic epithelium by luminal contents as                 livered to the surface epithelium. Since this is a major
previously reported in mice [23].                                site of u-PA production in the mucosa (P. R. Gibson,
    Mucosal alkaline phosphatase activity was un-                I. Birchall, O. Rosella, V. Albert, C. F. Finch, D. H.
affected by butyrate supplementation. Alkaline phos-              Barkla and G. P. Young, unpublished work), butyrate
phatase has previously been used as a marker of                  will have the best chance of exerting its modulatory
differentiation [6,7] because of its gradient of ex-              effects on that enzyme. In contrast, proliferating cells
pression from crypt to surface [24]. However, whether            are present in the lower portion of the colonic crypt
measurement of mucosal alkaline phosphatase ac-                  which, by virtue of the unidirectional flow of mucus
tivity is a sensitive or even valid marker of differen-           towards the lumen, would be exposed to lower
tiation remains uncertain (P. R. Gibson, R. Nov, M.              concentrations of butyrate. In support of this view is
Fielding, A. McIntyre, C. F. Finch, O. Rosella, J. M.            the observation that butyrate does inhibit abnormal
Mariadason, D. H. Barkla and G. P. Young, unpub-                 proliferation when the proliferative compartment is
lished work). While butyrate potently stimulates cell            expanded into the upper crypt, as observed in bile-
differentiation in cell lines [4], there is no evidence that      acid-injured colon [21] and in ulcerative colitis [14].
butyrate exerts this effect on normal colonic epithelial             The biological significance of reduced mucosal u-
cells. Normal crypt cells studied in vitro neither change        PA activity is uncertain. In the healthy subjects in the
their expression of alkaline phosphatase nor change              present study, a change in mucosal u-PA activity did
the rate of glycoprotein synthesis on exposure to                not correlate with a change in epithelial kinetic indices.
butyrate [6]. Furthermore, in colonic mucosa in organ            However, reduction of mucosal u-PA activity might be
culture, butyrate leads to a loss of mucosal alkaline            important in colitis or hyperproliferating epithelium
phosphatase activity, probably due to loss of mature             induced in rats by the ingestion of diets containing
surface epithelial cells by the induction of apoptosis           resistant starch or soluble fibre, where mucosal
[22].                                                            u-PA activity is elevated (P. R. Gibson, I. Birchall,
    In contrast to the lack of a butyrate-specific effect on       O. Rosella, V. Albert, C. F. Finch, D. H. Barkla and
epithelial kinetics or alkaline phosphatase activity,            G. P. Young, unpublished work and [29]). Its reduc-
butyrate enemas did significantly reduce mucosal                  tion in the surface epithelium might stabilize cell-
u-PA activity independently of an effect of the enema             substratum adhesion and thereby reduce the rate of
itself. This effect was similar to that observed in rats          epithelial death. This would, in turn, lead to a
ingesting wheat bran where mucosal u-PA activity was             reduction in the rate of epithelial proliferation. In
significantly reduced compared with that in rats fed              support of this contention, mucosal u-PA activity in
diets containing soluble or no fibre [13]. Wheat bran             the distal colon of rats directly correlates with pro-
leads to considerable elevation of distal colonic lumi-          liferative indices (P. R. Gibson, I. Birchall, O. Rosella,
nal butyrate concentrations [26]. Butyrate-mediated              V. Albert, C. F. Finch, D. H. Barkla and G. P.
reduction of mucosal u-PA activity also mimics the               Young, unpublished work). In colitis, the effect of
effect we have previously observed of butyrate on                 butyrate on u-PA activity might be one mechanism by
u-PA expression and secretion by isolated colonic                which it reduces inflammatory activity when given by
crypts [8].                                                      enema [14].
    The disparity of effects of butyrate on two different             In conclusion, rectal supplementation of butyrate in
end-points, epithelial proliferation and mucosal u-PA            humans with a histologically normal rectum has no
activity, might reflect either or both of two possibilities.      discernible effect on rectal epithelial proliferative
Firstly, the mechanisms by which butyrate acts on                kinetics beyond that induced by an enema itself nor on
these might differ. Butyrate-mediated stimulation of              differentiation, but inhibits mucosal u-PA activity.
epithelial proliferation is best seen in butyrate-starved        The findings mimic those observed with isolated
cells. Thus, stimulation is seen only when mucosa is             normal colonic crypt cells. The results demonstrate a
atrophic due to nutrient depletion. Recent evidence              potential mechanism – stabilizing cell-substratum ad-
has shown that butyrate inhibits cell proliferation by           hesion in surface epithelium – by which elevating distal
676                                                             P. R. Gibson and others

colonic luminal butyrate concentrations might be                               14. Scheppach W, Sommer H, Kirchner T, et al. (1992) Effect of
beneficial in patients with colitis or hyperproliferative                           butyrate enemas on the colonic mucosa in distal ulcerative colitis.
                                                                                   Gastroenterology 103: 51–6.
large bowel epithelium.                                                        15. Macrae FA, Kilias D, Sharpe K, Hughes NR, Young GP, MacLennan
                                                                                   R. (1994) Rectal epithelial cell proliferation: comparison of errors
REFERENCES                                                                         in measurement with intersubject variance. J Cell Biochem (Suppl.
                                                                                   19): 84–90.
 1. Cummings JH, Bingham SA. (1987) Dietary fibre, fermentation and             16. Yamada M, Mori M, Sagimura T. (1981) Purification and
    large bowel cancer. Cancer Surveys 6: 601–11.                                  characterisation of small molecular weight myeloperoxidase from
 2. Roediger WEW. (1980) Role of anaerobic bacteria in the metabolic               human leukemic HL-60 cells. Biochemistry 20: 766–71.
    welfare of the colonic mucosa in man. Gut 21: 793–8.                       17. Young GP, Rose IS, Cropper S, Seetharam S, Alpers DH. (1984)
 3. Bond JH, Levit MD. (1976) Fate of soluble carbohydrate in the                  Hepatic clearance of rat plasma intestinal alkaline phosphatase. Am
    colon of rats and man. J Clin Invest 57: 1158–64.                              J Physiol 247: G419–26.
 4. Young, GP, Gibson PR. (1995) Butyrate and the human cancer cell.           18. Coleman PL, Green GDJ. (1981) A sensitive coupled assay for
    In: Cummings JH, Rombeau JL, Sakata T, eds. Physiological and                  plasminogen activator using a thiol ester substrate for plasmin.
    clinical aspects of short chain fatty acids. Cambridge: Cambridge
                                                                                   Ann NY Acad Sci 370: 617–26.
    University Press, pp. 319–35.
                                                                               19. Sim PS, Stephens RW, Fayle DRH, Doe WF. (1988) Urokinase-type
 5. Hague A, Manning AM, Hanolon KA, Huschtscha LI, Hart D,
                                                                                   plasminogen activator in colorectal carcinomas and adenomatous
    Paraskeva C. (1993) Sodium butyrate induces apoptosis in human
                                                                                   polyps: quantitative expression of active and proenzyme. Int J
    colonic tumour cell lines in a p53-dependent pathway: implications
                                                                                   Cancer 42: 483–8.
    for the possible role of dietary fibre in the prevention of large
                                                                               20. Bradford M. (1976) A rapid and sensitive method for the
    bowel cancer. Int J Cancer 55: 498–505.
                                                                                   quantitation of microgram quantities of protein utilizing the
 6. Gibson P, Rosella O, Nov R, Young G. (1995) Colonic epithelium
                                                                                   principle of protein-dye binding. Anal Biochem 72: 248–51.
    is diffusely abnormal in ulcerative colitis and colorectal cancer.
                                                                               21. Velazquez OC, Zhou D, Seto R, et al. (1996) In vivo crypt surface
    Gut 36: 857–63.
 7. Gibson PR, Moeller I, Kagelari O, Folino M, Young GP. (1991)                   proliferation is decreased by butyrate and increased by
    Contrasting effects of butyrate on expression of phenotypic                     deoxycholate in normal rat colon: associated in vivo effects on c-fos
    markers of differentiation in neoplastic and non-neoplastic colonic             and c-jun expression. J Parenter Enteral Nutr 20: 243–50.
    epithelial cells in vitro. J Gastroenterol Hepatol 7: 165–72.              22. Mariadason JM, Catto-Smith A, Gibson PR. (1997) Butyrate
 8. Gibson PR, Rosella O, Rosella G, Young GP. (1994) Butyrate is a                increases permeability and alters epithelial kinetics and brush
    potent inhibitor of urokinase secretion by normal colonic                      border hydrolases activities in rat colon ex vivo. Gastroenterology
    epithelium in vitro. Gastroenterology 107: 410–19.                             112: A1033.
 9. Gibson PR, Rosella O. (1996) Abnormalities in the urokinase                23. Stragand JJ, Hagemann RF. (1977) Effect of lumenal contents on
    system in colonic crypts from patients with ulcerative colitis.                colonic cell replacement. Am J Physiol 233: E208–11.
    Inflammatory Bowel Diseases 2: 105–14.                                      24. Bell C, Williams C. (1979) Histochemical demonstration of alkaline
10. Gibson PR, Rosella O. (1995) Interleukin-8 secretion by colonic                phosphatase in human large intestine, normal and diseased.
    crypt cells in vitro: a response to injury that is suppressed by               Histochemistry 60: 85–9.
    butyrate and enhanced in inflammatory bowel disease. Gut 37:                25. Reference deleted.
    536–43.                                                                    26. McIntyre A, Young GP, Taranto T, Gibson PR, Ward PB. (1991)
11. Kripke SA, Fox AD, Berman JL, Settle RG, Rombeau JL. (1989)                    Different fibers have different regional effects on luminal contents
    Stimulation of intestinal mucosal growth with intracolonic infusion            of rat colon. Gastroenterology 101: 1274–81.
    of short chain fatty acids. J Parenter Enteral Nutr 13: 109–16.            27. Archer S, Meng S, Hodin RA. (1997) Molecular mechanisms
12. Harig JM, Soergel KH, Komorowski RA, Wood CM. (1989)                           underlying the effects of sodium butyrate on intestinal epithelia:
    Treatment of diversion colitis with short-chain fatty acid irrigation.         the role of histone hyperacetylation. Gastroenterology 112: A860.
    N Engl J Med 320: 23–8.                                                    28. Reference deleted.
13. Gibson P, Folino M, McIntyre A, Rosella O, Finch C, Young G.               29. Elliott R, Stephens RW, Doe WF. (1987) Expression of urokinase-
    (1995) Dietary modulation of colonic mucosal urokinase activity in             type plasminogen activator in the mucosal lesions of inflammatory
    rats. J Gastroenterol Hepatol 10: 324–30.                                      bowel disease. J Gastroenterol Hepatol 2: 517–23.

Received 24 September 1997/2 February 1998; accepted 2 February 1998

To top